Chapter 3

Capital Allocation

Accenture buys its way toward growth through a programmatic acquisition engine, and the AI-and-non-FTE pivot documented in Advanced AI Economics runs almost entirely through it. Over FY2022–FY2025 the company spent roughly $14 billion on 132 acquisitions, lifting goodwill from $13.1 billion to $22.5 billion with no impairment, while returning nearly $30 billion to shareholders — all funded from cash. FY2026 breaks that pattern: planned deployment of about $18.5 billion sits well above free cash flow, backed by the first real debt in the company's history.

Sources: goodwill and net cash from FY2025 Annual Report balance sheet [1]; FY2026 acquisition plan from the Q3 FY2026 earnings call [2]; dividend rate declared September 2025 [3].

The acquisition engine

Accenture describes its acquisition strategy the same way every year — "a disciplined acquisition strategy, which is an engine to fuel organic growth" — and the record supports the label more than the marketing usually does. The deals are numerous and small: 23 acquisitions for $1.5 billion in FY2025, an average of roughly $65 million each [4]. This is the tuck-in, capability-buying model — scale in high-growth areas, add skills, deepen industry expertise — not the transformational megadeal.

The pace is lumpy. FY2024 was the heavy year at $6.6 billion across 46 deals; FY2025 pulled back to $1.5 billion; FY2026 is now guided to about $9 billion, the largest program the company has run [5] [6].

Sources: FY2022 $3.4B/38 deals [7]; FY2023 $2.5B/25 [8]; FY2024 $6.6B/46 [9]; FY2025 $1.5B/23 [10]; FY2026E from the Q3 FY2026 call [11].

Every deal lands as goodwill, and the balance sheet shows it: goodwill has grown every year, from $13.1 billion in FY2022 to $22.5 billion in FY2025 — now 34% of total assets and roughly 72% of shareholders' equity [12]. The test of whether that price bought value is impairment, and here the record is clean: Accenture reports no goodwill impairment as of August 31, 2025 or 2024, with each segment's estimated fair value "substantially" exceeding its carrying value [13]. A services roll-up that has never written down goodwill while nearly doubling it is unusual, and it is the strongest evidence that the engine has been additive rather than value-destroying — with the caveat that an impairment is a lagging signal, disclosed only after a reporting unit's value has already fallen.

Source: derived from reported balance sheets, FY2022–FY2025 Annual Reports [14].

The FY2026 step-change

For most of its history Accenture returned most of its cash and spent the rest on tuck-ins, comfortably within its own free cash flow. FY2026 changes the arithmetic. Management now expects to deploy about $9 billion on acquisitions — explicitly "up from $5 billion" guided earlier — and to return at least $9.5 billion through dividends and buybacks [15] [16]. That is roughly $18.5 billion of planned deployment against FY2025 free cash flow of $10.9 billion [17].

Sources: FY2025 free cash flow [18]; FY2026 acquisition and capital-return plans from the Q3 FY2026 call [19] [20].

The gap is bridged by the balance sheet, and for the first time that meant borrowing. On October 4, 2024 an Accenture finance subsidiary issued $5 billion of senior unsecured notes across four tranches — 3.90% due 2027 through 4.50% due 2034 — for general corporate purposes including repaying commercial paper [21]. This is a company that carried essentially no long-term debt a year earlier; interest expense rose to $229 million in FY2025, up $170 million, on that issuance [22].

The debt is easily carried. There are no financial covenants on the notes, $229 million of interest is trivial against $10.8 billion of operating income, and even after borrowing Accenture holds roughly $6.3 billion of net cash [23] [24]. The significance is not leverage risk; it is what the borrowing signals. A management team that funded everything internally for two decades is now willing to lean on the balance sheet to keep both the buyback and a record acquisition year going at once — a choice, not a necessity.

Returns to shareholders, and what the buyback actually does

The dividend is the cleaner part of the story. Accenture raised the quarterly rate to $1.63 per share for FY2026 from $1.48 in FY2025, a 10% increase and a continuation of a steady, growing payout that reached $3.7 billion for the year [25] [26].

The buyback deserves a closer read. Accenture repurchased $4.6 billion of stock in FY2025 and the board added $5 billion of authority in September, bringing the total to $7.9 billion [27]. Yet the diluted share count barely moved: basic shares fell from 632.8 million in FY2022 to 624.9 million in FY2025, a decline of about 1.2% over three years despite roughly $13.5 billion of repurchases across the period. Most of the buyback offsets the dilution from equity-based compensation rather than shrinking the count — a familiar pattern for a firm that pays its workforce heavily in stock. The per-share compounding here comes mainly from earnings growth and the rising dividend, not from a shrinking share base.

Source: dividends and buybacks from reported cash flow statements, FY2022–FY2025 Annual Reports; FY2025 figures per the Letter to Shareholders [28].

Source: weighted-average basic shares from reported income statements, FY2022–FY2025 Annual Reports [29].

The read

The historical capital-allocation record is disciplined and, on the evidence, value-additive: small deals bought at the pace of organic growth, goodwill never impaired, returns on equity held in the mid-20s while the acquired base nearly doubled, and a growing dividend paid from a fortress balance sheet. That record is why the AI pivot's dependence on acquisitions is not, by itself, alarming — this is a management team with a demonstrated ability to buy capability and integrate it.

The thing to watch is the FY2026 departure from that template. Deployment now runs above free cash flow, the deals are getting larger and pricier as they move into non-FTE software (the OT-security platform management flagged as near-term margin-dilutive), and the company has taken on debt to sustain the pace. Most pointed: acquisitions are set to contribute about 1.5% of revenue growth this year, rising toward 2% into FY2027, against total growth guided to 3–4% in local currency — so roughly half of Accenture's already-slow headline growth is now bought rather than earned [30] [31]. The 10-K itself names the risk plainly: integration challenges are "magnified by the size and number of transactions we have executed" [32].

The evidence does not yet decide whether this scaling holds. What would change the read in either direction is concrete and checkable: a goodwill impairment, a slide in return on equity as the acquired base grows, or inorganic contribution rising while organic growth stalls — the signature of acquisitions masking a stagnant core — would confirm the bear's concern; a continued clean impairment record with returns intact through a $9 billion year would show the discipline scales.