Chapter 4

Buybacks and Debt

Gartner returns essentially all of its surplus cash one way: buying its own stock. There is no dividend. Across 2021–2025 it repurchased about $6.0 billion of shares — roughly equal to the free cash flow it generated — and shrank the diluted share count by nearly a quarter. The record shows real discipline on pace and a defensible funding structure, but the prices paid are the catch: every tranche since 2021 sits underwater at today's $142, and 2025's $2.0 billion was the largest and partly borrowed.

Sources: FY2025 Annual Report, Consolidated Statements of Cash Flows [1]; Note 8 — Stockholders' Equity [2]; Q1 FY2026 earnings call [2]. Share reduction derived from reported diluted share counts and 2026 guidance.

This chapter takes the cash flow shown to be durable elsewhere (Cash and the Float) and asks what management does with it — the largest single use, and the one a bull most needs to be right about at a depressed price.

A return policy with one instrument

Gartner pays no cash dividend and its credit agreement carries a covenant that would limit its ability to start one [4]. All capital returned to shareholders since 2015 has come through repurchases. The Board first authorized $1.2 billion in May 2015, then added $5.8 billion in increments between February 2021 and September 2025, and another $500 million in January 2026; roughly $0.75 billion remained at year-end 2025, refreshed to about $1.2 billion after the January addition [2] [3].

For a capital-light business the choice is rational rather than aggressive. Capital expenditure ran $115 million in 2025 — about 1.8% of revenue — so the business reinvests little and retains a high return on what it does deploy: management put rolling four-quarter return on invested capital near 27% [4], and reported return on capital employed has sat in the 25–30% range. A company earning those returns has few high-return internal uses for a billion-plus dollars a year, and acquisitions have been small: only tuck-ins since the 2017 CEB deal, and one of that deal's assets, the Digital Markets unit, was written down by $150 million in 2025 and put under agreement to be sold in January 2026 [5]. With reinvestment cheap and M&A deliberately modest, the buyback is the residual — which is also why the whole capital-return case rides on how well it is executed.

Funded by cash flow, timed with debt

The prior chapter flagged that 2025's buyback exceeded free cash flow and was part debt-funded. That is true, but the fuller six-year picture is less alarming than a single year suggests. Cumulative repurchases of about $6.2 billion over 2020–2025 came in just under the roughly $6.7 billion of free cash flow generated over the same span — the program is, across the cycle, funded by the cash flow it distributes.

Source: FY2025 and FY2022 Annual Reports, Consolidated Statements of Cash Flows; free cash flow derived as operating cash flow less capital expenditure [6] [7]. 2024 free cash flow includes a $300 million event-cancellation insurance recovery; underlying was near $1,083 million.

Two years break the pattern, and both are deliberate. In 2021 buybacks of $1.66 billion ran to about 132% of free cash flow alongside $600 million of borrowings; in 2025 buybacks of $1.99 billion reached roughly 169% of free cash flow, funded in part by $800 million of senior notes issued in November — $350 million due 2031 and $450 million due 2035 — with a portion used to repay the revolver [8]. The effect on net debt is easy to overstate. Net debt was about $1.26 billion at the end of 2020 and about $1.26 billion again at the end of 2025 [9]. The single-year "doubling" from $527 million to $1.26 billion runs off a temporarily low 2024 base, after the company had paid net debt down in 2023–2024 and then re-levered to buy the falling stock in 2025. What did change over the cycle is that gross principal rose about $1 billion to $3.0 billion, and the November notes price higher — 4.95% and 5.60% — than the 3.6–4.5% legacy coupons the earlier chapter documented [10]. So the funding concern is real but narrow: not a balance sheet being consumed, but a modestly larger, modestly more expensive debt load carried to accelerate repurchases when management judged the price attractive.

The price paid

Funding is the smaller question. The larger one is price. Repurchase intensity has moved inversely to the share price since 2022: as the stock climbed toward its 2024 peak, Gartner bought less — only 1.8 million shares in 2023 and 1.6 million in 2024 — and as the stock collapsed in 2025, it bought far more, 7.0 million shares. That is the right-shaped behaviour. The problem is that "less at the high" still meant paying up, and the high was very high.

Source: FY2021–FY2025 Annual Reports, MD&A repurchase disclosures; average price and discount derived from disclosed share counts and aggregate cost against the $141.61 price on 7 July 2026 [11] [12] [13].

At $141.61, the stock trades 32% to 61% below what Gartner paid in each of the last five years. The $2.0 billion spent in 2025 — the biggest year in the company's history — averaged roughly $286 a share, about twice the current price; even the fourth-quarter purchases, made as the stock fell, averaged $239.06 with about $745 million of authorization left at year-end [14]. On a mark-to-market basis, the entire multi-year program is currently in the red. The counter-cyclical instinct was correct in direction; the market simply kept falling past every level at which management acted. The genuinely cheap buying — 2025's fourth quarter, the $535 million spent in the first quarter of 2026 — is only now happening, and only after a 74% decline created the price.

The share count, and what the buyback must earn

For all the mark-to-market red, the mechanical result is a share count that keeps falling. Diluted shares dropped from about 90 million in 2020 to 75.6 million in 2025, and management guides to roughly 69 million for 2026 — close to a quarter fewer over six years [15] [16]. The first quarter of 2026 alone cut the count more than 4%, and management bought back roughly $2.4–2.5 billion of stock over the trailing twelve months [17].

Source: FY2025 Annual Report, Consolidated Statements of Operations, and Q1 FY2026 guidance of about 69 million diluted weighted-average shares [18] [19].

That shrinking count is the engine behind management's commitment to grow adjusted earnings per share at a compound rate above 12% over the next three years — the CFO named repurchases as "one of the bigger drivers" of that figure, alongside revenue and margin [20]. With contract-value growth near zero, the buyback is doing much of the per-share work. Sell-side estimates run with it: consensus centres on adjusted EPS near $13.70 for 2026 and a mean price target of $165, above the current $141.61.

Whether the buyback creates value reduces to one condition, and it is the report's through-line. If the cash flow proves durable, then repurchasing shares at $142 — well below every price paid since 2021, and against a business earning ~27% on capital — is straightforwardly accretive, and the more the company buys the more each remaining share is worth. If instead the contract-value stall and sub-100% wallet retention mark structural erosion, then management has used cash flow, and some borrowing, to concentrate ownership of a shrinking earnings base — buying more of an asset that is worth less. The buyback amplifies whichever way the franchise resolves; it does not decide it.

What would change the read is observable and cheap to monitor: the pace and average price of repurchases disclosed each quarter in Item 5, the trajectory of contract value and wallet retention that determines whether the earnings base is stable, and any move to slow buybacks or add debt to sustain them. A management that keeps buying heavily at these prices is signalling conviction in the cash flow; a sharp pullback would signal the opposite.