Indexes · Large cap
Russell 1000, The Large-Cap Rank Cut
Ranks 1 to 1,000 of the Russell ranking, the large-cap half of a mechanical cut. It covers the same ground as the S&P 500 and is built in the opposite way, which is the only thing about it worth knowing in detail.
What it is
The top thousand of the ranked list described on the Russell 3000 page. Eligibility rules are applied first (domicile, listing venue, security type, a minimum price and float), and whatever passes is ranked by market capitalisation. The largest 1,000 are this index; the next 2,000 are the Russell 2000.
Nothing else enters it. No profitability test, no sector balance, no committee. Because the parent page carries the ranking mechanics and the small-cap page carries the reconstitution, the useful subject here is the comparison every reader actually arrives with: how this differs from the S&P 500, given that both are quoted as "US large cap".
Where the two actually differ
| Property | Russell 1000 | S&P 500 |
|---|---|---|
| How membership is set | Rank on one date. | Eligibility gates, then a committee. |
| Profitability required | No. | Yes, at entry. |
| Reproducible from public rules | Yes, given the ranking data. | No, discretion is part of it. |
| When it changes | Wholesale, at the annual reconstitution. | Continuously, as events require. |
| Company count | 1,000 by rule. | 500 by selection. |
The row that matters most for anyone doing work rather than quoting is the third. A mechanical index can be rebuilt, so a study of "membership effects" is a study of a rule. A committee index cannot, so the same study is partly a study of the committee, and that is not a criticism of either, only a difference in what a result can be attributed to.
Banding, and why turnover was a problem worth fixing
A rank cut has one obvious pathology: a company sitting close to the boundary would cross it in most years on nothing but noise, and every fund tracking either index would have to trade the name each time. That is pure cost, borne by the funds' holders, in exchange for nothing.
Banding is the answer FTSE Russell adopted: a company stays where it is unless its rank has moved beyond a threshold around the cut-off. The boundary does not move, the willingness to cross it does. It is a good example of a mechanical index taking a deliberately non-mechanical decision for a practical reason, and worth knowing because it means membership is slightly stickier than a pure ranking implies. A name that "should" be in the large-cap index by this year's rank may still be in the small-cap one.
The equal-weight comparison
One practical use of a large-cap index deserves its own paragraph, because it costs nothing and answers the question this site keeps returning to.
Compare the capitalisation-weighted index with its equal-weighted version. When the weighted version outruns the equal-weighted one over weeks, the largest constituents are carrying it and the advance is narrow. When the equal-weighted version leads, participation is broad. It is the same conclusion the advance/decline line reaches by counting issues, arrived at in returns instead, and because it uses index data rather than exchange breadth counts, it is not affected by the funds and preferred shares that clutter an exchange listing.
The boundary is where the difference lives
Because both indices are dominated by the same very large constituents, everything that distinguishes them happens at the edge of the list, and the edge is where the two constructions behave least alike.
A company sitting near rank 1,000 moves between this index and the small-cap one when its rank crosses the boundary at reconstitution, subject to banding. The same company may not be in the S&P 500 at all, or may be added years later when a committee decides. So a study of "large-cap companies" over a decade is studying two different populations at the margin, and any conclusion drawn from the boundary (index-inclusion effects, small-versus-large rotation, the behaviour of newly promoted names) depends on which definition produced the sample.
Everywhere else the two are close enough that the choice barely matters. That combination is what makes the difference easy to overlook: it is invisible in the daily number and material in exactly the studies people run.
Reading breadth across a thousand names
One practical note, since this is a volume reference. A breadth count taken across this index is a count across a thousand large companies whose eligibility rules already excluded funds, trusts and other non-operating structures, which makes it a cleaner input than an exchange-wide count, and a narrower one.
Narrower matters. A thousand large companies are more correlated with one another than a full exchange listing is, so breadth readings taken here have a naturally tighter range: extremes are rarer and a modest divergence means more than the same figure would across a broader list. As everywhere, the threshold has to come from the same series it is applied to.
Why both indices exist
Because they serve different constituencies rather than competing for the same one. The S&P 500 is the reference most portfolios are measured against and the most widely quoted number in American markets. The Russell family is the standard for institutional style benchmarking, where a consistent, mechanical definition of "large cap" and "small cap" cut from one universe matters more than name recognition.
For a reader of this site the practical consequence is small and worth remembering: a breadth or performance figure quoted against "US large cap" could be either, and at the boundary they hold different companies. Naming the index is the same discipline as naming an indicator’s period.
Frequently asked questions
What is the Russell 1000?
The top 1,000 of the Russell ranking, the largest 1,000 eligible US companies by market capitalisation, taken from the same ranked list that produces the Russell 3000 and, beneath it, the Russell 2000. There is no selection and no committee: rank in the band on the ranking date and you are a constituent.
How does it differ from the S&P 500?
In construction rather than in coverage. This index takes twice as many companies and takes them mechanically; the S&P 500 takes fewer and applies eligibility gates — including positive earnings — plus a committee decision. They therefore hold different companies, particularly at the boundary, and both are weighted by float-adjusted market value so both are dominated by the same handful of very large names. Day to day they move almost identically; the differences are structural rather than visible.
Which is the better large-cap benchmark?
Neither, and the choice depends on what you want the benchmark to be. The Russell 1000 is a mechanical description of the large-cap market as it is, including whatever is unprofitable. The S&P 500 is a curated, gated list and is the more widely quoted of the two. If reproducibility matters, being able to rebuild the membership from published rules, only the Russell index offers it.
Does it hold unprofitable companies?
Yes, when they rank high enough by market value, because the rules never ask about earnings. That is the same property described on the Russell 2000 page and it is much less consequential here: at large-cap scale the number of persistently loss-making constituents is small. The property is the same; its effect is not.
What share of the US market does it cover?
The large majority of investable value, since capitalisation falls away steeply down a ranked list and the top thousand carry most of it. No percentage is quoted on this page on purpose. The figure drifts with the market and a number printed on a static page gets cited long after it stops being true. The structural statement is stable: this index and the Russell 3000 move almost identically because the extra 2,000 companies add little value weight.
When does membership change?
At the annual reconstitution, traditionally in late June, with new listings added quarterly. Companies move between this index and the Russell 2000 as their rank crosses the boundary, with banding to stop borderline names migrating every year on noise. Everything changes on one date, which is what makes reconstitution one of the highest-volume American sessions and a date to exclude from any volume reading.
Is there an equal-weighted version, and why does it matter?
Yes, as there is for most major indices, and the comparison is one of the more useful free readings available. When the capitalisation-weighted version outruns the equal-weighted one over weeks, the largest members are doing the work, the same conclusion an advance/decline line reaches by counting, arrived at through returns. It is the cheapest available check on whether an advance is broad.
Where do I get the constituent list?
From FTSE Russell around reconstitution, or from the daily holdings disclosure of any ETF tracking the index. Not reproduced here: 1,000 tickers are stale within weeks of the next reconstitution, and a current list applied to historical data introduces survivorship bias by silently dropping everything that failed or was acquired.