How Are Market Indices Constructed?

By EC Assets Research Team · Published 17 August 2026 · Updated 17 August 2026

Index Construction: The rules that turn a universe of securities into a single number: which constituents are eligible, how they are weighted, and when the list is refreshed. Weighting choice dominates - the same constituents with the same returns produce materially different index performance under cap, equal and price weighting.

What Index Construction Actually Determines

An index is a rule set, not an observation. Three decisions define it: which securities are eligible, how much weight each receives, and how often the answers are refreshed. Everything downstream — the benchmark a manager is judged against, the exposure a passive fund delivers, the risk premium a researcher measures — inherits those three choices.

The consequences are larger than the technical framing suggests. Weighting alone can produce a spread of more than ten percentage points a year between indices built from identical constituents with identical returns. A benchmark is therefore an active decision made once and then treated as neutral.

How It Works

Selection establishes eligibility: domicile, listing venue, minimum market capitalisation, free-float threshold, liquidity screens and, in some cases, committee discretion. Rules-based providers apply the screens mechanically; others retain a committee that may weigh representativeness and turnover against strict rules.

Weighting determines each constituent's influence:

$$w_i = \frac{\text{measure}_i}{\sum_j \text{measure}_j}$$

where the measure is float-adjusted market capitalisation, price, an equal share, or a fundamental quantity such as revenue or book value.

Rebalancing and reconstitution refresh the weights and the membership on a defined schedule. Cap-weighted indices are largely self-maintaining between reconstitutions, because a rising constituent's weight rises automatically. Every other scheme requires trading to restore its target weights, which is where index turnover originates.

Weighting schemes and their implicit positions

Float adjustment excludes shares not available to public investors — strategic stakes, government holdings, cross-holdings — so that index weights correspond to what an investor could actually buy.

Worked Example

A three-stock market, with one small constituent delivering the strong return:

Under the three main schemes:

• Cap-weighted: weights of 48.8, 48.8 and 2.4 percent give an index return of 5.85 percent.
• Equal-weighted: weights of 33.3 percent each give 16.67 percent.
• Price-weighted: weights of 76.9, 19.2 and 3.8 percent give 6.35 percent.

Identical constituents, identical stock returns, and a spread of nearly eleven percentage points. The equal-weighted index captures the small constituent's move because it holds an equal share of it; the cap-weighted index barely registers it. Neither is wrong — they are answers to different questions.

When It Applies (and Limitations)

Cap weighting is not neutral. It is frequently described as the market portfolio and therefore as assumption-free. It nonetheless embeds a systematic tilt: weights rise with price, so the scheme holds progressively more of whatever has appreciated. Concentration is a direct consequence, and it accumulates quietly until the top holdings dominate the index's risk.

Price weighting has no economic rationale. In a price-weighted index a high-priced share of a small company outweighs a low-priced share of a large one, and a stock split changes the index's composition without any change in the underlying business. The Dow Jones Industrial Average and the Nikkei 225 persist on this basis for historical reasons.

Equal weighting buys its tilt with turnover. Restoring equal weights means systematically selling what rose and buying what fell. That is a contrarian rebalancing rule with a real cost in transactions and capacity, and it is a large part of why equal-weighted returns look better on paper than after implementation.

Providers exercise discretion. Committee-maintained indices involve judgement in selection and timing. Foreseeable additions and deletions create the index effect: prices move ahead of the event as participants position for mandated passive flows, so the index buys after the move.

Regulatory caps distort weights. Diversification constraints under UCITS and similar regimes limit single-issuer weights, so a compliant index tracking a concentrated market deliberately deviates from pure cap weighting.

Backfilled history is not live history. Indices launched after the period they report have back-tested records shaped by rules chosen with knowledge of the outcome. Live-date disclosure is the check.

Why It Matters for Institutional Investors

The benchmark defines alpha. Excess return is measured against a specific rule set. A manager beating a cap-weighted index while running an equal-weighted process may be harvesting a size tilt rather than generating skill, and the attribution turns entirely on which index was chosen.

Passive exposure inherits the rules. An investor holding a broad index fund holds whatever concentration the weighting scheme has produced. When a handful of constituents reach a large share of index weight, a nominally diversified passive allocation becomes a concentrated one without any decision having been taken.

Tracking error has structural sources. Deviation between a fund and its index arises from replication method, rebalancing timing, cash drag, securities lending and withholding tax on dividends — not only from active positioning.

Turnover is a cost borne by holders. Reconstitution trades are executed by every tracker simultaneously. High-turnover schemes and predictable rebalancing dates transfer part of that cost to index investors.

Index choice is a governance decision. Because the benchmark determines both what passive money buys and what active managers are measured against, benchmark selection deserves the same scrutiny as manager selection. It is usually given far less.

References

1. Arnott, R. D., Hsu, J., & Moore, P. (2005). Fundamental Indexation. Financial Analysts Journal, 61(2), 83–99.
2. Perold, A. F. (2007). Fundamentally Flawed Indexing. Financial Analysts Journal, 63(6), 31–37.
3. S&P Dow Jones Indices. Index Mathematics Methodology. (https://www.spglobal.com/spdji)
4. MSCI. Global Investable Market Indexes Methodology. (https://www.msci.com)