What Proxy Statements Reveal About Buybacks and Dividends

Somewhere past the Summary Compensation Table in nearly every large company's proxy statement sits a short section that tells you more about a dividend's future than the dividend history page ever will: the list of metrics the board actually pays executives to hit. Total shareholder return, sometimes. Free cash flow, occasionally. And, more often than most income investors realize, earnings per share — a number a company can move without earning a single additional dollar of profit, simply by buying back enough stock to shrink the denominator.

Most dividend research stops well short of this page. Investors check the payout ratio, the free cash flow coverage, maybe the cash position — all worth checking, all covered elsewhere on this site — and then close the filing before reaching the compensation section at all. That's a gap, because the comp table answers a question none of those other numbers can: what is management actually paid to prioritize with the cash the business generates? A payout ratio tells you what a company can afford. The proxy tells you what its executives are being rewarded to choose.

This piece walks through four things a careful read of a proxy statement's compensation section reveals about where a dividend actually sits in a company's capital allocation priorities — and one real example, IBM, of what happens when a board admits its own pay metric had a buyback problem.

What the Comp Table Actually Rewards

Start with the Compensation Discussion and Analysis section, not the summary table itself. The CD&A spells out which financial metrics determine the annual bonus and the multi-year equity awards, and large-company boards lean on a fairly narrow set: revenue growth, operating margin, free cash flow, total shareholder return, and earnings per share, in some combination. The first four are reasonably resistant to financial engineering. EPS is not. A company that reduces its share count by 5% in a year gets roughly a 5% lift to EPS with zero improvement in the underlying business — no new customers, no better margins, nothing a competitor would need to worry about.

IBM's own proxy statements are a useful, disclosed example of exactly this tension. For years, the company's long-term incentive plan measured performance share units against operating (non-GAAP) EPS and free cash flow, with targets set at the start of each three-year period in a way that explicitly accounted for the effect IBM's own buyback program would have on the EPS figure — the company was, in effect, pre-loading the credit its repurchases would generate into the target itself. After sustained investor pushback over multiple proxy seasons, IBM changed course: it committed to stripping the impact of share repurchases out of the EPS number used to calculate incentive payouts, and added a relative total shareholder return modifier so the award couldn't be earned purely by outrunning a static internal bar. That's a company's own governance disclosures admitting, in effect, that its pay plan had been rewarding a lever unrelated to operating performance.

For an income investor, the read-through is direct. If a management team is paid on EPS growth and buybacks are the fastest, most controllable way to hit that number, the dividend is competing against a genuinely motivated internal incentive to spend cash on repurchases instead. That doesn't make buybacks bad — shareholder yield treats dividends and buybacks as two channels of the same return, and a well-timed repurchase at a depressed price can create real value. But a comp plan that only pays for EPS tells you the board hasn't asked management to weigh the dividend against the buyback at all. It's simply told them to hit a number, and left the capital allocation decision to whichever lever is fastest.

What the Ownership Guidelines Leave Out

Most large-company proxies also disclose stock ownership requirements — a rule requiring the CEO to hold shares worth some multiple of base salary, often five to ten times, funded from vested awards rather than open-market purchases. These guidelines genuinely align an executive's wealth with the stock price. What they say nothing about is how that price should be supported. An executive sitting on $30 million of required stock benefits identically whether the company grows a modest dividend for a decade or spends the same cash buying back shares at whatever multiple the market happens to be offering that quarter. Price appreciation is price appreciation, regardless of the capital allocation policy that produced it.

I'm less certain than I'd like to be about how much weight to put on ownership requirements alone as a signal, because a genuinely well-aligned executive can still have perfectly defensible reasons to prefer buybacks over dividend growth in a given year — a depressed valuation, a debt-paydown priority, a tax consideration for the shareholder base. Ownership size tells you the executive's incentives point roughly the same direction as yours. It doesn't tell you which lever they'll reach for when free cash flow shows up. For the broader mechanics of what a proxy statement covers beyond compensation — board composition, shareholder votes, governance structure — David Park's rundown of the document is the right starting point; this piece is narrower, focused specifically on what the pay structure implies about capital return.

What the Capital Allocation Priorities Language Reveals

Some CD&A sections go further than listing metrics — they state, in prose, how the board ranks dividends against buybacks and debt reduction. 3M is a recent, well-documented example of a company recalibrating that stated priority in real time. Following the spin-off of its health care business into Solventum, 3M cut its quarterly dividend by 53.6%, from $1.51 to $0.70 per share, beginning with the June 2024 payment — disclosed in an 8-K filed March 8, 2024 — and set a target payout ratio of roughly 40% of adjusted free cash flow going forward. That target ratio is exactly the kind of explicit capital allocation commitment that shows up again in the following year's proxy statement, giving shareholders a specific number to hold management against rather than a vague assurance that the dividend "remains a priority."

That specificity matters more than it might seem. A board that states a target payout ratio has given itself, and its shareholders, a number to check every subsequent quarter against — the same kind of coverage tracking this site has covered for reading 10-Q disclosures. A board that only says the dividend is "important to the company's capital allocation strategy," with no ratio attached, has committed to nothing checkable.

Here's why the distinction is worth a numerical example, not just a preference. Take a hypothetical $10,000 position yielding 3% today — $300 a year — growing its dividend 6% annually for fifteen straight years. By year fifteen, annual income has grown to roughly $678. Now take the same starting position, growing at the same 6% for four years, then absorbing a 50% cut in year five — roughly the shape of 3M's reduction — before resuming 4% annual growth from the smaller base. By year fifteen, that position pays out closer to $280 a year. Same starting yield. Same first four years. The cut alone accounts for more than half the year-fifteen income gap, and the smaller base keeps compounding against you for every year after it.

What Say-on-Pay Results and Shareholder Proposals Show

The proxy's advisory "say on pay" vote is worth a glance even though it doesn't bind the board. Support in the high 90s most years signals a shareholder base that's comfortable with how pay is structured, capital allocation incentives included. A meaningful multi-year decline, or a result below roughly 80%, is a visible signal that a growing share of institutional holders are uneasy with the pay design — and proxy advisory firms flag low support publicly, which tends to pressure the compensation committee to revisit the metrics the following year. IBM's own metric change didn't happen in a vacuum; it followed years of investor engagement that a declining say-on-pay trend would have made visible well before the plan actually changed.

Shareholder proposals occasionally target capital allocation directly, asking a board to commit to a specific payout ratio or to justify buyback timing against valuation. Most such proposals fail. But the percentage of votes one draws, even in defeat, tells you something a payout ratio never will: whether a meaningful slice of the ownership base is actively pushing the board toward the dividend side of the ledger.

None of these four signals — the comp metrics, the ownership guidelines, the stated capital allocation language, the say-on-pay trend — replaces the payout-ratio and free-cash-flow checks that should still anchor any dividend safety review. They add something those checks can't: a read on which direction management is actually being pushed, before the capital allocation decision shows up in a dividend announcement, an accelerated buyback authorization, or a proxy fight nobody saw coming three quarters out.

Key Takeaways

• The compensation section of a proxy statement, not just the payout ratio, reveals how executives are incentivized to weigh dividends against buybacks — EPS-based metrics reward share count reduction directly, while free cash flow and total shareholder return metrics don't.
• IBM's own proxy disclosures show the risk directly: operating EPS targets that accounted for buyback impact, followed by a stated commitment to strip repurchases out of the metric after years of investor pushback.
• A stated target payout ratio — like 3M's roughly 40%-of-adjusted-free-cash-flow commitment following its 2024 dividend cut — gives shareholders a checkable number; vaguer language about the dividend "remaining a priority" does not.
• Say-on-pay vote trends and shareholder proposals targeting capital allocation are early, publicly disclosed signals of investor pressure on pay-and-payout structure, often visible well before a payout decision changes.