Let’s pretend that FTSE is not an index calculator, but an active fund manager. Its investment philosophy is to invest in companies within a market, country or sector universe by selecting investments based predominantly on relative market size. In addition, potential investments are screened on investibility criteria: liquidity and free float. As a consequence, companies that have recent good relative performance tend to be selected, whereas badly performing companies are more likely to be sold.
We have now created our investment portfolio – we have also created the FTSE All-World, FTSE 100 or FTSE All-Share Information Technology Index, depending on the initial universe. So are index calculators just active fund managers? An investment philosophy described in this way is unlikely to win an active fund manager many investment mandates, but somewhat surprisingly many active fund managers fail to construct portfolios that perform better than this.
The reason for this is that an investment portfolio (or stock market index) constructed in this way captures materially all of the investment opportunities available in the selected market or universe. But some commentators still think this type of selection criteria introduces a degree of benchmark risk.
In analysing the risk associated with an investment strategy, it is necessary to establish a measure against which that risk will be calculated. Ideally, this will be a universe that is either risk- free or assumed to include all the portfolio risks. Peer group comparison, against a large enough and diversified peer group, may be a good proxy for a universe containing all the risks. Good performance against this universe may be an acceptable target. Hence measuring how your portfolio differs from the universe of portfolios should be a good measure of the risk being taken.
However, the whole universe may be assuming a common risk. For many years, UK fund managers maintained a lower exposure to the US market than historic performance indicates was optimum. It could be argued that the whole universe had a systematic risk. However, if the universe was considered to be the true comparative measure, the fact that it might have had an underweight position in the US (relative to a global index) is no more relevant than the fact that it had an underweight exposure to, say, classic automobiles – an asset class that also performed better than the universe.
Selecting a benchmark for comparison purposes does not, in itself, safeguard for the problems inherent in peer group comparison. The benchmark selected may or may not offer a better performance yardstick – it depends on the objective. If the objective is stated to be performance against an equity market or a group of equity markets, the index benchmark starts to look attractive. The question then is whether there is risk associated with the choice of benchmark.
Very few indices set out to include every security in a market place. They are normally restrictive in their initial selection and most screen for investibility. Indices need to be investible – there is little merit in including a company within an index that investors are unable to buy due to poor liquidity. There is also the question of materiality. Including many small companies within an index of predominantly large companies will often have no material impact. All they do is increase the cost to the portfolio manager of trying to benchmark against the index. In the FTSE All-World index, adding the next largest non-constituent company (even from one of the largest markets) would only increase the capitalisation of the index by about $500m to $23,500bn (less than 0.002%).
In the UK, the FTSE All-Share index contains 800 constituents out of the 1,660 in the market – just 48%. But it represents 98.6% of the total market capitalisation. Adding the next largest non-constituent company to the FTSE All-Share would increase the market capitalisation of the index by about £50m to £1,800bn. Even if this new company were to double in value, it would have no noticeable impact on the level of the index.
Broad-based indices like these are not selective and hence are not introducing any material degree of benchmark risk. However, other more focused indices like the FTSE Eurotop 100 or the DJ STOXX 50 can be materially different from the market as a whole. The key here is to ensure each investor is using an appropriate index for its needs.
Graham Colbourne is director of FTSE in London