Vendor enterprise value is computed off a stale market cap and can imply net debt on a net-cash company — float exceeding shares outstanding is the free tell

Claim

EV = market cap + debt − cash. Every term is available and correct in the feed. The composite is not, because the vendor computes it on a different refresh cycle than the quote.

Evidence — TTD, 2026-08-07 (the day after a −21% print)

Yahoo field	Value
`marketCap`	$6.527B → ÷ 470.1M shares = $13.885, the live price ✓
`enterpriseValue`	$7.324B
`totalCash`	$1.406B
`totalDebt`	$0.424B

The arithmetic tell: EV + net cash = $8.307B ÷ 470.1M = $17.67 — exactly the previous session's close. The entire −21% move is missing from the EV numerator.

Read naively, marketCap $6.54B against EV $7.32B implies ~$780M of net debt. The truth from the Q2 balance sheet is +$1,051M of NET CASH, with total debt of $434M that is 100% finance leases and zero borrowings. Corrected EV ≈ $5.44B — the feed overstated it by 34%, and every EV-based multiple with it.

Second instance the same day — CRTO

Found while assembling a declining-adtech comp set:

Yahoo enterpriseValue $1.61B on a company holding $348M cash against $109M debt (net cash +$239M). True EV ≈ $0.67B — a 2.4x overstatement.
Same issuer: float 97.10M against sharesOutstanding 50.24M. Float cannot exceed shares outstanding. The share-count block is internally inconsistent.

How to apply

Never quote a vendor EV on a name that has moved more than ~10% in a session. Recompute: live market cap + total debt − (cash + short-term investments).
The one-line check: (EV + net cash) ÷ shares. If it equals a recent close rather than the current price, the EV is stale by exactly that much.
Sign check first. If EV > marketCap on a company you believe holds net cash, one of the two beliefs is wrong. Resolve it from the balance sheet, not the feed.
float > sharesOutstanding is impossible and free to check. When it appears, distrust every share-count-derived field on that issuer, including EV and per-share metrics.
This trap runs in the company's favour — it makes EV multiples look worse than they are. That is the opposite direction from most vendor traps, and it means correcting it is part of building the bear case honestly, not just the bull case.
Also check cash. Feeds frequently report cash-and-equivalents without short-term investments, understating net cash. TTD's FY2025 "net cash $867M" reconciled only after adding $645M of short-term investments.

What would falsify this

A feed that recomputes EV on the same tick as market cap would not exhibit the lag. Check the (EV + net cash) ÷ shares identity once per vendor to find out which kind you have.

[[pitfall-vendor-forward-eps-is-stale-on-the-day-of-a-guidance-cut]] — same mechanism, same day, different field: a composite whose parts refresh on different clocks.
[[pitfall-yahoo-share-count-dual-class-fpi]] — TTD's sharesOutstanding of 427.0M is Class A only against 470.1M diluted, a separate 10.1% error on the same name.
[[principle-primary-source-beats-vendor]]

History