Using this approach (correctly, I hope, as disclosures were a bit different than those cited), I estimate MSFT's true SBC to be $18B and $13.9B in 2025 and 2024 respectively, vs. reported SBC of $12B and $10.7B.
Your 2025 analysis looks fine, but the 2024 number looks too low. Either way, you can see the significant difference in reported SBC versus reality. It is material.
Most investors have no idea, but once you see it, you can't unsee it.
One way I often ask investors and analysts to think about this issue is to assume the company in question was taken private. Since the company would most likely no longer be paying employees in shares of any kind, how do they keep employees who used to receive SBC? Simple. They would have to pay them in cash; else, these employes would leave. At this point, cash flows from operating activities would decline as base salary and bonuses are accounted for as operating activities. Thus, free cash flow would decline. Why should it be any different when employees are paid in shares and the company spends cash to repurchase shares to offset dilution?
I recently met the founder CEO of a very successful public company. Amonsg other things, we discussed stock-based compensation.
The business was highly cash generative, so I asked why SBC was still necessary.
He shrugged. “It’s standard practice. It aligns management with shareholders.”
The explanation felt thin, more habitual than considered.
“Let’s go back,” I said. “Before the IPO, after founding the business when cash was tighter, how did you fund the business?”
“Friends and family," he said.
“Did you dilute them every year by issuing yourself more equity?”
He looked surprised. “Of course not. I was grateful. They backed me early. They were people close to me and it would have felt like a breach of trust.”
“So why do it now?” I asked. “Why dilute shareholders just because they’re public and you don’t know them?”
He paused. No real answer.
His thinking had shifted. Now he saw this through a different lens. Whether that leads to a change of approach is yet to be seen
Really like the in-depth exploration here. I agree that that financial disingenuity should be addressed and accounted for by investors.
However, a key point that I see overlooked in the presentation as a long term tech insider in a more "traditional" company (where we're not paid RSUs so my bias may be limited here), is that these golden handcuffs have definitely worked for MAG7 and adjacent companies.
They tie up the best employees with the most transferable skills who have an outsized impact on the bottom line. As a senior dev, the arithmetic just doesn't make sense to move to a startup or a newer company where the long shot compensation may not materialize. In companies like ours, over the last decade, it has been impossible to find good engineers and researchers since if we don't pay them sufficiently imo. It's suboptimal capital allocation to not hire the best and haemorrhage business lines to Amazon.
I think it has clearly been tremendously profitable for MAG7 and adjacent companies to hire the best in the industry. With ROICs exceeding 40% for decades, large shareholders have seen tremendous growth in their investments.
Without the allure of these salaries, I think we'll see a repeat of the Yahoo and Google saga a lot more. For Google, it's safer and much more profitable to have a monopoly on the "input" human capital pipeline.
If MAG7 is not able to keep a hold on this pipeline, I believe these companies would not be nearly as formidable. Private Equity has tried the cost cutting playbook for decades and the results have been clearly suboptimal.
Your comments in relation to the rationale for paying SBC are all valid, yet they raise serious questions.
Pure economics dictates that labour should flow to where it may be applied most productively as the most successful businesses paying cash remuneration can afford to pay the highest salary.
But using equity as a form of currency to attract employees defeats this mechanism. It means that the Mag7 are locking up talent that could arguably be deployed to better effect elsewhere. Said differently, it limits the ability of lesser known innovators and entreprenuers from securing the labour that they need to succeed. Is that right?
Think about this. Despite pouring tens of billions of dollars into R&D for artificial intelligence, having over a decade head start, and maintaining a dedicated workforce of over 1,500 developeres focused exclusively on internal AI projects, Microsoft found itself trailing behind OpenAI, a relative newcomer in the field. This situation led to a critical moment following the launch of ChatGPT by OpenAI when CEO Satya Nadella questioned the effectiveness of Microsoft Research, sending a famous internal memo asking, ‘Why do we have Microsoft Research at all?’ Just imagine what those developers could have achieved elsewhere outside of the bureaucratic and inneficient culture at Microsoft! But they were all 'golden-handcuffed' through stock based comp. Is that right?
In any event, it is important to state that our podcast was not arguing against stock based comp per se, but more the way companies manage it and the way investors mis-read it when valuing a company.
Amazon is a good illustration of a more rational approach. Grants are tied to a dollar value rather than a fixed number of shares. That means the economic intent of the compensation remains stable regardless of what the share price does before vesting. Just as importantly, Amazon allows dilution to appear directly in the share count. Investors can see it clearly and decide whether the cost is justified by the company’s performance.
Meta represents almost the opposite approach. Grants are issued as a fixed number of shares, which means the realised value can expand dramatically if the stock appreciates before vesting. On top of that, the company spends very large sums on buybacks that effectively conceal the dilution created by those grants. The $42 billion spent on repurchases in 2025 is a good example. This labour cost does not appear on the income statement or in operating cash flows and FCF calculations, so does not figure in the valuation calculation of many investors. This is what we were pointing out. Their valuation models are deeply flawed.
Using this approach (correctly, I hope, as disclosures were a bit different than those cited), I estimate MSFT's true SBC to be $18B and $13.9B in 2025 and 2024 respectively, vs. reported SBC of $12B and $10.7B.
Your 2025 analysis looks fine, but the 2024 number looks too low. Either way, you can see the significant difference in reported SBC versus reality. It is material.
Most investors have no idea, but once you see it, you can't unsee it.
One way I often ask investors and analysts to think about this issue is to assume the company in question was taken private. Since the company would most likely no longer be paying employees in shares of any kind, how do they keep employees who used to receive SBC? Simple. They would have to pay them in cash; else, these employes would leave. At this point, cash flows from operating activities would decline as base salary and bonuses are accounted for as operating activities. Thus, free cash flow would decline. Why should it be any different when employees are paid in shares and the company spends cash to repurchase shares to offset dilution?
I recently met the founder CEO of a very successful public company. Amonsg other things, we discussed stock-based compensation.
The business was highly cash generative, so I asked why SBC was still necessary.
He shrugged. “It’s standard practice. It aligns management with shareholders.”
The explanation felt thin, more habitual than considered.
“Let’s go back,” I said. “Before the IPO, after founding the business when cash was tighter, how did you fund the business?”
“Friends and family," he said.
“Did you dilute them every year by issuing yourself more equity?”
He looked surprised. “Of course not. I was grateful. They backed me early. They were people close to me and it would have felt like a breach of trust.”
“So why do it now?” I asked. “Why dilute shareholders just because they’re public and you don’t know them?”
He paused. No real answer.
His thinking had shifted. Now he saw this through a different lens. Whether that leads to a change of approach is yet to be seen
Brilliantly stated.
Can I give you some feedback? I really enjoy your content but is there anyway you can get some glare free glasses? Hope that doesn't sound too weird.
Thank you for the feedback. I am delighted that you enjoyed it
Really like the in-depth exploration here. I agree that that financial disingenuity should be addressed and accounted for by investors.
However, a key point that I see overlooked in the presentation as a long term tech insider in a more "traditional" company (where we're not paid RSUs so my bias may be limited here), is that these golden handcuffs have definitely worked for MAG7 and adjacent companies.
They tie up the best employees with the most transferable skills who have an outsized impact on the bottom line. As a senior dev, the arithmetic just doesn't make sense to move to a startup or a newer company where the long shot compensation may not materialize. In companies like ours, over the last decade, it has been impossible to find good engineers and researchers since if we don't pay them sufficiently imo. It's suboptimal capital allocation to not hire the best and haemorrhage business lines to Amazon.
I think it has clearly been tremendously profitable for MAG7 and adjacent companies to hire the best in the industry. With ROICs exceeding 40% for decades, large shareholders have seen tremendous growth in their investments.
Without the allure of these salaries, I think we'll see a repeat of the Yahoo and Google saga a lot more. For Google, it's safer and much more profitable to have a monopoly on the "input" human capital pipeline.
If MAG7 is not able to keep a hold on this pipeline, I believe these companies would not be nearly as formidable. Private Equity has tried the cost cutting playbook for decades and the results have been clearly suboptimal.
Your comments in relation to the rationale for paying SBC are all valid, yet they raise serious questions.
Pure economics dictates that labour should flow to where it may be applied most productively as the most successful businesses paying cash remuneration can afford to pay the highest salary.
But using equity as a form of currency to attract employees defeats this mechanism. It means that the Mag7 are locking up talent that could arguably be deployed to better effect elsewhere. Said differently, it limits the ability of lesser known innovators and entreprenuers from securing the labour that they need to succeed. Is that right?
Think about this. Despite pouring tens of billions of dollars into R&D for artificial intelligence, having over a decade head start, and maintaining a dedicated workforce of over 1,500 developeres focused exclusively on internal AI projects, Microsoft found itself trailing behind OpenAI, a relative newcomer in the field. This situation led to a critical moment following the launch of ChatGPT by OpenAI when CEO Satya Nadella questioned the effectiveness of Microsoft Research, sending a famous internal memo asking, ‘Why do we have Microsoft Research at all?’ Just imagine what those developers could have achieved elsewhere outside of the bureaucratic and inneficient culture at Microsoft! But they were all 'golden-handcuffed' through stock based comp. Is that right?
In any event, it is important to state that our podcast was not arguing against stock based comp per se, but more the way companies manage it and the way investors mis-read it when valuing a company.
Amazon is a good illustration of a more rational approach. Grants are tied to a dollar value rather than a fixed number of shares. That means the economic intent of the compensation remains stable regardless of what the share price does before vesting. Just as importantly, Amazon allows dilution to appear directly in the share count. Investors can see it clearly and decide whether the cost is justified by the company’s performance.
Meta represents almost the opposite approach. Grants are issued as a fixed number of shares, which means the realised value can expand dramatically if the stock appreciates before vesting. On top of that, the company spends very large sums on buybacks that effectively conceal the dilution created by those grants. The $42 billion spent on repurchases in 2025 is a good example. This labour cost does not appear on the income statement or in operating cash flows and FCF calculations, so does not figure in the valuation calculation of many investors. This is what we were pointing out. Their valuation models are deeply flawed.
Probably one of the best breakdowns of stock based compensation I've heard hands down! Great Job.
Thank you for your kind words. They mean a lot.