Sally Bridgeland describes what makes a good benchmark index
The number of equity stock market indices available to pension fund investors has increased rapidly in recent years. Indices are also being used more and more – as the reference points for strategic investment benchmarks, passive investment products and derivatives contracts. The leading index providers are keen to have their share of the expanding European market. This survey looks at the current range of indices on offer and takes a step back to the qualities pension fund investors should be looking for in an index in the context of how the indices are used.
The availability of derivatives contracts is relevant to institutional investors wanting to implement changes in strategy or protect risks relative to a strategic benchmark. However, the main use of indices for pension funds are as strategic investment benchmarks.
At a fundamental level the benchmark is how the pension fund investors translate their investment strategy into a fund manager structure. The benchmark shows which asset classes are preferred for investment, either by stating them explicitly (eg, 50% equities; 50% gilts) or implicitly by referring to some peer group reference point. A tailor-made benchmark will usually be expressed as a combination of stock market indices.
The indices are chosen to represent the characteristics of the asset classes, but will be fundamental reference points for fund managers. Both the constituents and the constituents’ weightings which make up the index are important. The companies in the index will define the universe of stocks from which fund managers will construct their portfolios. Managers will tend to focus on taking positions which are underweight or overweight relative to the stock weightings defined by the index itself.
The way that indices are used in the portfolio construction process reflects investors’ requirements to manage risks and achieve a target level of outperformance relative to the benchmark indices. By managing the portfolio to control risks against the benchmark, fund managers try to reduce the risk of underperformance against the benchmark and therefore the risk of having a dissatisfied customer. But this has highlighted the importance of the choice of index. Investors and investment consultants cannot just assume that because an index is widely used as a benchmark that it is the best that could be available.
The work that Bacon & Woodrow and Barclays Global Investors carried out with FTSE International this year in launching the FTSE multinationals family of indices is an example of a search for a better approach. The launch of the new indices was driven by our research, which suggested that the inclusion of multinational companies within the FTSE All-Share index was increasing the riskiness of the index for UK pension fund investors. The FTSE All-Share is by far the most common index used by UK pension funds, yet the concentration of assets in a few stocks (including UK-incorporated oil and pharmaceutical giants) was running at an all-time high. Our work also indicated that liability risk management could be improved by splitting the All-Share roughly in half – with local UK companies in their own asset class and multinational companies included with other global companies in a new asset class of their own.
So what makes a good benchmark index? In the face of bright spotlights from a number of different directions, the requirements are quite onerous. The following list picks out the main issues to consider:
q How are the stocks weighted? If the stocks in the index are not weighted by the market capitalisation of the stocks available for outside investors then you can end up with the odd result that it is mathematically impossible for all investors to hold the index. This generally means that the benchmark index should be market-capitalisation weighted, but adjusted to reflect the free float of stocks available to investors and those which are locked up in cross-holdings by other companies. This approach is also consistent with the buy and hold strategy used by an index-tracking manager – here managers only need to trade when they reinvest dividends and when the equity structure of the companies in the index change.
q What information is available? Investors need regular information on the performance of price and total return indices and the index constituents. This data should have a long and high quality history.
q Is the index investible and liquid? The stocks in the index should be liquid and investible at a reasonable level of transaction costs. This means that investors will be able to replicate the portfolio underlying the index if they have sufficient assets and, in doing so, achieve very close to the index return.
q How broad is the coverage of the market? The primary purpose of the index (when used as a benchmark) is to reflect accurately the total economic activity of the relevant country, sector or region. From a purely theoretical perspective the ideal index should include every security. However, there is a trade-off between breadth and the investibility and liquidity of the smaller stocks. In practice, it is essential that securities from all important industries, sectors and regions are represented.
q Does the index have the same characteristics as the market it represents? To check that the benchmark index is representative of the asset class, the risk exposures of the index and the universe of securities it tracks can be compared – these should be reasonably similar.
q How concentrated is the index in the largest stocks? The index should represent a well-diversified portfolio, with stock selection risk a small proportion of the portfolio’s total risk.
q Are the index construction rules clear and consistent? The process of making changes to the index composition should be predictable and the number of changes should be kept low to minimise unnecessary and costly trading. The construction rules should be as objective as possible so that constituent changes can be anticipated by managers and handled in an efficient way. Judgmental rules may be necessary to provide a more accurate representation of the economy or some other desirable characteristic such as a more appropriate industry distribution or to limit the number of changes in the index constituents - but these should be kept to a minimum. All corporate events (such as dividend payments and stock splits) should also be treated consistently over time.
q Are sub-indices available? Indices that break down the index by sector, size and style of stocks provide a useful input to the strategic and tactical asset allocation process for active management.
q How does the index provider make its money? Investors may be concerned about the potential for conflicts of interest between the various activities of the index provider and may require an independent provider.
q How widely used is the index? Low transaction and transition costs are generally related to the liquidity of the stocks in the index. However, if an index has liquid futures contracts associated with it, and there are substantial opportunities for crossing of trades by the manager, transaction costs will be minimised. The existence of good crossing activities is directly related to the amount of assets managed against the index, the range of clients using the index and the number of managers with funds benchmarked against the index.
The survey answers provide some pointers to the best indices to be used for benchmarks for particular asset classes when matched up against these criteria. However, for a global equity strategy and fund manager structure it is also important to find a set of indices that cover the world without any overlap – or significant gaps. This continues the fundamental idea that the benchmark indices define the boundaries of the universe of stocks from which portfolios are constructed.
This is also an issue to consider when setting the benchmarks for a pan-European equity portfolio. The coming of the euro meant that, at least at face value, ‘Euro-zone’ companies have more in common with each other than those in the rest of Europe. This has prompted pension funds to consider a strategic allocation to Euro-zone companies and to require a representative Euro-zone index as a benchmark. However, this may not fit together neatly with other European country allocations and is not proving a popular strategy for those funds based outside Euro-zone, such as those in the UK.
Which ever approach is chosen, derivatives contracts are increasingly being used to make changes in strategic asset allocation. The narrower indices provide a basis for these contracts – for example, the London International Financial Futures and Options Exchange (Liffe) offers exchange-traded futures and options contracts based on the FTSE Eurostars, FTSE Eurobloc 100, FTSE Eurotop 100, FTSE Eurotop 300, FTSE Eurotop 300 ex UK, MSCI Euro and MSCI Pan-Euro indices. When switching to a new benchmark it makes sense to choose a derivatives contract that is as similar as possible to the benchmark index. Given the differing approaches adopted by the different index providers this generally means picking a narrower index from the same brand.
The income that index providers receive from derivatives contracts means that it is also in their best interests to provide a derivatives version of the broader benchmark indices.
The better-known indices can also form the basis of index-tracking products for individuals. Here, the appeal of a strong brand (such as the FTSE 100 in the UK) is more important than the other qualities of indices that make them more appropriate benchmarks for active management of pension funds.
Given the differing scope of coverage and construction approaches highlighted in the survey responses, the choice of index will have an impact on the returns and risks of any investment decisions which pay reference to the index. A careful and informed choice is required.
Sally Bridgeland is head of investment research at Bacon & Woodrow in London