How Benchmark Indices Are Built

The S&P 500 sounds like a simple idea, the 500 biggest US companies. It isn't quite that, and the gaps between "simple idea" and "actual published number" are exactly where index construction lives. Which 500 companies, weighted by what, updated on what schedule, adjusted for what — every one of those choices is a rule someone wrote down, and those rules determine what number a portfolio manager is measured against every single day.

The choices that build an index

Weighting. Most major equity indices are market-cap weighted: a company's index weight equals its market value divided by the sum of everyone's market value. That means the index's return is dominated by its largest names — if a handful of mega-caps rally hard, they can carry the whole index even if most constituents are flat or down. Alternatives exist: equal-weighted indices give every constituent the same weight regardless of size, and fundamental-weighted indices weight by revenue or book value instead of market price.

Float adjustment. Market cap weighting typically uses free-float market cap, excluding shares held by insiders, governments, or other companies that aren't realistically available to trade, rather than total shares outstanding. A company that's 60% owned by its founder contributes far less index weight under float adjustment than its raw market cap would suggest, because most index providers only want to weight by the shares an ordinary investor could actually buy.

Rebalancing. Indices are reconstituted on a fixed schedule, quarterly for the S&P 500, annually for the Russell indices, adding companies that now qualify and dropping ones that don't. The schedule matters because a stock's inclusion or exclusion can move its price mechanically, as index funds tracking billions of dollars are forced to buy or sell around the rebalance date regardless of their own view on the stock.

Corporate actions. Stock splits, spin-offs, mergers, and dividend payments all need mechanical rules for how they affect an index's level and its constituent weights, so that the index's return reflects real investment performance and not just an artifact of a company reorganizing its share count.

Worked example

Compare a market-cap-weighted and an equal-weighted version of a 3-stock index: Stock X ($800bn market cap, up 20% this year), Stock Y ($100bn, up 5%), Stock Z ($100bn, down 10%).

1. Cap-weighted return. Weights are 80%, 10%, 10%. Return $= 0.80 \times 20\% + 0.10 \times 5\% + 0.10 \times (-10\%) = 16\% + 0.5\% - 1\% = 15.5\%$.
2. Equal-weighted return. Weights are 33.3% each. Return $= \tfrac{1}{3}(20\% + 5\% - 10\%) = \tfrac{1}{3}(15\%) = 5.0\%$.

The same three stocks, the same three returns, produce a 15.5% "index return" under one weighting scheme and a 5.0% "index return" under another — purely from the construction rule, before any manager has done anything at all.

What this means in practice

A manager who is compared against a cap-weighted benchmark is implicitly being compared against a portfolio that's concentrated in whichever handful of stocks happen to be largest, which in some periods (heavy mega-cap leadership) makes beating the benchmark unusually hard for any diversified active manager, regardless of skill. Understanding index construction rules is what lets an analyst tell the difference between "the manager underperformed" and "the benchmark's own construction made this period unusually hard to beat."