For exchange-traded fund investors, understanding products' underlying indices and choosing the appropriate one to achieve the desired investment objective is a crucial piece of the investment process. The following article will help investors better understand the different nuances of index construction by highlighting the various methodologies used to construct an index as well as their specific advantages and disadvantages.
Index Definition
Before delving into the index construction, let’s take a step back first and define an index.
An index is a statistical measure representing the aggregated market value of its constituents. Each index represents a particular market, e.g. STOXX Europe 600 Index, or a portion of it, e.g. STOXX Europe 600 Financial Services Index, and therefore serves as a barometer for this market or industry. Indices can use different calculation methods which we will discuss in more detail in the next section.
Index Construction
The details of how the indices underlying ETPs are built are all too often overlooked. Weighting methodologies, rebalancing and reconstitution frequency, caps, and a host of other factors will all have implications for an ETP’s performance, its cost, and how it will fit within a broader portfolio.
First, let’s look at the four main weighting methodologies in more detail.
Market-Capitalisation Weighting
Float-adjusted market capitalisation weighting is the most common index weighting methodology and is firmly rooted in the tenets of modern portfolio theory. This method reflects the weighting of the security in relation to the aggregate size of its relevant market. Because closely held firms will have a smaller piece of their aggregate market capitalisation floated on public exchanges, the float adjustment serves to ensure the underlying liquidity of the holdings is superior relative to a pure market capitalisation weighting.
The easiest risk to spot in most indices is that of excessive concentration in a given constituent or number of constituents or perhaps in a specific sector. The MSCI Brazil Index is a prime illustration of concentration risk from the perspective of both individual constituents and sector representation as the index is very top-heavy, with its top-five components accounting for about 40% of its value. It is also highly concentrated from the perspective of sector exposure, with energy and materials firms representing almost half its value. Oil giant Petrobras and mega-miner Vale comprise 32% of the index’s value. In many instances, index providers have introduced caps on the weighting of individual constituents in order to avoid undue concentration. For example, the OMX25H index caps the value of any individual constituent at 10% of the index’s total. This prevents mobile handset giant Nokia from representing a massive portion of the index’s value.
Index-weighting methodologies can introduce certain biases that are also important to understand. For instance, market-capitalisation-weighted indices will by their very design tend to increase their allocation to those constituents that have increased the most in price and reduced their exposure to laggards over time. This introduces an inherent growth/momentum bias that is necessary to acknowledge.
During times of severe market dislocation, such as the technology bubble of the late 1990s, this inherent bias can lead to excessive concentration in the market’s most overvalued shares. Meanwhile, in the case of market-capitalisation- weighted fixed-income indices, market-capitalisation weighting can lead to excessive concentration in the obligations of highly leveraged, less creditworthy borrowers. A more detailed analysis of fixed income indices can be found here.
Price Weighting
Price weighting index components is also fairly simple. The weight of each underlying security is determined by dividing the price of the investment by the sum of the prices of all the index’s constituents. It is useful to illustrate this with a hypothetical example. We will build a price-weighted index consisting of four securities with the following prices: Security A (€10), Security B: (€20), Security C (€30), and Security D (€40). The sum of the prices of these four securities is €100. As a result, the weighting of security A in a price weighted index will be equal to €10 / €100, or 10%. Similarly, Security B will have a weighting of 20% (€20/€100), Security C will have a weighting of 30% and Security D 40%. As a result, the index constituent with the highest price also has the highest weighting and thus the highest impact on the index’s performance. The Dow Jones Industrial Average (DJIA), which tracks 30 major US companies, is perhaps the most widely recognised price weighted equity index.
Equal Weighting
In an equal-weighted index every security in the index gets the same weighting. For example, un a portfolio of four securities, each one will get a weighting of 25% in the portfolio. However, despite its simplicity, equal weighting has some drawbacks: (1) securities that comprise a large part of a given market’s total capitalisation are underweighted relative to a traditional market cap weighted index, while investments that represent only a tiny portion of a given market’s total capitalisation are overweighted vis-à-vis a traditional market cap weighted benchmark, (2) in order to maintain equal weightings these indices require frequent rebalancing--which can be costly.
Fundamental Weighting
Fundamental weighting typically involves looking at financial measures that are independent of security prices. Such measures might include book value, cash flow, revenues, earnings, dividends and the number of employees. This approach results in indices that tend to have a “value” tilt. This means that a fundamentally weighted index will have ratios of book value, earnings; dividends etc to market value are relatively higher than to similar market capitalisation weighted indices.
Price Return vs Total Return
Investors will often encounter two versions of the same index: a price return and a total return variant. The difference between the two lies in their treatment of income and/or dividend payments. Price return indices only measure the price changes of the underlying securities; they do not take into account any dividends or other forms of income that have been distributed. The total return indices, on the other hand, capture both price changes and the distribution of income. Moreover, investors have to distinguish between the net return and the total return. The net return excludes withholding taxes in the index calculation and is therefore a better (yet not entirely accurate) measure for an investors’ performance.
Rebalancing and Re-constitution
The frequency of index rebalancing and reconstitution also has important implications for ETP investors. The more often an index is rebalanced or reconstituted, the more costly it will be for an ETP provider to replicate that index. The expenses associated with these adjustments are ultimately borne by the end investor in the form of either a higher total expense ratio or poorer tracking performance. We discuss this in more detail in an article explaining the total cost of ETP ownership which can be found here.
Re-constitution is typically the result of periodic reviews of index constituents or corporate actions such as mergers, acquisitions, or bankruptcies. Reconstitution usually takes place after regular periodic reviews. Like re-balancing, re-constitution creates additional turnover within the affected indices, and again, results in embedded costs.
As the ETP market continues to grow, index providers have become ever more inventive. From more complex commodity futures strategies, to fundamentally weighted indices, to full-on active management, the diversity of strategies is always increasing and straying ever further from plain-vanilla equity and fixed-income indices. As this trend inevitably advances, it will become increasingly critical for ETP investors to have a firm understanding of the benchmark underlying their investment.
