Capital Allocation

Capital Allocation

Adobe returns essentially all of its cash to shareholders through buybacks, and the timing has run counter-cyclical: it spent the least at its 2021 price peak and the most — more than a full year's free cash flow — as the stock fell to multi-year lows, retiring about $45.7 billion of stock over the decade. It funds its AI response internally at 18% of revenue rather than by acquisition. The aborted Figma deal, which cost $1 billion for nothing, is the clearest evidence that the discipline is not guaranteed.

Sources: FY2025 Form 10-K, Consolidated Statements of Cash Flows and Stock Repurchase Program [1], operating-expense detail [2]; decade totals derived from FY2016–FY2025 cash-flow statements.

The Buyback Engine

Buybacks are the whole capital-return story. Over FY2016–FY2025, Adobe repurchased roughly $45.7 billion of its own stock — about 80% of the $56.8 billion of free cash flow it generated in the same span — and never paid a dividend. The cumulative effect sits on the balance sheet as $48.8 billion of treasury stock at cost, against which shares outstanding fell from 441 million to 413 million in the last year alone [3].

Source: FY2025 Form 10-K, Consolidated Statements of Cash Flows [4]; FY2016–FY2022 from prior-year cash-flow statements as reported.

The more useful pattern is the timing. Repurchases were smallest, relative to the cash coming in, in the years the stock was most expensive, and largest in the years it was cheapest. In fiscal 2021 the shares averaged about $561 and touched $688; Adobe spent $3.95 billion, roughly 57% of that year's free cash flow. In fiscal 2025 the shares averaged about $377, and it spent $11.28 billion [5] — 115% of free cash flow. In both fiscal 2024 and fiscal 2025 the buyback ran ahead of the cash generated, funded by drawing down the cash balance and by issuing $1.99 billion of new senior notes each year [6].

Sources: buybacks from FY2023–FY2025 Form 10-K cash-flow statements [7] and prior-year filings; average share price from market data; buyback/FCF derived.

The discipline is real but not absolute. Fiscal 2024 spending of $9.5 billion came at an average price near $525 — well above the fiscal 2022 and fiscal 2023 levels — so this is a company that leaned in as the multiple compressed, not one that waited for a bottom. The direction, though, is favorable for per-share compounding: retiring stock at roughly nine times free cash flow lifts free cash flow per share several points a year even if the business as a whole merely holds. Adobe's roughly $95 billion equity value is consistent with owner cash flow growing about 0% a year in perpetuity even on the conservative stock-comp-charged basis (an ~8.3% SBC-charged trailing free-cash-flow yield, down from a ~10.4% headline), yet because Adobe retires 4-7% of its shares each year by repurchasing at about 9x free cash flow, per-share owner free cash flow still compounds mid-single-digits on flat aggregate cash — so the stall the price embeds applies to the whole company, not to the owner of a share. As of November 2025, $5.90 billion remained under the $25 billion authorization the board approved in March 2024, which runs through March 2028 [8].

No Dividend

Adobe has never paid a cash dividend and states it does not anticipate paying one "in the foreseeable future" [9]. The entire return of capital therefore flows through repurchases. That choice is defensible for a business trading at a high-single-digit free-cash-flow yield — buybacks at that price are more accretive than a dividend, and they avoid committing to a payout the company would be reluctant to cut. The trade-off is that the value of the return is entirely a function of the price paid: there is no dividend floor, and a management team that repurchased stock indiscriminately at any multiple would be reducing value per share without it ever showing up as a reported loss. On the record so far, the price paid has trended the right way.

Build First, Buy Rarely

Away from buybacks, Adobe's default is to build. Research and development ran $4.29 billion in fiscal 2025, up from $3.47 billion in fiscal 2023, and has held at 18% of revenue throughout — this is where Firefly and the generative features discussed elsewhere were funded, without an acquisition [10].

Source: FY2025 Form 10-K, operating-expense detail; research and development held at 18% of revenue in each year [11].

The two exceptions to the build-first pattern define the range of the acquisition question.

Figma. In September 2022 Adobe agreed to buy Figma for approximately $20 billion, half cash and half stock — about the size of its entire fiscal 2024 and fiscal 2025 buybacks combined. Regulators in Europe and the UK objected, and in December 2023 the two sides terminated the deal. Adobe paid Figma a $1 billion termination fee, recorded in fiscal 2024 operating expenses and not tax-deductible [12]. The full deal-level cost is taken up in The M&A Record; what matters for capital allocation is that the same management that repurchases stock counter-cyclically was willing to pay a full-cycle price for a fast-growing rival, and only an antitrust intervention stopped it.

Semrush. The more recent deal is smaller and different in character. In April 2026 Adobe completed the acquisition of Semrush, a search-and-brand-visibility platform, for $1.87 billion, primarily in cash. About $1.25 billion of the price — two-thirds — was recorded as goodwill, and Adobe told investors the acquisition was not material to its consolidated results [13]. It slots into the marketing side of the business — the segment where search-engine and "generative-engine" optimization matter — and is a tuck-in, not a franchise bet. It is excluded from the fiscal 2026 guidance the company has issued.

The pattern that emerges: Adobe overwhelmingly builds, returns the rest through buybacks, and reaches for acquisitions rarely — but when it does reach, as with Figma, it has shown a willingness to pay prices that its own buyback discipline would never justify.

Stewardship and Alignment

The person setting this policy has been in place a long time. Shantanu Narayen has been chief executive since December 2007 and also chairs the board, a combined role balanced by a lead independent director [14]. Long tenure through the subscription transition is a point in management's favor; combining the chair and CEO roles is the standard governance caution, mitigated but not eliminated by the independent-director structure.

Alignment runs through pay, not ownership. Insider stakes are small: all directors and current executive officers together hold 803,767 shares — well under 1% of the company — with Narayen himself holding 438,975 [15]. The largest holders are the index complexes, Vanguard at 10.2% and BlackRock at 9.3% [16]. This is a professionally managed large-cap, not an owner-operator: management is incentivized by equity compensation — the same stock-based compensation the buyback partly offsets — rather than by a large personal stake. That is neither unusual nor disqualifying, but it means the counter-cyclical buyback record, not insider skin in the game, is the evidence that management's interests track a long-term owner's.

What Would Change the Read

On the evidence, capital allocation is shareholder-oriented and, on buybacks, counter-cyclically disciplined: Adobe has retired roughly a fifth of its share base by returning about 80% of a decade's free cash flow at a falling multiple, while funding its product response internally. The strongest fact against that read is the Figma episode — a $20 billion agreement at roughly 50 times revenue that ended in a $1 billion, un-shielded write-off. Two things would change the read. A return to large, richly-priced, dilutive acquisitions would show the Figma instinct was the rule rather than the exception. And a buyback that kept running at fiscal-2025 scale after a sharp re-rating — repurchasing heavily at a high multiple — would turn the current strength into the invisible value destruction that all-buyback policies risk. Neither is visible in the record through fiscal 2025.