• Company: Wise Group (Nasdaq: WSE / London: WISE)

• Market Cap: $12.4 bn

• EV: $10.6 bn

• Float: 69.8%

• Net Cash: $1.7 bn

• Return on Equity: >30%

• Organic Growth: >35%

• Addressable market to target: ~99%

• Normalized earnings multiple: ~16x

• PEG: ~0.5x

  ---

DISCLAIMER & DISCLOSURE: The author is invested in Wise Group at the date of publication but that may change. The views expressed are those of the author and may change without notice. The author has no duty or obligation to update this information. Some content is sourced from third parties believed to be reliable, but accuracy is not guaranteed. Forward-looking statements involve assumptions, risks, and uncertainties, meaning actual outcomes may differ from those envisaged in this analysis. Past performance is not indicative of future results. All investments carry risk, including financial loss. This analysis is for educational purposes only and does not constitute investment advice or recommendations of any kind. Conduct your own research and seek professional advice before investing.

Executive Summary

“This is one of the cleanest, most compelling investment set ups I’ve seen in a while”, James Emanuel

Wise is not a traditional business.

It doesn’t need branches, inventories, manufacturing plants, or large lending portfolios to grow. That matters.

We find ourselves in a world where nationalism trumps globalization. Tariffs, supply chain disruption, geo-political turmoil, inflation and rising capital costs dominate the news. Yet wise is largely insulated from the current economic environment.

Higher interest rates become a tailwind through increased net interest income.

Plus, this is not a business that can be disrupted by AI.

In fact, this is one of the most defensive counter-cyclical plays out there.

As the business scales, every new customer, transaction, and partner strengthens the platform.

The result is a rare combination of characteristics that investors rarely find in a single company: rapid growth, high margins, strong free cash flow generation, and a fortress balance sheet.

The business has counter-positioned itself in an industry that was ripe for disruption. It attacked one of the most profitable and customer-unfriendly corners of banking: international money transfers.

Competitors are unable to compete without cannibalizing their own business. Many have chosen a different route. Rather than compete directly, they partner with Wise and use its infrastructure because its transfer rails are faster, cheaper, and more efficient than their own.

The business has a founder CEO who is a missionary, not a mercenary.

He is unusually well aligned with shareholders. He draws a very modest salary, receives no stock options, and remains heavily invested through his own ownership stake.

His approach to running the company echoes that of Jeff Bezos, Warren Buffett, Mark Leonard, Jim Sinegal and Sam Walton.

The customer always comes first.

Scale economics are shared.

Profits are the outcome of delivering greater value, not the primary focus.

The network continues to deepen, creating a growing moat around the platform.

This business has all the characteristics that Nick Sleep would look for in a company.

Recent share price weakness has created an opportunity.

Sentiment deteriorated following reports that Belgian prosecutors were investigating money laundering activity involving third parties and that Wise was assisting with enquiries. No allegations have been made against the company to date, yet the share price is off about 20%.

At the same time, the familiar abhorrent practice of US lawyers trying to encourage shareholders to bring a class action law suit against any public company which experiences a drop in its share price has added noise and further soured market sentiment.

Yet under the surface, the underlying business remains exceptionally strong. Revenue continues to grow at more than 35% annually, and its all organic. When a business grows at this rate, it doubles its top line every two years. Layer in opertaing leverage, and earnings grow even faster.

Gross margins exceed 80%. Pre-tax margins are deliberately capped in the mid-to-high teens (more on this shortly). Cash conversion consistently exceeds 100%, supported by favourable working capital dynamics. The balance sheet holds substantial net cash, while the capital-light nature of the business means very little incremental investment is required to support future growth.

The quality of the business is equally impressive. Revenue is highly diversified across millions of customers. Retention rates are strong. Customer concentration risk is negligible.

Most importantly, Wise remains in the early stages of its journey. Despite its success to date, the company still serves only a tiny fraction of the estimated $43 trillion global cross-border payments market.

With approximately 1% market share, the runway for growth remains enormous.

High-quality businesses rarely become available at attractive prices. When they do, it is often because short-term concerns obscure long-term fundamentals. Wise appears to fit that description remarkably well.

As Buffett once said, "When it rains gold, reach for a bucket, not a thimble".

Contents

• The Company

• The Business Model

• The People

• Competition & the Moat

• The Numbers

• Mr Market

• The Money Laundering Issue

• The Bear Case

• The Bull Case

• What to Watch For

• Conclusion: Is Wise a Good Investment?

The Company

Very few companies are as misunderstood as Wise Group (Nasdaq: WSE).

Ask someone in the investment community about Wise. The response will be to question whether Wise can scale its volume fast enough to maintain profit margins while continuing to reduce its cross-border take rate.

No! That’s upside-down, back-to-front thinking.

This is a fundamental misunderstanding of how the business operates. It puts the cart before the horse. It’s all wrong.

As a company grows it becomes more efficient at producing goods or services. Scale delivers economic cost benefits.

Rather than capturing that value by allowing margins to expand, a few exceptional businesses return that value to customers in the form of lower pricing. Think Costco, Amazon or Walmart.

These companies have something else in common. They grow rapidly to become market leaders and they enjoy outsized success. The same golden threads running through all of these businesses is no coincidence. They are the product of management with outstanding business acumen.

Spotting these golden threads early in the journey of a company is the holy grail for any investor.

They were evident in the inaugural shareholder letter of Jeff Bezos at Amazon back in 1997. Anyone that spotted them at the beginning of the journey and invested would have turned $10,000 into almost $30 million today.

I look for these golden threads constantly, but they are exceptionally rare.

Wise Group is one of those rare exceptions.

Wise diligently focuses on maintaining PBT margins in the 15-20% range. If they start to creep above that range it reduces its take rate, thereby passing the economic benefits of scale to its customers in the form of lower pricing (take-rates).

This structural reduction acts as a customer acquisition engine. Cheaper costs drive higher transaction volumes. Higher volumes at the same margins result in ever increasing cash returns. Lower prices result in higher profits.

The exact same model has been very successfully deployed by GEICO, the Berkshire Hathaway owned insurance business. Rather than optimize combined ratios (an insurance proxy for margins) the company strives to reduce pricing to drive customer acquisition.

This is why the question being asked by most investors in the market is, as Charlie Munger used to say, ‘upside-down, back-to-front thinking’. Margins dictate pricing, not the other way around; they are never under threat.

Wise has grown it’s success running exactly the same playbook as Amazon, Costco and Walmart. The only difference is that Wise is only at the start of its journey.

“Fifteen years ago, we set out with a simple but ambitious goal: to make moving and managing money around the world as fast, simple and cost-effective as sending an email. We’ve come a long way since then. In the last financial year, we helped… move over $243 billion across borders instantly and at a fraction of the cost of traditional providers, saving our customers more than $3.3 billion in fees. Still, with $43 trillion moved across borders each year globally, we’re only getting started.”

Kristo Käärmann, co-founder and CEO at Wise

The other thing Wise has in common with these other uber-successful businesses is that it is not a single thing that makes it exceptional, but lots of small things in combination that are difficult, if not impossible, to replicate.

To the casual observer, Wise appears to be a sleek, consumer-facing payments application. However, Wise is more accurately defined as a financial infrastructure business operating under a consumer “wrapper”.

Wise began life as a transactional peer-to-peer “hack” (more on this shortly) and has transformed into a foundational settlement layer for the global financial system.

By relentlessly pursuing its differentiated model, Wise has created a widening operational moat that traditional incumbents are both technologically and economically unable to cross.

The business demonstrates exceptional unit economics, high returns on equity owing to having a capital light business model and a flywheel that fuels strong double digit growth rates. Even more remarkable is that 70% of customers are acquired via word-of-mouth recommendtions keeping marketing spend at a remarkably low 4% of income.

The company is domiciled in Jersey, one of the British Channel Islands. It recently transitioned its primary stock market listing to the Nasdaq (Ticker: WSE), retaining London as a secondary listing (WISE.L).

The Business Model

The ‘Hack’ That Spawned A Global Empire

Kristo Käärmann was a young Estonian living in London, working for Deloitte. He was paid in British pounds but serviced a mortgage in Euros on a property back home in Estonia.

Every month he would watch in horror as 5% or more of his hard earned income disappeared in opaque exchange rate markups and administrative fees imposed by his bank when exchanging the currencies. To Käärmann, it felt like an unavoidable tax imposed by the financial system which was entirely disproportionate to the service being provided.

It was pure exploitation and profiteering. It didn’t feel nice.

Then Käärmann met Taavet Hinrikus, another Estonian living in London who had the same problem. However, Hinrikus was employed at Skype and remunerated in Euros, although his living expenses were in British pounds.

Immediately it became aparent that there was an elegant solution to their problem.

Each month Käärmann could transfer his British pounds to Hinrikus, who would then reciprocate by transfering Euros back to Käärmann.

A transfer between personal accounts attracted no banking fees. They had achieved foreign exchange without any frictional costs.

It was a pure peer-to-peer swap that removed the necessity for international wire transfers entirely, providing the foundational logic for what would eventually become the foundations of the settlement network we are discussing today.

Because their methodology involved transfers, and it was a smarter way to operate, they decided to call the sytem ‘TransferWise’.

They raised a seed round of $1.3 million in 2012, and in its inaugural year, the company processed £8 million in transactions. Since then they haven’t looked back and by FY2025, that figure reached an astounding £145 billion.

This evolution from a simple P2P matching engine to a sophisticated global infrastructure business was punctuated by the 2021 rebranding from ‘TransferWise’ to ‘Wise’.

The renaming signaled the transition from a single-product remittance tool to a multi-product financial ecosystem and settlement layer for the global economy.

Today, Wise serves 16 million customers and manages over £20 billion in customer deposits.

Share

The People

Insiders own roughly 30% of Wise, which is reassuring for external shareholders. When management has meaningful skin in the game, incentives tend to be better aligned and capital allocation decisions tend to improve.

That alignment starts with founder and CEO Kristo Käärmann. His compensation structure is unusually shareholder friendly. He receives a modest salary of around £200,000, with no bonus and no stock-based compensation.

His economic interest comes almost entirely from his ownership stake, consisting of approximately 18% of the Class A shares and 47% of the Class B shares, currently worth around £3.3 billion. If shareholders do well, he does well. If they do poorly, he feels the pain alongside them. It’s a structure that resembles the approach taken by Warren Buffett at Berkshire Hathaway and Mark Leonard at Constellation Software.

In a world where many executives are paid handsomely regardless of outcomes, Käärmann sits in a very small group of genuinely aligned founder CEOs.

His influence extends far beyond ownership. Käärmann has shaped Wise's product strategy, culture, and long-term direction from the beginning. What stands out is his willingness to prioritise customer outcomes over short-term profitability. While much of the financial industry focuses on extracting maximum value from customers, Käärmann’s strategy involves deliberately and continuously lowering the company’s take rate to power growth and delight customers.

That philosophy is deeply embedded within the organisation. The company's internal values repeatedly emphasise transparency, humility, customer obsession, and long-term thinking. It is a mindset more commonly associated with founders such as Jeff Bezos, Sam Walton, or Jim Sinegal than with financial services executives.

Actions reinforce the rhetoric. Since Wise's IPO, Käärmann has not sold a single share. For a founder whose stake is worth billions, that is a notable signal. It suggests his focus remains on building long-term value rather than monetising success at the first opportunity.

The other co-founder, Taavet Hinrikus, is no longer involved in day-to-day operations but continues to hold a substantial Class B stake and remains an important shareholder.

The broader management team also appears strong. CFO Emmanuel Thomassin joined Wise after a decade as CFO of Delivery Hero, where he helped oversee a period of rapid growth and served on the Management Board. Before that he was CFO of MetaDesign. His background in scaling fast-growing businesses makes him a logical fit for Wise's current stage of development.

Chief Technology Officer Harsh Sinha has been with the company since 2015, having previously held senior roles at PayPal and eBay, while Chief Product Officer Nilan Peiris has been with Wise since 2014 and now oversees product, design, analytics, and sales across the business. Together they provide continuity across the key functions that underpin Wise's competitive position.

The non-executive board is equally impressive. Chairman David Wells previously served as CFO of Netflix. Other directors bring experience from organisations such as Euronext, Intercontinental Exchange, Morgan Stanley, Grab (the UBER of Asia), and Andreessen Horowitz. One particularly interesting appointment is Ingo Uytdehaage, CEO of Adyen, a dominant player in another corner of the payments ecosystem. Could there perhaps be a marriage of these two businesses at some point in the future?

What stands out is that the board does not appear to have been assembled simply to satisfy governance requirements. Many UK listed companies treat independent directors as a compliance exercise. Wise appears to have approached the process more strategically, bringing together individuals with relevant operating, financial, technology, and payments experience.

While Käärmann has no incentive package beyond his equity ownership, other members of the executive team participate in a Long-Term Incentive Plan (LTIP). The structure is thoughtfully designed. Awards are linked to a combination of total shareholder returns, transaction volume growth above 13%, and customer Net Promoter Scores (NPS).

The baseline threshold NPS is 63 and that qualifies for a 25% payout; the full 100% payout is only triggered by an NPS above 70. This reduces the risk of management pursuing volume growth or short-term profits at the expense of the customer experience. The message is clear: growth matters, but not at the expense of customer satisfaction.

The scale of these incentives is not insignificant. Under the LTIP, Thomassin, the CFO, who has a base salary of 2.5x the CEO, is able to achieve a maximum LTIP payout of 4x his base salary, which would take his total remuneration to 10x that of the CEO founder. That may appear unusual, but it demonstrates that Käärmann is not egocentric and that his decisions are all driven by what is best for the business. As a major shareholder he will benefit, alongside external investors, if his management team performs. Once again, there are directly parallels between Käärmann’s approach and that of Buffett, Leonard and Bezos.

The only governance issue worth flagging relates to Wise's dual-class share structure. Class B shares carry nine votes each, giving Käärmann approximately 48% of the company's voting power. At the time of the 2021 IPO, investors were told these enhanced voting rights would expire after five years, eventually returning the company to a one-share, one-vote structure. However, when Wise moved its primary listing from London to Nasdaq, shareholders were asked to prolong the arrangement for another decade.

This proved controversial. The co-founder, Taavet Hinrikus, publicly opposed the proposal through an open letter, arguing that "Wise should adhere to the foundational corporate governance principle of one share, one vote to protect its long-term business value and the interests of all owners equally." Despite the criticism, the resolution passed comfortably, but it did reveal a degree of tension that exists between major shareholders.

The dual-class structure gives Wise the stability and long-term orientation that often accompanies founder control, but it also reduces the ability of minority shareholders to influence governance. Whether that is a feature or a flaw will depend largely on one's view of founder-led businesses.

Competition and the Moat

Wise occupies a unique position within financial services. Unlike many of its competitors, it is essentially a pure-play cross-border payments business. The competitive landscape spans several categories, including incumbent banks, traditional money transfer operators, fintechs, and digital-first neobanks, but few companies are as singularly focused on reducing the cost and friction of international money movement.

What makes Wise particularly interesting is that it has counter-positioned itself against the industry. Traditional banks and money transfer providers have historically earned attractive economics from opaque foreign exchange spreads, hidden fees, and inefficient processes. Wise built its entire value proposition around eliminating those frictions. As a result, many incumbents face a dilemma. To compete effectively with Wise would require them to lower prices, increase transparency, and potentially cannibalise profitable revenue streams within their existing businesses. In many cases, they are either unable or unwilling to do so.

HSBC’s experience with Zing illustrates the challenge. Zing was specifically designed to compete with Wise and Revolut. Rather than relying on HSBC’s legacy infrastructure, it was built on a separate technology stack using the XYB cloud banking platform and operated under its own independent e-money licence. HSBC effectively recognised that its existing systems and business model were poorly suited to competing with digital-first challengers and attempted to build a dedicated competitor from scratch.

Despite the resources, brand recognition, and distribution advantages available to HSBC, the initiative failed. Zing was discontinued shortly after launch and ultimately shut down altogether. The episode highlights a recurring theme in financial services. Building a product that looks like Wise is relatively easy. Building an organisation whose economics, incentives, culture, and operating model are designed around permanently lowering prices for customers is considerably harder. That distinction helps explain why Wise continues to occupy a category of its own within the payments ecosystem.

To better understand Wise, one must first understand the “broken” system it disrupts.

The Society for Worldwide Interbank Financial Telecommunication (SWIFT) has governed cross-border movement for decades, but it remains an archaic messaging network rather than a true settlement layer.

If a sender’s bank lacks a direct relationship with the recipient’s bank, the funds must pass through a chain of “correspondent” or “intermediary” banks based entirely on trust. To mitigate risk of money being lost as it passes down the chain, the SWIFT system requires banks to maintain “nostro” (our money in your account) and “vostro” (your money in our account) balances across this chain to create pre-funded liquidity. Capital being tied up in this way justifies some of the fees involved. It’s a fragmented journey through a labyrinth of bureaucratic tollgates.

The process is also characterized by extreme opacity.

The traditional cross-border payments system is surprisingly inefficient. When a customer instructs a transfer, a choice must often be made as to who bears the fees. This means that if you send $100, either your account will be debited ~$105, or else the recipient will only receive ~$95. This is far from ideal.

To make matters worse, the final cost is often unclear at the outset because it depends on how many intermediary banks sit between the sender and the recipient.

Every intermediary bank in the chain extracts a fee for its role in the process. The SWIFT network does not move money itself, it is a messaging system used to update incongruent ledgers across different time zones, which creates delays of between 1 and 5 business days.

Importantly, banks using the SWIFT system don’t process customer FX trade individually. Instead, they net customer positions internally and only settle the residual obligations through the correspondent banking system, significantly reducing costs. Banks generally treat netting as an internal efficiency gain and so use this process to improve their margins.

Cross-border payments remain one of the industry's most profitable products, helping explain why incumbents have shown little enthusiasm for disrupting the status quo. Changing to the Wise model would require a total overhaul of their backend architecture and the cannibalization of their existing model which would be catastrophic to their margins.

Hamilton Helmer, in his book ‘7 Powers’, described this phenomenon as ‘counter-positioning’, and Wise is the perfect case study.

A leaked 2017 Santander memo revealed that if the bank were to match Wise’s pricing, it would destroy 84% of their money transfer profits. In this way, ‘counter-positioning’ becomes an incumbent’s dilemma, and it provides Wise with an almost impenetratable moat.

Global banks collect an estimated £200 billion annually in hidden retail cross-border fees, which ironically is good news for Wise. As Bezos used to say, ‘your margin is my opportunity’.

Wise approaches the problem from an entirely different angle. Rather than relying on correspondent banking networks, it maintains local liquidity pools in the countries in which it operates. By removing the chain of intermediary banks, the company eliminates much of the friction, delay, and cost embedded within the traditional system.

The impact is dramatic. Transfers that previously required several days can often be completed in seconds, and in many cases instantly.

Wise is also available around the clock, seven days a week. Many traditional financial institutions continue to operate within the constraints of banking hours and are closed on weekends and public holidays. In an increasingly digital world, customers expect money to move as easily as information. Wise was built for that.

Yet speed and convenience is not the only advantage Wise has. It’s most compelling advantage is cost.

According to the World Bank, the average digital remittance still costs approximately 6.5% of the amount transferred. By contrast, Wise reported an average cross-border take rate of just 0.52% as of March 2025. In other words, Wise operates with roughly a twelvefold cost advantage relative to the traditional system. Few businesses enjoy such a large and measurable pricing gap versus incumbent competitors.

Its huge competitive advantage is not an arbitrary marketing choice but a structural consequence of fourteen years of meticulous direct rail integration and regulatory licensing.

The regulatory environment is also moving in Wise’s favour. Policymakers increasingly expect transparency around fees and exchange rates, particularly for retail consumers. Hidden charges have come under greater scrutiny, and European regulators have already introduced measures designed to improve fee transparency. The more the industry moves toward clear pricing and customer disclosure, the quicker Wise will grow its market share.

In fact, the speed of growth at Wise is testament to this investment thesis.

However, the low-cost transfer service is only part of the story. Over time, Wise has evolved from a simple cross-border money transfer company into a much broader financial infrastructure provider. That evolution has improved unit economics, increased switching costs, and steadily widened the company’s competitive moat.

Three Phases of Evolution

The development of the business can be understood in three distinct phases.

Phase one was the original peer-to-peer model. Wise matched customers moving money in opposite directions and avoided the need for costly international transfers. The concept was elegant and highly disruptive, but it had limitations. Currency flows are rarely balanced. Large volumes of dollars move to Mexico, for example, but comparatively few pesos flow in the opposite direction. As Wise scaled, these directional imbalances became increasingly difficult to solve through customer matching alone.

This led to phase two. Recognising the limits of the peer-to-peer model, Wise began building a global regulatory and liquidity infrastructure. It secured local licences, established domestic banking relationships, and increasingly used its own balance sheet to provide liquidity. Rather than waiting for matching flows, Wise could complete transactions immediately while managing liquidity and foreign exchange exposure itself. The business evolved from a simple matching platform into an institutional-grade treasury operation capable of supporting significantly larger volumes.

The transition was not without challenges. The Nigerian currency crisis in 2016 exposed the risks associated with operating in emerging market corridors and highlighted the importance of maintaining deep local liquidity pools. Regulatory scrutiny during this period also forced Wise to invest heavily in compliance systems, controls, and risk management capabilities. While costly at the time, these investments have since become an important barrier to entry. Many smaller fintechs can replicate a user interface. Very few can replicate a global compliance infrastructure spanning dozens of jurisdictions.

The third and most important phase was the decision to bypass commercial banks altogether. In 2018, Wise became the first non-bank to gain direct access to the Bank of England’s Real-Time Gross Settlement system. This marked a significant milestone. Rather than relying on intermediary institutions, Wise began connecting directly to national payment rails.

Over the last decade the company has accumulated more than 80 licences and built direct connections across many of the world’s most important domestic payment systems.

Today it has direct participation in networks such as Faster Payments in the UK, SEPA in Europe, Pix in Brazil, Zengin in Japan, and payment systems across Australia, Singapore, Hungary, and the Philippines.

This infrastructure has required enormous investment. Wise now employs more than 1,000 engineers, a five-fold increase over the past six years. Their focus extends well beyond maintaining existing systems. They are continuously expanding direct payment connectivity, improving settlement speed, and reducing transaction costs.

The benefits are substantial. Direct integrations eliminate local intermediaries and significantly reduce variable costs. When Wise connected directly to the UK’s Faster Payments system, partner bank fees reportedly fell by a factor of nine. In the Philippines, direct access reduced costs by approximately 90%. As of 2025, nearly half of all Wise volume flows through these direct connections.

This creates a competitive advantage that is much harder to replicate than software. A competitor would need years of regulatory approvals, technical integration work, compliance investment, and relationship building to recreate what Wise has assembled. The moat is increasingly operational rather than technological.

WISE Platform

Perhaps the most interesting aspect of this evolution is how Wise has turned potential competitors into customers.

Incumbent banks and fintechs struggle to match Wise’s cost structure because doing so would often require dismantling parts of their existing operating model. Rather than competing directly, many have chosen to partner with Wise instead. Through Wise Platform, the company effectively offers its cross-border payment infrastructure as a service.

Banks, neobanks, accounting software providers, and enterprise platforms have integrated Wise's payment rails directly into their own products while maintaining their existing customer relationships.

For partners, the proposition is straightforward. They gain access to a faster, cheaper, and more reliable international payments solution without having to build the infrastructure themselves. In many cases, the fees charged to end customers remain unchanged, meaning the economic benefit is captured through improved margins and operating efficiency.

For Wise, the economics are even more attractive. Instead of acquiring every customer individually, the company embeds itself within the transaction flows of large financial institutions and software platforms. These relationships generate recurring volumes, larger average transaction sizes, and significantly lower customer acquisition costs. Partners effectively become a distribution channel.

This creates a powerful flywheel. Higher transaction volumes improve network efficiency and support further reductions in take rates. Lower prices attract additional customers and partners, which in turn generate more volume. Scale lowers costs, lower costs attract volume, and additional volume further strengthens scale.

The list of partners is increasingly impressive. It includes global banks such as Morgan Stanley, Standard Chartered, Itaú, UniCredit, and Raiffeisen Bank, alongside digital banks such as Monzo and Nubank, as well as software platforms including Xero and a growing number of enterprise resource planning providers.

Although Wise Platform currently accounts for only around 5% of total volume, up from 4% a year earlier, management expects that figure to reach 10% over the medium term. Given the size of the institutions now adopting the platform, that target appears achievable.

What began as a niche consumer-facing transfer service has evolved into a global payments infrastructure business serving both consumers and institutions. The moat is no longer simply lower prices or better technology. It is a dense network of licences, regulatory approvals, payment rail connections, liquidity infrastructure, compliance capabilities, and distribution partnerships that has taken more than a decade to build.

The largest opportunities still lie ahead. Markets such as the United States and India remain underpenetrated relative to their size, providing significant scope for transaction volumes to grow and for Wise’s operational advantages to become even more entrenched.

Beyond Money Transfers

Although Wise is best known for international money transfers, the business has evolved into something much broader. Today it generates revenue from three distinct sources, each reinforcing the others and strengthening the overall economics of the platform.

The first is the core transfer business discussed throughout this report. Customers pay a transparent fee, typically expressed as a percentage of the amount transferred, with pricing varying by currency corridor and payment method.

The second is Wise Platform, the infrastructure business that enables banks, fintechs, and enterprises to offer cross-border payments using Wise’s underlying network.

The third is the accounts business. Through personal and business accounts, customers can hold, receive, spend, and manage money across multiple currencies. These accounts generate revenue through card transactions, account-related services, and income earned on customer balances.

This distinction is important because the long-term value of a Wise customer extends far beyond a single transfer. Customers who adopt a full Wise account tend to remain with the platform longer, use a wider range of services, and generate multiple streams of revenue. Rather than earning solely from foreign exchange transactions, Wise participates in card spending, account activity, balance income, and recurring payment flows. Customer lifetime value increases significantly once an account becomes part of a user’s financial routine.

As a result, management increasingly views account adoption as one of the most important drivers of future growth.

This evolution has naturally raised questions about whether Wise is becoming a bank. The answer, at least for now, is no.

Wise has gradually moved closer to bank-like infrastructure in certain markets, but it remains primarily a payments and e-money institution. In 2025 the company applied for a U.S. national trust bank charter, a move that appears designed primarily to improve access to payment infrastructure and reduce reliance on intermediaries rather than to become a traditional deposit-taking bank. Direct connectivity to Federal Reserve payment rails would further reduce transaction costs and strengthen its competitive position.

The situation in the UK is similar. Wise continues to operate as an authorised e-money institution and has not formally applied for a UK banking licence, although reports suggest management has explored the possibility. This contrasts with competitors such as Revolut, which secured a full UK banking licence and is pursuing a strategy that includes lending, mortgages, and broader retail banking services.

Wise’s priorities appear different. The company does not need a banking licence to hold customer funds because regulators permit licensed payment institutions to safeguard customer money separately from their own balance sheets. This framework allows Wise to offer account functionality without taking on many of the risks associated with traditional banking.

• The personal account has become particularly attractive for internationally mobile consumers. Users can hold balances in more than 40 currencies, receive local account details in multiple jurisdictions, and spend globally through a debit card. By FY2026, Wise had grown its active customer base to approximately 19 million people.

  

  An increasingly important feature is the ability for customers to earn returns on idle balances. Traditional current/checking accounts often pay little or no interest, whereas Wise allows customers in eligible markets to access returns through money market funds and other investment products. This provides an additional incentive to retain larger balances within the ecosystem.

  

• The business account proposition may be even more compelling. Small and medium-sized businesses use Wise to manage international payroll, supplier payments, invoicing, treasury functions, and multi-currency cash management. Customers can receive payments locally in multiple jurisdictions, issue invoices in numerous languages, automate batch payments, and integrate directly with accounting software platforms.

  

For many smaller businesses, international payment costs are not merely an inconvenience but a genuine barrier to trade. If moving money across borders costs hundreds of basis points, profit margins can quickly disappear. Wise materially lowers that friction, making international commerce more accessible for smaller firms and expanding the company's addressable market in the process.

Business customers are also highly attractive economically. They tend to move significantly larger volumes than consumers and often become deeply integrated into operational workflows. Once payroll systems, accounting software, supplier relationships, and treasury processes are connected to Wise, switching becomes increasingly difficult.

Management appears well aware of this opportunity and has indicated that Wise Business will remain a major area of investment going forward.

The market opportunity is enormous. According to industry estimates cited by J.P. Morgan and the IMF, global cross-border payment volumes totalled approximately $195 trillion in 2024 and could exceed $320 trillion by 2032. The portion most relevant to Wise today, including consumer transfers, e-commerce payments, and SMB transactions, is estimated at roughly $43 trillion annually.

Wise's share of this market remains remarkably small. The company has captured approximately 5% of personal transfer volumes, less than 1% of SMB payment volumes, and has barely begun penetrating the large enterprise segment.

This leaves a substantial runway for future growth.

The retention dynamics are equally attractive. Customers who adopt Wise accounts tend to remain active for many years, producing low churn and steadily increasing cohort value over time. As new customers are added each year, the existing customer base continues to compound, creating a growing foundation of recurring activity.

The financial impact is already visible. Revenue generated from account holders has been growing at roughly 66% annually, compared with only 7% growth from transfer-only customers. In many respects, the investment case increasingly revolves around this transition from transactional users to fully engaged account customers.

Business adoption is growing particularly quickly. Business volumes have compounded at approximately 36% annually, comfortably ahead of the 25% growth rate achieved by personal customers. Given that the SMB market alone represents an estimated £19 trillion opportunity and Wise’s penetration remains below 1%, this growth appears far from exhausted.

Today, account holders generate roughly two-thirds of total group income. If current trends continue, that proportion is likely to increase further over time.

One final piece of the ecosystem is Wise Assets, an investment subsidiary that allows account holders to invest part of their balances into global equity index funds or government money market funds. Participation is entirely optional, but it represents another step in Wise’s gradual expansion from a payments company into a broader financial platform. We will examine this business in more detail in the numbers section.

Share

The Numbers

From a capital allocation perspective, Wise displays many of the characteristics associated with high-quality compounders. The business is asset-light, requires relatively little incremental capital to grow, and converts more than 100% of pre-tax profits into free cash flow.

Importantly, growth is not being purchased through aggressive customer acquisition spending. Marketing expenditure remains disciplined at approximately 4% of underlying income, reflecting the strength of the brand, word-of-mouth referrals, and the increasingly important contribution of distribution partners through Wise Platform. The company benefits from a business model where scale drives lower costs, lower costs attract more customers, and additional customers further improve scale.

Reported returns are already impressive. Wise generates a Return on Equity of roughly 30%, a figure that would be considered exceptional for most financial institutions. However, this arguably understates the true economics of the business. Regulatory requirements necessitate the maintenance of substantial cash balances and liquidity reserves, leaving the company with approximately £1.3 billion of net cash on what can fairly be described as a fortress balance sheet. While prudent, this excess liquidity depresses reported returns by inflating the equity base.

Viewed through the lens of invested capital rather than accounting equity, the underlying economics appear considerably stronger. Adjusting for excess cash, Wise’s Return on Invested Capital may be closer to 60%, highlighting the attractive economics of its infrastructure-led model.

Unlike traditional banks, which require significant balance sheet capital to support lending activities, Wise does not make loans. It generates growth primarily through software, payment infrastructure, regulatory licences, and network scale. The result is a business capable of earning exceptional returns without consuming large amounts of incremental capital.

These economics provide considerable strategic flexibility.

Strong free cash flow generation allows Wise to fund product development, expand into new geographies, secure additional regulatory licences, and deepen direct payment rail connectivity without relying on external capital.

It can continue investing heavily in technology and compliance while maintaining a debt-free balance sheet and avoiding the leverage-related constraints that often limit traditional financial institutions.

This combination of high returns, strong cash generation, limited capital requirements, and substantial reinvestment opportunities is rare. It is one of the reasons the investment case increasingly resembles that of a software or infrastructure business rather than a conventional financial institution.

Before diving into the numbers, a brief word of caution is warranted.

Historically, Wise reported its financial results under UK-adopted IFRS. However, following the transfer of its primary listing to the United States in Q2 2026, the company is transitioning to reporting under US GAAP. While the underlying economics of the business have not changed, accounting treatment differs in several areas, creating challenges when making like-for-like comparisons across reporting periods.

To ease this transition, management has restated FY2024, FY2025, H1 2025, and H1 2026 financial information under US GAAP, providing investors with a reasonable degree of continuity. However, the restatement does not extend back to the company’s IPO, meaning longer-term trend analysis still requires care. Historical comparisons spanning multiple years may reflect accounting changes as much as operational performance.

For consistency, this analysis primarily relies on the IFRS figures that management reported dating back to the IPO.

The key takeaway is straightforward: focus on the underlying economics rather than becoming distracted by accounting noise. The reporting framework may be changing, but the business itself remains the same.

REVENUES

One metric remains largely unaffected by accounting changes: revenue. On that front, Wise's performance has been exceptional. Between FY2021 and FY2025, revenue compounded at more than 40% annually, reflecting a combination of strong customer growth, increasing transaction volumes, and the successful expansion of the product ecosystem.

Today, Wise generates revenue from three mutually reinforcing streams, all of which are growing strongly:

1. SEND: core cross-border transfer business,

2. SPEND: card-related revenue, including debit card interchange, currency conversion activity, and account usage, and

3. HOLD & EARN: income generated on customer balances that are safeguarded within the Wise ecosystem.

The importance of this revenue mix is becoming increasingly apparent. What began as a transfer business is steadily evolving into a broader financial platform. As customers adopt accounts, cards, and balance products, revenue becomes more diversified and less dependent on transfer fees alone.

SEND

The SEND segment currently remains the largest of the three, and continues to be driven by customer growth, volume per customer, and take rate.

Active customers reached 13.4 million in H1 FY2026, up 18% year-on-year. Cross-border volume increased 24% to £85 billion, with business volumes growing particularly quickly at 35%. Notably, cross-border revenue increased by only 5% . The slower growth in revenue relative to volume reflects Wise’s deliberate strategy of reducing its take rate.

Importantly, volume per customer continues to rise, suggesting that existing users are entrusting a greater proportion of their financial activity to the platform.

SPEND, HOLD & EARN

The more interesting story lies outside the transfer business.

As of H1 FY2026, non-cross-border revenue, comprising SPEND and HOLD & EARN, represented 41% of total underlying income, up from 37% the previous year. This shift demonstrates that management’s diversification strategy is working.

Card revenue has been a major contributor. Growing account adoption has driven substantial increases in spending activity. Card and other revenue grew 54% in FY2024 and a further 45% in FY2025. During H1 FY2026 alone, customers spent more than £15 billion through Wise cards, generating £132 million of revenue, an increase of 28% year-on-year.

This growth reflects a broader behavioural shift. Customers are no longer using Wise simply to move money internationally. Increasingly, they are using it to hold, spend, receive, and manage money on an ongoing basis.

That behaviour has translated into substantial balance growth. Customer funds held within the platform have increased from approximately £3.9 billion in 2021 to almost £30 billion by FY2026.

Unlike a traditional bank, Wise does not lend these balances. There is no credit book, no mortgage portfolio, and no fractional reserve banking model. Instead, customer funds are safeguarded in a combination of highly rated bank deposits, government securities, and money market funds managed by institutions such as BlackRock and State Street. The objective is safety, liquidity, and regulatory compliance rather than credit creation.

These balances nevertheless generate income.

Wise earns a return on safeguarded funds and shares a significant portion of that return with customers. Management’s stated policy is to retain the first 1% of yield to cover operating costs and then return 80% of any additional yield generated. This approach is consistent with the company’s customer-first philosophy and stands in sharp contrast to many traditional financial institutions.

However, there is an important nuance that investors must understand.

Because Wise is not a licensed deposit-taking bank, customers must actively opt into yield-bearing products. Many choose not to. When this happens, Wise retains the full yield generated on those balances. As a result, reported earnings are often higher than the earnings that would be generated if every eligible customer elected to receive the returns available to them.

Management recognises this distortion and has provided underlying figures alongside reported results since FY2023. Investors should pay close attention to these adjustments when assessing valuation and profitability.

This creates two important sensitivities within the model.

The first is interest rates. Higher rates increase income earned on customer balances, while lower rates reduce it. The second is customer behaviour. If a greater proportion of customers opt into yield-sharing products, Wise will retain less of the interest income generated.

Interestingly, management appears comfortable with both outcomes. The company’s objective is not to maximise short-term profitability but to maximise customer value. Returning more yield to customers strengthens the proposition, deepens engagement, and increases the likelihood that larger balances remain within the ecosystem.

The strategy appears to be working. Customer balances increased 37% year-on-year to £29.4 billion by FY2026. Even as interest rates normalise from recent highs, these balances provide a valuable source of high-margin revenue and an additional layer of economic resilience.

It is useful to think of SEND and SPEND as ‘transaction revenue’ streams. HOLD & EARN is different. It is effectively an additional earnings layer generated from customer balances and will naturally fluctuate with interest rate cycles.

Taken together, Wise expects to generate approximately $2.5 billion of revenue during FY2026.

The composition of that revenue is becoming increasingly attractive. A growing proportion originates from accounts, cards, balances, and platform services rather than solely from transfer fees. This diversification strengthens customer relationships and reduces reliance on any single product line.

Geographically, Europe remains the largest contributor to revenue, reflecting Wise's origins and strongest market position. However, growth outside Europe is occurring at a faster pace. As more banks, fintechs, and enterprises adopt Wise Platform, international expansion is likely to become an increasingly important contributor to future growth. European volumes should continue to increase, but the larger opportunity lies in replicating that success across the rest of the world.

COSTS

Wise’s cost structure is relatively straightforward and can be divided into two broad categories.

The first is cost of sales, which includes transaction processing costs, banking fees, foreign exchange costs, payment network charges, and credit losses.

The second is operating expenses, which encompass employee compensation, technology development, compliance, marketing, and administrative costs. This represents the largest cost category.

The encouraging trend is that cost of sales has been declining as a percentage of revenue. As more transaction volume flows through direct payment rail connections and as higher-margin account and balance products become a larger part of the mix, gross margins continue to improve.

Operating expenses have increased meaningfully, but this should largely be viewed as investment rather than cost inflation.

During H1 FY2026, operating expenses increased 27% year-on-year, or 24% excluding listing-related costs. More than 1,000 employees joined the company during the period, technology spending increased 18%, and management announced plans to substantially increase marketing investment over the medium term.

These expenditures are directed toward expanding the platform, building capacity, strengthening compliance infrastructure, and accelerating customer acquisition. The benefits will likely be realised over many years, making traditional accounting treatment somewhat misleading. Under accrual accounting, these investments are expensed immediately despite creating value that may endure for a decade or more.

Viewed through that lens, the spending appears less like margin pressure and more like a management team aggressively pursuing a very large opportunity.

PROFITABILITY

Wise’s profitability has steadily improved as scale has increased.

Gross margins expanded from 62.6% in 2019 to 72.9% by FY2025. This ten percentage point improvement reflects the growing proportion of volume flowing through direct payment rail connections and the increasing contribution from higher-margin account-related revenue streams.

Yet management does not optimise for margin expansion in the conventional sense.

Like Costco, Wise uses scale efficiencies to reduce prices rather than maximise profitability. The goal is to maintain attractive returns while continuously improving the customer proposition.

Historically, management has targeted pre-tax margins of 15% to 20%. During FY2024 and FY2025 margins exceeded 20%, helped by elevated interest income generated on customer balances.

Part of this profitability boost reflects unclaimed customer yield. Until more customers opt into yield-bearing products and claim the returns available to them, Wise expects reported margins to remain somewhat elevated relative to underlying economics.

For FY2026 management lowered its guidance range to 13% to 16%, primarily to accommodate costs associated with the Nasdaq listing, infrastructure investments, and expansion initiatives. Despite these investments, underlying pre-tax margins remained approximately 16.3%, demonstrating the resilience of the business model.

CASH FLOWS & BALANCE SHEET

At first glance, Wise's cash flow statement can appear extraordinary as operating cash flow is many times larger than top line revenue (that would be impressive!)

In truth the vast majority of cash flowing through the business is customer money which shows up in working capital. It is not cash belonging to Wise and should not be included in Free Cash Flow calculations.

Management defines free cash flow as operating cash flow excluding working capital movements related to customer balances, less capital expenditure and lease payments. On this basis, free cash flow was £487 million in FY2024 and £615 million in FY2025.

After adjusting for the portion of interest income that management would like to return to customers, underlying free cash flow was approximately £247 million and £333 million respectively.

Even on this more conservative basis, Wise remains a highly cash-generative business.

Similar adjustments need to be made on the balance sheet. The actual Return on Assets, on Equity and on Capital are far higher when third party money is stripped out. Wise has very low capital intensity and so returns are very strong.

It should be noted that the customer cash sitting on the balance sheet introduces excellent negative working capital dynamics to the Wise business. Much of the company's growth is funded this way significantly reducing financing costs.

Similarly, regulatory capital requirements are funded through retained earnings rather than external financing. The result is a balance sheet carrying approximately £1.3 billion of corporate cash, negligible debt, and significant strategic flexibility.

This is ultimately the beauty of the model.

Mr Market

Wise went public in July 2021 at a price of 800 pence (approximately $10.80 USD). The timing was unfortunate.

The listing occurred just months before the peak of the post-pandemic market euphoria that culminated in November 2021. Investors initially valued the company at an eye-watering 266x trailing earnings, pricing in years of future success before that success had materialised (beware overpriced IPOs).

When market sentiment inevitably reversed, Wise was caught in the downdraft. By June 2022 the shares had fallen over 60% to approximately 308 pence despite the business itself continuing to execute well. The earnings multiple had contracted, but was still far from cheap at 102x.

Despite continued mutliple compression, those that bought the July 2022 dip have enjoyed a total return of 28% CAGR over the last four years.

However, the most interesting point is what has happened to valuation multiples.

Today, the shares trade around 829 pence, only marginally above the IPO price, but on much improved unit economics.

Meanwhile, the business has become larger, stronger, and more strategically important. Diluted earnings per share have exploded thirteenfold from 3 pence five years ago to 40 pence today.

In relation to the earnings multiple, the numerator may be the same as the IPO price, but the denominator has expanded significantly. This has caused multiple compression to levels that make Wise look incredibly cheap today.

Said differently, the excess valuation premium has now been fully flushed out and Wise sits at a critical inflection point for investors.

Let’s break it down into numbers and decompose shareholder returns into their underlying drivers.

In terms of share price, an investor at the over valued IPO is now back to where he started 5 years earlier, with a meagre 0.73% CAGR return. But how did he get there?

Over the period, revenue growth contributed approximately 13.7% annually to shareholder returns. Margin expansion added a further 9.5%. Share dilution was largely immaterial. The overwhelming drag came from multiple contraction. Investors who initially paid an excessive valuation experienced a roughly 21.7% annual headwind as the market gradually recalibrated expectations. That cancelled out all the heavy lifting that the underlying business had been doing.

In other words, shareholders did not suffer because the business disappointed. They suffered because they paid too much.

That distinction matters.

The valuation headwind that dominated the first five years of Wise’s public life appears largely exhausted. The company now trades on a debt and cash free basis at roughly 20x earnings that have been depressed by unusually high investment spending, including costs associated with the Nasdaq listing, accelerated hiring, infrastructure expansion, and growth initiatives. On a more normalised earnings base, the valuation falls closer to 16x.

That is a very different proposition from the triple-digit multiple investors paid at IPO.

Over the past twelve months alone, the market has been willing to value Wise at levels approximately 50% higher than today’s share price. More recently, sentiment has weakened following reports relating to a money laundering investigation being conducted by Belgian authorities. The allegations concern criminal activity that may have utilised Wise transfers as part of a broader scheme. Based on currently available information, the investigation appears directed at the underlying criminal conduct rather than the fundamental economics of Wise’s business. The company is cooperating with authorities and the long-term investment thesis appears largely unchanged. We discuss the matter in more detail in the risk section.

Whether the recent decline proves to be an opportunity remains to be seen. What is clear is that the valuation backdrop has changed materially.

For the first five years as a public company, multiple contraction acted as a powerful headwind. Going forward, it is more likely to be neutral and may even become a modest tailwind.

A simple thought experiment illustrates the potential.

Assume the share count remains broadly unchanged. Assume margins remain around current levels and do not expand further. Assume revenue growth slows modestly from recent rates. Finally, assume the market eventually values Wise at around 25 times earnings, hardly an aggressive multiple for a fast-growing, highly cash-generative platform business.

Under those assumptions, the shares could plausibly trade around 4,300 pence by 2031, equivalent to approximately $58.50 USD per share. That would imply annualised returns approaching 38%.

Importantly, this scenario does not require heroic assumptions. It does not require margin expansion, transformational acquisitions, or radical changes to the business model. It simply requires continued execution and a valuation that better reflects the quality of the underlying economics.

There are also several ways reality could prove more favourable than this base case.

Today’s compressed valuation multiples coincide with the period during which Wise has been deliberately reducing its cross-border take rates while scaling its infrastructure investments and moving to a Nasdaq listing. A compressed multiple on depressed earnings paves the way for an acceleration of shareholder returns as margins widen and multiples expand.

Wise Platform remains in its early stages and could accelerate transaction volume growth as additional banks and enterprises join the network. The accounts business continues to grow substantially faster than the transfer business and generates increasingly attractive economics. Direct payment rail connections continue to reduce costs and strengthen the moat. And as free cash flow accumulates, share repurchases become a realistic possibility.

Any combination of stronger revenue growth, modest margin expansion, or stock repurchases could materially improve shareholder outcomes.

Twenty sell-side analysts cover the stock with a mean consensus of ‘Outperform’, an average price target of 1167 pence ($15.75 USD) verses 818 pence ($11.04 USD) at the time of writing, and the highest price target is 1425 pence ($19.24 USD).

These analyst forecasts are not outlandish. The share price has traded up at those levels as recently as 2025.

Naturally, forecasting future returns with precision is impossible. The purpose of the exercise is not to predict a specific share price but to understand the relationship between business performance and valuation.

At today’s valuation, investors are no longer paying for perfection. They are buying a business with a long runway for growth, exceptional unit economics, a widening competitive moat, and a management team that has consistently executed. If those characteristics persist, the next five years could look very different from the first five.

Share

The Money Laundering Issue

On 1 June 2026, Wise disclosed that it was responding to enquiries from the Brussels prosecutor relating to its business. The company stated:

“We are currently working with the Brussels prosecutor to respond to queries about our business, as we routinely do with regulators and law-enforcement authorities. His office’s enquiries are still incomplete and no specific findings have been shared with us to date... We will continue to engage with the Brussels prosecutor’s office if and when any specific findings are made available to us.”

The market reacted swiftly. The shares fell by as much as 18% before recovering a significant portion of the decline.

The question for investors is whether this represents a temporary regulatory setback or evidence of a deeper problem within the business.

To answer that question, it is important to understand the tension at the heart of Wise’s operating model.

Wise’s competitive advantage is built on removing friction from international payments. Customers choose the platform because it is faster, cheaper, and simpler than traditional alternatives. Anti-money laundering compliance requires the exact opposite. Every additional review, escalation, investigation, and verification step introduces friction, increases costs, and slows transactions. There is an inherent tension between delivering a seamless customer experience and maintaining robust financial crime controls.

It is also worth recognising the complexity of the task. The more efficient, low-cost, and widely adopted a payments network becomes, the more attractive it inevitably becomes to criminals. Financial crime follows liquidity. Success itself creates part of the problem.

This risk is therefore unavoidable. Any institution serving millions of customers and processing hundreds of billions of pounds in transaction volume will occasionally find itself under scrutiny.

Large financial institutions routinely face investigations relating to anti-money laundering controls, sanctions screening, know-your-customer procedures, and financial crime monitoring. HSBC, ING, and ABN AMRO have all resolved major compliance matters through negotiated settlements and remediation programmes. Others have successfully defended themselves when challenged.

The broader point is that regulatory scrutiny is not unusual within financial services. It is part of the operating environment.

Cross-border anti-money laundering compliance is arguably one of the most difficult challenges in financial services because criminal organisations continuously adapt their methods. Prevention systems are always reacting to emerging threats rather than anticipating every possible variation in advance.

Regulators understand this reality. Their focus is typically not whether financial crime occurred, but whether the institution took reasonable steps to prevent, detect, and report it. The legal and regulatory distinction is important. Negligence, recklessness, wilful blindness, and systemic control failures are punished. The mere existence of criminal activity passing through a financial network is not, by itself, evidence of wrongdoing by the intermediary.

At present, no public allegations of wrongdoing have been made against Wise.

If anything, the current situation appears more likely to be a controls review than a fraud allegation directed at the company itself.

Wise has spent years building governance and compliance infrastructure designed to manage precisely these risks. Oversight is provided through dedicated risk and audit committees, while management continues to invest heavily in compliance capabilities.

During the company’s May 2026 Investor Day, CFO Emmanuel Thomassin noted that approximately one-third of Wise’s workforce is engaged in compliance-related functions. CTO Harsh Sinha also highlighted the increasing use of artificial intelligence across customer onboarding and financial crime monitoring processes.

That level of investment is significant and it is in direct response to Wise having previously faced regulatory actions relating to anti-money laundering controls, sanctions compliance, and customer due diligence in the past. These included a fine in Abu Dhabi, enforcement action linked to UK sanctions controls, and penalties from US regulators. None of these events materially altered the long-term trajectory of the business.

Criminals only need to succeed occasionally. Financial institutions are expected to succeed all of the time. It’s unrealistic. The asymmetry makes occasional failures inevitable. Mitigation is the best defense, explaining why compliance is not treated as a secondary function, but is increasingly becoming a core competency for Wise.

That does not mean investors should dismiss the current investigation. Compliance failures matter. Regulatory credibility matters. Reputation matters. However, context matters as well.

It should be noted that Wise experienced a number of senior departures within risk, audit, and compliance functions during October and November of 2025. These included executives responsible for internal audit, due diligence, risk management, KYC operations, and financial controls.

Without access to internal information, it would be speculative to draw firm conclusions. However, this is a business not a political body; senior people tend not to resign enmasse. A more reasonable interpretation is that accountability was imposed following identified weaknesses. If true, this would be consistent with management’s long-standing emphasis on compliance and governance rather than evidence of indifference toward regulatory obligations.

Ultimately, the investment question is straightforward.

If the Belgian matter evolves into evidence of widespread negligence or systemic disregard for compliance obligations, the investment case would need to be reassessed.

If, however, this proves to be a more conventional controls review followed by remediation, enhanced monitoring, and potentially a small financial settlement, then investors are likely to refocus on the factors that really matter.

Based on the information currently available, the latter outcome appears more likely than the former.

The Bear Case for Wise Group

The central bear case against Wise concerns take-rate compression. Critics argue that Wise is participating in a race to zero. More particularly, bears worry that if transaction volume declines, management could eventually face a difficult choice between accepting lower margins or having to increase prices.

While management frames its take-rate reduction policy as a deliberate strategy, bears argue that cross-border payments are becoming increasingly commoditised and that Wise has little choice but to keep lowering prices to remain competitive.

Under this view, today’s volume growth is simply masking tomorrow’s margin pressure.

Bears also challenge the durability of Wise’s competitive advantage.

They acknowledge that the company has built an impressive global payments network, but argue that much of its edge stems from inefficiencies that may not exist forever. Governments and central banks increasingly recognise the shortcomings of cross-border payments and are investing heavily in modernisation initiatives.

Projects linking domestic payment systems such as India’s UPI and Singapore’s PayNow demonstrate that policymakers are actively working to reduce friction within the global payments ecosystem.

If national payment systems become increasingly interconnected, bears argue that the value captured by intermediaries like Wise could gradually decline.

Taking the argument further, some investors believe that future payment infrastructure may eliminate the need for companies like Wise altogether.

Under this scenario, central bank digital currencies, real-time settlement networks, or other government-backed payment systems eventually provide consumers and businesses with direct access to low-cost international transfers.

In such a world, Wise’s infrastructure advantage could steadily erode.

Stablecoins represent another frequently cited threat.

Transaction volumes continue to grow rapidly, regulatory frameworks are becoming more established, and settlement can occur almost instantly at extremely low cost.

Bears argue that stablecoins solve many of the same problems Wise was originally created to address.

If businesses and consumers increasingly transact through blockchain-based payment rails, Wise could find itself competing against a fundamentally cheaper technology stack.

Governance is another concern.

Founder Kristo Käärmann retains effective control through the company’s dual-class share structure. While supporters view founder control positively, critics argue that minority shareholders have limited ability to influence corporate decisions.

The concern is not necessarily today’s governance but future governance.

Public shareholders are effectively betting that management’s judgement remains sound indefinitely.

Critics also point to Käärmann’s historical tax issue. In 2022 he failed to pay a personal tax liability on time, later attributing the matter to an oversight due to a super busy work schedule. The liability was settled, a fine was paid, and regulators ultimately took no further action. Bears argue that such incidents raise questions about judgement and oversight at the highest levels of the organisation.

Regulation presents an additional risk.

Wise operates across dozens of jurisdictions and processes enormous volumes of customer funds every day. Compliance requirements continue to grow more complex as governments tighten anti-money laundering, sanctions, and financial crime controls.

A significant compliance failure could result in fines, licence restrictions, reputational damage, or even the loss of key operating permissions.

Finally, bears question whether the market opportunity is as attractive as it appears.

Wise already dominates much of the independent cross-border transfer market. As the company becomes larger, maintaining current growth rates becomes increasingly difficult.

The law of large numbers eventually catches up with every company.

Under this view, investors are extrapolating current growth rates far into the future while underestimating the combined effects of competition, regulation, technological disruption, and market maturation.

Share

The Bull Case for Wise Group

The bull case begins with a simple observation. Moving money is not discretionary.

Consumers need to send money internationally. Businesses need to pay suppliers. Employees need to receive salaries. Companies need to manage cash across borders regardless of economic conditions.

Operating at the centre of essential financial activity, bulls argue that this creates a business with considerably more resilience than the market appreciates.

But they would also argue there is something even more attractive about the business model.

Wise is not a traditional company. It doesn’t require branches, inventories, manufacturing facilities, or large lending portfolios to grow. In a world increasingly impacted by tariffs, supply chain disruptions, geopolitical tensions, inflation, and rising capital costs, Wise escapes unscathed. Many businesses are fighting headwinds they cannot control, while Wise is largely insulated from them.

In fact, some of the very conditions creating pressure elsewhere can benefit Wise. Higher interest rates increase net interest income on customer balances, providing a tailwind at a time when many companies are facing higher financing costs and weaker demand.

Bulls would also argue that Wise occupies a position that is unusually resistant to artificial intelligence technological disruption. Over decades it has built infrastructure based on licenses and trust with a huge international customer base that no amount of vibe coding can replicate.

The result is a business that combines recurring demand, structural growth, and a degree of counter-cyclicality that is rare within the technology sector. In a difficult economic environment, Wise may prove to be one of the more defensive growth businesses available to investors.

In relation to take-rate reductions, the bulls say that the market misunderstands how the business actually operates.

Wise is not reducing prices because it is losing competitiveness. It is reducing prices because increasing scale continuously lowers the underlying cost of moving money through its network.

Volume growth is not coming at the expense of profitability. It is the mechanism through which profitability improves.

The company’s infrastructure has taken more than a decade to build and consists of regulatory licences, banking relationships, payment rail integrations, liquidity management systems, and local market expertise across dozens of jurisdictions.

Replicating that network would require substantial time, capital, and regulatory effort.

Bulls argue that this is far more difficult than critics assume.

The moat also extends beyond cost.

Network effects are becoming increasingly visible. Higher transaction volumes improve efficiency, lower costs, and allow Wise to offer even better pricing. Better pricing attracts more customers, which generates more volume.

The flywheel continues to strengthen.

Importantly, the business is no longer primarily a consumer remittance platform.

Wise Business and Wise Platform are becoming increasingly important contributors to growth. This shift matters because corporate payment flows are substantially larger than consumer transfers and considerably stickier.

Once payment infrastructure becomes embedded within treasury systems, payroll processes, invoicing workflows, and financial operations, switching providers becomes both disruptive and expensive.

Bulls believe this transition significantly improves the durability of future earnings.

The stablecoin threat is also frequently overstated.

Moving a stablecoin across a blockchain may be inexpensive, but customers ultimately live in a fiat currency world.

Money typically still needs to be converted into a stablecoin, transferred, converted back into local currency, and withdrawn into a bank account. Each step introduces costs, operational complexity, and counterparty risk.

Bulls would also point to the strength of the Wise card ecosystem.

For most customers, international payments are not just about moving money between bank accounts. They are about spending money while travelling, shopping online, or operating across multiple countries.

Wise allows customers to hold balances in different currencies and spend locally using a card wherever standard card payments are accepted. Foreign exchange is handled automatically at near-interbank rates with minimal fees.

Bulls argue this convenience is often overlooked when comparing Wise to emerging payment technologies such as stablecoins.

A traveller landing in Tokyo, Paris, or New York does not want to manage wallets, exchanges, and blockchain transactions. They want to tap a card and pay. For most customers, the winning solution is the one that makes international spending feel no different from spending at home.

Traditional banks offer a similar experience, but typically at unattractive foreign exchange spreads with exorbitant fees layered on top.

Bulls therefore argue that Wise is competing on more than just transaction costs. It is competing on simplicity, convenience, and user experience. It wins on all three.

Bulls say regulation is another tailwind.

Under Europe’s MiCA (Markets in Crypto-Assets) framework, anyone receiving stablecoins will need to complete KYC identity verification before converting those assets into a bank deposit. Similar regulatory regimes are emerging around the world as governments seek to prevent digital assets from being used for unlawful activities.

As different rules are introduced across different jurisdictions, managing cross-border payments through multiple wallets, exchanges, and compliance frameworks becomes increasingly cumbersome and time consuming for the average consumer or business.

This is where companies like Wise become relevant. Most customers have little interest in becoming experts in payment rules and regulations. They simply want their money to arrive quickly, safely, and at the lowest possible cost. A few clicks on a smartphone app and it’s done.

Bulls also argue that Wise is technology agnostic and so not tied to any specific payment rail. The company is fundamentally a settlement and ledger business, focused on moving money through whichever infrastructure offers the best combination of speed, cost, and reliability. Whether transactions ultimately settle through SWIFT, Faster Payments, SEPA, or a blockchain-based network is largely irrelevant.

This explains the company’s increasing focus on digital asset capabilities. Wise has no reason to resist technologies that improve efficiency. If stablecoins can reduce liquidity requirements, accelerate settlement times, or lower transaction costs, the company has every incentive to incorporate them into its network. The mission remains exactly the same as it has always been: move money across borders as quickly, cheaply, and seamlessly as possible.

The company benefits from improvements in payment infrastructure rather than being threatened by them.

On the topic of technology, bulls point to autonomous software agents executing complex payment workflows without human intervention and AI systems making payments directly to one another.

Wise is investing heavily in this future. Technology spending exceeded $250 million in 2025 and management has repeatedly highlighted artificial intelligence as a strategic priority.

Bulls also point to the strength of the recurring revenue model.

Revenue is spread across millions of customers globally with virtually no customer concentration risk. Customers who adopt Wise Accounts frequently increase their engagement over time through transfers, card spending, business payments, and account activity.

Customer balances continue to grow, card revenue is expanding rapidly, and account adoption remains strong. These trends suggest that customers increasingly view Wise as a financial operating system rather than simply a money transfer service.

The growth runway also remains enormous. The total addressable market is measured in tens of trillions of dollars annually while Wise’s share remains relatively small across most categories. Despite serving millions of customers globally, the company has only penetrated a fraction of the available opportunity.

Perhaps most importantly, growth remains overwhelmingly organic. Wise is not relying on acquisitions to drive expansion. Growth continues to come from product development, geographic expansion, infrastructure partnerships, and increasing customer adoption.

Businesses capable of sustaining strong double-digit growth without acquisition-driven rollups are relatively rare.

The US listing adds another potential catalyst. Greater visibility within the world’s largest capital market could attract additional institutional investors, increase liquidity, and potentially pave the way for inclusion in major equity indices over time.

Bulls believe the market continues to underestimate both the size of the opportunity and the durability of Wise’s competitive position.

If they are correct, today’s debate is not about whether Wise can continue growing.

It is about how much of the global payments ecosystem it ultimately captures.

Finally, Bulls say that Wise is a unique asset with no direct competition. It is becoming an increasingly important part of the global financial system. At approximately $10 billion of enterprise value it is large enough to be strategically important yet small enough to become an acquisition target. Rumour has it that JP Morgan has expressed an interest, but it may be a better fit for the likes of Adyen, Visa or Mastercard and be reminded that Adyen’s CEO sits on the board of Wise.

What to Watch For

• US Active Customer and Volume Run-Rate Post-Listing: Management stated the US listing will ‘help us accelerate our mission, helping to bring more of Wise to everyone in the US, as customers and as owners.’ Monitor whether North American volume accelerates significantly. The US represents the largest cross-border pool globally; market share capture here is key to accelerating Wise’s volume growth vector.

• The Balance of Take Rate vs. Operating Leverage: Track whether further fee reductions successfully trigger proportional volume growth. If the take rate contracts while volume growth flattens, it signals that price elasticity is weakening and scale-equity benefits are hitting a point of diminishing returns.

• Business Diversification: Watch the mix shift between personal and business volumes. A higher proportion of corporate volume expands cross-border income stability, but changes the average take-rate dynamics due to potential bulk transaction discounting.

• Regulatory and Compliance Costs: Dual listings carry material ongoing costs (regulatory compliance, dual reporting, investor relations). Accounting will move from IFRS to GAAP. The redomiciliation to Jersey introduces a new jurisdictional layer. The market is weighing the potential benefits of US investor access and brand visibility against the costs and complexity of the restructuring.

• Regulation: The European Union’s PSD3 framework extends open-banking principles further into cross-border payments. On one hand, this could benefit Wise by accelerating adoption of lower-cost account-to-account payment methods. On the other hand, easier access to banking infrastructure may lower barriers to entry for competitors. The regulatory environment is likely to remain a moving target for years to come. Wise's ability to adapt faster than rivals may prove an important source of competitive advantage.

• Margins: Management has historically targeted pre-tax margins in the 15% to 20% range. Recent guidance has been lower as the company absorbs the costs of the Nasdaq listing, infrastructure investments, technology spending, and accelerated hiring. The key question is whether these lower margins represent a temporary investment phase or a structural shift in management’s priorities. Investors should pay close attention to FY2027 guidance. A return to the traditional margin framework would suggest that recent pressures were largely transitory.

• Rates: Interest income is sensitive to central bank rate decisions; any material rate-cutting cycle would compress this revenue stream, while elevated rates will boost profitability. Investors should therefore distinguish between operational growth and interest-rate-driven earnings fluctuations when evaluating performance.

• Currency Exposure: Wise generates revenue across dozens of currencies and jurisdictions. This provides a degree of natural diversification, but it also introduces foreign exchange risk. Revenue and costs are not perfectly matched geographically, creating the potential for currency movements to affect margins and reported earnings. While currency fluctuations are unlikely to alter the long-term investment thesis, they can create meaningful volatility in reported financial results.

Conclusion: Is Wise A Good Investment?

Wise is one of the most interesting businesses operating in global financial services today.

Cross-border volume, customer acquisition, share of wallet and account balances continue to expand rapidly. Meanwhile, business customers are becoming a larger proportion of the mix and platform partners are aiding in the acceleration of growth.

At virtually every operational level, the company continues to execute superbly.

The opportunity remains vast. Wise has captured only a small fraction of the total addressable market, which itself is continuing to expand. The broader cross-border payments market continues to benefit from structural growth drivers including e-commerce, global trade, remote work, migration, and increasing international mobility.

What began as a fintech challenger has evolved into something more significant.

Wise increasingly resembles a piece of financial infrastructure.

Its competitive advantage is not based on branding alone. It is rooted in regulatory licences, payment rail connectivity, liquidity management, technology infrastructure, compliance capabilities, network scale, and cost leadership. These advantages reinforce one another over time and become increasingly difficult for competitors to replicate.

The economics are equally attractive. The business generates high returns on equity, substantial free cash flow, and requires relatively little incremental capital to grow. It continues to reinvest aggressively while maintaining a fortress balance sheet.

Management quality is another differentiator. Kristo Käärmann has consistently prioritised long-term value creation over short-term profitability. The company’s culture remains intensely customer-focused, and management incentives are generally aligned with shareholders.

The Nasdaq listing introduces a potentially powerful additional catalyst. Increased visibility, deeper liquidity, broader institutional ownership, and eventual index inclusion could all help narrow the gap between market value and intrinsic value over time.

That gap is difficult to ignore.

Sell-side analysts generally maintain price targets materially above the current share price. Either analysts are collectively too optimistic, or the market remains overly sceptical. Given the complexity of the business model and the tendency of markets to focus on near-term headlines, the latter explanation appears plausible.

Wise is not without risks. Regulatory scrutiny will remain constant. Interest-rate cycles will affect reported earnings. Competition may intensify. The dual-class share structure will continue to divide opinion.

Yet none of these concerns appear sufficient to undermine the fundamental strengths of the business.

The incumbents remain constrained by legacy systems and conflicting incentives. Wise continues to widen its infrastructure advantage. The platform becomes more valuable as volume increases. And the opportunity ahead remains substantially larger than the business today.

For investors willing to tolerate regulatory noise, founder control, and periodic market volatility, Wise appears to offer something increasingly rare in public markets: a long-duration organic compounder with a substantial runway for growth and a management team that understands exactly how to extend its competitive advantage.

That combination is difficult to find and even harder to replicate.

Share

Leave a comment

You may also be interested to read:

A Thought Exercise: Breaking The Market

You wake up to discover that you’ve travelled forward in time.

It’s June 1st, 2027.

The Nasdaq has fallen below 15,000. It’s down 34% in nine months. NVIDIA has been cut in half. Apple has lost a third of its value. Microsoft briefly traded below 20x forward earnings for the first time since the pandemic recovery.

What finally broke the market wasn’t war, inflation, or the Federal Reserve.

Ironically, it was the thing investors were most excited about: three mega-IPOs at the frontier of artificial intelligence and space technology.

For years, markets absorbed every macro shock imaginable: inflation, trade fragmentation, geopolitical conflict, energy shortages, rising deficits. Valuations expanded anyway.

The explanation was always the same. AI productivity gains justified everything.

Every quarter reinforced the narrative. AI infrastructure spending exploded. Data centre construction became one of the largest CAPEX booms in modern history. Investors convinced themselves that as long as intelligence became cheaper, productivity growth would sustain elevated valuations indefinitely.

Meanwhile, the market became increasingly concentrated.

By late 2026, the top ten companies in the S&P 500 represented almost 40% of the index. The market had become a concentrated momentum trade masquerading as diversification.

The problem was not technological progress. The problem was financial inevitability. Investors began treating rising valuations as a law of nature rather than a function of capital flows.

Then the structure changed.

Within roughly twelve months, three companies prepared to go public: SpaceX, OpenAI, and Anthropic.

Their combined implied valuation exceeded $4 trillion.

Even in a market conditioned to trillion-dollar companies, the scale was difficult to comprehend.

But the real danger wasn’t the valuations themselves.

It was that three new mega-cap destinations suddenly appeared for the same finite pool of speculative capital that had spent years concentrated in the Magnificent Seven.

The IPOs didn’t create new liquidity. They fragmented existing liquidity.

That forced a redistribution of capital across a larger set of mega-caps, compressing valuations across the sector.

For years, investors justified 35x, 40x, and sometimes 50x earnings multiples because there were relatively few businesses capable of absorbing enormous amounts of growth capital.

Then suddenly there were three more.

The first phase of the decline looked manageable.

Then the sell-off accelerated.

Retail investors began withdrawing money from index funds for the first time in years. Passive funds became forced sellers. The same mechanism that amplified the rally subsequently accelerated the decline.

The largest constituents absorbed the largest outflows.

Technology underperformed dramatically.

The concentration that had powered the bull market reversed direction and became a source of instability.

The market structure itself became pro-cyclical.

The excesses of the bull market became the catalyst that caused it to collapse under its own weight.

Share

Let’s dive in to the detail to explore how, and why, this scenario may play out.

The Market Backdrop

Overvaluation alone rarely causes a crash.

Despite years of inflationary pressure, fractured trade relationships, sanctions, geopolitical conflict, and rising deficits, markets have largely shrugged them off without materially resetting valuations.

It increasingly feels as though markets have become impervious to economics.

AI contains a contradiction few investors want to acknowledge. The productivity boom required enormous energy consumption (raising power prices for everyone) and devours huge volumes of CAPEX that must ultimately be passed down to consumers.

At the same time, the relentless displacement of people’s jobs is a slow bleed on the fabric of our economy.

Inflation moving up on supply side pressure; consumption falling due to buy side weakness; a tragic combination.

Yet markets ignore the obvious danger.

“This time is different” became the dominant psychology.

Then the IPOs entered the picture.

The Ticking Time Bomb

The scale of modern IPOs already looks radically different from previous cycles.

In the 1980s, the average U.S. IPO raised roughly $26 million, rising to around $151 million at the peak of dot-com mania. After the market crash the public’s appetite for tech listings waned. Private capital bridged the gap and companies stayed private far longer. By the 2020s, the average company was entering public markets fourteen years after founding rather than eight.

This brings us to today. Three mega-cap IPOs arriving simultaneously.

SpaceX is expected to debut around $1.75 trillion. OpenAI near $1 trillion. On 28th May, Anthropic reportedly raised $65 billion in a new funding round at a $965 billion post-money valuation, overtaking OpenAI’s latest reported valuation for the first time. A combined valuation rapidly approaching $4 trillion.

The valuations themselves aren’t the core problem.

The problem is structural.

The IPOs didn’t create new capital. They created new destinations for existing capital1.

Investors fund them by selling existing positions.

The investors attracted to narratives around the colonisation of Mars or artificial general intelligence will likely fund those purchases by trimming positions in incumbent technology leaders, such as Apple, NVIDIA, Microsoft, Meta Platforms, and Alphabet.

This is how the market may cannibalise itself.

That matters because technology has become the primary support structure beneath the entire index. Once finite capital is spread across a larger collection of mega-caps, multiples compress across the board.

Lower multiples in the leaders become the benchmark for everything else.

Then passive flows amplify the process.

Index funds don’t make valuation judgements. They allocate mechanically based on market capitalisation. Once the new IPOs qualify for inclusion, passive managers will be forced to reallocate capital, regardless of valuations.

NASDAQ’s Fast Entry rule, effective May 2026, potentially allows inclusion in the Nasdaq-100 after only fifteen trading days2. S&P Dow Jones Indices has also proposed rule changes that would make fast-track inclusion possible for SpaceX.

What does this mean for companies currently in these indices?

Because the U.S. indices are market-cap weighted, the largest incumbents are where that selling concentrates, and the pressure grows as each IPO's float expands into full index weight.

Re-ratings at the top of the market ripple down toward the base of the capitalization pyramid. Little or nothing escapes unscathed. Gravity pulls the entire index lower.

The IPO Roller Coaster

For the three companies entering the public markets for the first time, there is usually an initial climb followed by a steep drop.

IPO euphoria on a restricted float causes a launch price spike. Then entry into indices force passive funds to buy, but not before active managers and speculators front-run those inflows. Momentum traders follow. Retail investors pile in behind them. Everyone is having the ride of their lives.

The roller coaster soon reaches its apex and the rapid descent begins; this is where the capital destruction occurs.

Subscribe now

The Lock-Up Problem

IPO floats are often intentionally small. SpaceX was reportedly targeting an initial float of roughly 3% to 5% of total equity.

This is entirely by design. Tight supply combined with enormous hype creates artificial scarcity, pushing prices sharply higher in the early stages of trading.

But lock-ups don’t remove selling pressure. They delay it.

After the IPO, insiders are typically prohibited from selling shares for roughly 180 days. The goal is market stability. But once those restrictions expire, the tradable float can increase exponentially.

Supply floods the market.

SpaceX’s proposed structure appears particularly aggressive, replacing the traditional single lock-up expiry with staggered release tranches across several months.

That creates multiple cliff edges rather than one.

The mechanics become brutal because new supply arrives precisely when enthusiasm is highest.

The first sellers are often employees who have accumulated stock-based compensation over many years. Regardless of whether or not they are a true believer in the corporate mission, most will understandably choose to convert paper wealth into life-changing sums of cash. They don’t care about valuation; they care about liquidity. They’ll sell as soon as they can. Not might, they will.

Then come the early investors.

Some early SpaceX backers from 2008 (when the company narrowly escaped bankruptcy) are potentially sitting on returns measured in thousands of times their original investment. Their mandate is not ideological loyalty; it’s to generate and return capital to their limited partners. They’ll want to recycle capital into new opportunities. More heavy selling.

That creates two highly motivated seller groups entering the market simultaneously.

Then comes the third group: panic sellers afraid of being the last out.

This is where George Soros reflexivity takes hold.

As lock-ups expire, short sellers suddenly gain access to dramatically larger quantities of borrowable stock. Stock loan fees collapse. Short selling becomes cheaper and easier precisely as momentum weakens.

More shorting creates additional downward pressure. Falling prices trigger margin calls on leveraged long positions. Forced selling drives prices lower still.

The feedback loop intensifies.

Now the fundamentals catch up and they were never good enough to justify the IPO valuation anyway.

Momentum breaks.

Buyers step aside; few investors willingly try to catch a falling knife.

The bears will have taken control of the market.

Why This Could Spread Systemically

This dynamic isn’t theoretical.

Meta Platforms, then trading as Facebook, fell roughly 50% following its 2012 IPO lock-up expiry period.

The dot-com collapse provides an even stronger historical parallel. It wasn’t one company that broke the market. It was a rolling sequence of lock-up expirations and capital destruction across dozens of newly public technology companies.

This time may be worse because there are no isolated events.

SpaceX has published its S-1 detailing its IPO structure. There is reason to expect similarly staggered release schedules across OpenAI and Anthropic. From Airbnb and DoorDash to Reddit, staggered insider release mechanisms have increasingly become standard practice in technology IPOs. The same set of banks run the IPO process. The play-book is what it is.

That means the market may not face a single shock.

Instead, it may endure overlapping waves of insider selling spread across six to nine months.

SpaceX unlocks through the summer. OpenAI follows in autumn. Anthropic after that.

Before one wave fully clears, the next arrives.

Markets are path dependent.

What begins as orderly repricing gradually becomes exhaustion. Liquidity deteriorates incrementally. Investors stop viewing the pressure as temporary and begin recognising it as structural.

That psychological transition is where crashes can accelerate.

Subscribe now

The Seeds of Self-Destruction

The irony is extraordinary.

For years, analysts searched for an external catalyst to break the bubble: a recession, a banking crisis, a geopolitical shock, a Federal Reserve policy error.

Instead, the catalyst may come from within the bubble itself.

The speculative excess required to produce trillion-dollar IPOs may ultimately destabilise the valuations that made those IPOs possible.

A classic Minsky moment.

The market would likely watch the process unfold in real time: the filings, the hype cycle, the passive inflows, the lock-up expirations. Yet just like a nuclear reaction, once triggered it will be impossible to stop.

Technology underperforms. The IPOs collapse. Passive outflows accelerate. Multiples compress across the sector, then the broader market corrects sharply.

In hindsight, everyone suddenly claims the warning signs were obvious.

Everything makes sense looking backwards. The problem is that we are required to navigate forwards.

This is what inspired this thought exercise.

Share

Conclusion

This was simply a thought exercise. It was not intended to convince you of a particular outcome. Perhaps it has encouraged you to think a little more critically about the risks that may be hiding in plain sight.

As Professor Elroy Dimson observed, "risk means more things can happen than will happen." There is certainly no shortage of possibilities confronting investors today. Unprecedented levels of AI-related CAPEX, persistent inflation, unsustainable government debt, geopolitical tensions, and plenty more besides.

History shows that markets can climb for years despite serious underlying problems. But that doesn't mean those vulnerabilities should be ignored. More importantly, it doesn't mean investors should stop demanding compensation for bearing the risk.

Yet that increasingly appears to be the case. Valuations imply a degree of confidence that leaves little room for disappointment. The margin of safety that once protected investors from adverse outcomes has steadily eroded. Complacency has replaced caution.

What about the Efficient Market Hypothesis? Information is widely available, and markets possess mechanisms designed to promote price discovery and manage risk. Arbitrage, short-selling and hedging all serve important functions.

In theory, yes. In practice, markets rarely function so cleanly.

The problem is structural.

Bearish positioning becomes mechanically limited precisely when valuations are excessively stretched. So the ability of sceptics to counter excessive optimism diminishes when it is needed most and price discovery itself becomes distorted.

At the same time, passive capital continues to grow. The mandate of index funds is benchmark replication, not valuation discipline. They become forced buyers regardless of price. That creates highly inelastic demand at the exact moment enthusiasm is often at its peak.

Even sophisticated investors repeatedly fall victim to the “this time is different” fallacy. They convince themselves that SpaceX or OpenAI are exceptional cases; that early investors won’t sell because they believe in the mission. They’re wrong. Early investors believe in money.

The deepest irony is that the lock-up mechanism exists to create stability. Under normal circumstances, it usually succeeds.

But if multiple trillion-dollar IPOs arrive within the same market cycle, the mechanism designed to suppress selling pressure could instead concentrate it. The stabilisation mechanism itself may become the destabilising force. A pressure chamber quietly building energy beneath the surface ready to explode.

This dangerous situation may benefit pre-IPO shareholders whose restricted holdings appreciate just long enough to allow an orderly exit at inflated prices.

For everyone else, “take cover!”

Investors keep saying, “This time is different.”

They may be right.

Just not in the way they expect.

Leave a comment

Share

• Company: dLocal (NASDAQ: DLO)

• Market cap: $3.5 bn

• Enterprise value: $2.7 bn

• Net Cash: $800m

• Float: 30%

• EV/EBITDA: 6.9x

• Growth: strong double digit

DISCLAIMER & DISCLOSURE: The author is invested in DLO at the date of publication but that may change. The views expressed are those of the author and may change without notice. The author has no duty or obligation to update this information. Some content is sourced from third parties believed to be reliable, but accuracy is not guaranteed. Forward-looking statements involve assumptions, risks, and uncertainties, meaning actual outcomes may differ from those envisaged in this analysis. Past performance is not indicative of future results. All investments carry risk, including financial loss. This analysis is for educational purposes only and does not constitute investment advice or recommendations of any kind. Conduct your own research and seek professional advice before investing.

The Company & Its People

The company acts as a "toll booth" for payments in emerging markets, has a highly competent management team and is trading at arguably the most attractive valuation since its IPO.

dLocal was founded in 2016 by Andrés Bzurovski, Sergio Fogel, and Sebastian Kanovich, who recognized a fundamental problem: global companies like Amazon, Nike, or Uber found it nearly impossible to collect payments from or disburse funds to customers in countries like Brazil, Nigeria, or Indonesia. Each market had its own fragmented web of local payment methods, arcane regulations, currency controls, and banking infrastructure. The founders built DLocal to solve this by offering a single API integration that abstracted away all that complexity.

It is a financial infrastructure company headquartered in Montevideo, Uruguay, but listed in the US. The company went public on Nasdaq in June 2021 at the peak of the fintech mania, achieving a market capitalization of ~$20 billion.

Focus on that $20 billion number. today this company has an enterprise value of less than $3 billion. I’ll cover this in greater detail in the valuation section below.

The company’s formative years were shaped by the challenges of operating across the Global South: navigating hyperinflation in Argentina, Pix in Brazil, and mobile money in Africa. Every difficulty is an opportunity. Create a solution and that becomes a valuable service. That is very much the DNA that is now so deeply embedded in the business. An iterative process of evolution.

However, this was not the only challenge that the company face. It sustained a short-seller attack by Muddy Waters in late 2022, which alleged discrepancies in the company’s TPV and receivable disclosures. The allegation referenced related-party transactions, including a disclosed loan from the company to co-founder Kanovich and former President/COO Jacobo Singer.

The company’s Audit Committee, consisting solely of independent directors, oversaw an independent review with outside counsel and forensic accounting advisors, concluding that the allegations were entirely unfounded and unsubstantiated.

However, perception is reality and the company had suffered a degree of reputational damage. Investor pressure flowing from the controversy led to a restructuring of the board. It became necessary to transition from founder-led intensity to professional management. So, co-founder Sergio Fogel moved to a non-executive advisory role and high calibre independent directors arrived including Paco Ybarra (former Citigroup executive) and Nelson Mattos, (formerly of Google and IBM).

Most notably, in 2023 Pedro Arnt, formerly of MercadoLibre, joined initially in the capacity as co-CEO to work alongside founder Sebastián Kanovich, but ultimately took the reins and became the sole CEO. His aim was to help scale the business after it had already grown into a large, fast-expanding payments platform in emerging markets.

Arnt, who began his career at Boston Consulting Group, left MercadoLibre after more than two decades there, including roughly 12 years as CFO. He explained that after having achieved enormous success at MercadoLibre, he was stepping down to take on new challenges and to pursue new opportunities. He said he was attracted by dLocal’s huge addressable market, its high-growth/high-margin/high-cash-generation model.

dLocal offered Arnt a chance to apply his operating and finance experience in a newer company with a different growth profile. The company specifically valued his track record of building through scale in Latin American tech, which was central to the appointment.

Guillermo Lopez Perez was appointed CFO in November 2025. His background is primarily at American Express and Visa, and he holds an MBA from the University of Chicago Booth School of Business.

When a CFO is relatively new to office, I always like to know why the change occured and there seems to be nothing to be concerned about here. Diego Cabrera Canay served as CFO from December 2020, before the IPO, until March 2024. He was replaced by Mark Ortiz shortly after Pedro Arnt became CEO. As a former CFO himself, Arnt likely had a clear view of the type of finance partner he wanted alongside him. dLocal framed the change as part of a broader effort to strengthen financial strategy and support the company’s rapid global expansion, effectively a management upgrade aligned with scaling needs.

However, only a year later, Ortiz stepped down due to an unforeseen health issue, stating that his health required his full attention. This piqued my curiosity, but again, I found nothing untoward. The departure caught dLocal by surprise, with Jeffrey Brown appointed interim CFO while the company searched for a permanent replacement. That replacement was Lopez Perez. Ortiz remained available during the transition, which points to an orderly handoff rather than a breakdown in relations. There is also no evidence that Ortiz moved directly into another role after leaving dLocal, which lends credibility to the health explanation provided.

Executive compensation, including performance-based equity grants, is tied to TPV growth and gross profit targets, providing an insight into objectives of the business. Customer acquisition, retention and increasing wallet share are the focus.

The Business Model

dLocal’s economic engine is remarkably simple and powerful. It is a pay-in/pay-out platform. “Pay-in” means DLocal enables a global merchant to collect money from a local consumer. “Pay-out” means it enables that same merchant to pay local vendors, drivers, or sellers.

The company’s blended take-rate (the percentage of Total Payment Volume (TPV) retained as gross profit) is the primary economic driver.

Importantly, dLocal is not a bank and does not hold customer funds on its own balance sheet in the traditional sense. Instead, it operates under local payment institution licenses held across more than 60 markets.

This is also not a consumer-facing app; it’s business-to-business (B2B) infrastructure. DLocal provides a single, direct Application Programming Interface (API), one contract, and one platform, a model it calls “One dLocal” .

The business model’s key assumptions, are that:

• Global merchants (e.g., Amazon, Google, Uber, Western Union) will continue preferring a single vendor over building fragmented local infrastructure in every country in which it operates

• As long as cross-border e-commerce and digital services grow faster in emerging markets than in developed ones, DLocal’s addressable market expands.

• The company’s recent launches (BNPL Fuse, stablecoin settlement rails via a partnership with Stable Sea, and tokenized APMs) are attempts to embed itself deeper into the transaction flow, capturing more value and increasing switching costs for merchants (more on this in the Bull thesis).

Pedro Arnt joined dLocal as co-CEO in 2023 after spending roughly 24 years at Mercado Libre, including as CFO. That naturally raises the question of whether the two companies have any form of commercial relationship, but I could not find evidence of a direct connection. Strategically, the two companies occupy different layers of the payments stack. Mercado Libre’s payments arm, Mercado Pago, is largely a consumer and merchant ecosystem tied to its marketplace and fintech network. dLocal is more infrastructure-oriented, helping global enterprises access local payment methods across emerging markets. That said, there is likely indirect overlap. Mercado Pago is one of the dominant local payment methods in Latin America, and dLocal’s model is to aggregate local payment rails for international merchants. So a global merchant using dLocal could ultimately be accepting payments through Mercado Pago infrastructure in certain markets, even if Mercado Libre itself is not a dLocal customer.

Competition and the Moat

The critical nuance is dLocal’s economic moat is not built on processing Visa or Mastercard transactions, which are commoditized.

Instead, its value lies in integrating over 600 local payment methods that are entirely non-standard.

In Brazil, that means processing Pix, the instant payment system controlled by the central bank. In Peru, that means Yape. In South Africa, that means Payflex for buy-now-pay-later (BNPL). In Nigeria, that means Verve cards. In India, UPI. Then there’s Mexico which has become a massive growth engine for dLocal alongside Brazil.

dLocal earns higher margins on these alternative payment methods (APMs) than on standard card rails because the integration is more complex and the regulatory burden is higher.

dLocal operates in a “blue-red” ocean. The total addressable market for cross-border payments in emerging markets is vast and growing, but the competitive landscape is bifurcated. dLocal’s competitive positioning is best addressed by making some fundamental distinctions:

• Global competitors: The likes of Adyen, Stripe and PayPal (via Braintree) may be superior for standard card processing in developed economies with unified, credit-card-centric systems, they lack local payment method depth, in "High-Growth" emerging markets characterized by extreme fragmentation and non-traditional payment methods. A merchant using Stripe in Brazil can accept credit cards, but doing a Pix payout to a freelancer or processing a cash payment through Boleto is far more cumbersome. While these other global payment operators focus investment on growing in standardized payment markets, dLocal’s moat is its ability to differentiate itself from global competitors by solving for over 60 non-standardized markets.

• Local competitors: The likes of Ebanx in Brazil and Flutterwave in Nigeria offer solutions in their home markets, but are unable to offer single-contract multi-country coverage for 60+ countries. dLocal’s moat here is its aggregation; the “One dLocal” model. The unique sales proposition for its customer base is to simplify payment complexity with a one-size-fits-all solution.

• Cross-border emerging-markets payment services: This market is a smaller market and far less cleanly segmented. Companies operating in the space include AZA Finance, Fincra, Verto and Klasha. dLocal stands apart because it operates across multiple emerging-market regions globally, while most competitors are either more regionally concentrated or fragmented by customer need, such as FX and treasury management, remittances, or Africa-only payment flows. dLocal remains the broadest multi-emerging-market platform in the group.

The markets in which dLocal operates may not be G7 economies, but its solution enables access to over 71% of the world’s population.

dLocal targets regions that are "growing differently" rather than just "growing fast". : These "next frontier" economies (Africa, Asia and Latin America) are projected to drive over 60% of global GDP growth by 2030.

A country’s “unbanked” or “underbanked” status is a key indicator for dLocal’s entry.

dLocal acts as a bridge for global giants like Amazon, Meta, and Netflix to reach these billions of consumers who would otherwise be locked out of global commerce.

The difficulty of moving money across borders is significantly higher in dLocal's focus areas. These regions often face high regulatory hurdles and volatile local currencies which ordinarily make it difficult for international merchants to convert and repatriate their earnings.

dLocal’s raison d’être is to provide solutions for these problems.

There is always a risk that larger competitors may leverage their existing relationships, brand recognition, or financial resources to develop competing solutions, but to date all of their reinvestment in growth appears to be focused on developed economies operating on ‘standardized’ payment rails.

The moat for dLocal is its local density; the accumulation of payment method integrations, and 38 regulatory licenses (with 16 more in process) and operations across 60+ countries. This is not something that can be protected with a patent or exclusive contract; it is a product of time invested which is not easy to replicate, creating a cumulative operational first mover advantage.

More particularly, in emerging markets local payment methods are increasing in penetration year on year, at the expense of traditional payment cards, which strengthens the value proposition being offered by dLocal and widens its moat.

Numbers

This is a company growing at pace. Since IPO, free cash flow has compounded annually at over 58%. Revenue has grown even more quickly, indicating heavy reinvestment in growth. This is a capital light business (CAPEX on the bar chart below is barely visible). In terms of OPEX spending R&D consumes ~3% of revenue, while SG&A accounts for closer to 11%. Notably, SG&A had been running at ~25% of revenue prior to the IPO, so operating leverage is in evidence. After the initial platform is built, each additional transaction adds revenue with very little extra cost, allowing profits to scale rapidly with volume. Free cash flow is impacted by changes in working capital (evident in the 2024 numbers below), but this is purely down to timing of cash flows and nothing of concern. The business enjoys the operating efficiency that negative working capital delivers; their role as an intermediary naturally leads to a time lag between collecting from consumers and paying merchants, creating a working capital float. dLocal actively leverages the cash held within its operations, particularly in high-interest emerging markets alongside appropriate hedging of associated risk, to produce significant financial income that supplements its primary transaction fee revenue. This means that dLocal is able to operate with no debt and generate high returns on invested capital (ROIC), often above 30%. It doesn't require significant capital expenditure to scale, meaning growth doesn't automatically destroy working capital. Cash conversion is typically well over 100%.

The chart below shows how Total Payment Volume has increased since hte company’s inception back in 2016. This is clearly a business going from strength to strength.

The Q1 2026 provides a crystal-clear picture of dLocal’s financial reality: breakneck volume growth, healthy but compressing margins, and the strains of reinvestment.

• Volume (TPV): $47 billion, up 73% year-over-year (YoY). This is the sixth consecutive quarter of >50% TPV growth. This metric signals that dLocal’s core product is more deeply embedded in merchants’ systems than ever.

• Revenue: $336 million, up 55% YoY . Revenue is growing slower than TPV, which is the first critical tension.

• Gross Profit: Record $119 million, up 40% YoY, core business profitability remains healthy.

• Gross Profit Margin: Fell to 35% from 39% a year ago.

• Operating profit: $57M exc. prior years tax adjustments (+25% YoY)(Management clarified during the May 14, 2026 earnings call that the one-time tax adjustment was related to the historical tax treatment of an installment payment product in certain markets);

• Operating Cash Flow (before WC): $69m (+10% YoY). Underlying cash generation is strong, separate from timing issues.

• Net income: $52M exc. tax adjustments (+11% YoY)

• Corporate Cash: $452 million, a significant cash buffer to manage any short-term fluctuations.

The most important analytical insight from the numbers is the divergence between TPV growth (73%) and Gross Profit growth (40%). The take-rate (Gross Profit / TPV) is falling.

Management attributes this to a “mix-shift”. Not only is dLocal growing faster with large, established merchants who negotiate lower rates), but while ‘pay-ins’ via local credit cards generate premium take-rates, domestic instant-payment rails (Pix/UPI) compress the blended take-rate but bring unmatched volume.

This is not necessarily bad. It means dLocal is winning the largest, most stable merchants and growing market share in its key markets. The company is now trading margin percentage points for volume scale.

Operating expenses rose 58% to $62m (excluding the one-time tax adjustment), outpacing gross profit growth. This reflects the investment cycle in engineering and product headcount initiated in late 2025. Management openly states costs were heavier in H1 2026 but expects operating leverage to “flow through the P&L” in H2 2026 as that investment cycle ends . This is the key hinge point for earnings quality in the coming quarters.

The free cash flow yield is attractive but volatile; adjusted free cash flow dropped 63% in Q1 due to working capital timing. The recent cash flow pressure is partly due to dLocal fronting costs to onboard new "Tier 0" global clients (like Amazon, Netflix, and Meta) . While these clients operate on lower take-rates, their massive volume increases dLocal's strategic relevance and gross profit dollars, requiring upfront investment that temporarily impacts working capital. Management expects this to reverse in subsequent quarters.

Mr. Market

Mr. Market is deeply conflicted about dLocal.

On the one hand, analysts are bullish on the surface, with 8 out of 10 rating it a “buy” and a mean street price target of $17.65, implying >50% upside from the current $12 share price .

On the other hand, Mr. Market dislikes uncertainty, and he sees two things that make him uncomfortable: (1) a take-rate that is compressing faster than expected, and (2) operating expenses growing faster than gross profit. The narrative that DLocal was a hyper-scalable, "capital-light" machine with expanding margins is broken, at least for now. The market is pricing in a "show me" story. It believes in the volume but needs proof that management can convert that volume into sustainable, high-quality earnings without constant reinvestment.

In short, the market is pricing dLocal with skepticism. The sentiment is cautiously optimistic but highly trigger-happy on any sign of weakness in the numbers.

As of mid-May 2026, the stock trades at a forward P/E of approximately 17x. This is despite compounding its TPV at 88% CAGR since its founding in 2016, with net cash on the balance sheet, strong cash flows, a capital light model and strong management. This represents a significant discount to the payments peer average and a steep discount to its own historical valuation

Growth remains robust. In Q1 2026, TPV grew at 73% YoY, a sixth consecutive quarter of growth above 50%. Yet the charts below show how the market capitalization of the business and its share price have performed since its 2021 peak.

So, what’s going on here?

How does a business with such great economic fundamentals collapse in value?

This goes to the heart of the investment thesis.

The shares once traded close to $70 each, but today are hovering in the mid to $11 range. That’s a five year CAGR of -29.47%. But looking at the market cap or share price in isolation is misleading and will lead to the wrong conclusions. Instead, it is important to break it down into components, or factors, to understand what drove that price action. Only then is it possible to formulate reliable opinions on the path forward.

So, let’s look at the attribution break down of that negative 29.47% return:

It becomes clear that early investors learned the hard way that a great company at the wrong price makes for a poor investment!

From 2021 until 2026, revenue grew from $244m to over $1.2bn (~38% CAGR) and the share count reduced a little. Both of these factors are huge positive for shareholder returns. The good news is that these tailwinds continue unabated.

Almost all of the damage to the valuation can be traced to two factors:

• First, profit margins have fallen from the mid-30s to the mid-teens as the company aggressively invested for growth. This wasn’t the result of a weakening business model, but a conscious decision to sacrifice near-term profitability in pursuit of long-term expansion. Management has indicated that this period of elevated investment is beginning to moderate, with margins expected to improve in the second half of 2026 and continue expanding into 2027.

• Second, the valuation multiple has collapsed. The stock once traded at more than 290x earnings (what was Mr Market thinking?!?). Today it trades at around 17x earnings. Importantly, that multiple is being applied to earnings that have been temporarily depressed by heavy reinvestment. On a more normalised earnings run rate, the effective multiple is closer to the low teens. Factor in the company's fortress-like net cash balance sheet and the EV/E multiple is lower still.

The key point is that both major valuation headwinds have already done their damage. Margins appear to be near a cyclical low and may now be heading higher. Likewise, the multiple has already compressed from extreme levels to a far more modest valuation. If anything, the balance of probabilities suggests multiple expansion is more likely than further contraction over the next several years.

What might the next five years look like?

The good news is that top line growth remains robust, the share count is still contracting, margins have probably bottomed and multiple contraction is done. At current valuations, dLocal looks very attractive.

Assume that top line growth slows to 30% annually (prudently optimistic), investment spending moderates causing free cash flow to improve and margins to expand. the additional cash would enable a more aggressive rate of share repurchases, which would be highly accretive at current depressed valuations. All of this would result in a re-rating of the stock and perhaps a valuation multiple closer to 21. In this scenario, the 2030 share price would be over $70 (6.5x where it is today). The next five years could deliver over 42% CAGR total shareholder returns, with attribution across the factors as follows:

It is also worth noting that this scenario has a PEG of well below one.

dLocal has invested heavily onboarding platform clients including Amazon, Meta and Netflix. This bodes well for durable cash flows going forward and paves a long runway for growth. All of these customers offer subscription services with strong recurring revenue. Every time a customer pays the next periodic subscription fee for Amazon Prime, or Microsoft 365 or Netflix, dLocal takes a cut. It is the capital light toll-booth model that GARP investors crave.

How likely is this to play out?

Let’s dive in to the Bull and Bear cases.

Bear Thesis

The strongest bear case against dLocal is that the company is a victim of its own success.

The bears would argue that the company’s revenue is concentrated among a few large global merchants.

A renegotiated contract with a key customer or a merchant deciding to build internal capabilities could materially impact results.

Also, they say that the very merchants dLocal targets, the global giants, may be its biggest threat. Amazon, Google, and Uber didn’t get to their size by ceding strategic control. As a country like Brazil or Nigeria becomes economically important, these merchants may eventually seek to build their own local payment infrastructure, cutting out dLocal entirely.

Bears would say that what looks like 73% TPV growth today is just a temporary rental of dLocal’s license before the tenant buys the building.

Furthermore, bears would say that the take-rate compression is not just a “mix-shift” but a structural feature of the business. As low-cost, state-sponsored rails like Brazil’s Pix and India’s UPI proliferate, dLocal’s premium pricing may erode.

Gross profit / TPV ratio, having declined from 1.05% to 0.84% gross profit yield in one year, is quoted as evidence of this trend.

The bears will point to Q1 2026’s gross profit margin drop from 39% to 35% as the canary in the coalmine. If margins compress another 400 basis points, the company will be growing revenue but generating less absolute gross profit.

If emerging market e-commerce growth slows, dLocal’s TPV growth, which drives everything, would decelerate sharply, while fixed operating expenses would reverse the tailwind of operating leverage.

The company’s April 2026 appellate court win (dismissing a securities class action) removed one legal overhang, but operational regulatory risk remains. A single unexpected tax ruling or a sudden currency control law in a major market like Nigeria or Argentina could wipe out an entire quarter’s worth of profit.

Ultimately, the bears would argue the risk-reward is unattractive for a business trading at 15x forward earnings that faces both secular margin compression and geopolitical volatility.

One final bear point centres on General Atlantic (GA), an investment firm which still owns roughly 22% of dLocal and has been a cornerstone shareholder since 2019. The concern is straightforward. When such a large holder eventually exits, the market has to absorb a meaningful amount of supply.

Investors have already seen a preview of this dynamic. In September 2025, GA sold 15 million Class A shares in a secondary offering, alongside an option for another 2.25 million shares. The shares were priced at $12.75, a material discount to the previous day’s close. The stock fell almost 8% in a single session as the market digested the additional supply.

That episode reinforced fears that GA’s remaining stake could act as a structural overhang on the shares. Investors are now left asking when the next block might come to market, and at what price.

There is also a broader governance concern. GA’s incentives may increasingly shift from long-term value creation toward engineering an efficient exit. That does not necessarily imply malicious intent. Private equity firms eventually need liquidity. But public market shareholders have to recognise that a large financial sponsor may prioritise portfolio realisation, fund timelines, or tax structuring considerations over the long-term compounding potential of the underlying business.

This concern intensified after reports that GA encouraged dLocal to explore strategic alternatives, including hiring Morgan Stanley to evaluate a potential sale of the company to larger fintech firms or other private equity buyers. For some investors, that raises the possibility that the company’s future direction could be shaped more by shareholder liquidity objectives than by operating fundamentals.

To be fair, selling by a major shareholder is almost always interpreted negatively by the market, even when the motivation is entirely routine. Fund expirations, portfolio concentration limits, or simple rebalancing can all drive sales that have little to do with the underlying business quality.

Only time will tell whether GA ultimately proves to be a structural headwind to the investment case, or simply the source of an unusually attractive entry point.

Bull Thesis

The bull case argues that the market is misreading the Q1 2026 results entirely.

Bulls say that the 73% TPV growth is proof of accelerating adoption, not a problem to be solved. The take-rate compression is the natural, healthy consequence of mixing in massive, high-retention volume from merchants like Amazon and Microsoft.

Bulls say that the risk of dLocal’s largest customers creating the infrastructure to cut it out is overstated. Building infrastructure requires local regulatory licenses in dozens of volatile jurisdictions. While Amazon and others can easily build an API, do they want to negotiate and hold 38 separate local central bank licenses across emerging markets? This is less about technology and more about the friction of local regulatory compliance.

The bulls would argue that Gross Profit, up 40%, is the only number that matters for intrinsic value, and 40% organic growth is incredibly rare, highlighting that the runway ahead for dLocal stretches far into the distance and is a long way from being exhausted.

In terms of margin contraction, the current operating expense growth is a strategic choice to invest for the next decade, not a sign of a broken model.

Operating expenses grew faster than gross profit, delaying the operating leverage investors expected from a capital-light platform model. ‘Delaying’ being the operative word.

The bull case requires that the investment cycle ends as promised and that the second half of 2026 shows operating expense growth lagging gross profit growth.

When the investment cycle ends, the acquisition of sticky customers will pay huge dividends. Earnings will inflect significantly and the true earnings power of the business will become evident.

The bull thesis also rests on the “optionality” of new products. BNPL Fuse, stablecoin settlement rails, and tokenized alternative payment methods are not just dLocal marketing. Stablecoins aren't a gimmick for dLocal; they drastically lower the friction of merchant repatriation, solving the "trapped cash" problem in certain jurisdictions (e.g. parts of Africa, India and Argentina). These carry higher take-rates and can be cross-sold to the existing merchant base without incremental customer acquisition cost.

Furthermore, the expansion into Asia (Vietnam, Indonesia) and the Middle East (Oman, Kuwait, Qatar) opens up entirely new continents for growth as digital economies there formalize.

Subscribe now

What to Watch For

Where the bears see a company that has harvested all the low hanging fruit and is now past its peak performance, the bulls see a mispriced compounder.

The following indicators, all observable in future SEC filings, will determine which thesis prevails:

1. Watch the Q2 and Q3 2026 reports to see if the 35% gross margin was a trough or a new baseline. Stabilization or expansion here would be a powerful bullish signal.

2. Gross profit / TPV ratio: If the 0.84% from Q1 2026 stabilizes or expands, the bear case on structural margin erosion weakens. If it continues declining, the bull case falters.

3. Management promised operating expenses would moderate in H2 2026. If Q3 shows operating expense growth slowing below gross profit growth, the “investment cycle” narrative is validated.

4. New product growth, particularly the company’s commentary on BNPL and stablecoin adoption rates will offer a clue on the future trajectory of the business. Look for mentions of “take-rate uplift” or “merchant adoption rates” on earnings calls.

5. Client concentration risk remains high. Any announcement of a major renegotiation or loss of a top 5 merchant would damage the valuation.

6. The Q1 2026 adjusted free cash flow decline was attributed to a temporary timing issue. Q2 and Q3 filings will show whether a working capital reversal occures and whether cash flow normalizes.

7. The likes of Adyen, Stripe and Paypal could decide to aggressively acquire local licenses and build out APM coverage, or a consortium of local champions could form a cross-border alliance. For now, DLocal’s moat appears durable, but the company is stretching itself by moving into new products (BNPL, stablecoins) and geographies (Asia and the Middle East). This expansion, if executed poorly, could fracture focus and weaken its core value proposition.

8. The company generates cash. If the board authorizes a significant share buyback, it would signal confidence that the stock is undervalued, underpinning its valuation.

9. DLocal’s management has guided for 50-60% TPV growth and 22.5-27.5% gross profit growth for the full year 2026 . If they hit the high end of both ranges and demonstrate that operating leverage in the second half, the current valuation (~15x forward earnings) will look absurdly cheap in retrospect. If they miss, the stock may re-rate to downwards as it becomes a “volume, not value” story.

Share

Summing It All Up: Is DLocal a Good Investment?

DLocal is a high-quality business with a genuine, durable moat trading through a temporary, but painful, phase of margin compression and investment. The question for an investor is not whether the company will grow; it will. The question is whether the growth will be high-return, expanding-margin growth or low-return, capital-intensive growth.

The bear case is legitimate, while the bull case is plausible yet risky.

For an investor with a multi-year horizon and tolerance for emerging market volatility, dLocal offers a reasonable probability of asymmetric upside.

The price is discounting a low-growth, margin-eroding future. The reality, backed by $14 billion in quarterly TPV, is that dLocal remains the most important tollbooth for global e-commerce in the Global South. The toll rate is dropping, but the traffic is exploding. Owning dLocal is a bet that the traffic explosion wins over the toll rate decline.

At current prices, that isn’t an unreasonable bet.

Leave a comment

Share

You may also like to read:

https://rockandturner.substack.com/p/wise-the-holy-grail-investment (link goes live on 9th June 2026 at 14.00 ET)

There’s an argument circulating in the investment community. Every time I hear it, there’s a small part of me thinking something about it doesn’t hold up.

The argument goes like this: AI is not really eating software because the value still sits in the underlying data. Software companies remain the system of record. AI is just a layer on top, dependent on whatever sits beneath it.

The logic sounds reasonable. But I think it misses something important.

I ask you a simple question: “What’s eight multiplied by nine?”

You don’t calculate it. You don’t reach for a calculator. Your brain immediately says “seventy-two”. At some point in childhood, repetition burned that pattern into your memory. It stopped being arithmetic and became a reflexive recall function.

Large language models work in a surprisingly similar way.

When you ask an LLM to solve ‘eight multiplied by nine’, it also answers instantly.

If you picture the AI furiously multiplying eight by nine every time you or someone else asks, you’d be wrong. That would be absurdly inefficient. During training, the model has encountered ‘8 × 9 = 72’ so many times that the association becomes embedded in its weights. It becomes part of its neural fabric. When you ask the question, it remembers. Kind of the same way we do.

The model isn’t calculating. It’s recalling.

That distinction matters more than people realize.

Before AI, computing was algorithmic. A program running on a CPU required the computer to fetch data from source every time it ran. Data retrieval was part of the process. It would reach out across the internet, knock on an API’s door, and say, “Give me that piece of data, please.” And the data provider would say, “Certainly. That’ll be a fraction of a penny.” If the same fact was requested a million times, the provider monetized it a million times.

That model worked because software needed continuous access to the source of truth.

Now comes the era of the large language model. An LLM operates entirely differently. To be clear, it still burns an immense amount of computational power (FLOPs) during the query phase; every single token it generates requires billions of matrix multiplications across its weights. The AI then compresses information directly into its billions of algebraic parameters during training.

When a factual query is made, the answer emerges deterministically from those weights. It is statistical recall, not an API lookup loop. The model doesn’t know it has stored a discrete fact. It has simply seen the sequence so many times that the probability of “72” being the correct output has converged close to ‘1’.

Now imagine licensing a massive archive of historical stock prices to train a model. After training, the model can answer questions about Apple’s closing price on January 15, 2020 without querying a financial API. The information has effectively been absorbed into the network, synthetically memorized.

No API request. No toll booth. No recurring monetization.

The old model monetized a dumb algorithm that needed to look up the answer time and time again. AI learns and artificially remembers.

This explains that little voice casting a doubt in the back of my mind. AI is not as dependent on historic data as many imagine.

That is why I think “the data remains the moat” narrative is incomplete. In many cases, AI weakens the value of static data precisely because repeated retrieval is no longer necessary.

For companies built on charging for the same fact millions of times, that changes everything. It’s an existential crisis. It directly attacks the economic foundation of their business model. The uncomfortable question is simple: what exactly are you selling once the machine remembers?

Increasingly, the value is moving toward freshness, verification, and low latency.

An AI’s memory is a lossy compression library. You and I remember that eight times nine is 72, but if I asked you for the menu prices of a restaurant we visited in Omaha back in 2012, you’d look at me blankly. LLMs are also great at “8 x 9” because it’s ubiquitous, but, like us, they run into serious difficulty with rare or highly precise data.

Consider financial statement analysis. When trained on millions of rows of precise but similar numbers, details often blur. Because LLMs rely on statistical probabilities rather than exact database lookups, they can conflate digits or flat-out hallucinate.

That limitation matters enormously in fields like finance, law, medicine, and enterprise infrastructure, where a “near 1” statistical probability isn’t good enough. That lossiness is exactly why data providers still have some leverage.

That is why Retrieval-Augmented Generation (RAG) matters. In a RAG system, the model actively queries external sources during inference to retrieve verified or current information. The value is no longer the raw fact itself. The value is confidence that the answer is current and correct. It’s analogous to a lawyer checking a law book prior to offering counsel, rather than simply relying on memory.

The best run data businesses have recognized this stark new reality and are already pivoting.

Some are shifting toward large one-time licensing agreements for training data. Others are focusing on real-time information that cannot simply be memorized: live prices, inventory levels, sports scores, logistics data, market feeds and such like.

So, as an investment community, we should probably spend more time asking a simple question: which companies are adapting with humility, and which still believe the world will keep rewarding yesterday’s playbook?

Some management teams look at AI, changing consumer behaviour, shifting economics, and new competitive dynamics with genuine curiosity. They are willing to rethink assumptions, cannibalize parts of their own business, and accept that prior success does not guarantee future relevance. Those companies tend to evolve before they are forced to.

Others behave very differently. Years of dominance create institutional certainty. Processes harden. Incentives calcify. Management starts defending the existing model rather than questioning it. The company mistakes market leadership for permanence. That is usually when hubris begins to creep in. Management continues talking as though the old rules still apply.

And hubris is dangerous because it often looks rational in the moment. The incumbent still has scale. Still has margins. Still has customers. The decline rarely starts as a collapse. It starts as complacency. It happens slowly at first, then the cliff edge comes into view and it’s invariably far too late to change course.

AI has introduced an environmental shift. No question about it. Now, software companies need to evolve. It’s a Darwinian case of survival of the fittest; those that don’t evolve are likely to die out.

The difficult part for investors is separating temporary noise from genuine adaptability. Every company claims it’s embracing AI change. Yet very few actually restructure incentives, products, or business models for fear it will threaten their current cash flows.

The approaches of financial data businesses differ enormously. If you’re a thin UI layer on top of a system of record, you’re going to have to earn your keep. Dow Jones / Factiva is clearly adapting, packaging its content for GenAI use and licensing it for AI models and solutions. In stark contrast, many others still seem focused on defending legacy walled gardens. Bloomberg, FactSet, and LSEG are responding by embedding AI into their own product stacks rather than fully leaning into data licensing as a broader AI distribution opportunity.

Subscribe now

The companies most likely to survive major transitions are often the ones most willing to disrupt themselves early. The ones most at risk are usually the ones still explaining why disruption supposedly doesn’t apply to them.

So I’m curious,

• Which software/data companies do you think are showing real foresight and humility today?

• And which ones seem increasingly trapped by the assumptions created by their own past success?

Please leave your answers in the comments section below.

Leave a comment

Share

Attribution: this post was inspired by Harris Kupperman (‘Kuppy’) of Praetorian Capital.

Uber (UBER) was founded in 2009, riding the wave of the smart-phone revolution, and went public a decade later on May 10, 2019.

On the surface it appears to be a winner, but is it really?

Gross bookings and revenue are growing at around 20% a year, and costs are increasing much more slowly than the top line, which means that operating leverage is finally showing up in the numbers.

Yet, it took the company until FY2023 to break into the black. Finally, after 14 years, it stopped bleeding capital.

Profits are now compounding faster than revenue, and the business has evolved into something close to a transportation utility within the digital economy.

The numbers look encouraging. In 2025, Uber generated $5.6 billion of operating income on roughly $29.5 billion of invested capital, equivalent to a pre-tax return on capital of around 19%. That’s a respectable outcome, but it only captures a snapshot in time. It doesn’t paint the full picture.

What makes the story more complicated is the path it took to get there. Investors absorbed years of losses, relentless cash burn, and continuous dilution before the company finally produced meaningful returns.

Uber’s accumulated losses peaked at $32.8 billion in 2022 and it hasn’t worked that deficit off yet. At the end of FY2025, the net losses still stood at $10.4 billion.

What does this mean in real terms?

The essence of any investment is to turn one dollar today into more than one dollar tomorrow. It is about planting a seed and watching it grow into a wonderful tree. Yet throughout its 27 year life, Uber has destroyed way more capital than it has created.

And the story may not be finished. There is still scope for the company to reverse course and drop back into the red.

Consider this: autonomous vehicles may reshape the economics of the industry all over again. If self-driving technology becomes commercially viable at scale, it isn’t clear which platforms consumers will ultimately use to access that robo-taxi network. Uber may need to reinvent itself once more. If that happens, margins could compress, investment requirements could rise sharply again, and the cash burn cycle may restart.

There is something unsettling about a company achieving global dominance with Uber’s scale, brand, and global reach while consuming so much capital and struggling with profitability.

Uber is not an isolated case. This is a common phenomenon across the technology sector more generally.

Most people have used Twitter (now X) at some point.

It arrived at exactly the right moment. Social media was absorbing more and more of people’s attention and Twitter offered something differentiated from platforms like Facebook. The product spread in popularity as real-time information, open conversation, and distribution at scale all became compressed into a single network.

Twitter went public in 2013 at roughly $45 per share. Nine years later, Elon Musk acquired the company, taking it private, at approximately the same price.

In nominal terms, shareholders earned nothing across almost a decade of ownership. The company never paid a dividend and, after inflation, investors effectively generated negative real returns despite owning one of the most culturally influential platforms of the internet era.

There is little doubt the platform created enormous value for its users. It shaped politics, media, finance, culture, and public discourse for more than a decade.

But creating value for users and creating value for shareholders are not always the same thing. For an investor, if Twitter was a technology winner, what does losing look like?

Remember Napster? Long before Spotify existed, Napster fundamentally changed consumer behaviour around digital music consumption. Adoption exploded. Yet the business still ended with Chapter 11 bankruptcy protection having generated liabilities that far exceeded assets.

3dfx pioneered dedicated 3D graphics accelerators for PCs. Its Voodoo products helped define PC gaming in the late 1990s. Technological leadership didn’t save it from insolvency. Its assets were eventually sold to Nvidia.

Jawbone built highly desirable Bluetooth devices and fitness wearables. Venture capital funded aggressive expansion and at one point it achieved a private valuation above $3 billion. But costs outran the economics of the business. The company was eventually liquidated.

Pebble helped pioneer the modern smartwatch category and ran one of Kickstarter’s most successful campaigns. Demand was real. The product resonated. But revenue never kept pace with spending and the company ultimately failed.

The grave-yard of the tech sector is littered with failed businesses that had hugely successful products and services.

While a small number of tech companies generate extraordinary returns, the vast majority fail, at least economically.

That is the contradiction at the centre of the technology sector. The great dilemma for the investor. Invest for fear of missing out, and risk an unfavourable geometric skew, or simply sit it out and allocate capital to other sectors with a better risk/reward profile.

The underlying innovation often creates enormous societal value. Consumers gain better products, lower costs, faster services, and entirely new capabilities. But the investors who financed that progress frequently capture little of the economic upside themselves.

In many cases, their capital effectively subsidises the technological advancement and the true beneficiaries are not the innovators, but the second and third generation adopters of that technology.

Think about Xerox. It was responsible for pioneering many groundbreaking computer technologies, including the GUI, the computer mouse, Ethernet networking, laser printing, and the first personal computer (the Alto). However, it failed to effectively commercialize and capitalize on its key innovations, which were later adopted and popularized by other companies like Apple, Microsoft and Dell (second generation beneficiaries).

This is nothing new. Early investors in railroads funded massive track expansion in the 1800s and investing heavily in capacity during a speculative boom , driving returns down through competition, high fixed costs, and price wars. Many went bankrupt or earned poor returns despite creating invaluable infrastructure. The real winners came later; businesses that used that infrastructure. Without bearing the upfront capital risk, it becomes far easier to succeed economically. Companies like Sears and Standard Oil scaled nationally using cheap transport.

More recently in the 1990s the first wave of telecom firms laid excess fibre network capacity and many collapsed (e.g. Global Crossing). Later, internet companies like Google, Amazon, and Netflix thrived on that cheap bandwidth. Builders absorbed the capital risk; users captured the economic upside.

The pursuit of outsized returns can obscure the scale of risk involved when investing in new technology. Investing in this sector is often treated as a pathway to exceptional wealth creation, but the failure rates tell a different story.

Tech companies face materially higher risks than businesses in most other industries. Reported five-year failure rates are estimated at roughly 63%. For comparison, the next worst sector is construction, a cyclical and operationally difficult industry, yet the failure rate there is significantly lower (~34%).

That raises an uncomfortable possibility: technological advances may structurally favour consumers over investors. Sure, Uber and Twitter are winners having developed platforms that have gone viral creating huge amounts of utility for their users. But if this is what winning looks like, investors should ask themselves whether the game is worth playing.

Share

Leave a comment

DISCLAIMER & DISCLOSURE: The author holds no investment in RELX at the date of publication but that may change. The views expressed are those of the author and may change without notice. The author has no duty or obligation to update this information. Some content is sourced from third parties believed to be reliable, but accuracy is not guaranteed. Forward-looking statements involve assumptions, risks, and uncertainties, meaning actual outcomes may differ from those envisaged in this analysis. Past performance is not indicative of future results. All investments carry risk, including financial loss. This analysis is for educational purposes only and does not constitute investment advice or recommendations of any kind. Conduct your own research and seek professional advice before investing.

Why Should I Be Interested In RELX?

• Share price: £23.33 GBP

• Market Cap: £41.3bn GBP

• EV: £48.5bn GBP

• Float: 99.2%

• ROIC: 29%

• Dividend yield: 3.1%

• NTM P/E: 16.5x (historically closer to 30x)

• Share count reducing ~1.5% annually

• Potentially a major beneficiary of AI

RELX operates in markets where scale, data, and workflow integration matter. Across most of its divisions, it holds leading market positions built over decades. Its products are embedded into customer workflows and often support decisions tied to compliance, risk, research, and professional productivity.

That makes the spending highly resilient. For many customers, RELX is not discretionary. Its services are mission critical. Once integrated into daily workflows, switching costs become meaningful.

The result is a business with durable recurring revenue, strong pricing power, and highly predictable earnings.

While growth is strong:

The share price has weakened:

So, what’s happening here?

The Company

RELX is a global provider of information-based analytics and decision tools for professional and business customers and is headquartered in London. The company is now listed in London, having consolidated its dual Dutch listing back in 2018.

RELX began as a publishing business, but today it operates as a technology-enabled analytics company across four core segments: ‘Scientific, Technical & Medical’ (STM), Risk, Legal, and Exhibitions.

Its roots trace back to the formation of Reed Elsevier, created through the merger of a British publisher and a Dutch printing business. That publishing heritage still shapes parts of the culture, but the business itself has changed materially over the past decade.

Management has steadily repositioned RELX away from print and toward higher-margin electronic analytics and workflow tools, increasingly enhanced by AI and proprietary data. Print and print-related activities now account for less than 3% of revenue and continue to decline structurally. The transformation of the business mix has been significant.

The economic model has improved alongside that transition. Digital analytics products carry materially higher incremental margins than legacy print operations. As the revenue mix has shifted toward recurring decision tools and embedded workflow software, profitability has expanded. Adjusted operating margins increased from 30.5% in 2015 to 34.8% in 2025.

What stands out in RELX’s history is the consistency of its capital allocation and the patience behind its strategy. Management spent years investing in proprietary data assets, workflow integration, and technology infrastructure before the acceleration in organic growth became fully visible. The company did not optimize for short-term earnings at the expense of long-term positioning.

That matters because the key question for investors today is whether RELX is still early in monetising those investments, particularly through AI-enabled analytics and decision tools, or whether much of that value creation is already reflected in the current business and valuation.

Share

The Business Model

RELX makes money by selling proprietary data, analytics, and decision tools to professional customers who embed these products into their daily workflows. While the economics differ across segments, the common thread is the same: high switching costs, recurring usage, and products that customers rely on to perform core functions.

The business operates across five segments (four core in growth mode, and one legacy in decline):

1. Risk, the largest segment at roughly 37% of group revenue, sells fraud detection, identity verification, insurance underwriting analytics, financial crime (KYC and AML) and other compliance tools.

   

   

   Customers include banks, insurers, government agencies, and corporations. Approximately 61% of Risk revenue is transactional, linked to volumes such as insurance quotes or fraud checks, while the remainder is subscription-based. The segment operates at a 37.8% adjusted operating margin, the highest in the company. The key assumption that must remain true for the business model to work is that regulatory requirements around fraud prevention and identity verification continue to intensify, driving sustained demand.

   

   

2. STM - Scientific, Technical & Medical (roughly 29% of revenue) helps advance science and healthcare by combining high-quality, trusted scientific and medical information and data sets with innovative technologies to deliver critical insights that support better outcomes.

   

   

   It derives its income from journal subscriptions, open access article processing charges, and clinical decision support tools like ClinicalKey and ScienceDirect. Institutional customers pay for access to RELX’s proprietary content libraries and citation databases.

   

   

   The segment operates at a 37.1% margin. The critical assumption is that research institutions continue to value bundled access to RELX’s journals over alternative models such as open access or AI-powered research tools.

   

3. Legal, about 18% of revenue, sells legal research platforms led by Lexis+ in the United States, along with regulatory compliance tools and practice management software.

   

   

   This is predominantly subscription revenue. The segment carries the lowest margin at roughly 21%, reflecting its competitive position and historical underinvestment. The assumption here is that AI-enhanced products can drive continued share gains and margin expansion.

   The legal product, ‘Lexis’ is constantly evolving as the graphic below demonstrates:

   

   

4. Exhibitions, roughly 13% of revenue, earns money from exhibitor fees, sponsorships, and attendee registrations at trade shows.

   

   This is a transactional and exposed to economic cycles.

   

5. Print and print-related activities, now reported separately, is in structural decline. It contributed 1.8% of revenue in FY2025, down from 2.2% in FY2024.

In an April 2026 presentation, RELX outlined a strategy centred on sustaining strong growth in its Risk and Exhibitions divisions while improving the growth trajectory of STM and Legal.

The direction of travel is clear. RELX continues to shift the business mix toward faster-growing analytics, workflow, and decision tools, while legacy print operations become increasingly irrelevant. Print now represents only a small portion of the business and is being marginalized.

That transition matters because the higher-growth segments also tend to carry better economics. Risk benefits from proprietary data and embedded decision workflows. Exhibitions has recovered strongly with pricing power and attractive operating leverage. STM and Legal remain large recurring revenue businesses with scope for further digital monetisation and AI-driven product enhancement.

The broader theme is that RELX is steadily concentrating capital and management attention on businesses with stronger recurring revenue, higher switching costs, and better long-term margin potential.

The following charts show adjusted revenue growth (explained in the ‘Numbers’ section below), first by segment and then over time:

People

RELX is led by a management team that has remained notably stable throughout the company’s transformation from a publishing conglomerate into a data and analytics business.

CEO Erik Engstrom, formerly of McKinsey & Company, has led RELX since 2012. That kind of tenure is increasingly rare in large public companies. With the average CEO lasting only a little over four years, RELX stands out for having leadership willing to allocate capital and build strategy over multi-decade periods rather than quarter-to-quarter cycles.

That stability matters because RELX’s transformation required patience. The company spent years repositioning the portfolio, investing in proprietary datasets, digitising workflows, and building analytics capabilities before the financial benefits became fully visible. A shorter-term management team may not have maintained that level of strategic consistency.

The result has been a gradual but meaningful shift in both the quality of the business and its economic profile.

The cultural traits that matter are disciplined cost control, long-term investment horizons, and customer-centric product development.

RELX employs over 12,000 technologists and spends approximately £1.9 billion annually on technology, yet has still expanded margins. That combination is rare and suggests genuine operating leverage rather than simple cost-cutting.

The company’s incentive structure is reasonably aligned with shareholders. Management compensation is linked to adjusted earnings per share growth, cash flow conversion, and relative total shareholder return. Those are practical operating and capital allocation metrics rather than vanity targets.

Management explicitly states its objective of “managing cost growth below revenue growth” and has delivered on this consistently. This translates into improving operating leverage.

Another characteristic that stands out is the tone of management communication. RELX leadership has generally avoided promotional behaviour. Guidance has historically been conservative, with management tending to under-promise and over-deliver.

Management at RELX has demonstrated consistent capital allocation discipline and strategic patience over a long period of time:

Capital allocation priorities are sensible.

Organic investment and bolt-on acquisitions sit at the top of the list, which reflects the reality that RELX’s competitive advantage is built through data assets, workflow integration, and product enhancement over time. Management has consistently reinvested into the business rather than maximising short-term distributions.

Buybacks appear disciplined rather than mechanical. Repurchases have generally been tied to valuation and cash generation. The decision to increase buybacks to £2.25 billion for 2026, up 50% versus the prior year, is notable. It suggests management believes the current valuation remains attractive relative to the company’s long-term cash generation potential.

Leverage has also been managed conservatively. Outside of the temporary disruption during Covid-19, net debt has typically remained within a 2.0-2.5x range. Taken together, the framework reflects a management team focused on long-term compounding rather than financial engineering.

The one less attractive element, at least from my perspective, is the traditionally British and European preference for a rigid progressive dividend policy. Roughly 50% of adjusted earnings are distributed through dividends each year, regardless of capital allocation opportunities elsewhere in the business. It just makes no sense. (If you’re not familiar with my thoughts on dividends, please see this link.)

Share

RELX and AI

Rather than framing AI as a revolutionary narrative, management presents it as an extension of existing analytics and workflow capabilities that improves customer outcomes.

That distinction matters. RELX’s moat is not built on AI hype. It is built on proprietary data, embedded workflows, customer trust, and integration into mission-critical decision making

The company is focusing on using AI to extract and process data, before delivering it in a clean, standardized coherent format as required by its customers:

It is also leaning into generative AI to deliver workflow tools, tailored analysis, drafted documents and compliance with global regulations.

RELX is increasingly selling decision tools rather than raw information. Customers are paying for outcomes: lower fraud losses, faster legal workflows, better underwriting, improved clinical decisions, and higher productivity.

That distinction explains why RELX has been able to expand margins even while investing heavily in technology infrastructure and AI capabilities. The business is no longer monetising access to information alone. It is monetising improved decision-making embedded directly into customer workflows.

Subscribe now

Competition and the Moat

RELX has real competitive advantages, but the strength of the moat varies materially across its segments. The common thread is not content ownership alone. It is workflow integration, proprietary data, and the cost of disruption once customers embed RELX products into day-to-day operations.

Risk has the strongest moat in the portfolio. Customers use RELX’s analytics for fraud detection, identity verification, underwriting, and compliance. These tools are deeply integrated into operational systems at banks, insurers, and government agencies. Replacing them is not a simple vendor switch. It often requires rebuilding workflows, retraining teams, and revalidating compliance processes.

Regulation strengthens the moat further. Compliance departments need reliable and auditable data sources, and RELX has built trust over decades. Competitors such as Verisk, TransUnion, and Equifax remain credible players, alongside newer fintech entrants, but few combine RELX’s data breadth, regulatory relationships, and workflow integration at scale.

STM is more complicated. Historically, the moat came from ownership of prestigious journals and control of the academic publishing ecosystem through peer review and distribution. That advantage is gradually weakening as open-access publishing expands and AI-native research tools reduce the importance of accessing individual journal articles directly.

RELX’s response has been to move higher up the value chain. Through Elsevier, the company has focused increasingly on analytics, workflow, and discovery tools such as Scopus and ClinicalKey AI. The strategic objective is clear: become more valuable for research synthesis and clinical decision-making rather than simply monetising content access.

Competition here is intensifying. Traditional rivals such as Springer Nature now compete alongside AI-native research platforms like Semantic Scholar and Elicit.

In May 2026, Elsevier acquired Mytonomy, a provider of video-based patient engagement software used across hospitals and clinics. The deal expands Elsevier’s clinical workflow ecosystem by integrating patient education, surveys, and communication tools directly into electronic health record systems. Strategically, this matters because it pushes RELX deeper into workflow integration while offering clinically validated alternatives to unreliable internet and AI-generated medical information.

Legal sits somewhere in the middle. The US legal research market remains effectively a duopoly between RELX’s LexisNexis and Thomson Reuters through Westlaw. Both benefit from entrenched customer relationships, trusted legal databases, and integration into professional workflows.

At the same time, legal technology is becoming increasingly fragmented and competitive. AI-native firms such as Harvey, Legora, Everlaw, Luminance, and CoCounsel are growing quickly and targeting adjacent workflow opportunities.

It is probably more useful to think about RELX Legal as two separate businesses. The first is the legacy legal information business, where Lexis holds roughly one-third market share and continues to grow at high single digits. The second is the broader legal workflow and AI software market, which has a much larger theoretical TAM but lower barriers to entry and more fragmented competition from AI-native start-ups.

The benefit of competing with start-ups is that founders often have one eye on a relatively short-term exit. External venture capital funds research and development, many will fail, but of the few that succeed, RELX is well placed to swoop in and acquire them. It’s one way to de-risk the evolution of one’s business.

A recent example involved the April 2026 agreement to acquire Doctrine is strategically important because it expands LexisNexis deeper into continental European civil-law markets such as France, Germany, Italy, and Spain. Historically, LexisNexis has been strongest in common-law jurisdictions like the US and UK.

RELX has also pursued partnerships rather than trying to build every workflow internally. Its April 2026 partnership with Luminance extends its reach into contract drafting and legal workflow automation .This is part of RELX's strategy to expand its distribution footprint while keeping full content value within its core ecosystem.

Then came the May 2026 collaboration with OpenAI which aims to integrate generative AI more deeply across the LexisNexis ecosystem to support more sophisticated, personalized legal workflows and generative-AI tasks. This is part of RELX’s broader push to embed AI into its information product where text-heavy workflows are a natural fit for large language models.

Ultimately, combining OpenAI’s model capability with RELX’s proprietary legal content and workflow products, make tools like LexisNexis even more useful and more difficult to replicate. It is all about improving user adoption, pricing power, and retention rates.

Exhibitions has the weakest moat in the group. The business competes against players such as Informa in what is ultimately a relationship-driven and cyclical industry. Trade shows benefit from scale and brand reputation, but the pandemic demonstrated that these advantages are not impregnable. The segment remains valuable, but it is unlikely to be the primary long-term driver of RELX’s intrinsic value.

Numbers

Before diving into the numbers, it is worth addressing how RELX reports its financials.

The company emphasises “underlying” revenue and operating profit rather than strictly reported IFRS figures. These are adjusted, non-IFRS metrics.

In many businesses that would immediately raise concerns. Here, it is actually central to understanding the thesis.

At first glance, the reported numbers appear uninpressive. IFRS revenue growth was only 2.98% in 2024 and 1.65% in 2025. On the surface, that hardly looks like a high-quality compounder. But the reported figures contain several distortions that obscure the economics of the core business.

RELX makes adjustments to the audited numbers which strip out ‘noise’:

• Legacy print operations continue to decline structurally, falling roughly 21% in 2025. Since this part of the business is intentionally shrinking, management excludes it from underlying growth metrics so the performance of the digital analytics and workflow businesses is not masked by a runoff segment.

• Currency is another major factor. A significant portion of RELX revenue is generated in US dollars, while reporting is done in pounds sterling. Constant-currency adjustments therefore remove FX translation noise that has little to do with operational performance.

• Exhibitions adds another complication. The industry has a natural odd-even year cycle because certain large events occur biennially. That creates timing distortions in reported growth rates that do not necessarily reflect underlying demand.

• Then there are acquisitions and disposals, which can further distort year-over-year comparisons.

Once those factors are adjusted for, the picture changes materially. Starting with reported 2025 revenue of £9,590m and add back the revenue lost due to the Print decline (£109m), the stronger currency (£196m), and adjust for disposals (£91m), the “real” top line growth performance of the underlying business was roughly 7%.

That is a very different business from the one suggested by the headline IFRS figures. It better describes the performance of the four pillars in which an investor today pins his expectations for the future.

Underlying performance in 2025 was strong across the board. Revenue grew 7%, adjusted operating profit increased 9%, and adjusted EPS reached 128.5p, up 10% on a constant-currency basis.

More importantly, this does not appear to be a one-off. RELX has gradually accelerated organic growth over the past decade. Underlying revenue growth averaged roughly 3-4% between 2015 and 2019, but stepped up to a 7% average after 2021, as the business mix shifted further toward analytics, workflow software, and AI-enabled decision tools.

This is where the investment case becomes more interesting than the headline numbers initially suggest. Beneath the accounting noise is a business with improving growth, expanding margins, strong recurring revenues, and increasing exposure to higher-value decision tools embedded deeply into customer workflows.

The segment-level numbers show clearly where growth inside RELX is coming from and why the overall quality of the business continues to improve.

• Risk remains the standout division. Underlying revenue grew 8% in FY2025, broadly consistent with its multi-year growth profile. The segment is now operating at an annualised revenue run-rate of roughly £3.5 billion.

• Legal also continued to perform well. Underlying growth accelerated to 8.5% in the FY2025 versus 7% in 2024, helped by continued adoption of Lexis+ AI and broader legal workflow tools. This is viewed by management as key driver of fuutre growth.

• STM improved to 5% underlying growth, which is encouraging given the structural concerns surrounding traditional academic publishing. The segment continues to transition away from content monetisation toward analytics, workflow software, and clinical decision tools. The growth rate remains lower than Risk and Legal, but the direction of travel is positive.

• Exhibitions remains the smallest and most cyclical of the major divisions, yet still delivered 8% revenue growth. The segment benefits from strong event franchises and pricing power during healthy demand periods, although its economics and moat are materially weaker than the analytics businesses elsewhere in the portfolio.

Taken together, the mix shift is important. Growth is increasingly being driven by the higher-quality segments of the business: Risk, Legal, and analytics-driven STM. These are the divisions with stronger recurring revenue, embedded workflows, higher switching costs, and deep integration into customer operations. This translates into better pricing power and greater long-term margin potential.

The margin profile at RELX continues to support the argument that this is becoming a higher-quality analytics and workflow software business rather than a legacy publishing company.

Group adjusted operating margin expanded to 34.8% in FY2025, up from 30.5% a decade earlier.

Importantly, this margin expansion has not come at the expense of cash generation or capital efficiency. Cash conversion improved to 99% from 97%, while return on invested capital increased to 15.4% from 14.8%.

Those are strong indicators that RELX is scaling efficiently. As the revenue mix shifts toward digital analytics and embedded workflow products, incremental revenues carry materially higher margins and require relatively modest incremental capital investment.

The balance sheet also remains conservative. Net debt to EBITDA ended FY2025 at 1.93x, comfortably below management’s stated operating range of 2.0-2.5x. Despite the increased focus on acquisitions and AI investment, leverage remains controlled.

Share

Mr. Market

The market currently views RELX with deep scepticism. Shares now trade around 43% below their 52-week high and approximately 41% below the 200-day moving average.

The broad weakness suggests investors are treating RELX as a victim of AI-disruption rather than differentiating between the more vulnerable parts of the portfolio and the segments with genuinely durable competitive advantages.

The business today is objectively growing faster, generating higher margins, and producing better returns on capital than it was several years ago. Yet the valuation multiple has compressed materially.

RELX trades on roughly 17.5x forward earnings based on FY2026 estimated EPS of 127.5p, a significant discount to its two-year historical average multiple of roughly 30x.

Against a projected EPS CAGR of approximately 11.8% between FY2026 and FY2028, the implied PEG ratio is around 1.5x. That does not immediately screen as deeply undervalued, but it also doesn’t look demanding for a business with RELX’s quality profile, recurring revenue base, and improving margin characteristics.

Reverse engineering the earnings multiple on a sum-of-the-parts basis helps frame the valuation differently. On implied segment multiples, the picture appears more balanced:

• Risk at roughly 24x earnings arguably deserves a premium given its stronger growth, superior economics, and diversified analytics exposure relative to peers such as Verisk.

• Legal at around 17-18x appears broadly in line with comparable legal information and workflow businesses.

• STM at roughly 13x likely reflects both its leadership position and ongoing uncertainty around academic publishing and AI disruption.

• Exhibitions at around 12x is broadly consistent with businesses such as Informa.

• Print deserves the lowest multiple given its structural decline.

I generally dislike peer comparisons because they assume a level of homogeneity that rarely exists. Differences in competitive positioning, switching costs, scale advantages, and capital intensity matter enormously. But for a business as diversified as RELX, a sum-of-the-parts framework at least provides another way to think about valuation.

Viewed through that lens, RELX does not appear obviously cheap, but it also does not appear excessively expensive for a business with improving economics and accelerating organic growth.

Relative to peers, RELX still commands a premium on EV/EBITDA and EV/Sales. That premium appears justified given the company’s superior profitability profile, with EBIT margins around 35% versus a peer median closer to 27%. Meanwhile, the free cash flow yield of roughly 6.4% is broadly in line with peers.

The market appears to be pricing RELX as though the recent acceleration in growth will fade and the business will revert toward its older, slower-growth profile, rather than its present reality as an embedded analytics and decision-tools platform..

What stands out more is the growing divergence between market pricing and sell-side expectations. Despite the weakness in the share price, analyst sentiment remains overwhelmingly positive. Sixteen analysts still maintain broadly constructive ratings. Morgan Stanley recently downgraded the stock to equal weight, but not even that was bearish; it reduced its price target from £33.20 to £29.70, which still remains above the current share price of £24.50.

In fact, consensus targets remain materially higher. The average analyst target of roughly £36.39 (down from £44 six months ago) still implies upside of around 47% from current levels.

At today’s valuation, expectations embedded in the stock are not particularly demanding. A forward P/E of roughly 17.5x for a business generating ~7% organic growth, 35% operating margins, and nearly 100% cash conversion would normally imply a premium multiple rather than a discounted one.

The interesting part is how sensitive long-term shareholder returns become if the current growth profile proves sustainable.

If RELX can maintain roughly 7% annual revenue growth over the next five years, pushing sales toward approximately £13.5 billion by FY2031, while margins expand modestly toward 36-37%, the earnings profile changes materially. Add in continued share repurchases reducing the share count by around 1.5% annually and the current valuation begins to look far less demanding.

Under those assumptions, even if the earnings multiple remains broadly unchanged around current levels, total shareholder returns could exceed 70% over five years (11-12% CAGR) with dividends included.

If the market eventually re-rates RELX closer to its longer-term historical multiple in the high 20s, the upside becomes significantly larger. In that scenario, total shareholder returns could exceed 150% over five years (21% CAGR).

Those are clearly optimistic outcomes, but they illustrate the asymmetry in the thesis. A business with durable recurring revenues, expanding margins, disciplined capital allocation, and embedded workflow products does not need extraordinary assumptions to generate attractive returns if the market eventually regains confidence in the durability of growth.

The bear case is the opposite. Revenue growth slows materially. Margins compress as AI competition intensifies and investment requirements increase. Free cash flow weakens, limiting the pace of buybacks and reducing balance sheet flexibility. In that environment, valuation multiples could compress further and shareholder returns may ultimately disappoint despite the current drawdown in the share price.

That is ultimately the core debate surrounding RELX.

If the recent acceleration in underlying growth reflects a cyclical peak driven by temporary AI enthusiasm and post-pandemic normalisation, then the market’s scepticism is justified.

But if RELX has genuinely improved its structural growth profile through workflow integration, analytics monetisation, and AI-enhanced decision tools, then the current valuation may still be underestimating the quality and durability of the business.

That is why the distinction between cyclical acceleration and structural improvement matters so much here.

So with that framework in place, it becomes easier to examine both the bear and bull cases in greater detail and understand the logic underpinning each side of the debate.

Bear Thesis

A sophisticated bear case against RELX is not that the company is low quality. It’s that the market is mistaking a cyclical growth acceleration and AI enthusiasm for a durable structural improvement.

The first concern is Risk. At a roughly £3.5 billion revenue run-rate, sustaining 8% growth requires adding close to £280 million of incremental annual revenue. The larger the segment becomes, the harder that task gets. Around 61% of Risk revenue is transactional, leaving the business exposed to slower insurance activity, weaker lending volumes, or softer financial markets. A cyclical downturn could therefore produce a sharper slowdown than investors currently expect. Competition is also intensifying in areas such as digital identity, fraud prevention, and compliance technology, where fintech challengers are often more agile than large incumbents.

The second concern is STM and the long-term impact of AI on academic publishing. Open access continues to gain traction, while AI research tools increasingly allow users to synthesise information without directly accessing journal content. If researchers can obtain reliable summaries through large language models trained on scientific literature, the perceived value of journal subscriptions could gradually weaken.

RELX’s own disclosures acknowledge risks around changing payment models and research integrity. Bears would argue that STM’s recent 5% growth represents a temporary stabilisation rather than a durable recovery.

The third concern is Legal. Growth accelerated to 8.5% in FY2025, but sceptics argue this may reflect a catch-up period after years of underinvestment rather than a permanently higher growth profile. Industry data already suggests legal technology spending growth may be slowing in 2026. If the legal market normalises, the longstanding duopoly between LexisNexis and Thomson Reuters may simply revert to slower mid-single-digit growth. The segment’s relatively modest 21% margin could also imply that competition is stronger and pricing power weaker than many bulls assume.

The fourth concern is margins. RELX has benefited from years of operating leverage, but AI may alter the economics of the industry. Large cloud providers and AI infrastructure companies possess deeper technical resources and substantially larger balance sheets. If RELX is forced to accelerate investment merely to defend its position, the historical pattern of steady margin expansion could reverse.

The fifth concern is cybersecurity and reputational risk. The March 2026 LexisNexis data breach exposed approximately 2GB of data, including more than 21,000 customer accounts and hundreds of thousands of user profiles, including government-related users. For a business built around trust, proprietary data, and professional workflows, reputational damage matters.

To add balance, the company stated that the compromised information largely involved legacy and deprecated data from before 2020, with no active passwords or social security numbers exposed. That distinction is important when assessing the true severity of the breach. Still, the incident may increase cybersecurity spending, invite regulatory scrutiny, and create some pressure on customer retention within Legal.

Finally, there’s valuation. A forward P/E of 17.5x is not obviously cheap if group growth ultimately slows toward 4-5%. The £2.25 billion buyback programme also assumes continued strong cash generation. If growth disappoints, buybacks may need to slow or leverage could drift above the company’s preferred range, weakening the investment case further.

The bear thesis is therefore relatively straightforward. Risk slows toward 5-6% growth. Legal normalises toward similar levels. STM drifts back toward 3-4%. Margins stop expanding or begin compressing. In that scenario, RELX starts to resemble a slower-growing information services company rather than a premium analytics compounder.

Share

Bull Thesis

The bull case is that the market is fundamentally misreading what RELX actually is.

RELX is not simply a publisher or software interface layer. It owns and curates proprietary professional datasets that are deeply embedded in regulated industries. In an AI-driven world, that data becomes more valuable, not less.

Large language models are only as useful as the quality of the information they can access. RELX controls significant volumes of structured legal, scientific, medical, and risk-related data that are difficult to replicate. More importantly, the company does not merely aggregate information. It organises, verifies, structures, and continuously refines it using domain expertise built over decades.

In scientific publishing, Elsevier manages the peer review process behind prestigious journals such as The Lancet. In legal services, LexisNexis controls extensive archives of case law, statutes, and commentary integrated directly into professional workflows.

That creates leverage within the AI ecosystem. AI providers either need licensed access to trusted professional datasets or risk operating with lower-quality information. RELX therefore remains in control of assets that AI systems increasingly depend upon.

Workflow integration strengthens the moat further. Lawyers use LexisNexis for legal drafting and research. Banks rely on RELX Risk products for fraud detection and compliance screening. Researchers depend on Elsevier’s databases and workflow tools. Once embedded, switching becomes expensive, disruptive, and operationally risky.

Customers are not paying purely for raw data. They are paying for trust, compliance, reliability, customer support, workflow integration, and validated outcomes. Even if AI lowers software development costs, building enterprise-grade alternatives for regulated industries remains difficult.

RELX is also embracing AI rather than resisting it. The company already spends heavily on technology and has integrated machine learning and natural language processing into products for years. Tools such as Lexis+ AI and Nexis+ AI build on that foundation using RELX’s proprietary closed datasets rather than generic internet-trained models.

This is important because running AI systems on proprietary workflow and usage data can create reinforcing advantages over time. As customers interact with the platform, RELX accumulates additional behavioural and workflow intelligence, strengthening switching costs further.

The bull case ultimately rests on the idea that growth has structurally improved rather than temporarily accelerated.

Risk has consistently delivered 7-9% underlying growth for years since the pandemic period. The drivers are secular: fraud prevention, anti-money laundering requirements, digital identity verification, and regulatory complexity. Those trends are unlikely to disappear in weaker economic conditions.

STM may also be adapting faster than sceptics assume. The business is shifting steadily toward analytics, workflow software, and AI-enabled research tools. Management commentary increasingly suggests generative AI functionality is being embedded across much of the revenue base rather than treated as a standalone product initiative.

Legal may represent the largest opportunity. Lexis+ AI appears to be gaining traction, while management has expanded workflow functionality materially since launch. The segment’s relatively low operating margin of around 21% also implies substantial operating leverage potential if adoption continues to scale.

Importantly, Legal does not need to dominate the entire legal AI ecosystem to create value. It simply needs to deepen workflow integration and monetise its existing customer base more effectively.

Margins also support the bull case. RELX has already expanded operating margins by more than 400 basis points over the past decade while simultaneously increasing technology investment.

Capital expenditure remains disciplined relative to revenue, suggesting the company is not entering an uncontrolled AI spending race.

Valuation ties the thesis together. If RELX can sustain roughly 7% organic growth while continuing to expand margins modestly, earnings growth could remain comfortably in the low-to-mid teens. Businesses with that combination of recurring revenue, pricing power, workflow integration, and cash generation typically trade at premium multiples rather than discounted ones.

The key question is therefore whether RELX is still being valued as a legacy publisher with cyclical growth, or whether it is actually evolving into a higher-quality analytics and workflow platform whose best years may still be ahead of it.

What to Watch For

1. The H1 2026 results, expected in July 2026, will likely be the most important near-term test for RELX. They should provide the first real indication of whether the stronger 2025 growth profile represents a structural improvement or a cyclical peak.

2. The most important metrics will be segment growth rates.

   Risk needs to sustain growth near 8%. A slowdown toward 5-6% would materially weaken the broader investment case because Risk is now the primary driver of both group growth and cash generation.

   Legal needs to hold in the 8-9% range rather than reverting toward historical mid-single-digit growth rates. If Lexis+ AI and workflow tools continue gaining traction, it would support the argument that Legal has entered a structurally stronger growth phase.

   STM is arguably the most strategically important swing factor. Growth moving above 5% toward 6-7% would suggest the transition toward analytics and AI-enabled workflow tools is working. A slowdown toward 3-4% would reinforce concerns around open access and AI commoditisation.

3. Margins matter just as much as growth. Any expansion above 35% would indicate that RELX continues to generate operating leverage despite rising AI and technology investment. A plateau or decline in margins, even modest, would suggest the economics of the business are becoming more competitive and investment-intensive.

4. Acquisitions also remain an important part of the story. RELX completed five acquisitions totalling roughly £270 million in 2025 and has remained active in 2026. Historically, management has integrated acquisitions well and used bolt-on deals to deepen datasets, expand workflows, and strengthen customer relationships. Done properly, acquisitive growth can reinforce both top line growth and competitive positioning.

5. The £2.25 billion buyback programme is another important signal. Completion of the programme while maintaining leverage within the company’s preferred 2.0-2.5x range would confirm confidence in the durability of cash flows. Any reduction, delay, or material increase in leverage would likely signal that management sees a softer outlook developing.

6. Competitive dynamics should also be monitored closely. Product developments from Thomson Reuters, adoption trends among AI-native legal software firms, and the pace of open-access adoption in STM all have implications for RELX’s moat and long-term pricing power.

7. The March 2026 LexisNexis data breach also remains worth monitoring. Even if the exposed information largely involved legacy data, any evidence of customer churn, reputational damage, or regulatory escalation would matter given how central trust and data integrity are to the investment case.

8. Foreign exchange exposure is another variable investors should not ignore. RELX reports in pounds sterling but operates globally, with significant revenue exposure to North America and operations spanning more than 180 countries. Currency fluctuations can therefore materially affect reported numbers even when underlying operating performance remains strong.

9. Regulatory developments may also become relevant. Proposed FCC rules targeting offshore call centres could create some operational pressure within parts of Risk, although the ultimate impact remains uncertain.

Share

Summing It All Up: Is RELX A Good Investment?

The central tension in RELX is the gap between what the business has achieved over the past several years and what the market believes is sustainable.

Ultimately, the investment case rests heavily on the Risk segment. If Risk can continue growing near 8%, the broader economic model likely remains intact. The cash generation from that division gives management room to continue investing in Legal and STM while maintaining shareholder returns. If the Risk segment slows materially, the entire thesis weakens.

The broader theme underlying RELX is also important.

The company increasingly monetises decisions rather than information. Businesses that help professionals make better decisions faster tend to develop stronger pricing power, higher switching costs, and deeper workflow integration. Businesses that merely distribute information are becoming commoditised.

RELX appears to understand that distinction and has spent years repositioning itself accordingly.

The market, however, still seems uncertain whether to value RELX as a legacy publisher facing disruption or as a higher-quality analytics and workflow platform with durable competitive advantages in the AI era.

The next two to four quarters will likely determine which interpretation proves correct.

Leave a comment

Value the Essence of the Trade, Not its Packaging

I stood on a stage before an auditorium of Wall Street’s finest. You know the type. Sharp suits, big egos and salaries that are difficult to justify. These are the players in the modern financial system; individuals who maneuver billions of dollars, yet rarely experience personal loss because they’re trading with other people’s money.

I opened with a simple question: “What is it that you actually trade?”

The room answered with confidence. Equities. Fixed income. Structured products. Asset backed securities. All technically correct. All fundamentally wrong.

They were naming the vehicles, but missing the engine entirely. Equities, bonds and swaps are merely the shells that house a more fundamental element. They were describing the topology of the trade, not its essence.

The only thing being traded by anyone in that room, or in any market across the globe, is ‘risk’. Always has been.

Risk is the DNA of any financial instruments. Strip away the product names and the Bloomberg terminals, and risk is all that remains.

Every transaction reduces to the same exchange. You are paying a price today to assume uncertainty about tomorrow, with the expectation that you are being compensated for it.

When you buy a stock, you are not really acquiring a slice of a business in the way people like to imagine. You are underwriting the ‘risk’ that future cash flows, discounted appropriately, justify the price you paid.

The same logic applies to bonds. When you trade debt, you’re balancing credit ‘risk’ against duration ‘risk’.

Different label, same underlying capital market trade. A business needs money and those providing it are required to accept risk; the only question is how that risk is packaged.

While everyone in the room understood risk, their perception of what they were trading revealed the markets biggest blind spot. Focusing on the wrapper rather than the element contained within explains why most investors misprice assets.

At this point, most investors will instinctively push back. So let me ask you this: for pricing purposes, do you use a Discounted Cash Flow model, a Capital Asset Pricing Model, simple ratios and multiples, or peer comparisons?

While these approaches are neat and tidy, the thinking isn’t.

The tools feel rigorous. The process feels disciplined. These frameworks produce clean numbers. That’s part of the appeal. But the clarity is deceptive because the treatment of risk inside these models is inconsistent.

Across each of these approaches, risk is either hidden, misclassified, or ignored entirely. I’ll break down how this happens in each case shortly.

For now, focus on the inconsistency. The entire exercise is about pricing risk. We’ve established that is the core trade.

So what does it mean if the framework you rely on doesn’t properly incorporate risk?

At best, the output is incomplete. At worst, it’s misleading.

What is Risk?

Risk means more things can happen than will happen1. That single sentence contains more practical wisdom than most investment textbooks. The future is not a fixed path. It is a distribution of outcomes.

Those who perform well over time don’t need to know what will happen next. They need to know how outcomes are skewed. A casino doesn’t predict individual spins. It prices the game so the expected value sits in its favour. That’s the model. It’s not prediction, it’s positioning.

As Howard Marks puts it, “Never confuse possibility with probability“, then ask, “What is a safe price to pay to participate?”

That framing changes everything.

Markets don’t price risk. They price the perception of risk. The gap between perceived risk and actual risk, that’s the trade. When they diverge, you get mispricing. That is the only place an edge can exist.

Most participants don’t operate there. They chase narrative. They anchor to price. They react to momentum. They feel analytical, but they avoid the hard part, which is interrogating downside risk properly. Narrative without risk-framing is just speculation dressed in analysis.

These are cognitive shortcuts, and the market is ruthlessly efficient at extracting money from people who rely on them. Most assume that because the logic of their thesis is sound, the trade must be too.

The correct question is not “will this go up?” It’s “what happens if I’m wrong?”

That is where asymmetry lives. If the upside is meaningful and the downside is contained, the trade works even if you are wrong more often than you are right.

Discomfort is not risk. Confidence is not safety. Those distinctions are where most capital gets lost.

When you trade with a salary and a bonus structure, trading becomes process. Something one does to earn a living. When trading with someone else’s capital, the personal stakes are low, so too is the intellectual rigour. In contrast, when you trade your own capital, risk becomes visceral. It’s a tightening in the chest. An intellectual exercise that consumes your thinking 24/7. It shapes behaviour. It forces discipline.

Once you internalise that you are trading risk, the rest of finance starts to look different. Including valuation.

Markets are not grading your reasoning. They are aggregating the risk assessments of every participant simultaneously, and the price reflects that collective judgement at any given moment. The edge, if it exists at all, lies in identifying where that collective judgement is systematically wrong. Where fear has overpriced danger, or where optimism has buried it.

That’s the only game worth playing.

The Evolution of Packaged Risk

Valuation began with a simple exchange: something today for something uncertain tomorrow.

Over five thousands years ago, in Mesopotamia, loans were made in grain. A farmer might borrow grain at planting season and agree to repay a larger amount after harvest. That extra amount was essentially a price of the loan.

It was compensation for the lender taking risk, waiting for repayment, and giving up present use of resources in exchange for future return. Higher uncertainty required higher return.So, even at the dawn of civilisation, the relationship between time, risk, and value was already understood. Thes ideas became the foundation of modern asset pricing theory.

The Code of Hammurabi, circa 1754 BC, imposed legal price limits on lending. Grain loans carried higher rates than silver loans. The reasoning was obvious. Grain could spoil. Harvests could fail. Silver was stable. Different risk profiles, different required returns.

That principle never changes.

What did change was the system built around it. Several dominant religions deemed demands for interest to be immoral, sinful. In 325 AD, the Council of Nicaea prohibited commercial lending, although attitudes have since changed. Yet Islamic Sharia law still prohibits charging interest even today.

The tension between theological ideology and economics is that progress requires credit. Farmers need to invest in seed, merchants need working capital, and governments need money. The economy could not function without lending, so finance did what it always does when blocked by rules: it adapted.

During the Renaissance, Italian merchants developed increasingly sophisticated contractual workarounds. The Medici became one of history’s great banking dynasties by mastering these methods. Through partnership structures, trade finance arrangements, and repurchase-style agreements, they found ways to price capital while respecting the letter, if not always the spirit, of canon law.

Finance was becoming more systematic, turning into something measurable and transferable. Compound interest was increasingly understood and Fibonacci introduced methods that made these calculations practical.

Debt now existed with varying maturities, giving rise to the beginnings of a yield curve.

But something more profound happened in the early 1600s. The Dutch East India Company (VOC - Vereenigde Oostindische Compagnie) needed to raise capital and issued the first widely traded bonds to the general public. More significantly, the VOC is also widely considered the first company to issue publicly traded shares, effectively creating the first modern initial public offering (IPO). Now investors could choose between different terms for the provision of capital; fixed contractual claims (debt) or residual equitable claims (shares). It became a trade off between certainty and upside potential; risk had simply been repackaged.

This explains why for much of the eighteenth and nineteenth centuries, valuation centred on dividend yield and asset backing. It reflected market reality. Stocks were often viewed as uncertain versions of fixed-income securities rather than long-term compounding machines, so investors expected higher income in return.

Once again, the the link between price and risk is in evidence.

Over time, this view evolved. Irving Fisher reframed valuation around two variables: the stream of future payments and the rate used to discount them. This brought equities closer to a unified theory of asset pricing.

The 1930s marked a turning point. John Burr Williams formalised the dividend discount model. The logic was clean. A stock is worth the present value of the cash it will return to shareholders. This explains why dividends were considered sacrosanct in the early 20th Century (less so today with many of the best companies not paying dividends at all).

This thinking still underpins value investing today, but it subordinated risk from being a direct driver of price to being an indirect consideration within future cash flow forecasting. That is problematic.

Everything changed, yet nothing really changed. Risk is still the esssence of every trade.

Valuing Risk Today

What is the future worth today, given time and risk?

Modern finance attempted to formalise that question. In some areas it succeeded. In others it created the illusion of precision.

Discounted cash flow2 is a good example.

Railroad companies were among the first to compare the upfront cost of projects with the present value of future cash benefits.

If future discounted returns exceeded cost, the investment made sense.

Businesses also began considering residual value: the worth of an asset at the end of its useful life.

At this stage, such techniques were designed to be applied to projects rather than for valuing companies.

At a project level, it works. You know the cost of a machine. You can estimate its cash flows. The asset has a defined life and a terminal value. This can all be modelled.

But at the corporate level, it breaks down.

Most DCFs derive the majority of their value from terminal assumptions. Run a standard model and you will often find that well over half the valuation comes from cash flows beyond the explicit forecast period. Sometimes close to eighty percent.

That means the model is not really about the next five years. It is about everything that comes after, which is precisely the part you cannot forecast with any reliability.

Small changes in assumptions drive large changes in output. Growth moves slightly. Discount rates shift. Valuation swings materially. The model looks rigorous. The sensitivity tells you otherwise.

The discount rate itself is another weak point. Cost of debt is observable. Cost of equity is inferred. It rests on assumptions about risk that are often circular.

Then there’s the elephant in the room that no-one speaks about. Businesses don’t have fixed lives. They adapt, decline, reinvest, or disappear. Treating them like depreciating assets introduces a mismatch between model and reality.

DCF remains useful as a framework. It forces you to think about cash generation and capital allocation. But as a tool for precise valuation, it overstates what can be known.

That is why the more useful application is often inversion: a reverse DCF.

Start with the market price. Work backwards. Ask what assumptions must be true to justify it.

This turns valuation from prediction into interpretation. You are no longer pretending to know the future. You are assessing whether the market’s implied view of the future is reasonable.

If a mature business needs to compound at high double-digit rates for a decade to justify its price, the conclusion is straightforward. Not because your model is precise, but because the assumption is improbable.

Reverse DCF doesn’t remove uncertainty. It relocates it. That is an improvement.

Other models struggle with the same problem from a different angle.

The efficient market hypothesis argues that all available information is already reflected in an asset’s price. What began as an academic framework in economics eventually spilled into corporate finance, shaping models such as the Capital Asset Pricing Model (CAPM).

CAPM attempts to define risk using volatility (beta) as a proxy. This is convenient. It is also wrong. Historical price volatility is not the same thing as future investment risk.

A stock that has fallen significantly in value can become less risky if the excess valuation is removed. Beta will often increase as risk falls. The CAPM model confuses behaviour with outcome.

That misclassification matters because it feeds into the discount rate, which feeds into valuation.

If your definition of risk is flawed, your valuation will be too.

So, while a DCF marginalizes risk, CAPM entirely misclassifies it.

Multiple Shortcuts

Markets love shortcuts, and valuation is no exception. Multiples such as price-to-earnings, EV/EBITDA, price-to-sales, and price-to-book became popular because they offer something every investor wants: speed.

Rather than building a full forecast, the idea of being able to compress a business into a single ratio and compare it instantly against peers looks appealing.

That convenience explains their popularity, but it also creates false confidence. There is no consideration of risk built into multiples. Multiples are not valuation, they’re pricing. And that issue isn’t just semantic, it is foundational.

A 2x price-to-sales multiple can mean very different things depending on the underlying economics. For a business earning 20% net margins, it implies a 10% earnings yield. For one earning 5%, it implies just 2.5%. The multiple is the same, but the reality is not.

Companies within the same sector can have very different unit economics, and those differences drive valuation. A multiple observed elsewhere in the industry is not automatically transferable. This is why peer comparisons often mislead.

A multiple is a compressed expression of assumptions. Growth, margins, capital intensity, durability. All embedded. None explicit.

Peter Lynch tried to bring growth into the framework with the PEG ratio, linking the P/E multiple to expected growth. A company capitalized at 15x earnings and growing at 15% would sit at a PEG of one, which he deemed acceptable.

The issue is that PEG treats all growth as equal. An 8% grower at 8x and a 40% grower at 40x both produce a PEG of one, yet the underlying risk is very different. High growth is harder to sustain, especially as scale increases and the law of large numbers starts to constrain expansion.

Neither raw multiples, nor PEG, includes the essence of valuation: risk.

Multiples may be useful as a starting point. They are dangerous as a conclusion.

Used carefully, they may be useful screening tools. Used lazily, they are nothing more than valuation theatre.

Conclusion

All of this leads back to the same issue.

Risk is either hidden, misclassified, or ignored entirely.

DCF pushes it into assumptions about the future. CAPM reduces it to volatility. Multiples bypass it altogether.

Yet risk is the only thing that matters.

So what does a practical framework look like?

It starts with abandoning false precision.

Financial statements tell you what happened. They don’t tell you why, or whether it will persist. That requires judgement.

Management quality matters. Incentives matter. Capital allocation matters. These do not fit neatly into models, but they drive outcomes.

Different assets require different tools. Predictable cash flows lend themselves to structured models. Equities do not. They are adaptive systems exposed to competition, regulation, macro conditions, and human decision-making.

Forecasting decades into the future with confidence is not realistic.

A more grounded approach is to focus on drivers.

Then ask a simple set of questions. What drove returns historically? Can those drivers persist? Are they already reflected in the price?

If margin improvements led to multiple expansion and in combination they drove most of a company’s past returns, it is not reasonable to assume those forces will persist at the same rate. Both are subject to limits. When both margins and multiples are stretched, the drivers of future returns narrow. The burden shifts to revenue growth, which is typically slower and more constrained. If growth disappoints, multiple contraction becomes a real possibility, and even stable operating performance can translate into weak or negative shareholder returns.

That’s a different risk profile entirely, not captured where a share is priced to perfection based purely on past performance.

Everything comes back to the same point.

Markets aggregate views. Prices reflect collective judgement. Your job is not to out-model the market. It is to identify where that collective judgement is wrong.

The numbers matter, but numbers alone are never enough. Financial statements can tell you what happened. They rarely tell you why it happened, or whether it can happen again. That is where judgement enters the process.

Valuation is more art than science.

Is risk overstated? Or is it understated?

That is the only edge available, and it starts with being clear about what you are actually trading. It’s Risk!

Further reading on the topic of Risk:

If you enjoyed this, you’ll almost certainly enjoy these:

Share

Leave a comment

It has been argued that the primary driver of sky high tech-sector valuations has been a huge reflexivity loop which may be about to break, as revealed in the hyperscaler Q1 earnings this week.

First, let’s understand how the reflexivity loop works, then we’ll explore why it may be breaking.

Companies repurchase their own shares, primarily to offset dilution from stock based compensation, regardless of whether the share price is cheap or expensive. Blind buying at any price creates artificial demand for the stock -▷ this elevates the share price -▷ which leads to FOMO -▷ which creates momentum -▷ then comes more blind buying. And so the cycle repeats, year after year.

Deutsche Bank Global Research called out the issue in 2019, prompting me to write about it back then (below is an excerpt from that post):

Think about the reflexivity loop in terms of a chain reaction which drives equity valuaitons higher.

Now, why might this process be about to go into reverse?

This week the hyperscalers published their Q1 results for 2026.

Below is the cash flow statement from Meta.

Note that Meta repurchased no shares in Q1 2026, versus $13 billion in the corresponding period a year earlier.

Let’s take a look at Alphabet. Stock repurchases ran to over $15 billion in Q1 2025. This year? Nothing.

The same thing was seen in the numbers for the other hyperscalers.

So the key link in the chain reaction described above seems to have broken.

The story doesn’t end there.

Now lets look at the issuance of debt.

Last year for Alphabet in Q1 2025, issuance of debt and repayment of debt almost perfectly balanced at $4.5 billion. In other words, they were simply rolling their debt obligations. But this year, we see a net issuance of $30 billion!

This explains why buybacks are stopping. There is no surplus capital. More particularly, if these firms are levering up heavily to finance AI expansion, repurchases may be stunted for a long time to come.

The cash is all going on CAPEX spending which has exploded.

Although revenues and operating cash flows are climbing, they lag CAPEX spending which compresses Free Cash Flow. Just look at this statement from Amazon’s recent 10-Q:

What does this do to FCF multiples?

Maybe multiples explode higher… Or maybe they remain the same, but this will require market caps to collapse.

Let’s turn our attention to margins.

Costs are rising. Margins will fall.

History is clear. In 1929, average US stock market margins peaked at 8.9 percent. They collapsed in the Great Depression and didn’t recover that prior peak until 2007, when margins touched 9.4 percent alongside a 22 times multiple. That cycle ended in the Global Financial Crisis and another sharp contraction.

Today, margins are hovering around an eye-watering 13 percent while the market trades near 24 times earnings. Part of the lift came from the 2017 Tax Cuts and Jobs Act (TCJA), which slashed corporation tax from 35 to 21 percent. Most came from index concentration. A handful of mega-cap hyperscalers now drive close to 40 percent of a 500-stock index, and their capital-light models have inflated aggregate margins. That tailwind is fading. Costs are rising. Capital light is turning capital intensive. When margins reach the summit, there is only one way forward; down.

Margins down means earnings down. Earnings set valuations. Lower earnings drag returns and pressure prices. Multiples compress alongside fundamentals. These forces multiply. Re-rating risk is high.

What happens if the market becomes a net seller of the hyperscalers in the short term when there is no artificial demand from buy-backs to prop up the market?

Oh, and one more thing. If share repurchases stop, but stock based compensation continues, there’s a tidal wave of dilution coming over the horizon. What does that do for FCF per share with a rapidly falling numerator and an expanding denominator?

Could this be the beginning of the price correction that many have been predicting for a while?

Have valuations become too stretched?

I don’t have the answers, but I like to think that I ask the right questions.

What do you think?

Leave a comment

Share

DISCLAIMER & DISCLOSURE: The author holds no investment in ServiceNow at the date of publication but that may change. The views expressed are those of the author and may change without notice. The author has no duty or obligation to update this information. Some content is sourced from third parties believed to be reliable, but accuracy is not guaranteed. Forward-looking statements involve assumptions, risks, and uncertainties, meaning actual outcomes may differ from those envisaged in this analysis. Past performance is not indicative of future results. All investments carry risk, including financial loss. This analysis is for educational purposes only and does not constitute investment advice or recommendations of any kind. Conduct your own research and seek professional advice before investing.

Why Should I Be Interested in ServiceNow?

ServiceNow Inc. (NYSE: NOW) has established itself as a primary architectural pillar for global enterprise digital transformation, evolving from its historical roots in IT Service Management to become what management characterizes as the “AI control tower” for business reinvention.

Today ServiceNow is one of the largest enterprise software companies in the world, helping consolidate workflows across different departments into a single proprietary cloud-based platform, the Now Platform, that allows organizations to orchestrate and automate complex workflows across disparate departmental silos.

This “platform-of-platforms” approach acts as a unifying layer over legacy systems of record such as ERP, CRM, and HCM systems, enabling ServiceNow to maintain a durable, mission-critical presence in the world’s largest organizations.

Brief History

Fred Luddy, former CTO of Peregrine Systems, founded ServiceNow (then Glidesoft) in 2003, later incorporating it in 2004. It began as a one-person operation focusing on providing IT services via a SaaS model.

Renamed in 2006, as Service-Now.com, it grew rapidly through SaaS ITIL-based solutions. Sequoia Capital invested heavily, with Series D funding in 2009 helping accelerate expansion.

Frank Slootman (who later went on to run Snowflake) was named CEO in 2011, taking over from founder Fred Luddy, and shifting the focus to development.

ServiceNow went public in June 2012 (NYSE: NOW) with an IPO that raised $210 million, valuing the company at over $2 billion.

Under subsequent leaders; John Donahoe (starting 2017) and later Bill McDermott (2019), the company expanded from ITSM into HR Service Delivery, Security Operations, and Customer Service Management, moving toward AI-powered workflow automation.

Revenue increased from $13 million in 2007 to $86 million in 2010. Fast forward to today and the company is at a run rate of closer to $13 billion in revenue.

As of 2026, ServiceNow serves over 8,100 enterprise customers, including 85% of the Fortune 500 in addition to many smaller enterprises, meaning little in the way of customer concentration risk.

The Visionary Leader

CEO Bill McDermott is currently in the driving seat. He is an impressive character whose history is defined by a consistent pattern of entrepreneurial ambition.

McDermott grew up in a working-class family on Long Island. He demonstrated his business acumen at the age of 16 when he bought a local deli, the Amityville Country Delicatessen, for $7,000, funded using savings accumulated from working three part time jobs. Running the deli provided McDermott with his foundational training in business and customer service, and the profits from the venture helped pay for his undergraduate education at Dowling College.

After earning an MBA from Northwestern University’s Kellogg School of Management, McDermott spent 17 years at Xerox, famously talking his way into a door-to-door copier sales job and rising to become the company’s youngest-ever corporate officer. His leadership legacy was later cemented during a 17-year tenure at German software giant SAP (now Europe’s most valuable company), which saw him rise through the ranks to the top job, where he oversaw a market value increase from approximately $39 billion to over $160 billion.

Since joining ServiceNow in 2019, he has presided over a period where the company more than tripled its annual revenues.

Share

The ServiceNow Value Proposition

ServiceNow’s value proposition here is that it acts as a single workflow platform that connects IT, HR, finance, security, and other functions so work does not get trapped in separate departmental silos. In practice, that means fewer handoffs, less duplicate work, and more automated, end-to-end processes across the enterprise.

A concise way to put it is that ServiceNow helps organizations replace fragmented point solutions and disconnected databases with one shared platform and data model, so teams can see the same work, use the same workflows, and coordinate more easily.

ServiceNow’s economic engine is fundamentally driven by a subscription-based software-as-a-service (SaaS) model, which accounts for almost all of its revenue. This model provides exceptional financial visibility, bolstered by a historical gross renewal rate that has consistently tracked at 97% to 98%.

The company’s core products are organized into four primary workflow families:

1. Technology

2. Employee

3. Customer

4. Creator

Each of these has now surpassed $1 billion in Annual Contract Value (ACV).

Technology workflows automate the lifecycle of IT services and infrastructure, providing the “system of record” for the entire digital estate. Customer workflows connect front-office engagement to back-office operational teams to resolve issues end-to-end. Employee workflows streamline internal requests such as onboarding and payroll through a unified AI-driven portal, while Creator workflows offer low-code tools that allow business units to build their own custom applications.

The company’s competitive advantage is anchored by its proprietary Configuration Management Database (CMDB), which acts as a single, unified data model mapping every piece of hardware and software within an enterprise. This single architecture eliminates the integration complexities associated with fragmented point solutions and creates significant data gravity.

As more data flows into the CMDB, the platform’s AI and automation become increasingly accurate, creating a barrier that point solutions cannot easily replicate.

High switching costs further protect this moat, as migrating to a competitor would require untangling thousands of integrations and retraining an entire workforce. This dynamic is evidenced by the fact that 90% of customers utilize multiple workflows on the platform, and it typically takes roughly four years for a customer to grow from $1 million to $5 million in ACV.

Additionally, ServiceNow leverages an ecosystem advantage by shifting approximately 35% of its implementation work to elite partners like Deloitte and Accenture, who build dedicated practices around the platform.

Strategic Shift

A major strategic shift occurred in late 2025 and early 2026 as ServiceNow moved away from its historical “build-over-buy” philosophy to spend approximately $11 billion on three major acquisitions: Armis, Veza, and Moveworks.

• Armis was acquired for $7.75 billion, a premium valuation of 23x ARR, to close a critical visibility gap regarding unmanaged devices such as factory robots and medical equipment.

• Veza was acquired for approximately $1 billion to address identity governance for AI agents, which now outnumber human identities by roughly 82 to 1.

• Moveworks, acquired for $2.85 billion, serves as the conversational AI front-end for the platform.

When integrated, these acquisitions create a “See-Decide-Act” loop where Armis discovers a vulnerable device, a ServiceNow workflow creates a remediation ticket, and the system automatically isolates or fixes the issue.

These moves have effectively tripled ServiceNow’s security Total Addressable Market (TAM), moving it into direct competition with dedicated cybersecurity vendors.

The timing of this strategic shift may be perfect. Anthropic’s new model, Claude Mythos, has not been generally released because Anthropic says it is unusually strong at finding and exploiting high-severity software vulnerabilities, especially in operating systems and web browsers. But this and more powerful AI models are coming and so improving cyber-security will be a key objective for every enterprise in the years ahead. ServiceNow appears to have positioned itself in the right place at the right time.

Armis is considered a leader among Cyber physical-systems security (CPS) protection platforms, as demonstrated by this Gartner magic quadrant chart.

Why isn’t Crowdstrike on this chart? Because it doesn’t offer CPS. If you need to secure a, mostly, standard IT environment, CrowdStrike is a top choice. If you have a complex network with a large amount of IoT, OT, or unmanaged devices, Armis is superior, particularly when paired with CrowdStrike for protection. The two services are complimentary and it is for this reason that the two companies are collaborating.

In short, ServiceNow faces a complex competitive landscape where it is both a leader in some areas and, following the acquisitions, a challenger in others.

Share

Avoiding the SaaS Apocalypse by Leveraging AI

Many argue that AI agents will displace enterprise workflow software incumbents.

That view underestimates why these systems exist. Enterprise workflows are built to provide audit trails, enforce compliance, and ensure predictable process execution. These are not optional features. They are requirements. Agentic AI, by design, does not reliably deliver them.

First, it is important to distinguish between a System of Record and a System of Engagement. These are often confused. A company like SAP specializes in deep, data-centric transaction handling (system of record), whereas ServiceNow shines at automating cross-functional processes and enhancing user experience (system of engagement). While SAP runs core business operations (ERP). ServiceNow orchestrating workflows.

Ultimately, AI’s value is not replacing the likes of SAP, ServiceNow, or Salesforce, but may instead be making them programmable, usable, and far more accessible. The winners will be the platforms that first tap into transformation budgets by clearly reducing implementation risk, cost, and timelines, then expand into everyday operations as the trusted control layer for work. Over time, they will peel back the legacy UI and replace it with composable workflows, governed automations, AI-assisted actions, and lightweight apps.

Said differently, ServiceNow is the necessary system of engagement, the rails, on which AI agents can run. In this respect, ServiceNow is likely to benefit from the adoption of AI solutions rather than becoming a victim of AI (as many suggest).

Importantly, ServiceNow is leaning into the AI revolution. The current phase of ServiceNow’s product strategy focuses on the monetization of Generative AI through the “Now Assist” product line. By integrating AI natively into workflows rather than treating it as a “sidecar” application, the company is attempting to pivot from seat-based pricing to value-driven and consumption-based models.

This transition is critical because AI agents that resolve 90% of requests autonomously could potentially lead to “seat compression,” reducing the number of human licenses required by customers. To offset this, ServiceNow sells GenAI-enabled “Pro Plus” and “Enterprise Plus” SKUs at a significant price premium.

Management has raised its 2026 AI-specific ACV commitment to $1.5 billion, noting that 50% of net new business is now sourced from non-seat-based models like tokens and connectors.

Financials and the Future

The financial performance has always been strong with sequential top line growth (25.3% CAGR over the past decade) translating into growing gross profits at almost an 80% margin, and operating income which is a less impressive 13%.

Financially, ServiceNow delivered a “beat and raise” performance in Q1 2026, reporting total revenues of $3.77 billion, which represents a 22% increase year-over-year.

Subscription revenue growth was impressive after accounting for a 75-basis-point headwind caused by delayed deals in the Middle East due to regional conflict1.

Following these results, management raised the full-year 2026 subscription revenue guidance to a range between $15.735 billion and $15.775 billion.

Despite these strong metrics, the stock experienced a 17.7% drop following the earnings report, driven by analyst concerns regarding margin pressure from the Armis integration and uncertainty about the quality of organic versus acquired growth. The Armis integration alone contributed to a 75-basis-point headwind on operating margins and a 25-basis-point headwind on subscription gross margins. These headwinds are anticipated to continue for several years due to ongoing integration costs before synergies and cross-selling benefits feed through.

Share

Competition

In the IT Service Management market, it is the dominant player with a 44.4% market share in ITSM according to Gartner, larger than its nearest competitors combined. However, it faces “flanking” maneuvers from Atlassian’s Jira Service Management, which is gaining traction in the mid-market due to faster setup times. In the customer service space, it competes with Salesforce, arguing that service requests are only meaningful when connected to underlying IT infrastructure.

ServiceNow also faces competition from Workday in employee experience and potential encroachment from hyperscalers like Microsoft and Amazon as they move up the stack into automation and workflow orchestration.

Microsoft has Power Platform, Copilot, and Viva (employee experience) which it bundles aggressively. Given Microsoft’s enterprise reach and AI integration, the threat should not be taken lightly.

Is ServiceNow a Good Investement? Strong Story, Tough Valuation

The long-term investment thesis for ServiceNow still rests on one core idea: its position as a mission-critical AI orchestration layer for global enterprises. If AI adoption across large organisations continues accelerating, ServiceNow is well placed to benefit as a workflow and automation backbone.

Financially, the balance sheet remains strong, although it has $2.5 billion in convertible senior notes falling due between 2026 and 2029). Against this backdrop, the company holds $5.2 billion in cash and short-term investments, giving it flexibility to fund growth initiatives while supporting an aggressive $5 billion share repurchase programme, including a $2 billion accelerated buyback intended largely to offset dilution (more on this shortly).

There are genuine risks. Geopolitical instability could weigh on enterprise spending, while AI-led productivity gains may eventually reduce seat-based software demand. Even so, ServiceNow’s 97% subscription revenue, 98% renewal rate and $27.7 billion RPO backlog provide a meaningful buffer and strong visibility into future revenue.

CEO McDermott has recently claimed that ServiceNow not only achieves the magic SaaS ‘Rule of 40’, but that it is the only software company achieving 55. He assumes ~21% topline growth and ~34% operating margins, but there are two issues here:

1. Top line organic growth has been in steady decline from over 30% in 2021 to just over 20% in 2025. Acquired growth will help prop this up in 2026, but at a cost to margins.

2. ServiceNow’s non-GAAP operating margins may be ~34%, but real operating margins are closer to ~13%. When a company spends over $10bn in buy backs over 5 years and the share count doesn’t reduce (it actually increases) then that’s the real cash cost of stock based compensation. It is the mechanism through which dilution is absorbed. Ignore the number quoted by the company in cash flow from operations, that’s non-cash fiction. Treating stock-based compensation as a harmless non-cash add-back in cash flow from operations obscures the underlying economics. If investors instead recognise the true cash outlay required to offset dilution, both free cash flow and operating income look meaningfully weaker. On that basis, underlying margins appear closer to the low-teens than the mid-30s being promoted. A significant portion of investor capital has been diverted away from reinvestment and shareholder returns, instead being used to enrich insiders. This matters.

   (Authors note: If a company burned $10 billion on acquisitions that delivered no tangible benefit to shareholders or the business, investors would be calling for heads to roll. Yet burn the same $10 billion on repurchases that achieve nothing, and scarcely anyone seems to notice. Go figure!)

So, does ServiceNow really achieve the Rule of 40, or even 55 as it claims? Or is it closer to 33 and falling?

The upcoming Financial Analyst Day on 4 May 2026 should be important. Investors will be looking for clearer evidence around monetisation of Now Assist, integration progress on recent acquisitions, and whether management can sustain or reaccelerate growth.

One of the more interesting recent developments came in February 2026, when CEO William McDermott purchased $3 million of stock personally. In total, five senior executives cancelled their 10b5-1 selling plans: William McDermott, Gina Mastantuono, Nicholas Tzitzon, Jacqueline Canney, and Russell Elmer (The CEO, President & CFO, Vice Chairman, Chief People & AI Enablement Officer, and Special Counsel.)

That matters because insider buying is relatively uncommon in SaaS companies. Why? Senior executives are typically compensated heavily in stock-based awards, which often function as deferred salary and are sold once vested. To avoid allegations of insider trading, executives usually rely on pre-arranged 10b5-1 plans that set out instructions for brokers to automatically sell shares, regularly, over time.

If insiders want to buy shares in the open market, they are required to cancel those plans and wait six-months before purchasing, or else they are liable to forfeit any profits that they make between the last price at which they sold, and the lower price at which they repurchased the shares. So when multiple executives halt 10b5-1 planned automatic sales and the CEO starts to buy personally, it naturally attracts attention.

That said, insider buying should not be interpreted blindly. Management incentives are not always aligned with outside shareholders. Executives still hold large unvested equity packages, which will have fallen sharply in value alongside the stock price. They cancel the 10b5-1 plan because they no longer wish to mechanically sell at what they perceive to be depressed prices. They may start buying ‘small quantities’ relative to their far larger holding and net worth, to signal confidence in an attempt to support the share price and prevent it from sliding further. This is rational behaviour, it doesn’t necessarily mean that the stock is a screaming buy at today’s prices.

So while insider activity is notable, it is not conclusive.

Even after a sharp sell-off from around a peak $211 (adjusted for the Dec 2025, 5-for-1 share split) to roughly $90, valuation remains difficult to justify. A market capitalisation near $93 billion against earnings of $1.7 billion implies a multiple of roughly 54x earnings. This is still a very expensive stock by almost any conventional measure; consider other SaaS peers, Microsoft is capitalized at 26x, Salesforce at 23x and Adobe at 14x (while these are not all in direct competition with ServiceNow, on an opportunity cost basis for investors, they remain relevant).

At prior highs, investors were paying well above 100x earnings; what were they thinking? The draw-down may not be due to the AI eats SaaS narrative at all; it may instead be a long overdue price correction.

Many investors may still be anchored to those previous prices and now view the stock as “cheap” simply because it has fallen. That can create persistent volatility, especially when sentiment rather than fundamentals drives ownership and the mindset is more speculator than intelligent investor.

From a capital efficiency perspective, ServiceNow generates around 15% return on equity, with around half of that produced by leverage. Against that backdrop, a premium valuation of around 8x book value looks demanding. An earnings yield below 2% requires sustained high growth to justify it, well beyond the 20% top line growth that the company has achieved in recent years.

How does a software company with gross margins near 80% struggle to produce operating margins above 13%? While some OPEX reflects growth investment, investors ought to reasonably view this as evidence of weak cost discipline and/or excessive employee compensation.

Many bulls cite adjusted free cash flow of $4.6 billion (FCF margin 44%) for FY25. But if you treat buybacks used to offset dilution as a real economic cost, which it is, true owner earnings are closer to $1.9 billion, equating to a properly adjusted 14.5% FCF margin. It’s respectable, but does it justify the company being capitalized at $93 billion (49x true FCF)?

Then there’s the M&A strategy and the Armis acquisition to consider. An $11 billion spending spree is aggressive (although preferable to burning $10.4 billion on buybacks that don’t reduce the share count, as mentioned above).

Strategically, the Armis deal appears eminently sensible. Cybersecurity and asset intelligence fit naturally into ServiceNow’s broader enterprise workflow ecosystem. But paying $7.75 billion for a business with just over $330 million of ARR looks expensive. Even if strategically sound, the amortization of this investment is likely to take years to justify, even if there are no issues encountered during the integration of the two businesses, which can never be assured.

ServiceNow is clearly a high-quality business with strong products, excellent management credibility, sticky customers, and meaningful AI optionality.

But quality does not automatically equal value.

The debate today is not whether ServiceNow is a good company; it clearly is. The real question is whether investors are still paying too high a price for that quality, particularly while stock-based compensation, dilution, and unanswered questions around the execution of its new M&A strategy remain unresolved.

CEO Bill McDermott is an unmistakable figure, his pompadour hairstyle and dark glasses as recognisable as Jensen Huang’s leather jackets. He is also prone to promotional language and grand claims, which has arguably helped fuel investor enthusiasm and, at times, an overextended valuation. As recently as April 2026, he declared: “There is no company in the enterprise software industry that is operating at the Rule of 55, that’s only ServiceNow.” As discussed earlier, that claim is open to challenge. Shares supported too heavily by narrative rather than fundamentals are always vulnerable to reversal; perhaps helping explain the 65% drawdown.

For now, the verdict remains open.

The $11 billion strategic pivot will require flawless M&A integration and a successful shift of customers towards consumption-based pricing. If executed well, the combination of asset visibility (Armis), identity governance (Veza), and core workflow automation could place ServiceNow at the centre of enterprise IT, security, and AI operations.

This may prove to be the moment ServiceNow became the defining enterprise software platform of the AI era, or else it may be remembered as the moment ambition outran execution. Only time will tell.

What’s your view? Please share it in the comments section below.

Leave a comment

Share

For decades, the promise of modern capitalism rested on a simple, positive-sum bargain: work hard, contribute to growth, and in time you could own a piece of the kingdom. Effort, discipline, and patience were meant to convert labour into capital. That was the social contract.

But if you compare the twenty years leading into the Great Financial Crisis with the twenty years that followed, it becomes clear that something fundamental changed. We moved from an era of aspiration into an era of entrenchment. The system no longer appears structured to create new winners. It appears increasingly focused on protecting the old ones.

The Great Financial Crisis was the line in the sand.

Before the crash, wealth was still largely connected to economic activity: building companies, expanding industries, producing goods, taking commercial risk. After the crash, the priority shifted from creation to preservation. By pinning interest rates near zero for more than a decade, central banks built a protective moat around existing asset holders.

That policy did not usher in a renaissance of factories, wage growth, or broad-based ownership. Instead, it made capital extraordinarily cheap for those who already had it. Cheap money flowed into finite assets; property, equities, land, infrastructure. The roof over your head became less a place to live and more a financial instrument.

From that emerged what can only be described as a Toll-Bridge Economy.

In a healthy capitalist model, workers become consumers, and consumers fuel growth. In the current model, participation itself has become monetised. Life increasingly feels like a chain of unavoidable tolls; subscriptions, insurance premiums, platform fees, service charges, utility costs, rent extraction.

You’re no longer living your life; you’re paying a subscription fee for the right to exist within the digital and physical infrastructure owned by a shrinking group of “super-rich elite.”

This helps explain the widening wealth gap of the last forty years. The engine that creates wealth has changed.

In the two decades before 2008, wages and wealth still moved in rough tandem. Imperfectly, yes, but recognisably. A worker could save, earn interest, buy a modest home, and slowly climb. The ladder still existed.

The post-crisis era replaced that ladder with an accelerator for the few and a treadmill for everyone else.

By suppressing rates to zero and printing new money, policymakers effectively signalled that the easiest route to wealth was no longer labour, innovation, or enterprise, but ownership itself. Hold assets, borrow cheaply, watch valuations rise. Those who already owned stocks and property were rewarded. Those dependent on wages watched the entry price to middle-class life drift further out of reach.

This created a K-shaped society.

One side, the Asset Class, sees net worth expand through passive appreciation.
The other, the Service Class, sees labour steadily cannibalised by the rising cost of existence.

The most visible symbol of this shift is the explosion of the billionaire class.

When the first global billionaire rankings appeared in 1987, there were just 140 billionaires worldwide. Adjusted for inflation, a billion dollars then would be roughly three billion today. Even using that higher threshold, the growth is astonishing. By 2026, there are more than 3,000 billionaires globally. This 21x increase is a total transformation of how much of the world's "paper wealth" is held at the very top.

In the late 1980s, the combined net worth of the world’s billionaires was around $295 billion. By 2026, it exceeds $16 trillion. While ordinary households struggle to keep pace with food, energy, rent, and mortgage costs, elite wealth has compounded at extraordinary speed.

The system has become a high-speed elevator for those who own the infrastructure of the world, while everyone else takes the stairs.

When wealth becomes concentrated in too few hands, the core engines of middle-class security are transformed into extraction tools. Housing becomes rent-seeking. Education becomes debt-financed access. Healthcare becomes recurring liability. Ownership becomes inheritance.

The system stops rewarding the climb and starts rewarding the altitude.

Nowhere is this clearer than housing.

In the two decades before the GFC, median home prices in many developed economies often sat around three to four times average annual income. Today, ratios of six or seven are increasingly common, and far worse in major cities.

When shelter rises faster than wages, labour is devalued. People work more hours merely to pay the same tribute to an asset owner.

For the first time in modern memory, a permanent renter class is emerging; people who participate fully in the economy yet are mathematically locked out of owning any meaningful stake in it.

Even savings have been folded into the machine.

Through pension schemes, workplace funds, and passive index investing, workers’ capital is automatically funnelled into the largest listed corporations. That creates a permanent bid under asset prices, regardless of fundamentals. Savings of the workers help inflate the valuations that price them out of ownership.

It is a feedback loop of elegant brutality.

When a society decides asset prices must never fall, this becomes a decision that the next generation must never rise.

The divide between those who work for a living and those who own for a living risks becoming structural, permanent, and politically explosive.

And this is where economics spills into politics.

When the ladder of mobility breaks, populations begin searching for someone who promises to smash the system. The result is often not reform, but the concentration of power in a new form.

Look across the world; established autocracies, populist democracies, technocratic states. Different branding, same instinct. Centralise authority. Bypass friction. Govern by decree. Protect insiders.

In a traditional democracy, the process of law-making is intentionally slow, designed to provide a check against the whims of the powerful. But in our current era of stasis, global leaders are increasingly bypassing these "democratic frictions" through executive orders and personal decrees.

For a figure like Trump, who is himself a billionaire scion of the asset-owning class, the rhetoric of "the people" serves as a powerful aesthetic for what is essentially the consolidation of power. His administration’s pivot toward a cabinet of tech oligarchs and loyalist "vassals" isn't an accident; it is the natural political extension of a world where wealth is no longer generated by broad-based competition, but protected by proximity to the throne.

Perhaps this helps explain why the old globalisation consensus has frayed so badly and why geopolitical conflict is on the rise.

When growth no longer feels shared, cooperation fades. Nations turn inward. Trade becomes weaponised. Hostility rises.

The deeper issue is not that capitalism failed. It is that capitalism without mobility hardens into aristocracy.

History suggests that systems built on excessive concentration rarely remain stable forever. When too much wealth, power, and opportunity calcify at the top, pressure accumulates beneath the surface until release becomes unavoidable. How this era ends is the open question. It could come through democratic reform; aggressive taxation, housing resets, anti-monopoly action, and a deliberate redistribution designed to rebuild the middle-class ladder before it snaps entirely. It could arrive through technological disruption that unexpectedly destroys old monopolies and creates new paths to ownership. But history also offers darker endings: social unrest, political extremism, class conflict, currency crises, or war; moments when societies choose destruction because renewal was denied too long. Empires rarely collapse because they run out of wealth; they collapse because they hoard too much of it. If the current order refuses to widen prosperity voluntarily, then change will eventually come involuntarily. The only uncertainty is whether it arrives through reform or rupture.

Share

Why Do I Burn Capital When Risk/Reward Skews My Way?

[Accreditation: This post was inspired by a discussion with my friend Mauricio who writes under the pseudonym Polymath Investor. Thank you for the inspiration amigo!]

Let’s play a coin-flip game.

You start with a $100 of capital.

The coin is fair: an equal chance of landing on ‘heads’ or ‘tails’.

If you win, your capital will increase by 50%. If you lose, your capital reduces by 40%.

So after the first flip, you’ll either have $150 or $60. Up $50 or down $40.

Are you willing to play?

At first glance, it looks attractive. You make more on a win than you lose on a loss, while chances are 50:50.

Most people would say “yes”.

Most people would lose money.

What? Why?

It’s all about understanding markets, compounding, managing risk and some very simple math.

Russian Roulette Explains it Best

Russian Roulette is a ‘game of chance’ that involves placing a single bullet in one of six chambers of a revolver. The player spins the revolver (hence the ‘roulette’ reference), points the gun to his head, and pulls the trigger.

The player has five-in-six chance (83%) of survival, a win, and a corresponding 17% chance of losing (certain death).

The odds are favourable, but the risk is unacceptable. If you lose, there is no chance of ever making a recovery. Even if the incentive to play was a cool $1 billion US dollars if you survive, are you taking the chance?

As this demonstrates, what ought to determine your decision to play is not the size of the prize, but the extent of the possible loss and how difficult it will be to recover from it.

In the world of investing, this is the part that most people miss. They become captivated by upside potential while giving too little weight to downside risk, drawdowns, and the mathematics of recovery.

Perhaps that is because, in the stock market, it is not their life at stake. But financially, the consequences can still be equally severe.

Share

LTCM: When Genius Ignores Path Dependency

Long-Term Capital Management (LTCM) was an investment firm established in 1994 and stacked with elite minds, including Nobel Prize winning mathematicians and economists.

Early returns were extraordinary.

Year 1 saw annualized returns of over 21% (after fees). Year two and three were even better at 43% and 41% respectively.

But in 1998, its fourth year of operation, it lost $4.6 billion. Inadvertently, it had been playing Russian Roulette in a financial context and it met an abrupt and sudden death.

Brilliance didn’t save them from mathematics.

So what can we learn from this?

Back to the Coin Flip

Be reminded that you begin with $100, if you win, your capital increases in value by 50%. If you lose, your capital reduces by 40%.

Let’s keep this simple and consider only two flips of the coin. There can only be four possible combinations from two flips:

1. Win/Win [CAGR 50%, two years, capital value = $225]

2. Win/Lose [+50% to $150, -40% to an end capital value = $90]

3. Lose/Win [-40% to $60, +50% to an end capital value = $90]

4. Lose/Lose [CAGR -40%, two years, capital value = $36]

Total of all outomes: $225 + $90 + $90 + $36 = $441

Average (mean) result: $441 / 4 = $110.25

If I start with $100, and the average outcome appears to be $110.25 after only two flips of the coin, that looks great, right?

Wrong!

Think about the median outcome. Three of the four paths leave you below where you started.

You have a 75% chance of losing.

That’s the trap.

Finance is full of this mistake: assuming that favourable averages for the market as a whole automatically translate into favourable long-term results for each individual investor.

The winning path, on which you have only a 25% probability of landing, delivers an outsized compounded return that pulls the overall average higher.

The gain from that single favourable path exceeds the combined losses from the other three possible paths, which is precisely why the arithmetic average appears skewed to the upside. But it’s a misleading signal. Do you want to accept a three-in-four chance of losing money?

It’s like Russian Roulette all over again, but this time with a modified chamber. Rather than only one bullet in six chambers of the revolver, you are playing with three bullets in only four chambers!

Do you still want to play?

Although a 50% upside versus a 40% downside on a 50:50 coin flip may look asymmetrically favourable, the real problem is that you are thinking in percentages rather than monetary outcomes.

Why does that matter?

If someone offered you $50 for a win while risking $40 for a loss, you would take that bet all day long. Over a large number of coin flips, heads and tails tend to balance out. So, for every $40 loss, you collect $50 on a win. That is a genuinely favourable skew.

But that is not what is happening when the outcomes are expressed as +50% and -40%.

Percentages are relative to the changing value of your capital. They do not offset each other in a straight-line way.

For example, a 50% loss requires a 100% gain just to get back to where you started. Likewise, if you gain 100% on one flip and then lose 50% on the next, you end up exactly where you began.

This is why percentage outcomes can be deceptive. A gain and a loss of similar size are not equal in economic terms.

So, if you are measuring returns in percentages, the upside must be materially larger than the downside for the odds to be truly skewed in your favour (see chart below):

The Basic Math

Let’s break this down in simple terms looking at the geometric mean as a proxy for median1. Below are four scenarios (A to D) that demonstrate why chasing large gains is what kills most investors portfolios and why both Howard Marks and Warren Buffett advise that the focus should be on managing downside risk, after which the upside will look after itself.

(Scenario A)

The example we considered above (Win: +50%, Loss: -40%):

• Chance of a win = 0.5; Chance of a loss = 0.5

• Impact on capital base of a win = 1.5x

• Impact on capital base of a loss = 0.6x

• Expected value per flip = (0.5×1.5) + (0.5×0.6) = 1.05

• Since 1.05 > 1, we would expect a positive arithmetic outcome.

• But the geometric mean is: √(1.5×0.6) = √0.9 = 0.949

• Since 0.949 < 1, this reveals that the most probable outcome will be negative.

Essentially, we see negative geometric growth despite having a positive arithmetic expectation.

The arithmetic outcome is dominated by a vanishingly small number of exceptionally lucky sequences in which the player strings together multiple consecutive wins early on, allowing gains to compound strongly.

This is visible on a distribution chart through the pronounced long tail. Capital can only fall to zero, creating a hard floor on the downside, yet there remains a small probability of compounding upward to extremely large values.

That long right tail is what pulls the simple average higher and creates the misleading impression (a ‘false positive’ signal that the game offers a favourable expected return).

This is why people confuse positive expectancy with positive wealth creation.

They are not the same thing when compounding is involved.

It is the median outcome that represents what happens to the “typical” player, and what happens most often, is a loss of capital: a negative result.

As the number of flips increases, this situation becomes more pronounced; the median will slowly but surely pull your capital balance towards zero, while the mean will creep higher.

(Scenario B)

Digging deeper reveals that people chasing high returns, and risking larger drawdowns on their capital to do so, are on a fools errand (as LTCM learned the hard way).

Let’s keep the ratio of benefit to detriment unchanged, moving from +50:-40 in the last example, to +100:-80 (Win: +100%, Loss: -80%)

• Chances of a win or loss are still 50:50

• New impact on capital base of a win = 2.0x

• New impact on capital base of a loss = 0.2x

• Expected value per flip: (0.5×2.0) + (0.5×0.2) = 1.1 (the positive expectation has increased due to the higher win percentage lengthening the tail)

• But the outcome has become way more volatile. Bigger win vs. bigger loss potential.

• The geometric mean now: √(2.0 × 0.2) = √0.4 = 0.632

The geometric mean has fallen dramatically, and this remains the most likely outcome. This is a super dangerous game to play.

(Scenarios C and D)

So instead of chasing outsized returns, what happens if we focus on mitigating the downside (managing risk).

Again, we’ll keep the win:loss ratio the same, but take it in the other direction. Our initial +50:-40 first becomes +25:-20, and then +12.5:-10.

(Win: +25%, Loss: -20%)

• Expected value per flip: 0.5×1.25 + 0.5×0.8 = 1.025

• Geometric mean: √(1.25×0.8) = √1.0 = 1.0

• A geometric mean of 1.0 is the breakeven point: zero growth and zero loss.

Now take a look at the ‘break-even’ bar chart above. You’ll see that +25% and -20% are indeed the balancing point.

So, we’ve moved from two guaranteed loss scenarios to break-even. Now we seem to be travelling in the correct direction.

If we reduce the loss percentage still further, are we able to push into a guaranteed sinning strategy?

(Win: +12.5%, Loss: -10%)

• Expected value per flip: 0.5×1.125 + 0.5×0.9 = 1.0125

• Geometric mean: √(1.125×0.9) = √1.0125 ≈ 1.0062

• Positive geometric growth. This is the only scenario of the four studied with consistent long-term growth!

It becomes evident that the game is not won by maximising upside.

It is won by controlling downside.

Large drawdowns damage compounding more than large gains help it.

That’s why risk management matters more than return targets.

Share

Kelly Gets It: The Magic Formula

The Kelly Criterion, popularised by J. L. Kelly and later championed by Ed Thorp, grew out of Claude Shannon’s Information Theory. At its core, it is a framework for maximising long-term compound growth.

Its most important lesson is simple: never risk ruin. Don’t bet the farm, no matter how attractive the opportunity appears. Survival comes first. This is where the Russian Roulette analogy sits.

So, Kelly’s central question was straightforward:

“What fraction of my capital should I allocate to the next opportunity, based on the probability of success and the payoff if I am right, relative to the loss if I am wrong?”

That works neatly in games of chance. A coin toss has known odds. A roulette wheel has fixed probabilities. The variables are measurable.

Equities are different.

When buying a stock, there is no objective probability of success. No formula can tell us the precise odds that a company will outperform, rerate, or compound over time. At best, we are making informed judgements. That makes the pure Kelly formula more elegant in theory than useful in practice for investors.

More particularly, if the win vs. loss payout skew is unfavourable, why stake any capital at all? Better opportunities exist elsewhere. In that sense, Kelly appears to be about forcing a bet and sizing it accordingly, which is something a rational investor would never do.

So, stepping out of the world of academia, how can investors use a derivative of the Kelly principle when investing?

Rather than pretending to know precise probabilities, we build scenarios.

Before investing, we develop a bull case and a bear case. In each scenario, we estimate what the business may be worth relative to today’s share price. From that, we can calculate:

• the potential percentage gain if the thesis plays out

• the potential percentage loss if it doesn’t

Those are practical, investable variables.

And once we have them, we can begin to think the Kelly way.

For algebraic purposes, we’ll call a win ‘X’ and a loss ‘Y’. For a 50/50 game, the breakeven condition for geometric growth is:

(1+X)(1-Y) = 1

Which may be rearranged to define X in terms of Y at breakeven, as follows:

1 - Y + X - XY = 1

X - Y - XY = 0

X(1-Y) = Y

X = Y/(1-Y)

Now we can suplant the numbers in our scenarios.

In scenario C, when the drawdown on a loss (Y)=20%: to achieve breakeven X would need to be 25% (20%/80%), which was exactly the win percentage, hence the breakeven outcome.

The only scenario that yields positive geometric growth is D, when the drawdown on a loss (Y)=10%. To achieve breakeven X would need to be ≈11.11% (10%/90%). Since the gain on a win was 12.5% (greater than 11.11%), the typical player actually makes money over time.

In short, to win, you need X > Y/(1-Y)

In the world of investing, this explains why volatility drag is real, why loss ratios matter, and why the understanding risk is essential for long-term growth.

Moral of the story

Many investors are playing financial Russian roulette.

They see favourable averages and ignore path dependency, volatility, and downside risk.

That’s how capital gets destroyed.

As Howard Marks always counsels, ‘Look after the downside and the upside will look after itself’.

Always focus on the risk!

Subscribe now

Share

Further Reading: More ‘Risk’ Content

If you enjoyed this, you will also like these:

Kinsale Capital Group (KNSL) is an insurance company with a difference. Since its 2016 IPO, the stock has surged over 2,000%, powered by a staggering 41% EPS CAGR that has left the rest of the industry in the dust.

As the only publicly traded pure-play in the Excess & Surplus (E&S) market, Kinsale has carved out a massive advantage by insuring the “hard-to-place” commercial risks that standard carriers are too slow or too rigid to handle.

At the heart of this success is a proprietary, browser-based technology platform built from scratch over seven years; a structural moat that allows Kinsale to quote complex risks in hours while legacy competitors often take weeks.

By strictly centralizing all operations in a single Richmond, Virginia headquarters and rejecting the industry trend of regional branch offices, founder Michael Kehoe has engineered a high-efficiency machine with a 1,000-basis-point expense advantage over its peers.

Kinsale’s strategy is built on radical selectivity: the firm processes nearly a million submissions a year but ultimately binds only about 7% of them. Unlike almost every other major player in the E&S space, Kinsale refuses to delegate underwriting authority to third-party agents (MGAs), keeping every pricing and claims decision in-house to ensure an airtight feedback loop and industry-leading risk selection. This disciplined, data-driven approach has produced a combined ratio consistently in the mid-70s, an “inverse profit margin” that is essentially unparalleled in the property and casualty world.

Leading this charge is Michael Kehoe, a visionary owner-operator who founded the company in 2009 amidst the wreckage of the financial crisis and still maintains a significant personal stake today. Inspired by legendary leaders, including Jack Welch and Andy Grove, Kehoe has successfully translated his “hard way” philosophy into a culture of speed and precision, drawing frequent comparisons to insurance success stories like Progressive and GEICO for his ability to gain market share across all phases of the insurance cycle.

In this exclusive hour-long interview, we sit down with Kehoe to go inside the “Kinsale Machine.” We explore how the firm maintains its massive performance gap over the industry, the mechanics of its conservative reserving culture, and the roadmap for reaching a double-digit share of the $135 billion E&S market. Whether you are a value investor or a growth seeker, this is a masterclass in how disciplined execution and proprietary tech can turn a “boring” niche into a world-class compounding engine.

Share

To better understand the E&S Insurance Capital Cycle, please see:

For further reading on Kinsale Capital, Rock & Turner analysed the business early in 2025 and then Alexander at Slow Compounding wrote up a 100 page deep dive at the beginning of 2026, both links are below and worth reading:

Slow Compounding

Deep Dive - Kinsale Capital Group (Part 1)

In 5 articles, you’ll get a deep dive of Kinsale Capital Group (~100 pages…

Read more

5 months ago · 48 likes · 8 comments · Alexander

The Excess & Surplus (E&S) market has always operated as the industry’s safety valve. It is where non-standard risks meet flexible, high-conviction capital, and where pricing is built case by case rather than forced into rigid actuarial templates. That flexibility is the point.

This market follows a rhythmic, supply-driven cycle, but it is important to note that the E&S market is not monolithic. It’s a complex portfolio of micro-niches, each of which experience its own cycles: a "hard" market might persist in catastrophe-exposed property or specific cyber niches, even as a "soft" phase takes hold in the casualty segment.

In the U.S. today, we are witnessing a synchronized descent into a soft market across many core E&S lines. However, this is not a traditional softening. It is being hyper-accelerated by a volatile cocktail of trapped liquidity, inflation, and the unprecedented velocity of AI enhanced data driven underwriting, creating a cycle that is moving faster, and perhaps more dangerously, than any we have seen before.

Subscribe now

The Anatomy of the Soft Cycle

Soft markets in E&S are not driven by weak demand. The underlying risks are increasing, not declining. Climate volatility, litigation trends, and cyber threats all point in the same direction; demand is strong. The pressure instead comes from the supply side. Capital enters after a period of strong returns, and competition to deploy that capital pushes pricing down. Margins compress, and loss ratios expand.

Here’s a short clip from an interview with Michael Kehoe, CEO and Chairman of the only listed pure-play E&S insurer in the U.S. He explains how excess capital is driving mispricing across the market. That mispricing sets up future losses and, ultimately, capital destruction. Disciplined operators respond by stepping back. Firms like Kinsale avoid chasing inadequately priced risk and allow weaker behaviour to run its course. As a result, growth slows during the soft phase of the cycle. That is a feature of underwriting discipline, not a weakness. (Source: Full Interview with Michael Kehoe)

The 2026 ‘Soft Market’ Drivers

The current phase traces back to the profitability seen between 2022 and 2025. That performance attracted new forms of capital, and with them, new incentives.

⁜ The MGA Explosion

Managing General Agents (MGAs) have proliferated and changed how risk is distributed. They act as agile intermediaries that deploy institutional capital. An MGA has delegated authority to underwrite, price, and bind risks on behalf of an insurer, along with handling claims. Most of their earnings are generated in commissions and service fees, with only a small element being based on the underwriting profitability of the risk being placed. As such, the focus tilts towards volume rather than margins.

⁜ Alternative Capital Inflows

Alternative capital has scaled. Insurance-Linked Securities (ILS) and catastrophe bonds have broadened the investor base. This capital is not always anchored to long-term underwriting returns. It is often seeking risk diversification, which makes it more sensitive to relative returns elsewhere. When it enters, supply expands quickly.

⁜ The Reinsurance Effect

Reinsurance has reinforced the dynamic. Lower reinsurance costs increase capacity at the primary level. Carriers can offload risk more cheaply, which encourages additional underwriting. Competition then pushes pricing lower, often beyond what underlying loss trends justify.

Subscribe now

The Complex Macro Environment

The current economic landscape adds a layer of complexity to this narrative.

※ Gated Funds

Private Equity (PE) firms have been “gating” their credit funds, which ought to signal tightening capital. In practice, it can have the opposite short-term effect. Insurers under pressure to maintain liquidity may write more business to generate premium inflows. That response supports softer pricing, even as the broader system is becoming more constrained. It is a temporary release valve, not a solution.

※ Interest Rates

Elevated interest rates are distorting underwriting discipline. Higher yields increase the returns earned onfloat (the money held between receiving premiums and paying claims). This investment income acts as a subsidy, allowing them to tolerate poor underwriting results and keep premiums artificially low. The subsidy is real, but it isn’t permanent.

※ Inflation

The Iranian war and consequential energy crisis has revived the specter of inflation. While the market feels abundant today, these forces are creating a “coiled spring” effect. By competing on price during an inflationary spike, insurers are writing policies today that will be settled at tomorrow’s inflated prices. Balance sheets can appear healthy while embedded liabilities are understated. This is most acute in casualty, where the time between underwriting and settlement can span years. The industry may currently be in a phase of Reserve Inadequacy1.

Subscribe now

The Catalysts

These conditions rarely unwind gradually. The adjustment tends to be abrupt.

A large shock can remove excess capital in a single step. A $100 billion catastrophe is enough to deplete the surplus that has been supporting aggressive pricing.

The more structural risk sits in pricing itself. If the premiums written in 2025 and 2026 prove inadequate relative to future claims, capacity contracts quickly. Balance sheets come under pressure, and pricing resets.

Reserve weakness can accelerate the process. Once it becomes clear that prior-year underwriting does not support expected liabilities, carriers are forced to reprice entire books, often at once.

At the same time, the marginal capital in the system is not permanent. When alternative asset classes offer better risk-adjusted returns, capital from private equity and insurance-linked securities can exit just as quickly as it entered. That withdrawal leaves a gap in capacity that the remaining market cannot immediately fill.

When capital tightens, the market hardens.

Subscribe now

The Speed of the Modern Cycle

What is changing is the speed of this process. The traditional cycle stretched over a decade, with long feedback loops between loss experience and pricing. That lag has shortened materially.

In previous decades, insurers relied on quarterly actuarial “look-backs.” It could take years for deteriorating loss trends to show up in pricing. Today, data moves faster. Underwriters have access to near real-time signals, and pricing decisions adjust accordingly. Platforms such as Tractable and Cytora are part of this shift, enabling rapid changes in risk selection and pricing at a granular level.

The effect is a more reactive market. Carriers no longer blindly drift into unprofitability over extended periods. They adjust quickly, sometimes abruptly. Entire segments can reprice or shut down within weeks.

The current environment reflects the early, more comfortable phase of the cycle. Capital is available and pricing is competitive. But the same forces supporting this phase are also increasing the likelihood of a sharper correction.

The cycle has not disappeared. It has accelerated.

Share

Share

Companies start with purpose: a founder, a vision, a reason to exist. But founders leave and often, the soul leaves with them.

In come the agents; career managers. No vision, no north star to guide them, no passion.

Purpose gets replaced by process. Decisions are delegated to committees. Nobody leads; everyone manages to a script.

Soon, the company forgets why it exists. Bureaucracy creeps in. Meetings multiply. Progress slows.

Gone are the days of innovation, entrepreneurialism and of building something special. Now it’s about extraction, sucking the blood of the business. Personal interest now trumps collective corporate responsibility.

It’s no longer about the company, but the parasites within it. Who’s accountable for the future?

Not the managers; CEO tenures average just over 4 years, they’re already polishing their CVs.

Not the investors; the average holding period is a mere 156 days, they’ll soon move on to the next opportunity.

Everyone’s just passing through. A chapter of the story. No cares about how it ends.

Leadership becomes a revolving door. Tenure is brief and succession is outsourced.

Strategy talks about “the long-term”, yet management only thinks in quarters.

We’ve seen this movie before. We know how it ends. Stock price climbs. Executives cash out. The narrative collapses. Someone else is left picking up the pieces.

Smash and grab; that’s the playbook. Get in, juice the numbers, get out. “We’ll be gone before it all blows up”. That’s the real strategy

Poor incentives lead to poor outcomes. The smiles are fake, the decks are polished, but the mindset is clear.

For the senior management it’s less about value creation, and more about value extraction. Stewardship is symbolic, wisdom is absent.

Accounting games mask economic reality: “Adjusted earnings”, “non-GAAP”, “EBITDA”. Bullshit earnings as Munger called them. Financial reports gradually drift into fiction.

Inconvenient truths excluded, dilution masked with overpriced repurchases dressed up as a return of capital. Each tactic works, briefly, then the damage sets in. What looked good on paper was destructive in practice.

It looks healthy until it breaks. A bit like treating the symptoms, while the disease spreads unseen. It works, until it doesn’t.

Management cashes its chips and runs for the exit. The company limps on. Predators circle.

It’s all about today, not about tomorrow. No patience, no intergenerational vision, no trust, no future.

Collapse comes suddenly when the music stops playing.

Hope Is Not Lost - Some Do It Differently

Not every company plays this game. Some think long-term, acting with moral integrity.
They invest wisely, allocate capital carefully and treat culture as an asset. They have patience and humility. Respect for employees, suppliers and customers.

Rather than focusing on the next quarter, they extend their time horizons. They play the long-game. It’s all about resilience, adaptability and optionality.

And guess what?
When they win, shareholders win too!

A great company needs a great leader, but they are exceedingly rare.

Singleton, Buffett, Bezos, Leonard, Jobs - you know the names. If you want great outcomes, invest in great people.

In all other cases? Caveat Emptor! (Buyer Beware!)

The Straits of Hormuz Oil Market Price Crisis

In a “normal” world, oil usually costs more in the future because you have to pay for storage and insurance. But when the world is screaming for oil immediately and supply is short, it pushes up near term prices, inverting the price curve (a market in ‘backwardation’).

When the spot price (what you pay today) is significantly higher than the price for delivery three months from now. It’s the market’s way of saying: “I don’t care about next month; I need those barrels of oil today.”

The Chain Reaction & Ticking Time Bomb in Oil Markets

The problem is that this situation can very easily spark a chain reaction in which the market quite literally eats itself and we have what is known as a ‘blow-off top’.

The Oil Trader

Imagine you’re an oil trader, importing oil and selling it to enterprise customers. You have a long physcial position at sea and a short futures position to lock in the financial viability of your shipment1. But the Straits of Hormuz are closed and your oil is delayed. Your futures contract is coming up for expiry, but the physical oil contractual obligtation is still unfulfilled.

You have no choice but to roll the futures position, but this involves buying back your current expensive futures contract at the short-end of the curve (at a massive loss) and selling a new one further down the curve where prices are lower.

The result?

Not only is the financing of the roll very expensive, but desperate buying to “close out” expiring futures contracts pushes the front-end price even higher while the sale of longer dated contracts exacerbates the backwardation.

You are effectively throwing gasoline on a fire you would ideally like to put out.

The Oil Supplier

Now, you might think, “If I’m an oil supplier and I see the price spiking every day, why would I sell now? Shouldn’t I wait for tomorrow’s even higher price?”

While it seems logical to hoard your oil (“diamond handing”), physical suppliers (like a midstream company or a storage provider) usually do the opposite for three reasons:

• The “Bird in Hand” Rule (negative carry): If the spot price is $110 and the three-month future is $90, you are essentially losing $20 a barrel by waiting. In the physical world, “Cash is King.”

• The Tax on Storage : Holding oil in backwardation is financially suicidal. You are paying to store an asset that the market tells you will be worth less tomorrow.

• Inventory Depletion: Suppliers drain their tanks to the “bottoms” to capture record-high short-term prices.

A physical supplier who "holds out" for a higher price tomorrow is effectively taking a massive unhedged speculative bet. Most physical businesses are low-margin and high-volume; they aren't in the business of diamond handing oil. They sell when the market screams "I need it now."

Once the oil supplier’s tanks are almost empty and the supply-demand imbalance creates more competition for what little oil is available pushing short-dated pricing higher still.

The Oil Producer

A producer (like an oil drilling company) can’t just flip a switch to pump more of the black liquid. Their supply is relatively inelastic. This means that higher prices fail to remedy the situation; producers can’t just swoop in to flood the market with new production, no matter how attractive the price may be.

The Oil Consumer

To make matters worse, when a market is in steep backwardation, the physical consumer (the refiner) stops buying "extra" inventory. They buy only what they need to run the refinery today because any barrel they buy and keep for next week will technically be "cheaper" next week (according to the cost curve).

But if every refiner is running with zero margin of safety and a shipment is delayed, they have to jump into the spot market and pay ‘any price’ just to keep the refinery from shutting down.

The ‘Paper’ Oil Market

It should be noted that despite all this activity, the volume of actual physical barrels traded each day is tiny compared to the enormous amount of ‘paper’ trading happening in the oil market; futures, options and other derivatives that are financially rather than physically settled.

Most market participants are either speculating or hedging on futures price moves, not physically moving oil from one place to another.

This tends to amplify the backwardation as speculative interest is largely momentum driven.

Margin requirements are a genuine cost of doing business and these too can exacerbate the backwardation at the front-end of the curve as speculative shorts are squeezed out of the market, forced to buy in an effort to cover a short position that is becoming too expensive to finance.

The Blow-Off Top in Oil Markets

The confluence of these events results in what is known as a “blow-off top”, a vertical price spike.

The market quite literally eats itself from the inside out.

Why Doesn’t ‘Force Majeure’ Apply?

Force majeure is a clause in a contract that frees parties from liability or obligation when an extraordinary, uncontrollable event, such as a war or natural disaster, makes performance impossible.

Force majeure is difficult to invoke in oil contracts because the market itself has not stopped functioning. Oil is still being produced, transported, and sold. A disruption in one route or region does not eliminate supply altogether. Unless a contract explicitly ties delivery to a very specific source, such as a named field or terminal in the Middle East, it is hard to argue that performance has become impossible rather than simply more expensive or inconvenient.

Even if the contract specifies supply from the Middle East, the case remains weak. Flows continue through alternative routes, including limited shipments via the Strait of Hormuz and pipeline networks that bypass it entirely. The system adjusts rather than shuts down.

But let us assume, academically, that force majeure were successfully claimed, the broader impact on the market is unchanged. The supplier may be relieved of the obligation to deliver, but the pain simply shifts to another party. The underlying demand does not disappear. The buyer still needs the oil and must replace those barrels in the open market, typically at the front of the curve where prices are highest. The economic burden is not removed. It is transferred.

At the same time, as described, most of the trading that drives price is the financial layer of the market. In ‘paper’ oil markets, force majeure has no role. Participants with short positions must still close or roll them regardless of physical disruptions.

The International Energy Agency’s (IEA) Response

Although we have seen a steepening over the first quarter of 2026, reserves have not yet been depleted. The International Energy Agency (IEA) has released a record amount of emergency oil reserves and urged immediate demand restraint.

The IEA released 400 million barrels from strategic stockpiles (an unprecedented move in terms of its magnitude), but it warned that supply-side action alone would not be enough if the disruption continued.

While the IEA remains ready to release more if conditions worsened, but if the situation continues and if reserves are ever completely depleted, the price spike seen so far may pale into insignificance.

Ultimately, the price would go parabolic until it hits a level so high that factories shut down and people stop driving. This would be the only way for the market to be brought back into equilibrium, but the economic damage is huge.

More particularly, the crisis is affecting more than oil, including gas, fertilizers and petrochemicals.

Like dropping a pebble in a pond, the ripples extend outward in every direction. This is why global recession often follows an oil crisis.

Reopening the Strait of Hormuz is the key to restoring energy market stability.

Share

DISCLAIMER & DISCLOSURE: The author holds an investment in Ashtead Technology at the date of publication but that may change. The views expressed are those of the author and may change without notice. The author has no duty or obligation to update this information. Some content is sourced from third parties believed to be reliable, but accuracy is not guaranteed. Forward-looking statements involve assumptions, risks, and uncertainties, meaning actual outcomes may differ from those envisaged in this analysis. Past performance is not indicative of future results. All investments carry risk, including financial loss. This analysis is for educational purposes only and does not constitute investment advice or recommendations of any kind. Conduct your own research and seek professional advice before investing.

Why Should I Care

If you’re a GARP investor, then you’ll be hard pressed to find a better opportunity in the market than this one right now.

Imagine a business that came public in 2021 and has compounded revenue at 41% annually since. Earnings per share have grown at 37% over the same period, without any help from buybacks. Returns on invested capital sit at 22.7%. Growth has been driven by a disciplined roll-up strategy that continues to expand the opportunity set.

On the surface, this is exactly the kind of company the market usually rewards. High growth, strong returns, and a clear path to reinvest capital at attractive rates. Forward expectations suggest that this trajectory is not slowing down. If anything, the runway into the end of the decade still looks strong.

You would expect a premium valuation, right?

Instead, the business trades at just 7.9 times trailing adjusted earnings.

The numbers look exceptional. The valuation implies the opposite. The question is simple: “Which one is wrong?”

Let’s dig in to the detail.

The Backstory

Ashtead Group plc didn’t begin as a global scale operator. It started in 1947 as a village plant hire business in Surrey, UK, serving local construction demand with a limited asset base and a simple rental model.

For decades it remained small. At the point of its 1984 management buyout, the business was generating roughly £1 million GBP in revenue. Yet what followed from that base has been one of the more aggressive compounding stories in British corporate history.

The inflection point came in 1990 with the acquisition of the US business, Sunbelt Rentals, for £17.5 million. The price understated the strategic significance. That deal shifted the company’s centre of gravity to North America, placing it into a deeper, more fragmented rental market with higher long-term growth potential. Over time, that decision came to define the business. It embarked on a mission to consolidate the industry through strategic acquisitions: a roll-up strategy (Brad Jacobs style).

It’s growth has been astounding, achieving a 20% CAGR on its share price over the past decade, and a 13% CAGR over the last 30 years.

The playbook was executed with consistency. Capital was recycled into a steady stream of bolt-on acquisitions, building density across local markets while expanding into higher value verticals such as climate control and disaster recovery. The model evolved from general equipment rental into a broader infrastructure support platform. Scale improved utilisation, purchasing power and customer stickiness.

The Sunbelt acquisition mattered. It was the pivot that saw the business move from being exclusively UK focused, to now seeing more than 85% of group revenue (exceeding $10 billion), alongside the vast majority of profits, generated in the US. It fundamentally became a North American earnings engine and is the second-largest equipment rental company globally, with a market capitalisation north of $30 billion.

It explains why, as of March 2026, the group has rebranded to Sunbelt Rental Holdings, with its primary listing moved to the United States under the ticker SUNB, alongside a secondary listing in London.

But this is not a story about Ashtead Group, the focus here is a smaller, less visible business that it spawned: Ashtead Technology Holdings (AT.L), listed in London.

It began as a small, non-core division within the broader group, but was eventually spun out. The rationale was simple. It sat outside the parent’s core equipment rental focus, even if the underlying model shared clear similarities.

Today, the spun out Ashtead Technology Holdings is deploying a remarkably similar playbook to its parent, growth through roll-up acquisitions to gain a dominant market position, albeit in the sub-sea specialist equipment space. It is now enjoying similarly impressive compound growth rates.

The sub-sea business was founded in 1985 in Aberdeen, a city shaped by the development of the North Sea oil and gas industry. Ashtead Technology grew within a specialised offshore ecosystem where reliability and technical capability were essential.

Its focus was subsea electronics rental. This was not generalist equipment hire. The inventory consisted of highly specialised tools such as sonar systems, ROV sensors, and hydrographic survey equipment used for underwater inspection and data collection. These assets required calibration, maintenance, and technical support. They also had to integrate seamlessly into complex offshore operations, which raised the bar for service quality and expertise.

From the outset, the business operated in a different economic environment to its parent. Demand was linked to offshore energy cycles rather than construction activity. Asset complexity was higher, utilisation patterns were different and customer relationships were more embedded in operational processes. Over time, that divergence became more pronounced, setting the foundation for what would eventually become a standalone business with its own strategy and trajectory.

The separation came in 2008. Backed by Phoenix Equity Partners, management executed a buyout at roughly £95.6 million. Independence removed any strategic constraint. Capital and attention could be directed solely at offshore energy, where equipment complexity, utilisation and service intensity all compound differently to general rental. It shifted the centre of gravity from being a division to being a focused operator.

The next inflection came in 2016 when Buckthorn Partners and APICORP took ownership. This is where the model started to evolve in a more deliberate way. The business moved beyond pure rental into integrated subsea solutions. That shift sounds incremental but it changes the economics. Rental is transactional. Solutions embed you deeper into customer workflows, extend duration and increase switching costs. The strategy that followed was straightforward in design and demanding in execution. Build out capability through acquisition, extend geographic reach and layer services around a core fleet.

The acquisitions in that period were carefully crafted. TES Survey Equipment Services established a foothold in the Middle East. Forum Subsea Rentals was more consequential, scaling the fleet and consolidating market position. Welaptega Marine added inspection and asset integrity, moving the company closer to critical maintenance workflows. Aqua-Tech Solutions strengthened exposure to the Gulf of Mexico. Each deal added a piece of the same puzzle. More equipment, more services, more proximity to the asset lifecycle.

By the time the company came to market in November 2021, the shape of the business had already changed. The IPO on AIM (London Alternative Investment Market) raised £52 million and valued the company, post money, at around £129 million. The stated use of capital was to delever the balance sheet and continue acquiring. The more important backdrop was demand. Offshore wind was moving from concept to capital cycle. Subsea infrastructure was no longer just oil and gas. The same capabilities began to apply to installation, inspection and maintenance of renewable assets.

The acquisition cadence didn’t slow after listing. WeSubsea and Hiretech extended capability into dredging and decommissioning. ACE Winches added heavy-duty lifting and mooring, which sits directly in the critical path of offshore wind installation. Seatronics and J2 Subsea, acquired from Acteon Group, expanded the fleet materially and deepened technical capability in ROV tooling. The pattern is consistent. Increase fleet density, broaden services, and move closer to mission-critical activity.

What followed was a period of rapid scaling. Revenue quadrupled post-IPO, and the mix shifted meaningfully towards renewables. By October 2025, the company moved to the Main Market London Stock Market with a premium listing. That step was less about optics and more about access. Deeper liquidity, broader institutional ownership and a platform that can support continued consolidation in a fragmented market.

Allan Pirie, CEO, has led the business for over 15 years, across private equity ownership, the IPO and the uplisting. That kind of tenure is not common in roll-up stories. Ingrid Stewart, CFO, joined ahead of the IPO and has been central to capital allocation and deal execution. Beneath that, much of the senior team has remained in place. That stability matters because the strategy relies on integration discipline and domain expertise. This is not financial engineering. It is operational accumulation.

The result is a business that looks very different from where it started. From a niche rental arm to a scaled subsea technology platform embedded across the energy infrastructure lifecycle. It now sits in the flow of both traditional offshore energy and the build-out of offshore wind. That dual exposure is the outcome of a long, consistent strategy applied across multiple ownership cycles. Growth in both segments ebbs and flows, so having exposure to both smooths the growth profile of Ashtead Technology. In FY25, growth in renewables was impacted by the Trump administration policy of favouring of oil and gas over renewable energy generation: “Drill, baby drill!” is Trump’s mantra.

Given geo-political issues in the middle east preventing the free flow of oil through the Straits of Hormuz, offshore activity is likely to pick up in the year ahead. Ashtead Technology is well placed to ride that tailwind.

The Numbers

The company has ~650 employees with market leading domain knowledge and expertise, enabling it to gain the trust of customers to undertake complex underwater engineering and project execution. Complimentary acquisitions over time have resulted in the business now being able to offer a flexible service and an equipment selection to meet the wide ranging needs of customers across the full lifecycle of a subsea energy infrastructure programme. It is now, in many respects, a one-stop shop with a comprehensive offering that others are unable to match. This helps to increase the TAM of the company (forecast to grow at 6% CAGR through 2029, see chart below). Ashtead Technology is also expanding its range of proprietary in-house designed, engineered, assembled and operated technologies. Collectively, this provides Ashtead Technology with unrivalled capability that cannot be bought by competitors, positioning it well for future growth. This is a business with a deep moat, as demonstrated in its strong unit economics.

Increasing backlogs underpin future performance and provide good forward visibility.

The FY25 earnings continue to impress. Revenue growth was 22% on a constant currency basis, 21% in real terms (19% acquired, 3% organic and a 1% headwind due to adverse FX movements). It generated a 29.1% adjusted1 EBITA margin and 10% adjusted EPS growth. Strong strategic and operational momentum continues and the growth prospects look impressive. Importantly, second half revenues were 5% up on the first half, so momentum shows no sign of abating.

Meanwhile the balance sheet looks stronger as leverage is reduced to a mere 1.3x. Financing costs were higher than normal in FY25 as the result of recent acquisitions, but the deleveraging of the balance sheet ought to normalize interest expense in FY26, adding another boost to expected earnings.

The company is very disciplined and runs working capital at 16% of revenue, targetting 15%. Growth is funded primarily from operating cash flows, augmented with debt on a revolving credit facility.

The results speak for themselves. In the 4 year period 2022-2025, revenue has grown at 41% CAGR, adjusted EBITA at 44% CAGR and adjusted EPS at 37% CAGR.

This brings us to the company’s capital allocation policy. Reinvesting in organic growth and seeking out new bolt-on acquisition targets remains the focus, as it should be. Maintaining balance sheet strength is also a core objective.

The only real criticism of management, and it is a very British one, is the self-imposed constraint on capital allocation through a progressive dividend policy. It limits flexibility at exactly the moments when flexibility matters most. There are two arguments that can be made here:

1. A dividend is not a neutral act. It is a partial liquidation of the balance sheet. It should only be paid when the business has exhausted its reinvestment opportunities. Returning cash to shareholders while simultaneously relying on debt to fund operations is economically incoherent. It amounts to borrowing to pay a dividend.

2. The opportunity cost is obvious. With adjusted EPS of 49.4 pence and a share price of 380 pence, the business is capitalized at 7.9 times earnings despite delivering revenue and earnings growth in excess of 35% CAGR. That is a material disconnect from intrinsic value. In such situations, capital should be directed toward repurchasing shares. Buybacks at this valuation are not cosmetic. They are accretive, improve capital efficiency, and concentrate ownership at a discount to intrinsic value. Contrary to popular belief, and to that of management, repurchases should never be construed as being interchangeable with a dividend (click to learn more).

Management have been lobbied by shareholders on this point year after year, but refuse to change their position.

Is Ashtead Technology Holdings A Good Investment?

After reaching peaks above 800p in mid-2024, the share price saw a dramatic correction, falling as low as 346p by mid-2025.

The company reported significant project delays in the first half of 2025, particularly in the US market, which dampened investor confidence. This was caused by the new Trump administration and its shift away from investment in renewables.

Changes in trade policies and tariff uncertainties caused some clients to pause expansion plans for survey work and rental gear. This too was a temporary headwind.

As such, there is no structural issue within the business, merely a change in the macro environment.

Tariffs have since been deemed illegal by the Supreme Court, while Ashtead is technology agnostic: Its equipment is used both for offshore renewables (principally wind), and also for the oil and gas sector.

This means that neither of these issues is really an issue. They don’t impact the long-term prospects of the business and the FY25 figures support that.

After a massive rally from 2021 to 2024, the stock was aguably priced for perfection. It was way too expensive at its peak. When the growth story shifted from "explosive organic growth" to "steady acquisition-led growth," a correction was inevitable, but has the pendulum swung too far the other way?

On the face of it, the stock looks mispriced and way too cheap at the moment. It is arguably due a re-rating. The share price has been moving higher, slowly, off its 2025 low, but it should have much further to go.

That said, balance matters. There are risks that need to be acknowledged.

Customer concentration is one. The business has meaningful exposure to a relatively small group of large clients, primarily energy majors and offshore contractors. Layer onto that the inherently cyclical nature of offshore energy, and you have a setup where a downturn could pressure both utilisation and pricing at the same time.

However, context is important. The industry appears to be in an upward phase of the cycle, with strong visibility likely through to the end of the decade. And while Ashtead is reliant on a small number of customers, the relationship runs both ways. These customers are not incentivised to own large fleets of specialised subsea equipment that sit idle between projects. Renting on a flexible, pay-as-you-go basis is economically rational. That is the core of Ashtead’s value proposition.

The competitive landscape reinforces this positioning.

• Teledyne Marine operates primarily as a manufacturer and innovator. It builds high-spec components that underpin subsea operations, but it is not a rental platform. It supplies the tools rather than managing their deployment.

• Oceaneering is the closest scaled competitor, but its model is structurally different. It is vertically integrated, offering bundled solutions through its own ROV fleet and personnel. Ashtead takes the opposite approach. It provides a broad catalogue of equipment sourced from multiple manufacturers, allowing customers to select the best tool for each job without being locked into a single ecosystem.

• The evolution of Acteon Group is also instructive. Its 2024 acquisition by Buckthorn Partners and One Equity Partners marked a strategic shift. Acteon is becoming more focused on engineering-led services such as foundations, moorings, and decommissioning, while divesting rental-oriented assets like Seatronics and J2 Subsea. Ashtead moved in the opposite direction, acquiring those businesses to expand its equipment base and deepen its rental platform. Acquisitions like Seatronics and J2 Subsea consolidate these capabilities into a broader, more integrated offering that is difficult to replicate organically. The result is two companies that still compete, but with increasingly different strategic focuses and priorities.

• Below these big names sits a fragmented layer of specialists such as Subsea Technology & Rentals, Kongsberg Maritime, and EIVA. Each focuses on narrow capabilities across survey, positioning, or sensing. This fragmentation is exactly what enables Ashtead’s roll-up strategy.

All of this points to a clear conclusion. Ashtead has carved out a distinct niche within the subsea market. Its model combines breadth of equipment, capital efficiency for customers, and a scalable acquisition strategy.

That shows up in the numbers. Strong margins indicate pricing power. Sustained growth indicates demand. Together, they suggest durable earnings power.

The question is not whether the business is good. It is for how long will the market continues to misprice it. Either a re-rating will occur, or else it would come as no surprise if Ashtead Technology becomes the subject of a takeover attempt. It just looks too cheap.

This post was not intended to be a deep dive, but instead an introduction to this very interesting opportunity. If you want to investigate further, I recommend the following two Substack posts which do a wonderful job of diving deeper into Ashtead Technology:

Undervalued and undercovered

25% ROIC, 40% CAGR, PE of 7 — Why Is This Still So Cheap?

25% ROIC, 4x in revenues since IPO 4 years ago, reinvesting heavily to grow further. Strong M&A strategy plus pricing power and organic growth in double digits. And a PE of barely above 7, not to mention strong visibility…

Read more

7 months ago · 37 likes · 3 comments · Hugo Navarro

Compound with René

Deep Dive: Ashtead Technology ($AT)

Every now and then, a stock gets punished not because of deteriorating fundamentals, but because the market simply doesn’t know what to do with it. Ashtead Technology feels like one of those cases. Despite delivering blistering growth, fat margins, and exceptional returns on capital, the market seems to be looking the other way. And that disconnect is exactly what caught my attention…

Read more

a year ago · 22 likes · René Sellmann

Share

Why are dividends and share repurchases often presented as two ‘equivalent’ mechanisms for returning capital to shareholders?

I recently took this matter up with Professor Aswath Damodaran, whose recent post on dividends and buybacks really fired me up.

When a company raises capital, it has two broad options: debt or equity. Debt creates a fixed-term contractual claim. Equity creates a beneficial ownership claim that can exist indefinitely. Because equity is perpetual and subordinate, it is structurally more expensive. Interest payments on debt are tax deductable, whereas equity distributions are not. On every measure, all else equal, debt capital is preferable to issuing equity.

However, in the early stages of a company’s life, issuing equity is often necessary. Cash flows are uncertain, often limiting the availability of debt finance. Financial flexibility matters more than cost, so founders surrender equity in return for growth capital.

But once the business matures and can fund itself internally, the logic changes. Surplus capital should be used to repay the capital that financed the business in its early phases of life.

With debt, this principle is widely accepted and legally obligatory. Debt can be rolled, but contractually its purpose is repayment over time. As borrowing needs decline, leverage falls and financial risk reduces. This is disciplined capital allocation.

Share

Why should equity financing be treated differently?

A share repurchase is economically the retirement of equity capital. It reduces the ownership claims outstanding and increases the proportional interest of remaining shareholders. In that sense, it is the equity equivalent of paying down debt. The purpose is balance sheet optimization and capital structure discipline.

If shares are repurchased below intrinsic value, the transaction is accretive to investors. If repurchases occur above intrinsic value, shareholder equity is destroyed. Timing therefore matters.

This is where the repayment of equity capital differs from debt. Debt follows a schedule and carries an obligation. Equity doesn’t. Buybacks are an option, not a requirement. They should be executed opportunistically, not as a routine or simply to offset dilution regardless of price.

A good capital allocator pounces when the opportunity presents itself, but at all other times sits on his hands. That frames how stock repurchases should be approached and why they require judgment. They are not inherently good or bad. The outcome depends entirely on price and timing.

A dividend, by contrast, is a partial liquidation of the balance sheet without altering the capital base. The dividend capital allocation lever ought to always be the last to be pulled.

Any company with a rigid dividend policy, that puts the payment of dividends as a priority, is tying the hands of its management. This ought to be a red flag for any right thinking investor.

Said differently, a dividend is a one-time payment to the investor, which deprives the company of capital for reinvestment. Moreover, it creates an expectation that it will be repeated, acting as a constant drag on growth. Meanwhile, a share buyback, done properly, increases the equitable claim of remaining shareholders in perpetuity - it is the gift that keeps giving, year after year, and at no additional cost to the company.

Dividends and buybacks are therefore not interchangeable tools. One adjusts the capital structure by reducing outstanding ownership claims, while the other distributes surplus cash when the company has no other use for it. They serve different economic purposes and should be deployed under entirely different circumstances.

Enough Theory! Give Me An Example!

The mechanics are simple, but the implications are often missed.

Imagine you are a CEO looking to acquire a business for $20 million, which is capitalized at 12.5x earnings, offering an 8% earnings yield. Your own shares are trading at $2, and you believe that valuation is generous relative to intrinsic value. Instead of using debt, you decide to issue equity and lean into that overvaluation. You issue 10 million new shares and raise the full $20 million needed for the acquisition.

Now introduce a shock. A macro event hits, war in the middle east disrupting oil supplies, markets broadly sell off, and your share price falls to $1. At this level, you know that your company is trading below intrinsic value.

At that point, you have a choice. One rational response is to reverse the earlier transaction. You repurchase the same 10 million shares in the market, which now costs you only $10 million.

The net effect is that the acquisition, which was priced at $20 million, has only cost $10 million in real terms. That price is equivalent to only 6.25x earnings, boosting your yield on invested capital from 8% to 16%.

This is the core idea. Capital structure is not fixed. CEOs of public companies can use the liquidity of capital markets to their advantage. In practice, very few have the commercial accumen to understand this, which is why so few do it.

This is not hypothetical. It is exactly what Warren Buffett did with General Re in 1998. He used Berkshire’s elevated valuation at the time to fund the acquisition, then benefited as that relative valuation shifted over time.

Henry Singleton, the man Buffett described as the greatest capital allocator of all time, also did it this way.

Although this example involved M&A, acquiring another business is not a pre-requisite. The CEO may instead simply decide to buy back more equity in his own company, which is tantamount to M&A (albeit with lower execution risk).

Either way, repurchases are capital allocation decisions that exploit mispricing. They are not simply a mechanism for distributing excess cash.

Framing buybacks as a ‘return of capital’ is where the confusion starts. Done properly they are accretive to shareholder returns, but using this ‘return of capial’ language encourages investors to treat them as equivalent to dividends, yet they’re not remotely similar in substance or in form:

• A dividend is a partial liquidation of the balance sheet through a direct transfer of capital from the company to its shareholders: it weakens the business.

• A buyback, done well, is an investment decision: it strengthens the business.

Share

If you want to learn more on how capital allocation will impact your investment returns, please see:

Share

DISCLAIMER & DISCLOSURE: The author holds no small position in LSEG at the date of publication but that may change. The views expressed are those of the author and may change without notice. The author has no duty or obligation to update this information. Some content is sourced from third parties believed to be reliable, but accuracy is not guaranteed. Forward-looking statements involve assumptions, risks, and uncertainties, meaning actual outcomes may differ from those envisaged in this analysis. Past performance is not indicative of future results. All investments carry risk, including financial loss. This analysis is for educational purposes only and does not constitute investment advice or recommendations of any kind. Conduct your own research and seek professional advice before investing.

The London Stock Exchange Group has fallen around 30% over the past year significantly underperforming its US and European peers.

Zooming out further, it is clear to see that while other exchanges have seen their share price returns dip sharply, LSEG appears to have fared the worst.

It seems that companies listed in London are somehow cursed with negative market sentiment.

Should LSEG move its listing to the US?

Wow, just imagine that!

Of course that will never happen, so let’s explore what’s going on. Has LSEG been over sold? Is this a good investment opportunity?

Excutive Summary

London Stock Exchange Group (LSEG) is a poorly understood business. Although it began as a market place for buying and selling shares, equity trading now accounts for only 5% of revenues.

Today, LSEG sells the tools that banks, hedge funds and trading firms all depend on: market data, indexes, risk intelligence, trading workflows and clearing infrastructure.

Every trade, portfolio decision and compliance check runs through some combination of these services.

The market’s current concern is that generative and agentic AI will weaken the economics of financial terminals and workflow products. The more likely outcome is the opposite. As AI becomes embedded inside financial workflows, the number of data queries, model checks, and compliance validations increases. Each of those interactions requires reliable, structured, proprietary data. That plays directly into LSEG’s core strengths.

The company’s strategy reflects this shift. Through its long term partnership with Microsoft and the “LSEG Everywhere” approach, LSEG is embedding its data directly into the tools professionals already use such as Excel, Teams and Copilot. The goal is simple: when AI runs inside financial workflows, it runs on LSEG data.

The result is a systemically important infrastructure business that benefits from the growth of financial data usage.

Tell Me More

London Stock Exchange Group has spent the last several years transforming itself from a regional exchange into a global financial data and infrastructure provider.

The Refinitiv acquisition was the turning point. That deal reshaped the company’s revenue mix and pushed the business toward recurring data and subscription income.

Today roughly 73% of group revenue is recurring, which gives the company unusually high visibility into cash flows (peer recurring revenues for comparison: Nasdaq 77%, Euronext 55%, ICE 51%, Deutsche Boerse 36%, CBOE 27% and CME 19%).

What sits underneath that revenue base is a set of assets that are deeply embedded in the functioning of global markets.

London Clearing House (LCH), the group’s clearing house, arguably its crown jewel, dominates the clearing of over the counter interest rate swaps. SwapClear processes more than 90 percent of cleared OTC interest rate swaps globally and holds around 80 percent market share across the major currencies. The product is embedded in dealer balance sheets and risk management systems, supported by long term bank partnerships that extend well into the next decade. Clearing infrastructure is extremely difficult to replicate because it depends on liquidity, regulatory approval and network effects between large banks.

Tradeweb (TW) - [LSEG holds a 51% stake] - provides a different but related source of strength. The platform sits at the center of the long term shift toward electronic trading in fixed income markets. Average daily volumes now exceed $2.2 trillion. The platform benefits when markets become more volatile and when trading shifts away from voice broking toward electronic execution. The underlying trend has been steady for more than a decade and continues to expand across asset classes. The chart below demonstrates the Tradeweb average daily volume (ADV) growth trajectory across asset classes.

FTSE Russell contributes another important layer to the group. The index business generates high margin intellectual property revenue tied to the continued growth of passive investing. Its benchmarks support more than $1.4 trillion in ETF assets and continue to grow as asset managers increase their use of indexed strategies, particularly in fixed income. Alongside this, the Risk Intelligence division provides regulatory and compliance data services that have been growing at roughly 10% annually, supported by increasing global regulatory complexity.

Despite these strengths, the market has recently begun to treat LSEG as an artificial intelligence loser. The concern is that generative AI could weaken the value of traditional financial terminals and workflow products. Investors worry that large language models might replace parts of the information and research layer that financial data providers sell to institutions.

This narrative has weighed heavily on the share price. LSEG has significantly underperformed both the FTSE All Share and its US information services peers.

Yet the underlying exposure to AI disruption is often misunderstood. The most vulnerable activities are research and advisory workflows. Those areas represent a minority of the company’s business. The Workflows segment accounts for about 22 percent of group revenue, but the part most exposed to AI substitution represents roughly 7 percent of total revenue.

Much of the workflows business sits inside trading environments rather than research desks. Traders rely on extremely low latency data, tightly integrated execution systems, and infrastructure that must operate with near zero error tolerance. Generic AI tools cannot substitute for that environment. A trader operating in fixed income or equities generates several million dollars of revenue per year for a large bank. A small reduction in terminal subscription costs provides little incentive to risk operational disruption or regulatory issues.

The company’s data assets also represent a significant barrier. LSEG aggregates information from around 575 exchanges globally. That number is substantially higher than most competing platforms. The group also maintains a historical tick database that stretches back to 1996 (peer comparison: 2008 for Bloomberg and 2013 for FactSet). For machine learning and market modelling, long and consistent historical datasets are extremely valuable. Replicating that history is effectively impossible for new entrants.

The company’s partnership with Microsoft reflects this reality. Rather than attempting to compete directly with AI platforms, LSEG is integrating its data infrastructure with Microsoft’s cloud and productivity ecosystem.

The first stage of the strategy is the migration of roughly 350,000 users onto the Workspace platform, which will gradually incorporate natural language search and AI assisted workflow tools. Over time, the platform is expected to support agent based automation that can assist with complex financial analysis tasks while still relying on LSEG’s proprietary datasets.

The infrastructure behind this is substantial. LSEG processes around one billion data ticks per second and uses specialised timing systems to synchronize market data with extreme precision.

The group is also migrating approximately 33 petabytes of data to Microsoft Azure. This shift reduces the need for internal capital investment while improving accessibility for clients and developers.

A key part of the strategy is distribution. By embedding LSEG data inside Microsoft 365 applications such as Excel, Teams, and Copilot, the company places its datasets directly into the daily workflow of financial professionals. The goal is to ensure that AI tools still rely on LSEG as the underlying data source. If successful, AI does not displace the company’s data franchise. It increases the number of data queries and creates opportunities to charge for additional usage.

LSEG strategic partnerships began with Rogo and Databricks, but since expanded to include Snowflake, Microsoft and Athropic (Claude). In August 2025, the company partnered with Rogo, an AI platform used by investment banks and a Microsoft partner, allowing Workspace users to search, analyse and generate insights directly from LSEG datasets using AI models and agents. These tools can automate tasks such as preparing investment banking pitches and analysing market trends. A month later, LSEG announced a partnership with Databricks to make its data available natively within the Databricks platform through Delta Sharing. This allows financial institutions to combine LSEG’s raw tick and reference data with their own internal datasets and rapidly deploy AI agents for analytics, risk management and trading workflows. Together, the partnerships expand the ways LSEG’s data can be embedded into AI driven financial applications.

Share

The Numbers

Financially the company remains strong. Adjusted EBITDA margins are approaching 50% and organic growth remains solid, see chart below:

Yet the market values the company at a large discount to comparable US information services firms. The shares currently trade at roughly 18 times forward earnings, compared with multiples above 30 times for companies such as S&P Global and MSCI.

The gap becomes even wider when adjusting for LSEG’s 51 percent stake in Tradeweb. Removing that stake implies that the core LSEG business trades closer to 15 times expected 2027 earnings1. That multiple sits near the bottom of the exchange and financial data peer group.

The share price and the EPS numbers appear to be moving in opposite directions.

At current prices the market is effectively discounting several negative scenarios at once. One interpretation is that earnings expectations will be cut materially. Another is that workflows growth will stall while other data segments slow. The most extreme interpretation is that the market assigns little or no value to the workflows business at all, despite the company holding roughly 9 percent share of the global terminal market.

Management appears to view the valuation as unusually low. The company is “returning capital”2 through a £2.5 billion share buyback programme.

The company has a progressive dividend policy (so typically British!). This is a shame because at the current valuation, surplus capital applied to buy backs would be far more accretive for shareholders than dividends. Unfortunately, old-fashioned and misinformed British investors continue to demand dividends (a partial liquidation of the balance sheet), then they wonder why British companies underperform their international peers!?! In any event, the payout ratio is high, averaging ~60% of earnings, yielding ~1.6%.

Is LSEG A Good Investment?

The central issue for investors to ponder is whether the AI disruption narrative is correct.

The underlying numbers suggest that the exposure is limited and that AI is more likely to be a tailwind for the business than a headwind.

Progress on the Microsoft integration, the rollout of AI enabled tools and clearer disclosure around user growth could all help rebuild market confidence.

At present the shares appear to trade materially below the value implied by the company’s assets and long term cash generation. Some analysts are suggesting that the shares are trading at 66 pence on the pound, which implies ~50% upside from current levels around 80 GBX per share.

If the market agrees and gradually recognises this, the current valuation gap could close. Is LSEG at its current depressed price a good investment opportunity? Only you can decide. This is not investment advice, purely educational material.

Share

The digital economy was sold as weightless. Data would float in the cloud. Software would scale without friction. But that story has now met a hard physical constraint: Electricity.

After two decades of flat U.S. power demand at roughly 4,000 TWh per year, the system is shifting. AI and hyperscale data centers have altered the slope. Demand is now projected to rise roughly 20 percent by 2030. For the first time since the electrification boom of the 1920s, the binding constraint on industrial expansion is not capital or raw materials. It is the grid’s ability to deliver power.

By the end of 2025, at least five U.S. data center campuses had reached gigawatt scale. Each one requires the equivalent output of a large nuclear plant. Yet transmission infrastructure remains slow to expand. New high voltage lines can take a decade to permit and build. Gas turbine supply chains are effectively sold out through 2030. Even if generation capacity were available, the wires often are not.

Some operators have considered diesel backup just to keep servers running. That alone tells you how tight the system has become.

The instinctive response is to build more generation. Investors have piled into nuclear and Enhanced Geothermal Systems. But that assumes the core problem is insufficient production. Increasingly, it is not. It’s interconnection.

No company understands this tension better than Berkshire Hathaway Energy (BHE), part of the Berkshire Hathaway (BRKA) group. With roughly $380 billion on its balance sheet and decades of utility operating experience, it should be positioned to capitalize on rising baseload demand. Instead, it suspended three major geothermal projects near California’s Salton Sea: Black Rock, Morton Bay, and Elmore North.

That decision sits with Greg Abel who recently succeeded Warren Buffett as CEO at Berkshire Hathaway.

Abel is not a capital allocator removed from operations. He is a utility operator. Raised in Edmonton and a 1984 graduate of the University of Alberta, he began at PricewaterhouseCoopers before joining CalEnergy (a geothermal power producer) in 1992. After CalEnergy acquired MidAmerican and Berkshire took control, Abel rose to lead what became Berkshire Hathaway Energy. He oversaw large scale wind, solar and grid investments and has effectively run Berkshire’s non insurance operations since 2018.

His first shareholder letter in February 2026 emphasized capital discipline over speculative expansion. Producing power without a secure path to deliver it destroys returns. The Salton Sea suspension reflects that logic.

The issue was not geology or drilling capability. It was transmission interconnection delays, permitting complexity, and the absence of firm offtake agreements backed by the state. California’s grid is congested. The Salton Sea sits at a transmission dead end. Moving large volumes of incremental power to Los Angeles or Phoenix would require multi billion dollar line upgrades. No single merchant developer wants to underwrite that risk.

Utilities remember the telecom fiber buildout of the late 1990s. Companies such as Global Crossing and WorldCom spent more than 500 billion dollars laying fiber. Only a fraction was utilized. The industry collapsed under debt tied to unused infrastructure. Today’s grid operators see echoes of that cycle in the AI buildout and are keen not to repeat the same mistakes.

Time horizons are misaligned. AI servers depreciate over five years. Transmission towers are designed to last half a century or more. Utilities cannot justify ten billion dollar line investments for customers whose business models may reset within a decade, or an AI boom that may become a bust.

Under current regulation, if a large data center fails, stranded transmission costs often flow back to residential ratepayers. Politically, that is untenable.

There is also concern about circular financing. Capital flows from large technology firms into AI ventures, which then consume cloud capacity from the same firms, which then use that demand to justify further infrastructure buildout, looks a lot like a house of cards.

Policy has started to respond. The 2025 Ratepayer Protection Pledge under the Trump administration effectively told hyperscalers to fund their own grid impacts. Data centers increasingly sign take or pay contracts, committing to pay for reserved capacity whether they use it or not. Letters of credit and cash deposits now cover up to half of certain infrastructure costs. Risk is shifting from regulated utilities to technology companies.

In parallel, some operators are building a shadow grid. Behind the meter natural gas plants and small modular reactor proposals bypass the public system altogether.

Transmission strategy is also evolving toward smaller high voltage direct current links that can scale incrementally rather than sinking capital in to multi state megaprojects.

All of this reframes geothermal’s role in the AI era.

AI workloads require continuous, high density power. Hyperscalers such as Google, Microsoft and Amazon seek 24/7 carbon free energy, which makes cloud infrastructure easier to sell because their customers account for indirect emissions embedded in their cloud usage.

Solar and wind suffer from intermittency that necessitates expensive battery storage or natural gas backup. Far from ideal. It translates to a low capacity factor (the ratio of actual energy output to potential output). For solar this is ~25% as its unusable at night, while wind is weather dependent, so closer to ~35%. In other words, to generate the same annual energy as a 1 GW nuclear plant (which runs continually and has close to 100% capacity factor), a developer would need to build four 1 GW solar farms or three 1 GW wind farms.

Why not just use nuclear power, which is green and delivers reliable baseload power? Well, it has its drawbacks. It’s expensive to build, projects take a very long time to deliver and there is also the uranium feedstock issue, not to mention the radiation risk (think Chernobyl and Fukushima Daiichi nuclear disaster).

Is there another solution?

Geothermal offers baseload reliability, close to 100% capacity factor, without nuclear’s waste or risk profile. It is also relatively easy and cheap to build.

This video-graphic demonstrates how geothermal works (simple, clean and promises unlimited 24/7 power):

Traditionally, geothermal was a “geographic lottery” prize, available only in rare spots where tectonic activity brought steam or hot water near the surface. However, the advent of Enhanced Geothermal Systems (EGS) has changed the narrative. By borrowing horizontal drilling and hydraulic fracturing techniques from the oil and gas industry, companies can now “create” geothermal reservoirs in hot rock where fluid was previously absent and it can increasingly be sited closer to where the power is actually needed.

EGS transforms geothermal from a niche resource into a scalable, high-tech utility.

On paper, that makes geothermal well suited for data centers.

So why is Greg Abel not powering ahead with his geothermal infrastructure?

The Salton Sea projects represented 357 megawatts of potential carbon free capacity. They sit within a Known Geothermal Resource Area with highly saline brines rich in lithium. But electricity sales into a congested California grid offer modest regulated returns. Lithium extraction from the same brine offers commodity upside.

BHE is pivoting toward Direct Lithium Extraction. The brines beneath the Salton Sea may hold enough lithium to supply roughly 40 percent of U.S. battery demand. Lithium can be shipped globally. Electricity cannot. By prioritizing minerals over electrons, BHE shifts from a regulated return profile to a higher return commodity model while side-stepping immediate transmission bottlenecks.

This highlights a bifurcation taking place within the geothermal industry.

Lets consider the complex corporate structure of Berkshire Hathaway.

BHE subsidiaries include NV Energy (Nevada) and PacifiCorp (which serves Utah, Wyoming, and Oregon). In these states, the role of BHE is primarily that of the Utility Provider, not the Merchant Developer. In the world of regulated utilities, BHE makes a guaranteed rate of return by building and maintaining the “wires”, the transmission lines. From a corporate strategy standpoint, it is often more profitable for BHE to let third-party developers (like Fervo Energy or Ormat Technologies (ORA)) take the massive technical and financial risk of drilling new geothermal wells. Once those developers find heat, BHE signs power purchase agreements and earn regulated returns on transmission and distribution assets. The capital at risk profile is very different.

By contrast, the Salton Sea projects are a “Merchant” play for BHE. Here, they own the land and the mineral rights. They aren’t the utility; they are the miners and the generators. In California, overlapping federal, state and local jurisdictions extend timelines and amplify uncertainty. The suspensionof its projects in California reflects a cold calculation: until the state fixes the interconnection “final boss” and creates a financial floor for lithium-integrated geothermal, BHE would rather sit on the sidelines than bleed cash into a congested grid.

Is Geothermal Power A Good Investment?

The broader lesson from this narrative is structural. Transmission has become the scarce asset. Heat in the ground is less binding than copper in the air. Said differently, grid ready sites command a premium over heat ready sites.

The suspension at the Salton Sea is not a verdict against geothermal. It’s a case study in capital discipline under physical constraint. Under Abel’s leadership, Berkshire Hathaway is signaling that it will not fund generation growth without synchronized transmission and contracted demand.

As AI pushes electricity demand higher, the winners will not simply be those who find a source of reliable power. They will be those who secure deliverability, align incentives across regulators and customers, and structure contracts that protect against stranded assets.

Geothermal remains technically attractive for the data center era. Whether it scales depends less on drilling success than on whether the grid and regulatory framework evolve quickly enough to connect supply with demand.

As an investor, if someone pitches you a geothermal project as a ‘get rich quick’ investment thesis, don’t jump in without asking all the right questions first! Alternatively, trust someone who knows the energy space exceptionally well: perhaps investing in Greg Abel at Berkshire Hathaway is a safer play in the energy space.

On a side note, Berkshire has restarted share repurchases under its long-standing intrinsic value framework. The company disclosed that it began buying back both Class A and Class B shares on 4 March 2026, ending a pause that had lasted roughly 18 months. As always with Berkshire, there is no preset amount or schedule. The filing makes clear that repurchases can occur in the open market, through negotiated block trades, or via 10b5-1 plans, and that activity can stop at any time. The approach remains consistent with the policy Buffett put in place: buy shares only when the price is below a conservative estimate of intrinsic value.

The restart comes as Greg Abel assumes full responsibility for capital allocation. At the start of 2026 he took over the day-to-day decision making around how Berkshire deploys its large cash balance. In his first shareholder communication as CEO he reiterated that buybacks remain an important tool, but only when they clearly increase per-share value. Berkshire has historically used repurchases opportunistically rather than mechanically, often stepping in when the stock’s price to book ratio drifts toward levels management views as a discount to intrinsic value. The absence of buybacks over the past six quarters reflected Buffett’s reluctance to repurchase shares at higher valuation levels.

Abel has also reinforced the message with personal capital. In early March he reportedly bought around 15 million dollars of Berkshire stock, coinciding with the restart of the repurchase program. It follows a much larger purchase in 2022, when he spent roughly 68 million dollars acquiring 168 Class A shares. The timing is clearly meant as a signal of alignment and conviction. For shareholders, the broader implication is that buybacks will remain a flexible capital allocation lever - Berkshire Hathaway has been looking for an elephant sized acquisition for deployment of its $380 bn cash pile, maybe it just needed to look in the mirror!

Share

Long before became immortalized in the movie The Big Short, he was dissecting accounting gimmicks on MSN Money during the Dot-com era. Today, he’s raising a different alarm. Tech companies, he argues, are stretching depreciation schedules and excluding stock-based compensation as if it were imaginary. He argues that parts of the technology sector are overstating economic reality which explains their astronomical valuations.

On that point, he’s not wrong. The distortions are very real.

Where things get more interesting is the next part of his warning: that passive investing is amplifying these distortions, storing up a “fat tail” event that could rival 2000 or 2008.

That’s the part worth examining.

Something structural has certainly changed, but has it increased or decreased the probability of a market panic?

Nassim Taleb’s critically aclaimed book, ‘The Black Swan’, focuses on the extreme impact of rare and unpredictable outlier events. The types of occurence that cause stock market crashes.

But what used to be a dangerous black swan, may now be relatively tame.

The distinction being made by Peter Lynch in this quote matters. The task of an intelligent investor is to identify a high quality business within the market and allocate capital to it. The performance of the rest of the market becomes academic.

But that is exactly what’s changed.

Participants are no longer investing in single stocks. They are allocating capital to indexes instead.

The S&P 500 today behaves less like a means to measure the performance of a basket of 500 businesses and more like a securitized asset in its own right. It’s become a liquidity vehicle. Some may argue that it is no longer an index; it’s become the product. It’s a single, scalable instrument that absorbs savings flows from retirement accounts, sovereign funds, and systematic allocators.

They call it “Passive” investing, but this is a misnomer. People are actively allocating capital to this new synthetic asset. It’s not investing based on unit economics, it becomes a momentum strategy. If enough people buy a momentum driven synthetic asset, there is no limit to where it’s price may be driven, just look at Bitcoin as a perfect example. “Intrinsic value? What’s that?”

Most 401(k) contributions buy the index automatically. No one asks whether software multiples are stretched or whether margins are peaking. The allocation is mechanical. Market-cap weighted. Indifferent. That mechanical bid changes everything.

The rise of passive investing has altered how capital behaves under stress. Capital doesn’t flee the market, it moves within it.

In previous decades, a “panic” meant everyone ran for the exits simultaneously, liquidation was universal, and correlations spiked as everything moved together. Today, investors are blind. They invest passively. They don’t analyze corporate fundamentals and they don’t track macro-economics. Capital no longer evaporates. It rotates. It is algorithmically reallocated within the market. It’s a cold, mathematical process that prevents the “feedback loop of fear” that defined panics such as those seen in 1929, 1987, 2000 and 2008.

New passive funds continually flow into the system by virtue of 401(k) plans, and capital isn’t withdrawn from the system, it’s redistributed according to a formula. Passive flows act as a structural bid. Not because investors are confident, but because the system is automated. This creates a permanent synthetic floor for the index.

Recently, some software stocks were crushed by 50% or more while the S&P500 barely flinched.

Passive investing has become a shock absorber for the market, but it comes at a cost. To the casual observer watching headline indices or volatility metrics, it looks orderly. Beneath the surface, it’s carnage.

In 2008, a spike in the VIX reflected existential fear. Capital was leaving equities. The plumbing was in question. Today, volatility spikes can coexist with modest index drawdowns. They often reflect localized stress, hedging flows, or sector-specific liquidation rather than systemic exit.

Liquidity has pooled at the index level, so a stock that drops out of an index or loses its “momentum” status can experience a liquidity vacuum, leading to those 50% drops witnessed in the recent sell off. The “panic” hasn’t disappeared; it’s just been localised to the stocks that the index no longer loves.

In that environment, dispersion widens. Passive flows create a "protected class" of index constituents. The rest of the market, those not in the index, increasingly are left to fend for themselves.

There’s another layer of complexity: concentration.

If mega-cap stocks dominate index weight, they anchor index stability. Weakness in smaller sectors barely registers in headline metrics. In 2000, when tech collapsed, it dragged the index with it. Today, unless the largest names crack simultaneously, concentration can mask severe internal damage.

This creates a dangerous optical illusion.

A 10% drawdown at the index level can hide 50% drawdowns across large swaths of growth equities. Risk feels contained. In reality, it’s concentrated.

So is Burry right?

If his argument is that accounting distortions exist, yes. If his argument is that passive investing has altered market structure, absolutely. But the direction of the effect may be more nuanced than a simple “bigger crash is coming.”

Passive flows appear to dampen systemic collapses by preventing wholesale capital flight. They turn panics into rotations. They provide a synthetic floor at the aggregate level. They don’t eliminate risk, they redistribute it.

The market absorbs stress differently than it did in 2000 or 2008. Broad index collapses are less frequent. Dispersion across individual stocks is higher. Risk is increasingly concentrated in a small group of large companies as more capital tracks the same index.

Instead of broad multi-year index drawdowns, we may experience sharper, more violent single-name collapses. Instead of universal fear, we get sectoral panic events. Instead of everything falling together, dispersion explodes.

What do you think? Have your say in the comments box below.

Share