Market Capitalization (Market Cap)

Reviewed 12 August 2026 · Sourced from the SEC's filing rules and investor education pages, FINRA, and the published index methodologies of S&P Dow Jones Indices and FTSE Russell

What the number actually is

Two screeners, same company, same minute. One says the company is worth $3.0 billion, the other $3.2 billion. Neither is broken. A banker pricing that company for a takeover would say $5.2 billion and would also not be wrong. The founder, asked what she could actually get for every share she owns, would give a fourth number lower than all of them.

All four start from the simplest calculation in finance.

The SEC's own investor education site defines it exactly that way. Market capitalization is “the value of a corporation determined by multiplying the current public market price of one share of the corporation by the number of total outstanding shares.” One price, one count, one multiplication. No adjustments, no discounts, no judgment.

That plainness is why market cap is the first column on every screener, this site's included. It is a genuinely useful number, and the one new investors misread most often.

Everything that goes wrong comes from two facts sitting inside that formula. The price is the price of one share — the last one that happened to trade. And the count is not a fact, it is a choice, because a public company has three or four defensible share counts that do not agree.

Those three denials are the page. Take them in order.

It is not what the company would cost to buy

A headline says a company is “worth $3 billion.” The natural reading is that $3 billion is roughly what it would take to own the whole thing. It is not, and the gap is not small.

Buying every share gets you the company's equity, and equity comes attached to whatever the company owes. Debt does not evaporate at closing; you now control a company that has to service it. Cash runs the other way: buy a company sitting on cash and you get the cash, so the real outlay is smaller than the sticker. The construction that accounts for both is enterprise value.

Market cap prices the equity. Enterprise value prices the whole capital structure — what the operating business is worth regardless of who financed it. Two companies can carry an identical market cap and be nothing alike.

Notice which way the distortion runs. Debt makes a market cap look small next to the business it controls. Every valuation ratio built on Ridgeline's market cap makes it look cheaper than Kettle; swap in ratios built on enterprise value and the ranking flips. That is why the valuation stage of Fundamental Analysis works through the balance sheet before it lets you touch a price multiple.

It is not the company's money, and it is not a price you could get

The company never received it

This is the most common misunderstanding among people new to investing, so here it is bluntly. A $3 billion market cap does not mean the company raised $3 billion, holds $3 billion, or has ever touched $3 billion.

A company receives money from its own shares only when it sells newly created shares and books the proceeds — in its initial offering, and in any later follow-on. That is the list. Every trade after that is secondary: one investor buying from another. The money moves between two brokerage accounts, and the company is not a party to it.

So a company can go from a $400 million market cap to $4 billion without a dollar of that increase reaching the business. And a market cap can be halved in a month while the company's cash, revenue and payroll are exactly what they were the day before.

A rising market cap does help indirectly: the company can issue new shares at a better price and borrow against a thicker equity cushion. And buybacks run the mechanism backwards — the company spends its own cash to buy shares out of the market, shrinking the count on the right-hand side of the formula.

And it is not a price anyone could get

Market cap values every share at the price of the last share that traded — a trade that may have been for a hundred shares. The formula extrapolates that price across the entire outstanding count, a quantity that could never transact at it.

Anyone who has watched an order book knows why. At any instant only a limited number of shares are bid for near the current price; sell more than that and you take the next bid down, then the one under it. Trying to sell 100 million shares into a market that trades 800,000 a day is not a transaction, it is a demolition. That is why large blocks get negotiated off-exchange at a discount and why insider sales are pre-scheduled in tranches. Depth matters as much as price, and volume is a rough proxy for it.

Which share count — and where the real one lives

Here is why two honest screeners print different numbers for the same company at the same moment. “Shares outstanding” is not one number. Which one you pick moves the market cap by low single digits on a normal company, and by much more on one that pays its staff heavily in stock.

The dilution gap surprises people, so here it is with numbers. Under US accounting standards the diluted count uses the treasury stock method: assume the options are exercised, assume the proceeds buy shares back at the current price, count only the difference.

Where the authoritative count lives

Not in a screener — in the company's own filing. The Form 10-Q and Form 10-K cover pages require the issuer to “indicate the number of shares outstanding of each of the issuer's classes of common stock, as of the latest practicable date.” That is a signed filing, it is the freshest count the company has, and it is free on the SEC's EDGAR. The balance sheet in the same filing gives the count as of period end — a slightly older figure, normal rather than an error.

Two things move the count quietly. Stock-based compensation adds shares every period, so unless the company buys back as many as it issues the count drifts upward — market cap can grow while each share represents a slightly smaller claim on the business. And a company may have more than one class of common stock, usually with different voting rights, sometimes with only one class listed, and the listed classes can trade at different prices. Valuing a multi-class company means pricing each class and summing, not applying one class's price to every share in existence.

Full market cap against free float

If market cap is a per-share price times a count, a fair question is which shares are even for sale. A founder holding 22% is not offering it. Neither is a government, a family trust, or the private equity firm with two board seats. Those shares exist, they count toward market cap, and they are not part of the tradable market.

S&P Dow Jones Indices defines an Investable Weight Factor as “(available float shares)/(total shares outstanding)” and weights its indices by float-adjusted market cap: price × total shares × IWF. Its float adjustment methodology — published, and one of the few primary sources in this whole subject — strips out long-term strategic holders: officers and directors, private equity firms, corporate cross-holdings, employee share plans, governments and sovereign wealth funds, and any individual holder disclosed at 5% or more. It states plainly that “S&P Dow Jones Indices uses a 5% minimum threshold for strategic blocks.” Composite 1500 constituents must also have an IWF of at least 0.1.

FTSE Russell does the same on the Russell US indexes, stating that it “adjusts its index membership weights for float shares to accurately represent the investable universe.” Its reconstitution ran once a year, after the close on the fourth Friday in June; FTSE Russell announced a move to a semi-annual schedule beginning in 2026.

This idea has a regulatory cousin with legal consequences. Under SEC Rule 12b-2 a company's filing obligations turn on “the aggregate worldwide market value of the voting and non-voting common equity held by its non-affiliates” — public float — measured on the last business day of the most recently completed second fiscal quarter. Cross $75 million and the company becomes an accelerated filer; cross $700 million and it becomes a large accelerated filer, with faster deadlines and a heavier audit requirement. A float-based market cap measure is not only an index convention. It sets how much disclosure the company legally owes you.

The size tiers are a convention, not a rule

Large cap, mid cap, small cap, micro cap, mega cap. These sound like defined categories with edges. They are not. No regulator defines “large cap.” No statute, no rule and no filing turns on it. What exist are a few published sets of boundaries that mostly agree, plus the round numbers everybody repeats without saying whose they are.

FINRA publishes one set, with the caveat most articles drop: “The delineation between each group can vary, but generally, you'll see them broken down like this.” Those are FINRA's words about FINRA's own table, reproduced here.

Now the reason this page treats the tiers as convention rather than fact. The people who actually build large-cap and small-cap indices mostly do not use fixed dollar cutoffs at all.

FTSE Russell assigns by rank. On rank day every eligible stock is ranked by total market capitalization in descending order and the largest 4,000 become the Russell 3000E universe; the Russell 1000 takes the large end of that list and the Russell 2000 the rest. There is no dollar line; the boundary sits wherever the thousandth company falls that year.

S&P Dow Jones Indices does use dollar ranges, and revises them. Its methodology states company market capitalization ranges with an effective date, because they are “reviewed at the beginning of each calendar quarter and updated as needed to ensure the ranges reflect current market conditions.” As of the ranges effective 1 July 2025, carried in the April 2026 edition: S&P 500, US$22.7 billion or more; S&P MidCap 400, US$8.0 billion to US$22.7 billion; S&P SmallCap 600, US$1.2 billion to US$8.0 billion. Check the current PDF before quoting those — a figure quoted without its date will quietly go wrong.

And membership is a policy question, not just a measurement: from 31 July 2017 multiple share class structures were ineligible for the S&P Composite 1500, and on 17 April 2023 S&P announced they would again “be considered eligible candidates.” Nothing about those companies changed on that date.

How size changes the way a stock trades, and how indices use it

None of what follows is a reason to prefer one size over another. It is mechanics.

Liquidity and the spread

Smaller companies have fewer shares available and fewer people trading them, which shows up as a wider gap between the best bid and offer and a price that moves further on an ordinary-sized order. A wide spread is a cost paid on entry and again on exit, and it never appears on a statement as a fee. Volume is the crude reading on it.

Coverage, disclosure and index inclusion

Analyst coverage thins going down the size scale, so fewer people publish estimates. And under Rule 12b-2 a company below the $75 million public float line is a non-accelerated filer with longer deadlines, while a smaller reporting company may scale back certain disclosures. Less information exists about small companies, by design of the rules and by the economics of the research business. Whether that is opportunity or hazard depends on whether you can do the work yourself, which is what Fundamental Analysis is for.

And because index thresholds are published, crossing one is an event: a company entering a widely tracked index must be bought by every fund tracking it, at a size set by its float-adjusted cap, on an announced schedule. Leaving works in reverse. That flow has nothing to do with the business.

Volatility, and the argument about the size premium

Smaller companies are typically less diversified, carry shorter operating histories and are repriced on thinner volume, so their prices move more. That much is mechanical. The claim built on top of it is not. That small-cap stocks earn a persistent long-run return premium traces to Rolf Banz in the Journal of Financial Economics in 1981 and entered practice through Eugene Fama and Kenneth French's three-factor model in the early 1990s. It has been contested ever since — Horowitz, Loughran and Savin, “Three analyses of the firm size premium,” Journal of Empirical Finance 7(2), 2000, is one well-known challenge. The clearest statement of where the argument stands comes from the paper that defends the premium. Asness, Frazzini, Israel, Moskowitz and Pedersen, “Size matters, if you control your junk,” Journal of Financial Economics 129(3), 2018, opens by listing the charges: the premium has been accused of “having a weak historical record, being meager relative to other factors, varying significantly over time, weakening after its discovery, being concentrated among microcap stocks, residing predominantly in January, relying on price-based measures, and being weak internationally.” Their argument is that those objections dissolve once you control for firm quality. A live dispute with named participants on both sides — not a settled fact.

Cap weighting, and what it does mechanically

Most broad indices and most index ETFs weight their holdings by float-adjusted market cap. The consequence is arithmetic: as a company's price rises its market cap rises, so its share of the index rises with no trade required, and as it falls its weight falls. A cap-weighted index therefore concentrates into whatever has gone up and thins out of whatever has gone down, automatically and continuously.

That is neither a flaw nor a virtue. It is why such an index needs very little turnover — much of why these funds are cheap — and why a handful of very large companies can dominate an index nominally holding five hundred. Weighting is a choice, though, not a law: the Dow Jones Industrial Average is price-weighted, so a share price rather than a market cap sets each member's influence (see Dow Theory). And market cap is nominal, so comparing one across decades without adjusting for inflation compares two different dollars.

What is sourced here and what is not

Three kinds of claim appear above. The difference matters more than the content.

What trips people up

• Reading market cap as an acquisition price. It prices the equity. Debt comes with the company and so does the cash; enterprise value accounts for both.
• Thinking the company has the money. It received proceeds only in its offerings.
• Treating it as a realizable amount. It is one share's price times a count that could not be sold at that price. And multi-class companies need each listed class priced and summed, not one class's price applied to every share in existence.
• Comparing a diluted market cap against a basic one. On a company that pays heavily in stock the gap is real money — 6% in the example above. Check which count your data source used before comparing two companies.
• Expecting index weight to match market cap. Indices weight by float-adjusted cap under published rules that exclude closely held blocks — a 0.78 float factor is a 22% haircut.
• Quoting a size-tier dollar range as if it were defined. No regulator defines the tiers. FINRA publishes ranges and says they vary, FTSE Russell uses rank instead, and S&P's dollar bands carry an effective date and get revised.
• Assuming a cap-weighted index is diversified because it holds hundreds of names. Weighting by market cap means the largest members carry the outcome — the mechanism working as designed, not a malfunction.
• Assuming small caps outperform because a factor model says so. The defenders' own paper opens by listing eight objections to the premium.
• Comparing market caps across decades without adjusting for inflation. A billion dollars in 1990 and a billion now are not the same billion.

The figures on this page are checked against the source that publishes them, and dated. Published rates move after the release named above — the linked source always carries the current number. This page explains a term; it does not recommend a product.