Equal Weight vs Cap Weight: One of Them Is a Momentum Strategy

Most people file cap weighting under "neutral" and equal weighting under "a choice." That is backwards, or at least it is not obviously right, and the cleanest statement of why came from Seth Cogswell on Behind the Ticker in five words: "Cap weighting is basically a momentum play."

He is describing an arithmetic fact rather than making an argument. A cap-weighted index holds more of whatever has gone up, automatically, continuously, with no decision required from anyone. That is the definition of a momentum position. It just does not feel like one, because it arrives as the default.

Once you see both schemes as active choices with different embedded behaviour, the comparison gets much more useful than "which one wins."

What each rule actually does

Cap weighting sets each holding’s weight in proportion to its market value. A company worth $3 trillion in an index whose members total $50 trillion gets 6% of the fund. Nothing needs to be traded to maintain this. If the stock doubles, its weight rises on its own, because the numerator and the denominator both moved. The scheme is self-maintaining, which is why cap-weighted funds have famously low turnover and famously low costs.

Equal weighting gives every member the same weight. In a 500-stock index that is 0.2% each, regardless of whether the company is worth $4 trillion or $8 billion. This does not maintain itself. As prices move, weights drift apart, so the fund has to rebalance, typically each quarter, selling what has risen and buying what has fallen.

That rebalance is the entire difference. One scheme lets winners run without limit. The other systematically trims them.

Why cap weighting behaves like momentum

Cogswell’s four consecutive sentences on the show are the clearest short explanation of this trade-off available anywhere, so they are worth taking in order.

First, the historical claim: "if we compare equal weighting to cap weighting, equal weighting up until the last decade had outperformed their cap weighted counterparts over every single rolling decade in history, except for very recently."

Then the reason: "The reason why equal weighting outperforms cap weighting over the long run is because it takes advantage of mean reversion."

Then the mirror image: "Cap weighting is basically a momentum play."

And then the condition under which the momentum position wins: "we haven’t seen a bear market, we could argue, in 16 years, which has really favored cap weighting as people just pile into the same things, driving that wave larger and larger."

Read as a set, that is a complete theory. Equal weight harvests mean reversion and therefore does well when prices oscillate around fundamentals. Cap weight rides trends and therefore does well in a long, uninterrupted trend led by a narrowing group of names. Neither scheme is a free lunch, and each has a regime it quietly prefers.

Cogswell’s rolling-decade claim is his, and it deserves the caveat that results like it depend on the index, the period and the treatment of costs. The mechanism behind it does not depend on any of that.

The reconstitution trade nobody sees

There is a second, subtler cost inside cap weighting, and Rob Arnott of Research Affiliates has spent a career on it.

When an index adds and drops members, the funds tracking it must trade. Arnott put numbers on what those trades look like. On the buy side: "You’re buying stocks typically at twice the market multiple after they’ve doubled in the last couple of years." On the sell side: "You’re selling stocks typically at half the market multiple after they’ve underperformed by about 7,000 basis points on average."

He restated the same pair more compactly in a later conversation: "On average, stocks kicked out are kicked out at half the market multiple," and "Stocks that are added are added at twice the market multiple."

Buying at twice the market multiple and selling at half of it is not a strategy anyone would write down. It is a by-product of a weighting rule that defines membership by size.

Arnott also pushed back on the survivorship story people tell about additions: "for every Tesla or NVIDIA that gets added to an index, there’s a dozen or more companies that are hot bubble stocks popular, beloved, and then go on to crater." We remember the additions that kept climbing. The index owned all the others too.

The flip side, and the reason Arnott is interested in any of this, is that a scheme which rebalances against price has what he calls "a rebalancing alpha" available to it. Equal weighting is the simplest possible version of that.

Concentration, and why this is being asked now

The question gets asked in every cycle where the top of the index gets heavy, and it is being asked now for a reason.

Raymond Bridges described the situation plainly: "There’s 10 companies overall in both S&P and the NASDAQ 100, S&P 500 that really drive the index." Catherine LeGraw of GMO observed the same thing from the allocator’s chair, noting that many model portfolios "have drifted from model and have gotten pretty top-heavy U.S. growth and specifically market cap-weighted growth."

That drift is the point worth internalising. An investor who chose a cap-weighted index fund a decade ago and changed nothing has become far more concentrated than they were, without ever making a decision to be. Mannik Dhillon of Victory Capital referred to it as simply the issues "with concentration that have existed" in core exposures.

Arnott, characteristically, went further and offered a view: "I view the magnificent seven as somewhat of a bubble." That is his opinion and you can take it or leave it. The concentration itself is not an opinion, it is a number you can look up on any index fact sheet in about a minute.

What equal weight actually costs

An honest comparison has to include what you give up, because the case for equal weighting is usually made without it.

Turnover. Rebalancing quarterly means trading. That is direct cost, and in a taxable account outside an ETF wrapper it can mean realised gains. Cap weighting requires almost none of this.

A size tilt you did not necessarily ask for. Giving the 400th largest company the same weight as the largest is a systematic bet on smaller companies. When large caps lead, equal weight lags for that reason alone, independent of any mean-reversion argument.

Long stretches of underperformance. Elena Khoziaeva of Bridgeway put the factor investor’s honest disclosure in one sentence: "when the two factors are out of favor and for a long period of time, we will tend to underperform." That is said about small-cap value, and it applies with equal force here. A scheme that leans against the leaders will trail while the leaders lead, and that can run for years.

It is not a hedge. Equal weight is still fully invested in the same 500 companies. In a broad decline it declines. What it changes is which companies drive the result.

The framing that is actually useful

Both schemes are rules, neither of them is neutral, and each carries an implicit view of how prices behave.

Cap weighting says: let the market’s aggregate judgement set my weights, and accept that my concentration will rise when a few names win. It is cheap, tax-efficient, low-turnover, and it will feel wonderful in a trend and uncomfortable at the turn.

Equal weighting says: I will systematically trim what has run and add to what has lagged, and accept higher turnover and long stretches of trailing the index for it.

The right question is not which one is better. It is which embedded behaviour you want to own for the next ten years, and whether the one you currently hold got there by decision or by drift.

This is educational content and not investment advice. It is not a recommendation regarding any security. Investing involves risk, including possible loss of principal. Past performance does not guarantee future results.