Financebotresearch desk研究台

IT › analyze

IT · Analyze

HOLD Technology

Date: 2026-08-04 | Price: $185.79 (+19.0% on the print) | Market cap: $12.44B Prior file: analyze-2026-06-11.md at $155.84 · +19.2% since

Headline: the gate is not satisfied. It is half satisfied.

Watchlist.md recorded this print as "🟢 GATE SATISFIED — needs /analyze." That call was made from the CEO's language on the call. It does not survive contact with the written test.

The gate, as set on 2026-06-11, was: "positive net CV additions 2 consecutive quarters."

Quarter What we know Gate progress
Q1 CY2026 Contract Value growth +1% YoY, net quarterly addition −$56M — recorded in the June file ❌ negative
Q2 CY2026 Total CV growth improved 70bps vs Q1; CV ex-Fed +3.3%; CEO: "Contract Value growth accelerated again" ✅ likely positive — quarter 1 of 2
Q3 CY2026 not yet reported ⏳ the quarter the gate actually turns on

Q1's net CV addition was negative. A streak of two cannot begin before Q2. The earliest the gate can clear is the Q3 print in early November. The watchlist called the race at the halfway mark, and the stock has already paid out +19% on it.

Two further qualifications on the CV number itself:

  1. "Ex-Fed" is carrying weight. The clean-sounding +3.3% is the number excluding the Federal book. Total CV growth is materially lower — improving 70bps off a +1% base puts it near +1.7%. The ~160bps gap is the shrinking Federal business, and it is not disclosed separately in the summaries.
  2. It is a growth rate improvement, not a level. +1.7% total CV growth is still the weakest CV print in Gartner's public history outside 2020.

Financial health — the part the print did not mention

This is where the analysis diverges hardest from the tape. Gartner's reported EPS is rising fast. Its business has not grown in four years.

Fiscal year 2022 2023 2024 2025 4yr change
Revenue $5.48B $5.91B $6.27B $6.50B +18.6%
Operating income $1.11B $1.11B $1.16B $1.18B +6.3%
Operating margin 20.3% 18.8% 18.5% 18.2% −210bps
Net income $808M $882M $1.25B $729M −9.8%
Diluted EPS $9.96 $11.08 $16.00 $9.65 −3.1%
Diluted shares 81.07M 79.68M 78.34M 75.61M −6.7%

Operating income has gone from $1.11B to $1.18B in four years while revenue grew 18.6%. Every dollar of incremental revenue arrived at a lower margin than the dollar before it. That is the whole quantitative story, and it is not visible in any adjusted EPS figure.

⚠️ The FY2024 EPS of $16.00 is not a usable comparison base. Net income ($1.25B) exceeded operating income ($1.16B) that year — the one-second diagnostic from [[pitfall-unrealized-equity-marks-break-headline-pe]] fires. Something below the operating line (tax and non-operating items) produced roughly a third of that year's earnings. Any "EPS fell 40% in 2025" or "EPS is recovering" framing anchored on $16.00 is measuring an artifact. FY2025's $9.65 is the honest number.

Free cash flow — good business, flat trajectory

2022 2023 2024 2025 FY26 guide
OCF $1.10B $1.16B $1.48B $1.29B —
Capex −$108M −$103M −$102M −$115M —
FCF $993M $1.05B $1.38B $1.18B ≥$1.185B

FCF margin is ~18% and capex is a rounding error (1.8% of revenue) — this is a genuinely asset-light business and always has been. But FY26 FCF is guided to $1.185B against FY25's $1.18B. That is +0.4%. Flat. Alongside it, adjusted EPS is guided to ≥$14, up ~24% YoY.

FCF flat, EPS +24%. The entire gap is the share count.

The buyback is a leveraged recapitalization

2022 2023 2024 2025
Buybacks $1.04B $606M $735M $1.99B
Buyback ÷ FCF 105% 58% 53% 169%
Debt issued $0 $0 $274M $800M
Shareholders' equity $228M $681M $1.36B $320M

In FY2025 Gartner spent 169% of free cash flow on its own stock and issued $800M of new debt in the same year. Book equity fell from $1.36B to $320M; P/B is now 199×. In Q2 alone it retired 3.6M shares for $547M — share count 70.85M → 66.95M, −5.5% in two quarters — and raised the authorization to ~$1.2B.

This is not a criticism of the capital allocation in isolation: buying a 19%-FCF-margin subscription business back at 10× FCF is defensible arithmetic. It is a statement about what the EPS growth is made of, and how long it can last. Revenue is guided down, operating income is flat, equity is nearly exhausted, and the buyback is now partly debt-funded. The EPS line is being manufactured, and the machine that manufactures it has a visible end.

📌 Data flag: Yahoo reports EV $11.83B, implying $0.6B net cash. The balance sheet shows $1.72B cash against $3.35B debt — $1.63B net debt. Correct EV is ~$14.07B. Every EV-based multiple from that vendor field is ~16% too low on this name. All EV figures below are computed from the statements.

Moat

The moat is real and I am not disputing it. Magic Quadrant is a genuine standard-setter: enterprise buyers cite it in procurement documents, vendors organize product roadmaps to move quadrants, and no competitor has a substitute with comparable authority. Gross margin is 69% and has been stable. Research is ~98% subscription with high wallet retention. This is a category-defining asset.

The stress test — how would a rival attack? Not by building a better Magic Quadrant; that took thirty years of neutrality capital. They attack the unit of consumption. Gartner sells seats that entitle a human to ask an analyst a question. Generative AI attacks precisely that metaphor: if a CIO's own model can synthesize vendor comparisons on demand, the value migrates from access to the analyst to the proprietary data underneath, and Gartner prices the former.

The evidence that this is already happening is in the table above: revenue +18.6% over four years, operating income +6.3%. Gartner is spending more to hold the same profit. That is what a moat looks like in the year the toll road gets a parallel route — still standing, still collecting, but no longer able to raise the toll.

Note the counter-evidence, which is real: core subscription products grew ~7% in the quarter, and management attributes CV strength to demand for AI-related insight. Being the neutral referee in a confusing technology transition is a good place to stand. The Conferences segment grew ~12% FX-neutral same-conference — people still show up. Consulting fell $156M → $142M (−9%), but Consulting is the low-quality third of the business and shrinking it is not a loss.

Evergreen rating: 6/10 — durable but no longer compounding. The franchise survives. The pricing power that made it a compounder has stopped showing up in operating income.

Valuation

FY26 guidance: revenue ≥$6.375B, EBITDA ≥$1.57B, adjusted EPS ≥$14, FCF ≥$1.185B.

⚠️ Revenue is guided to decline. $6.375B against FY2025's $6.50B is −1.9%. Q2 revenue was $1.68B, −0.6% YoY. The "beat" everyone traded was a beat on a shrinking top line.

Method Input Result
P/E on guided adj EPS $185.79 ÷ $14.00 13.3×
EV/EBITDA $14.07B ÷ $1.57B 9.0×
P/FCF $12.44B ÷ $1.185B 10.5×
EV/FCF $14.07B ÷ $1.185B 11.9×
FCF yield 9.5%
Graham IV √(22.5 × 9.65 × 0.93) $14.58 — meaningless; BVPS is $0.93 by construction
Bogle expected return 0% dividend + ~0% organic growth + ~5.5%/yr share retirement ~5.5% + multiple change

Graham is inapplicable here and I am not going to pretend otherwise — a company that has bought its own equity down to $320M cannot be valued on book. Bogle is the honest model, and it returns ~5.5% annually from share retirement alone, with nothing from the business.

Fair value: $155–195, central ~$172. That is 11–14× guided adjusted EPS — the low end for a no-growth business, the high end crediting the subscription model and 69% gross margin.

At $185.79 the stock sits in the top quartile of that range. Consensus target is $160.38 — 13.7% below spot. The sell side, which upgraded nothing today, agrees.

Verdict — 🟡 HOLD · conviction [6.0], unchanged

Not an upgrade. The gate did not clear, and the price consumed the upgrade before it could.

The three things that are true at once:

  1. The business got slightly better. CV growth accelerated for the first time in over a year, core subscriptions grew ~7%, conferences are strong. This is the first genuinely encouraging print in four quarters and it deserves to be recorded as such.
  2. The business is still not growing. Revenue guided −1.9%, operating income +6.3% over four years, FCF guided flat, and the EPS growth everyone bought is share count.
  3. The price already paid for outcome (1) and ignored (2). +19% in a session, +19.2% since the June file, +49.5% off the 52-week low, and 13.7% through the consensus target.

The June file gated an upgrade on evidence. The evidence is half delivered. Raising conviction now would be rewarding a stock move rather than the test that was written down — exactly the discipline the gate existed to enforce.

Conviction stays [6.0]. Entry raised $130–145 → $140–160 to reflect the improved CV trajectory, but that is below spot — the name is not actionable here. Trim 13× fwd.

What would change this: a Q3 print in early November showing a second consecutive positive net CV quarter and total CV growth (not ex-Fed) above 3% would clear the gate properly and justify [7.0]. Conversely, a Q3 that re-decelerates while the buyback continues at 169% of FCF would be a [5.0], because the EPS support runs out at a computable date.

A portfolio-specific passage was removed from the public build.

Sources