Capital Allocation

Across 2021-2025 Gartner repurchased about $6.0 billion of stock (~103% of cumulative free cash flow) at a blended average near $280 a share, funding 2025's record $2.0 billion program partly with $800 million of new ~5.3% senior notes, into a stock that fell to about $142 by mid-2026; yet executive pay is tied to Contract Value, EBITDA and revenue rather than earnings per share, and the CEO's 2025 compensation actually paid was negative $7.0 million against a $19.2 million grant-date total, so the buyback ran ahead of the market price without being an EPS-target mechanism — and it was executed while GTS contract value was flat at $3,910 million and total contract value only 1% higher.[1][2][3]

A buyback that shrinks the share count is often the arithmetic a management team runs to hit an earnings-per-share target. This one is not that. Earnings per share appears nowhere in Gartner's pay plan: the annual bonus is set on revenue and EBITDA, the long-term equity award on Contract Value, and the executives directing the repurchases are measured on those operating metrics, not the per-share figure a buyback lifts [4][5]. In 2025 the CEO's realized pay went below zero as the shares fell — his stock awards were marked down by more than his whole year's grant [6]. What the repurchase is, then, is a bet: that the cash flows being retired are worth more than the market's mid-2026 price for them.

The bet is, so far, underwater. The 21.5 million shares retired over five years cost about $6.0 billion and are worth roughly $3.1 billion at $142 — a paper shortfall near $3 billion. Only the ~$535 million repurchased in the first quarter of 2026, near $142, sits below the ~$187 a share the reverse-DCF treats as the no-growth floor (Valuation Reset); 2025's ~$284 average sits above it. Whether the repurchasing compounded or destroyed per-share value is not settled by the record — it turns on the durability question the report is built around. If contract value re-accelerates, retiring a fifth of the shares below their eventual worth reads as conviction; if it fades, it reads as borrowing to buy a shrinking stream.

A cash machine, by design

The subscription model that drives the Insights franchise also makes Gartner unusually cash-generative. Clients pay in advance, the business owns few hard assets, and capital spending has held near $100–115 million a year — under 2% of revenue [7]. The result is free cash flow (operating cash flow less capital expenditure) that has exceeded net income in every one of the last five years, and by a wide margin in years without one-off items.

Source: FY2025 Annual Report (Form 10-K), Consolidated Statements of Cash Flows and Operations [8]; figures computed (FCF = operating cash flow − capex).

Source: FY2025 and FY2022 Annual Reports (Form 10-K), Consolidated Statements of Cash Flows [9][10]; FCF computed as operating cash flow − capex.

The two years where conversion tops 150% — 2021 and 2025 — are the paid-in-advance model working in Gartner's favour and against reported earnings: deferred revenue and working-capital swings lift cash above profit, and in 2025 a $150 million non-cash goodwill charge (discussed below) depressed net income without touching cash [11]. The cash generation is real and consistent; the quality of the engine is not in question here. How that cash is deployed is what the rest of this chapter examines.

One lever: buybacks

Gartner pays no dividend and has made almost no acquisitions in recent years — 2024 and 2023 acquisition spend was $2 million and $4 million [12]. Practically all discretionary cash goes to repurchases. Across 2021–2025 the company bought back about $6.0 billion of stock against roughly $5.9 billion of cumulative free cash flow — it returned, in effect, every dollar the business produced, and in three of five years spent more than it earned in cash.

Source: FY2025 and FY2022 Annual Reports (Form 10-K), Consolidated Statements of Cash Flows (purchases of treasury stock) [13][14]; FCF computed.

The mechanical effect has been exactly what buybacks are meant to do: concentrate ownership. Diluted shares fell from 92.1 million in 2018 to 75.6 million in 2025, and to 70.0 million by the first quarter of 2026 — a fifth of the company retired in eight years [15][16].

Source: FY2025 Annual Report (Form 10-K) [17] and Q1 FY2026 Form 10-Q [18]; pre-2021 figures from prior filings, historical series.

The price paid

A buyback creates value only when the shares are bought below what they turn out to be worth. On that test the record is, so far, unflattering. The approximate average price Gartner paid rose every year from 2021 to 2024 as the stock climbed, peaking near $460 a share in 2024, then eased to about $285 in 2025 as the price fell. Blended across the five years, roughly $6.0 billion bought about 21.5 million shares at an average near $280 — against a share price of about $142 in early July 2026.

Sources: shares repurchased from each year's Form 10-K MD&A [19][20][21][22][23]; cash outlay from the cash-flow statements [24]; average price computed.

Two things keep this from being a simple indictment. First, buybacks are not marked to market like a trading position — the shares are retired for good, and the payoff depends on Gartner's cash flows over years, not on the July 2026 quote. Second, the timing within 2025 was better than the headline suggests: the stock opened the year near $540, and the $2.0 billion spent that year at an average near $285 means the bulk of it was deployed in the second half, when the price sat in the $220–290 range, not at the January high [25]. Management leaned into the decline rather than buying the top. The company kept going in early 2026, repurchasing a further $535 million in the first quarter [26].

Every level so far has proven to be above the next one. At today's price the 21.5 million shares bought over five years are worth roughly $3.1 billion, against about $6.0 billion paid. Whether that was value created or destroyed is not yet decided — it turns on the same durability question the report is built around. If the contract-value stall is cyclical, retiring a fifth of the shares at an average well below their eventual worth will read as conviction; if it is structural, it will read as feeding cash into a falling business.

Debt now helps fund the buyback

For most of the last decade Gartner's repurchases were funded from cash flow. That changed at the margin in 2025. To sustain a $2.0 billion program while free cash flow was $1.2 billion, the company drew down cash and, in November 2025, issued $800 million of new senior notes — $350 million at 4.95% due 2031 and $450 million at 5.60% due 2035 [27]. That new money costs roughly 5.3%, against a blended ~4.0% on the older 2028–2030 notes — Gartner is now borrowing at a higher rate, in part, to buy its own shares [28].

The leverage this adds is still modest. Gross debt rose to about $3.0 billion, but against $1.7 billion of cash and roughly $1.4 billion of adjusted EBITDA, net debt is around $1.3 billion — under one turn of EBITDA — and interest is covered close to nine times by operating income [29][30]. The balance sheet is not stretched. What has changed is the posture: management is confident enough in the cash engine to lever up and repurchase into a 70%-plus drawdown, with $0.7 billion of authorization remaining at year-end and a further $500 million added in January 2026 [31].

Source: FY2025 Annual Report (Form 10-K), Note 6 Debt and Statements of Operations [32][33]; adjusted EBITDA and net debt computed.

The one deal outside the core

The buyback discipline stands in contrast to Gartner's one visible operating bet outside the core. The Digital Markets business — the software-listings operation reported in the "Other" segment — was written down by $150 million in the third quarter of 2025, after "ongoing weakness in the market" forced a cut to its long-term earnings forecast [34]. In February 2026 the company sold it outright for about $105 million net of cash, booking a small $6 million gain over its impaired carrying value [35]. The episode is small relative to the $6 billion of buybacks, and exiting a non-core, decelerating asset is reasonable housekeeping — but it is a reminder that Gartner's returns come from the research franchise, not from capital deployed elsewhere, and that the impairment (not any cash loss) is what pulled 2025 reported earnings down.

The read

Gartner is an unusually cash-generative business that has chosen to return essentially all of that cash through buybacks, and the mechanical result — a fifth of the shares retired, EPS supported even as net income wobbles — is real. The price paid is what the record cannot yet settle. The average paid over five years is roughly double the current quote, the largest program yet was executed while the core engine stalled, and 2025 marked the first time debt was raised to help fund it. Set against that: leverage remains under a turn of net EBITDA, the 2025–26 buying was concentrated at derated prices rather than the peak, and the shares are permanently gone. Whether this compounded or eroded per-share value cannot be judged from the buyback record alone — it depends on whether the cash flows being bought prove durable. The signal to watch is straightforward: a re-acceleration in contract value would validate the aggressive repurchasing; continued erosion would leave management having borrowed to buy a shrinking stream.