Playbook › Playbook
After a spin-off, vendor consensus and reported actuals sit on different bases for at least a quarter
Claim
A spin-off creates three different EPS figures for the same quarter — GAAP (consolidated), pro-forma (continuing operations only), and adjusted — while vendor consensus continues to carry the pre-spin basis until analysts republish. Any comparison that crosses those bases is meaningless, and the financial wires make it routinely.
The damage is not merely analytical. The price moves on the artifact, which means the mispricing is real even though the miss is not — so the error is also an opportunity, provided you can tell which one you are looking at.
Evidence — SPGI Q2 2026, reported 2026-07-28
S&P Global spun off Mobility Global (NYSE: MBGL) on 2026-07-01, one day after quarter-end. Q2 2026 therefore produced four separate EPS numbers:
| Basis | Q2 2026 EPS | What it includes |
|---|---|---|
| GAAP diluted | $4.12 (+18%) | Consolidated, incl. Mobility |
| Pro-forma | $4.08 (+26%) | Continuing ops, ex-Mobility |
| Adjusted | $4.83 (+23%) | Continuing ops, ex-deal amortisation |
| Vendor consensus | $4.75 | Pre-spin basis, never refreshed |
Wire coverage compared the pro-forma $4.08 against the unadjusted $4.75 consensus and reported a 14% EPS miss. The stock fell 5.2% intraday.
The correct comparison — adjusted $4.83 against the $4.49 adjusted consensus — was a 7.6% beat, on revenue of $3.678B pro-forma (+11% organic constant-currency) versus $3.64B. Ratings grew 17% with transaction revenue +25%; Indices posted a 13th consecutive record quarter; adjusted operating margin expanded 200bp to 54.3%. The quarter was a clean beat on both lines and was reported as a double-digit miss.
The same event corrupted the forward figure in the same direction — Yahoo's implied forward EPS of $20.24 was the pre-spin FY26 basis against a post-spin price, showing SPGI at 20.2x when the true multiple on guided post-spin EPS of $17.50–17.75 was 23.2x.
How to apply
- On any post-spin print, find the company's own pro-forma table first. The press release reconciles the bases; the wire story does not.
- Never read a headline beat/miss in the quarter a spin completes, or the quarter after. Recompute it yourself from the release.
- Check the effective date against the quarter-end. SPGI's spin landed one day after quarter-end, so Q2 revenue of $4,146M still included Mobility while the forward guide excluded it — the trailing and forward figures were on different bases within one document.
- Trailing vendor fields (EPS ttm, BVPS, P/B, EV) stay contaminated until discontinued-ops treatment flows through, typically a further quarter.
- A large single-day move on a post-spin print is a prompt to check the basis, not a signal.
What would falsify this
Vendor feeds republishing consensus on the post-spin basis on or before the first post-spin report date, making the naive comparison correct. Worth re-testing at the next spin.
History
- 2026-08-12 — created from the SPGI
/analyze. Related: [[pitfall-spinoff-carveout-balance-sheet-persists-in-vendor-feeds]] (the balance-sheet mirror of the same event), [[pitfall-divested-segment-corrupts-multiyear-cagr]], [[pitfall-vendor-forward-eps-is-the-wrong-fiscal-year]], [[principle-primary-source-beats-vendor]].