Equal-weight vs. cap-weighted S&P 500: what each one is and what the data say

Cap-weighted is the usual S&P 500, the one quoted in the news. Each company weighs what it is worth in the market: if one is worth twice another, it weighs twice as much. As prices move, weights adjust by themselves, without buying or selling. S&P uses float-adjusted market cap, that is, only the shares that trade freely.

Equal-weight is the same companies with the same weight each, about 0.2%. Between rebalances, the ones that rise end up weighing more, so the official index (S&P 500 Equal Weight) resets them every quarter, in March, June, September and December. That means selling a little of what has risen most and buying a little of what has fallen most, four times a year.

Sources: S&P 500 Equal Weight Index FAQ and its methodology (S&P Dow Jones Indices).

The 500 companies added up to $69.9 trillion. Twenty companies made up half; the 250 smallest, half of the index by count, weighed 8.2% together. NVIDIA and Apple together (14.7%) weighed more than the 250 smallest combined. In the equal-weight index, each of those 250 weighs as much as NVIDIA.

Sectors shift the same way. Technology was 38.0% of the cap-weighted index and 17.0% by number of companies (85 of 500). Industrials went the other way: 7.3% by market cap, 15.2% by count. Holding equal weight means, in practice, less technology and more industrials, utilities or real estate.

These figures use total market cap, not the float-adjusted cap S&P uses, so the index's official weights are somewhat different. The order of magnitude is the same.

Over 23.4 years they ended almost level: 0.27 points a year in SPY's favor, with slightly less volatility and a slightly shallower maximum drawdown, the one that bottomed in March 2009. They moved very similarly day to day (correlation 0.957), but not year to year:

RSP came out ahead in 10 of the 22 full years. Its best years against SPY were 2009 and 2010, coming out of the financial crisis, and 2022, the year rates rose; its worst, 2023 and 2024, when a few very large companies pulled the index up. In 2026, through September 11, they were nearly even: 13.10% RSP, 12.85% SPY.

Here equal weight leads by 1.7 points a year, but almost all of it comes from one stretch: from late 1998 to late 2009 the calculation returned 5.9% a year and SPY 0.6%, because the dot-com bust hit the largest companies hardest. From 2010 to 2019 they were level (13.1% and 13.0%); from 2020 to August 2026, SPY led (15.3% against 12.4%).

Looking at 10-year windows, starting every month, equal weight came out ahead in 140 of 212. Its best was March 2000 to March 2010, 7.6 points a year ahead; its worst, one of the most recent, May 2016 to May 2026, 3.2 points behind. The eight market stretches, one by one, are in the guide to investing in the S&P 500 from Spain.

• Size. Equal weight gives the index's mid-sized companies the same weight as the giants. When the largest do better, as in 2023 and 2024, cap weight wins; when they do worse, as in 2000-2002, equal weight wins.
• Rebalancing. Resetting weights each quarter means selling some of what has risen and buying some of what has fallen. It helps when prices reverse and costs when the winners keep winning.
• Sectors. Less technology and more industrials or utilities: the result also depends on which sector leads each period.
• Concentration. With 40% in ten companies, what those ten do decides much of the cap-weighted index. In equal weight, no company weighs much more than 0.2% for long.

None of these reasons says which will return more going forward. They explain why the winner has changed from one stretch to the next, not which stretch comes next.

The model portfolios split money equally among their holdings. Comparing them with the cap-weighted S&P 500 would mix two effects, which companies are picked and how the money is split, and you could not tell which one does the work. So the benchmark for the equal-split version is equal weight, and for the cap-weighted version, SPY.

RSP does not exist before May 2003 and the history starts in 1998, so the engine calculates equal weight itself with every company that was in the S&P 500 each month, including those that later left. Where both exist, the calculation beats RSP by 0.57 points a year — the calculation pays no fee and no trading costs — and they move almost identically (correlation 0.999). Details in the methodology; how to read the volatility and drawdown in these tables, in the Sharpe and drawdown guide.

What is the equal-weight S&P 500?

The same S&P 500 companies, but each with the same weight: about 0.2% of the total at every quarterly rebalance. In the regular index each company weighs what it is worth in the market, and on September 18, 2026 the ten largest added up to about 40%.

Which has returned more, equal weight or cap weight?

It depends on the period. With the real funds, from May 1, 2003 to September 11, 2026, RSP (equal weight) returned 11.24% a year and SPY (cap weight) 11.51%. With this site's long calculation, from 1998 to 2026, equal weight returned 10.37% against SPY's 8.68%. Almost all of that lead was built between 2000 and 2009.

Is the equal-weight S&P 500 riskier?

With these funds' data, somewhat more volatile and with somewhat deeper drawdowns: from 2003 to 2026, RSP had 19.72% annual volatility and a 59.93% maximum drawdown, against 18.49% and 55.20% for SPY. In exchange, it depends less on a few companies.

What is RSP?

The US exchange-traded fund that has tracked the equal-weight S&P 500 since May 2003, with a 0.20% annual fee. SPY tracks the cap-weighted index, at 0.09%. They are cited here as yardsticks, not as a suggestion to buy.

Why does this site compare its portfolios with equal weight?

Because its portfolios split money equally, and comparing them with a cap-weighted index would mix two things: which companies are picked and how the money is split. The cap-weighted version of each portfolio is compared with SPY.

• Some figures are real and some simulated. RSP and SPY are real funds, with their fees inside (0.20% and 0.09% a year). The 1998-2026 equal weight is a calculation before costs: with costs it would have come out somewhat lower.
• In dollars and before taxes. For anyone investing in another currency, the exchange rate adds or subtracts.
• The weights are from one day. September 18, 2026, using total market cap. They change every session.
• One window and one market. With another start or end date, the winner can be the other one; the table by year shows it.

More guides: how to invest in the S&P 500 from Spain, how to read Sharpe and drawdown, what the P/E ratio is. All of them in the guides. What an amount invested in SPY with dividends reinvested would have become, from any start date, in the S&P 500 calculator.