Indexes · Broad market
Russell 3000 Components
Not a selection but a cut: the largest 3,000 eligible US companies, ranked by market value. Both the large-cap and small-cap Russell indices are portions of this list, which is why the ranking rules belong here rather than on either of them.
Eligibility first, then the ranking
Two steps produce this index, and the first is easy to overlook. FTSE Russell applies published eligibility rules (domicile, listing venue, security type, a minimum price, a minimum available float), which exclude structures that are not operating companies. Whatever passes is then ranked by market capitalisation, and the largest 3,000 form the index.
That eligibility filter is the quiet reason the Russell universe is a cleaner list than a raw exchange count. An exchange lists closed-end funds, preferred shares, exchange-traded products and multiple share classes alongside operating companies; this list is mostly the companies. It matters directly for anything computed across it, a breadth count over the Russell universe and one over an exchange are counting different populations.
What "broad market" does and does not mean
The phrase is used loosely, so it is worth pinning down. This index covers the large majority of American investable market value, and a minority of the listed companies.
| Segment | Status |
|---|---|
| Ranks 1–1,000 | Russell 1000. Most of the index's value sits here. |
| Ranks 1,001–3,000 | Russell 2000. Twice the company count, a small share of the value. |
| Below rank 3,000 | Outside the index. Thousands of listed companies, covered by the Microcap index or by nothing. |
| Ineligible structures | Never ranked at all, funds, certain trusts and partnerships are excluded before the ranking. |
The last two rows are what separates this from an exchange listing. A count of "how many US companies advanced today" gives a different answer over this universe than over the NYSE list, and the difference is not noise: it is the deliberate exclusion of everything that is not an operating company, plus a cut-off at rank 3,000.
Why the parent and the large-cap segment move together
A reader looking at both indices on the same day will often find them almost indistinguishable, and the reason is worth stating because it is the clearest illustration of what capitalisation weighting does.
The 2,000 smallest constituents are two-thirds of the company count and a small fraction of the value, so their combined influence on a value-weighted parent is modest. A day on which the small companies did something quite different from the large ones will barely register in the Russell 3000, which is precisely the situation breadth measures exist to reveal, and precisely what a weighted index cannot show. If the question is what most companies did, the answer is in the counts, not in this index.
One ranking date, one enormous session
Because the whole family is rebuilt from a single ranking, every change lands at once. Companies move between segments, enter from the microcap tail, or leave the universe, and every fund tracking any Russell index must reflect all of it by the close.
The result is one of the highest-volume American sessions of the year, composed almost entirely of obliged trading. Every caution on the Russell 2000 page about reading volume across that date applies here and applies to the large-cap segment too, since all three indices reconstitute together.
Using it as a screening universe
Beyond benchmarking, the parent list has a practical use the segments do not: it is a defensible starting universe for a screen. Its eligibility rules have already removed the structures that are not operating companies (the funds, trusts and products that clutter an exchange listing), which is the first filter every screen on this site recommends applying anyway.
Two cautions come with it. The cut at rank 3,000 excludes the microcap tail, so a screen built on this universe is silently a screen on companies above a certain size; and membership is a year old for most of the year, so a recently listed company may not be in it. Both are acceptable provided they are stated, which is the same standard this site applies to any filter.
Why the parent is the page to read first
Because the two segment indices have no rules of their own. Every property of the Russell 1000 and the Russell 2000 (the absence of a profitability test, the annual wholesale change, membership following price, the microcap tail sitting outside) is a property of this ranking, inherited whole.
That is the useful shape to keep in mind when comparing index families. The S&P indices are selections with gates and a committee at each level; the Russell indices are one ranked list with a line drawn through it. Neither is the market. Both are rules, and the rule is what you are quoting whenever you quote the number.
Frequently asked questions
What is the Russell 3000?
The largest 3,000 eligible US companies by market capitalisation, ranked on a set date. It is not a curated selection. It is a cut at rank 3,000 of a ranked list, and the Russell 1000 and Russell 2000 are simply the top and bottom portions of it. Everything the two segments do follows from the ranking that produces this parent list.
How much of the US market does it cover?
The large majority of investable market value, and considerably less than all of the listed companies. Capitalisation falls away steeply down a ranked list, so the first few hundred names account for most of the value while thousands of listed issues rank below 3,000 and sit outside the index entirely, in the Russell Microcap Index, or nowhere. "Broad market" describes the value covered, not the company count.
What makes a company eligible?
Rules published by FTSE Russell covering domicile, the listing venue, security type, a minimum price and a minimum available float. They exclude structures that are not operating companies — closed-end funds, certain trusts, limited partnerships in some cases — which is what makes the Russell universe a cleaner list of companies than a raw exchange count. Eligibility comes first; the ranking is applied to whatever passes.
Why does it matter which universe an index is cut from?
Because two indices built from different parents are not comparable even when they use the same rule. A small-cap index cut from a universe that excludes non-operating structures holds different things from one cut from a raw exchange list, and any breadth or performance figure inherits that difference. It is the same discipline as knowing which exchange a breadth count came from.
Is the Russell 3000 itself worth tracking?
As a benchmark, yes. It is the standard broad-market reference for US equities and there are funds tracking it. As a reading, it behaves almost exactly like the Russell 1000 on any given day, because capitalisation weighting means the largest names dominate and the 2,000 smallest contribute little. If you want to see what the small companies did, the segment index is the one to look at.
When is it rebuilt?
At the annual reconstitution, traditionally in late June, from a ranking taken some weeks earlier, with new listings added quarterly between reconstitutions. The whole family moves at once, which is what makes reconstitution one of the highest-volume American sessions, and a date to exclude from any volume-based reading rather than to interpret.
Where do I get the constituent list?
From FTSE Russell around reconstitution, or from the daily holdings disclosure of any ETF tracking the index. This page does not reproduce it: 3,000 tickers are stale within weeks and a static list is a maintenance liability. The same survivorship caution applies as everywhere, a current list used on historical data quietly removes everything that failed.
How does it compare with the S&P Composite 1500?
They cover similar ground by different routes. The S&P Composite 1500 is the 500, 400 and 600 combined (three committee-selected, profitability-gated indices), while this is a mechanical ranking with no earnings test and no discretion. The Russell parent holds more companies and a wider range of quality; the S&P composite holds fewer and applies a gate at the door. Neither is the market, and both are frequently described as if they were.