Indexes · Capped weighting
Nasdaq 100 and Its Concentration Caps
The one index in this section whose methodology deliberately departs from what the market did. When its largest members grow too large, weight is taken from them and given to companies that did not grow.
A rule that overrides the market
Every other index here lets the market decide the weights. Value-weighted indices give influence in proportion to size, and the Dow gives it in proportion to share price, but in both cases the number that comes out is a consequence of prices rather than of a committee’s view about diversification.
This index adds a step. When a member exceeds its ceiling, the excess weight is removed and shared among the others in proportion to their own weights. That is a deliberate departure from what happened, and it is worth understanding as a feature rather than as a flaw: the index exists to be tracked by funds, and funds have diversification limits to respect.
| Member | Market weight | Capped weight |
|---|---|---|
| Member C | 11.0 % | 14.9 % |
| Member D | 8.5 % | 11.5 % |
| Member E | 6.0 % | 8.1 % |
| Member F | 4.0 % | 5.4 % |
| Others (94 members) | 20.0 % | 20.0 % |
| Member B | 22.5 % | 20.0 % |
| Member A | 28.0 % | 20.0 % |
What the caps are for
Funds in several jurisdictions face diversification rules limiting how much of their assets may sit in any one holding, and how much may sit in the holdings above a given size. An index dominated by two or three enormous companies cannot be tracked by such a fund at all.
So the caps are a tracking requirement rather than a measurement improvement, and stating it that way resolves most of the argument about them. As a description of what the largest Nasdaq non-financials did, a capped index is worse than an uncapped one. As an instrument that regulated funds can actually replicate, it is the only version that works.
The special rebalance, and why it shows up in the volume data
The interesting operational consequence is that the caps can bind between scheduled reviews. When they do, the Nasdaq-100 methodology allows an unscheduled reweighting, a special rebalance, which is announced in advance and executed by every tracking fund at the same reference price.
That concentrates enormous volume into single sessions in the affected members, which matters for two things on this site. It puts an event session inside any volume baseline that spans it, raising the denominator for weeks afterwards. And it is one of the four causes that decide a monthly volume ranking. A company can top such a list entirely because its index weight was adjusted.
Two definitional choices worth noticing
No financial companies. The exclusion is in the rule, and it is not a small thing: an index without banks or insurers responds differently to interest-rate news than the market does, and part of what looks like a technology tilt is the absence of the sector that would have offset it.
One exchange. Membership requires a Nasdaq listing, so the index is a selection from a venue rather than from the American market. A company of identical size listed on the NYSE is not eligible. That is the same property the Composite has in a more extreme form, and it is the reason both are measures of a listing rather than of an economy.
Neither choice is a defect. Both are reasons the index is quoted for things it does not measure, and between them they explain most of the difference in behaviour between this index and a broad float-weighted one over any given quarter.
Capping is not equal weighting
The two are regularly conflated, and they are different in kind. An equal-weighted index gives every member the same weight regardless of size, so it answers "what did the typical member do". A capped index is value-weighted everywhere except at the ceiling: it keeps the ordering of the market and truncates the top of it.
The consequence is that a capped index behaves like a value-weighted one most of the time, and departs from it only when the caps bind. In a market where no member approaches the ceiling, the cap does nothing at all and the index is exactly a capitalisation-weighted index of its members. In a highly concentrated market it becomes a hybrid, and the degree of hybridity changes as prices move, which is an unusual property for a benchmark to have.
That variability is the honest criticism of the construction, more than the departure from market weights. A measure whose method effectively changes with the level of concentration is harder to reason about over a long history than either of the pure alternatives: a comparison of this index across two decades spans periods when the caps were inert and periods when they were doing a great deal of work, and nothing in the level says which.
Reading a capped index alongside a breadth count
One consequence of capping deserves its own note, because it cuts against the usual argument for breadth data. In an uncapped value-weighted index, a rise carried by two enormous companies can lift the index while most members fall, the situation the breadth section exists to make visible.
Capping mutes that, slightly, by holding the dominant members to a ceiling and giving weight to the rest. It does not remove the problem: a hundred members with a fifth of the index in two of them is still concentrated, and a breadth count across the exchange still answers a question the index cannot. But it is the one construction in this section where the methodology partly compensates for the effect, and that is worth knowing before treating all value-weighted indices alike.
Frequently asked questions
What is the Nasdaq 100?
An index of around a hundred of the largest non-financial companies listed on the Nasdaq exchange, weighted by market capitalisation with a modification: concentration limits. Membership is decided by a published rule based on size and liquidity, reviewed annually, which makes it more reproducible than a committee selection, and the exclusion of financial companies is a definitional choice with real consequences for what the index represents.
What is a concentration cap?
A ceiling on how much of the index any single member, or any group of the largest members, may represent. When the market pushes a member above its ceiling, the excess weight is taken away and redistributed across the other members in proportion to their own weights. The result is an index that deliberately does not reflect what the market did, which is the point of it.
Why would an index want that?
Because funds tracking it face diversification requirements. Regulated funds in several jurisdictions must limit how much of their assets sit in any one holding, so an uncapped index dominated by a handful of enormous companies becomes untrackable for the funds that are its main users. The caps exist to keep the index investable, not to measure anything better.
What does capping cost?
Accuracy as a measurement. An uncapped capitalisation-weighted index answers "what happened to the value of these companies"; a capped one answers a modified version of that question, and the modification grows with concentration. The figure on this page computes a redistribution: capping two dominant members moves several percentage points of weight to companies that did not grow, which then determines part of the index return.
Does it need a special rebalance?
Yes, and this is the practical detail worth knowing. Alongside the scheduled reviews, the methodology provides for a special rebalance when the concentration limits are breached, so a period of strong performance by the largest members can force an unscheduled reweighting. Those events are announced and they concentrate large volumes into single sessions, exactly as an ordinary rebalance does.
How is it different from the S&P 100?
Three ways, all definitional. This index applies concentration caps and the S&P 100 does not, so their treatments of a dominant member diverge. This one excludes financial companies and is drawn from one exchange; the S&P 100 is drawn from the S&P 500 across venues. And membership here follows a published rule while the S&P selection is a committee decision. Similar counts, different constructions.
Is it a technology index?
Not by definition, and largely so in practice. Nothing in the rule mentions sectors — it excludes financials and takes the largest remaining Nasdaq listings — but the companies that chose a Nasdaq listing and grew largest have mostly been technology businesses. That is a consequence of listing history rather than a design intent, and it is why the index is quoted as a technology proxy while its methodology says nothing about the subject.
Where does it fit alongside the Composite?
They share a venue and nothing else. The Composite includes everything listed on the exchange with no selection at all; this index takes a fixed count of the largest non-financials and adjusts their weights. Quoting a level from one against a level from the other is a category error, and the Nasdaq page sets out why the Composite is the more unusual of the two.