How a stock index is built, and why the method changes what it tells you

An index is a rule for combining many prices into one number. The rule is a choice, and different choices produce genuinely different answers about the same market.

Market capitalisation weighting

The most common method. Each company counts in proportion to its total market value, so a company worth ten times another has ten times the influence.

The argument for it is that it reflects the actual investable opportunity: it is what you would hold if you owned a slice of everything. It also requires little trading to maintain, because weights adjust themselves as prices move.

The objection is that it mechanically holds more of whatever has gone up. In a concentrated market that means a handful of companies dominate, and an index of a hundred names can behave like an index of five.

Equal weighting

Every constituent counts the same regardless of size. This gives a much better picture of how the typical company is doing, and it is often the more useful number when you want to know whether a rally is broad.

It requires regular rebalancing, which costs money, and it tilts toward smaller companies, which behave differently from larger ones.

Price weighting

The oldest method and the hardest to defend. Companies count in proportion to their share price, which is an arbitrary number: a company can halve its share price with a stock split without anything changing about the business.

The Dow Jones Industrial Average still works this way, which is why it can diverge from broader measures for reasons that have nothing to do with the economy.

Why this matters when you read the news

When one index is up and another is down on the same day in the same market, weighting is usually the explanation. If the cap-weighted index rises while the equal-weighted one falls, a few large companies did well and most did not. That is a materially different day from the headline.

This is general education and not investment advice.