7.1 Equity Index Construction Methodologies

7.1 Equity Index Construction Methodologies

1. Why Weighting Scheme Is an Investment Decision

Selecting an index is often treated as a neutral, administrative choice. It is not. The weighting methodology determines the factor exposures, concentration, turnover, and capacity of the resulting portfolio. Two indices covering the identical universe of securities can deliver materially different returns and risk purely because of how they weight constituents.

For a consultant, this matters in two places: benchmark selection in the investment policy statement, and passive implementation, where the choice of index is the entire investment decision.

2. Price Weighting

A price-weighted index weights each constituent by its share price. The index value is the sum of constituent prices divided by a divisor that is adjusted for splits, spin-offs, and constituent changes to preserve continuity.

$\text{Index} = \frac{\sum_{i=1}^{n} P_i}{\text{Divisor}}$

The Dow Jones Industrial Average and the Nikkei 225 are the surviving major examples.

The defining flaw: influence is determined by share price, which is an arbitrary artifact of how many shares a company has chosen to issue. A company trading at $600 per share has ten times the index influence of one trading at $60, regardless of which is the larger business.

Worse, a stock split mechanically reduces an issuer's index weight without any change in the company's value. A 2-for-1 split halves the price and therefore halves the issuer's influence on the index — a pure accounting artifact driving a change in portfolio exposure. This is the single most-tested weakness of price weighting.

Implicitly, a price-weighted index corresponds to holding one share of each constituent.

Worked Example: The Effect of a Split

A three-stock price-weighted index contains stocks priced at $90, $60, and $30. The divisor is 3, so the index value is $(90 + 60 + 30)/3 = 60$.

The $90 stock splits 3-for-1, becoming $30. To preserve continuity, the divisor is reset so the index value is unchanged at 60:

$\frac{30 + 60 + 30}{d} = 60 \;\Rightarrow\; d = \frac{120}{60} = 2.0$

Before the split, the largest company represented $90/180 = 50%$ of index influence; afterward it represents $30/120 = 25%$. The business did not change; its index weight halved.

3. Capitalization Weighting

A capitalization-weighted index weights each constituent by market value — share price multiplied by shares outstanding.

$w_i = \frac{P_i \times Q_i}{\sum_{j=1}^{n} P_j \times Q_j}$

The S&P 500, MSCI ACWI, FTSE 100, and Russell 1000 are cap-weighted. This is the dominant global standard for identifiable reasons:

• Macro-consistency. Cap weighting is the only scheme that all investors can hold simultaneously, because it mirrors the aggregate market portfolio. Every other scheme requires someone to take the other side.
• Self-rebalancing. As prices move, weights adjust automatically. No trading is required to maintain the weighting, which produces very low turnover, high capacity, and minimal transaction costs and taxable gains.
• Theoretical grounding. It corresponds to the market portfolio of the CAPM.

Float adjustment. Modern cap-weighted indices weight by free-float market capitalization, excluding shares not available to public investors: closely held founder and family stakes, government holdings, cross-holdings, and strategic corporate stakes. Without float adjustment, an index would direct investors to buy shares that cannot be purchased, creating tracking problems and price distortion. This matters most in markets with heavy state ownership.

The principal criticism is that cap weighting is momentum-implicit: a security's weight rises as its price rises, so investors hold progressively more of appreciating assets. If a security becomes overvalued, a cap-weighted investor holds proportionally more of it. This is the core of the "noisy prices" argument for alternative weighting, and it manifests as concentration risk when a small number of very large constituents dominate an index.

4. Equal Weighting

An equal-weighted index assigns identical weight $1/n$ to every constituent regardless of size.

Systematic consequences:

• Small-cap tilt. Relative to cap weighting, equal weighting massively overweights the smallest constituents. In the S&P 500, the smallest company receives the same weight as the largest.
• Value tilt. Rebalancing back to equal weight systematically sells constituents that have appreciated and buys those that have declined — a mechanical contrarian rule.
• Turnover. Weights drift with prices, so periodic rebalancing is mandatory. This generates turnover, transaction costs, and — critically for taxable clients — realized capital gains that a cap-weighted index does not produce.
• Capacity constraint. Buying equal dollar amounts of large and small companies places disproportionate demand on the least liquid constituents, limiting the strategy's scale.

Equal-weighted indices have often outperformed their cap-weighted counterparts over long historical periods, but the correct interpretation is that this is largely compensation for size and value factor exposure plus rebalancing, not evidence that equal weighting is a free improvement.

5. Fundamental Weighting

Fundamental weighting — associated with Robert Arnott and Research Affiliates — weights constituents by accounting measures of economic footprint rather than by market price: sales, cash flow, book value, and dividends, often blended.

The argument: if market prices contain error, then cap weighting overweights overvalued securities and underweights undervalued ones, creating a systematic drag. Weighting by fundamentals severs the link between price and weight, so pricing errors do not propagate into portfolio weights.

The counter-argument — and candidates should be able to state it — is that fundamental weighting is simply a value factor tilt in different packaging. When a stock's price falls while its sales and book value are unchanged, fundamental weighting buys more of it. That is the definition of a contrarian value strategy, and its returns are largely explained by established value and size factors rather than by a distinct "indexation" insight.

Like equal weighting, fundamental weighting requires periodic rebalancing and therefore incurs turnover and tax consequences.

6. Comparative Framework

7. Reconstitution and Index Effects

Indices periodically reconstitute, adding and deleting constituents according to published rules. Because index-tracking assets must trade to match, reconstitution creates predictable flows.

The classic index effect is the price appreciation of a security in the window between the announcement of its addition and its effective inclusion, as trackers are forced to buy. The effect has diminished substantially as index providers moved to less predictable rules, staggered implementation, and as arbitrageurs began anticipating changes — but front-running of reconstitution remains a documented cost borne by index investors, and it is a legitimate criterion when evaluating index methodologies.

Annual Russell reconstitution in late June is the most prominent scheduled example and reliably generates one of the highest-volume trading days of the US year.

Consulting implication: an index with mechanical, widely telegraphed rules is more transparent but more susceptible to front-running; one with buffer zones and staggered implementation reduces turnover and trading costs at the price of a looser fit to its stated universe. Neither is universally superior, and the trade-off should be an explicit part of benchmark selection.