A closer look at
the buyback.

Stock-based compensation is a cash expense paid with a lag. Until you adjust for it, the headline buyback yield is an accounting illusion.

→  Good morning

When a company announces a fifty billion dollar share repurchase program, the headline implies that capital is being returned to shareholders. However, the math is not that simple. If the same company is issuing new shares to employees as compensation at a rate comparable to what it is buying back, the net effect on share count is negligible. What looks like a capital return functions as dilution management.

Stock-based compensation is the most important topic in corporate finance that most investors still get wrong. It is treated as a non-cash expense, which is technically true, but economically meaningless. The cash comes out later, when the company buys back shares to prevent the dilution that SBC creates. The accounting timing hides the real economics. Stock-based compensation is a cash expense, paid with a lag, using the buyback program as the payment mechanism.

Monday's piece on Texas Instruments praised its buyback program for retiring 60 percent of shares outstanding since 2004. The number is real and the discipline is real. Part of what makes TI remarkable is that its SBC runs at roughly one to two percent of revenue, small enough that the retirement math is dominated by buybacks rather than offset by dilution. Most large-cap American companies cannot say the same.

The hidden connection

Why buybacks don't always reduce shares.

The foundational empirical work comes from Daniel Bens, Venky Nagar, Douglas Skinner, and Franco Wong's 2003 paper in the Journal of Accounting and Economics, titled “Employee stock options, EPS dilution, and stock repurchases.” They found that executives increase the level of their firms' stock repurchases under two conditions: when the dilutive effect of outstanding employee stock options on diluted EPS is rising, and when reported earnings are tracking below the level required to sustain the historical EPS growth rate. Importantly, the paper finds no association with actual option exercises, which strengthens the earnings-management interpretation. The buyback is being used to manage reported diluted EPS, not as an independent view on whether the stock is attractive to own.

This distinction matters because the mechanism distorts the meaning of the buyback. An allocation decision based on intrinsic value has different economics from a mechanical offset designed to prevent reported share count from drifting upward. The first creates value when the share price is below intrinsic value. The second destroys value when the share price is above intrinsic value, because the company is forced to retire shares at whatever price happens to be trading, not at a price that reflects opportunistic timing.

Kathleen Kahle, writing in the Journal of Financial Economics in 2002, sharpened the picture further. Her paper, “When a buyback isn't a buyback: Open market repurchases and employee options,” found that firms are more likely to announce repurchases when executives hold large numbers of options outstanding and when employees hold large numbers of options currently exercisable. Once the decision to repurchase is made, the amount repurchased is positively related to total exercisable options held by all employees, not just by management. The market appears to recognize the motive. Repurchase announcements at firms with high levels of non-managerial options draw less positive reaction. The treasury function has become, in part, a mechanism for funding employee compensation through the public float.

The math

What the adjusted yield reveals.

Michael Mauboussin and Dan Callahan, again in Morgan Stanley's Counterpoint Global Insights, have advanced the most useful framework for thinking about this in practice. The concept is “adjusted buyback yield.” The calculation takes gross buybacks and subtracts shares issued for stock-based compensation, equity-funded acquisitions, and equity offerings. What remains is the net capital return to shareholders, measured as a percentage of market cap. For most large-cap technology companies, the adjusted yield runs materially below the headline.

Consider a company that announces $20 billion in annual buybacks against a $2 trillion market cap. The headline yield is 1.0 percent. Now consider that the same company issues $12 billion in stock-based compensation each year. The buyback, net of SBC dilution, is $8 billion. The adjusted yield is 0.4 percent. The reported number overstates the actual capital return by a factor of roughly two and a half.

The hyperscaler pattern illustrates the point. Alphabet's stock-based compensation ran at roughly six to seven percent of revenue in 2023 and 2024, on the order of twenty-two to twenty-three billion dollars against revenue approaching three hundred fifty billion. Meta and Microsoft sit on a similar spectrum, with SBC programs that consume a meaningful fraction of operating cash flow. The buyback programs at each of these companies are real, and so is the economic pressure from SBC that offsets them. The observation is about how headline buyback numbers should be read, not a criticism of any of the underlying businesses, which are among the most productive generators of free cash flow in the global economy.

The contrast with lower-SBC businesses is instructive. Texas Instruments and the industrials covered in Monday's piece run SBC at roughly one to two percent of revenue. Each dollar spent on repurchases there translates, more or less directly, into a dollar of retained shareholder value. The high-SBC programs cannot claim the same arithmetic.

The Gap

1.0% → 0.4%

Net of stock-based compensation

A $20 billion buyback on a $2 trillion market cap looks like a 1.0 percent capital return. Net of $12 billion in SBC issuance, the net return to shareholders is $8 billion, a 0.4 percent yield. Investors who read headline figures without adjustment will price the company at roughly two and a half times its actual capital-return discipline.

The incentive

Why this keeps happening.

The persistence of this accounting divergence has an incentive explanation. Daniel Bergstresser and Thomas Philippon, in a 2006 Journal of Financial Economics paper titled “CEO incentives and earnings management,” examined the relationship between executive compensation structure and the use of discretionary accruals to manipulate reported earnings. Their finding was direct. Accruals-based earnings management is more pronounced at firms where CEO compensation is more closely tied to the value of equity holdings. During years of high accruals, CEOs exercise unusually large amounts of options and insiders sell substantial quantities of shares.

The direct subject of the Bergstresser and Philippon paper is accruals, not buybacks. But read alongside Bens and Kahle, the picture is consistent. Executives whose compensation is heavily weighted toward equity have documented incentives to manage reported earnings. Buybacks that mechanically offset SBC are one of the available levers. The program props up reported EPS, supports the share price, and allows equity-based compensation to vest at favorable prices. The shareholder benefit of this configuration is limited. The executive benefit is substantial.

The governance implication is significant. A board that approves a buyback program without explicitly evaluating it against a valuation framework is approving dilution management, not capital allocation. A board that does evaluate it against a framework will, in most cases, conclude that the opportunistic approach described in yesterday's ROIC piece is superior. The most thoughtful corporate boards have begun disclosing SBC-adjusted buyback metrics alongside headline figures. Most boards have not.

The public disclosure question is under-appreciated. A company that reports both headline and adjusted buyback yields in its investor materials is making a choice about transparency. A company that reports only the headline is making a choice too. Neither choice is neutral. Readers should treat adjusted-yield disclosure as a signal about how a board thinks about capital allocation.

The discipline

How to read the numbers.

Each of the four papers that anchor this memo points in the same direction. Bens and Wong show that buybacks are timed to manage reported diluted EPS rather than to buy undervalued stock. Kahle shows that the size of a repurchase scales with the options exercisable across the entire employee base. Bergstresser and Philippon show that equity-weighted CEO compensation is associated with more aggressive earnings management generally. Mauboussin and Callahan supply the framework for correcting the headline number. The analytical conclusion is not subtle.

The practical framework that follows has four components when evaluating any buyback program.

i.

Subtract the dollar value of stock-based compensation from gross repurchases to arrive at the net capital return to shareholders.

ii.

Compare net buybacks to free cash flow to assess whether the program is a meaningful capital return or a marginal offset.

iii.

Evaluate whether repurchase timing correlates with valuation discipline or with mechanical dilution management.

iv.

Weight adjusted yield over headline in any valuation work, and look for boards that disclose both.

Most investors do not perform any of this analysis. Most research reports quote headline buyback yields without adjustment. The gap creates opportunity for allocators who do the work. Companies with strong adjusted buyback yields tend to be underpriced relative to their headline peers. Companies with weak adjusted yields tend to be overpriced relative to what their actual capital-return discipline justifies. The spread between the two groups has been one of the more stable return anomalies of the past decade.

In Closing

The accounting treatment of SBC as non-cash was the original sin. The buyback as quiet funding mechanism is the consequence. Both are now part of how the largest companies in the world return capital, and the adjustment is not optional if you want to read the numbers correctly.

The week has moved from divestitures to spinoffs to return on invested capital to buyback dilution. Each topic is a piece of the same puzzle. Capital allocation is easier to discuss than to measure, and the measurement problems are where most investors get it wrong.

Tomorrow closes the week with the synthesis: the eight CEOs from Thorndike's The Outsiders whose track records prove the framework works in practice, and what their example demands of the companies we own today.

Marques

Filed from Chanhassen  /  04.23.26  /  07:30 CT