Why Float Numbers Differ Across Sources

Direct answer: Most of the time, two float numbers disagree because they come from filings dated on different days. One tool reads last quarter's 10-Q. Another already pulled in an 8-K from last week. Throw in estimates versus reported counts, shares issued between filings, and different ways of handling restricted stock, and both numbers can be right at once.

You check a small-cap's float in two places. Two answers. So which one's right? Usually neither is wrong. They're just built from different filings on different dates. Let me show you why that happens, and how to actually read the gap instead of shrugging at it.

Why do two tools show different float numbers?

Four usual suspects, roughly in the order they'll trip you up:

1. Different filing dates. Float isn't a fixed number. It's a snapshot. One source is reading the last 10-Q cover page. Another already updated off a fresher 8-K or prospectus. Same company, different "as of" date, different number.
2. Estimated vs. reported. Some numbers come straight off a filing. That's reported. Others are modeled, with insider holdings subtracted using assumptions. That's estimated, and estimates drift away from the filed count.
3. Shares issued between filings. An ATM, warrant exercises, note conversions. Any of these can add shares between two filing dates. A source that hasn't caught the newer filing still shows you the older, smaller count.
4. Different restricted-share treatment. Float is outstanding minus restricted and insider shares. But sources draw the "restricted" line in different spots. Lockups, affiliate holdings, recently registered shares. So their float can differ even off the same outstanding count.

Which number should I trust?

The one that's sourced and dated. You want to see which filing it came from and when that filing is from. A number with no date attached, you simply can't check. The freshest verifiable filing usually wins. But you're not crowning a winner here. You're figuring out what you're actually holding. A count with a visible source and date beats a rounder number floating free every time.

Is a float disagreement a red flag?

Sometimes. Read it right, though. When two well-sourced numbers don't match, that gap is telling you something. Usually it means fresh issuance landed between filings: an ATM drawdown, a conversion, warrants exercised. That's information. Not noise.

Guardrail, plainly: this is rare. In practice you're looking at something like ~3 out of 60 names. Nowhere near half the market. On most tickers, float sources agree closely. When they don't, that's the handful worth a second look. Don't walk away convinced every float number is suspect. The mismatches are the exception, and that's exactly why they're worth reading.

That mismatch is what we flag as an amber "source split." It's not a verdict. It's a heads-up that the underlying filings don't line up and the number is moving.

See it yourself

You can rebuild any float number straight from SEC EDGAR. No tool required:

1. Outstanding shares — cover page of the latest 10-Q or 10-K. Grab the "as of" date on the cover. That's your baseline timestamp.
2. Recent issuance — scan 8-K filings and prospectuses (424B5, S-1, S-3) since that date for ATM sales, conversions, or warrant exercises that added shares after the 10-Q.
3. Restricted/insider shares — ownership and registration filings tell you what to subtract from outstanding to get to float.
4. Do the math — outstanding, minus restricted, adjusted for anything issued since. Now put your figure next to whatever a tool shows you.

If your dated reconstruction lands somewhere else, you've just found why, and which filing the other source missed.

This is the FloatVerify edge. Every figure is sourced and dated, so you see the exact filing and its date behind each number. When two sources diverge, we don't paper over it. We surface the source split so that gap works for you instead of hiding as a silent error.

FAQ

Why does my float number not match another site's? Almost always different filing dates, or estimate-vs-reported. Check the "as of" date on each. Data, not advice.

Does a float mismatch mean the data is broken? No. Usually it means shares changed between filings. The gap is often real information about recent issuance.

How often do float sources actually disagree? Rarely. Roughly 3 in 60 in practice. Most names agree closely, and the disagreements are the exceptions worth a look.

What's a "source split"? Our amber flag for when the underlying filings don't line up, so a float number is uncertain or moving. A prompt to check, not a conclusion.

Which float number is the real one? The one you can trace to a dated filing. No source and no date, no way to verify it.

CTA

FloatVerify — float, dilution and cash-burn on US small-caps, every figure sourced and dated. When two sources diverge, we show you the gap. → floatverify.com

---

Informational only. Not investment advice.