In the light of economical and political integration in Europe, asset allocation along country lines seems to lose impact. Moreover the introduction of the euro, which implies the elimination of currency risk within Euro-land, has strongly increased the demand for benchmarks representing the European stock market. To cope with this demand a growing number of Europe-based stock indices have been developed over the past few years. To appreciate the importance of benchmarking, it seems important to briefly summarise success factors for constructing an equity index.
Such an index has two main functions:
q Benchmarking: The intention is adequate coverage of the aggregate stock market or some component of the stock market over a specific period. In this sense the index can be used, for instance, as a benchmark for performance measurement.
q Tradeability: In this case the construction of the index is more orientated on the objective to serve as an underlying for stock index derivatives or as a basis for index funds.
Considering these two functions some factors can be identified, that determine the success of an index, measured as the overall market’s acceptance. These are:
q Representativity: The structure and constitution of the index should reflect the covered market as well as possible.
q Continuity: The index should not be subject to frequent changes or adjustments.
q Actuality: The index calculation should be based on “real” (transaction) prices to guarantee the tracking.
Investibility: This aspect is in particular relevant for the remainder of this article, as the European stock market per se is still a national segmented market. The investibility should guarantee that no restrictions exist for foreign investors, concerning for instance market entrance or free float.
International comparability: International investors in particular create a demand for indices that are comparable in a global context.
q Index derivatives: This item is mainly relevant for institutional investors, who are interested in hedging instruments in order to reduce their market risk exposure.
It is obvious that these success factors require different, often irreconcilable features of indices and we therefore conclude that a perfect index does not exist, as no index is able to serve all purposes equally well.
In recent years several indices covering European stocks have been developed – although no common European stock exchange yet exists – to satisfy investors’ needs.
To meet the sometimes contradictory requirements index providers have differentiated between so-called blue-chip indices and broad-market indices. The first group of indices contains approximately 50–200 constituent stocks each, representing a suitable percentage of the total market capitalisation.
The latter group consists of a substantially greater number of stocks – from 300 to over 700 stocks – to provide a better mapping of European stocks’ development. The broad market indices are therefore intended to serve as a performance benchmark, by stressing the representativity function mentioned above.
Furthermore we can distinguish between Euro-land indices and indices covering all of Europe. The latter group covers all developed national stock markets in Europe – that is, except for the Eastern European emerging markets. In the rest of this article these are referred to as ‘European indices’. The other category concentrates on the national stock markets of member countries of the European monetary union, so called Emu-indices (10 countries in general, as Luxembourg is excluded due to the lag of data distribution).
Table 1 gives an overview of some index families, following two criteria: covered universe and blue-chip versus broad-market indices.
The construction of the Dow Jones Stoxx index families refers to a criteria matrix, along market capitalisation and sector grouping, with a representation of 80% for each criteria. The broadest index, DJ Stoxx, has 652 constituent stocks from 16 countries. The DJ Euro Stoxx 50, an Emu-based blue-chip index contains 50 companies from nine countries, whereas the DJ Euro Stoxx, with 326 stocks represents 10 Emu member countries.
The FTSE Eurotop indices pursue another strategy, in that the selection criteria for the broad-market indices differ from that for the blue-chip indices. The FTSE Eurotop 100 contains 100 titles ranked by the importance of the home country’s stock market (in terms of market capitalisation). The broader FTSE Eurotop 300 consists of the 300 largest European stocks, in terms of capitalisation only. The Emu-based FTSE Eurobloc 300 is a subset of the FTSE Eurotop 300 and has just 149 constituent stocks, whereas the FTSE Eurobloc 100 has the indicated 100 (60% as a subset of the FTSE Eurobloc 300 and 40% to represent Emu sectoral structure).
The FT/S&P Actuaries World Europe index includes companies from 14 countries and represents with 722 stocks the broadest market index. The FT/S&P World Europe Eurobloc is a subset of the broader index consisting of 342 stocks. The selection criteria concentrate on the market capitalisation of sectors in each constituent country, in order to represent 85%.
The S&P Euro Plus consists of 200 European stocks, except for the UK, whereas the S&P Euro is a subset with 158 titles. The selection is again based on market capitalisation.
Finally the blue-chip indices from MSCI remain to be mentioned. The MSCI Pan-Euro contains 236 stocks from 16 countries, whereas the MSCI Euro has 130 constituent stocks. The applied selection criteria is also based on the market capitalisation.
Table 2 contains some summary statistics for the European indices. (It does not contain all the indices mentioned above because of the lag in the availability of data.) One can see that over the past couple of years the blue-chip indices show a better performance. The results do not depend on the choice of universe – the Emu-based blue-chip indices also outperformed the broad-market indices.
The empirical analysis was performed according to the market model in respect to a reference index – the MSCI Europe and MSCI Emu were chosen. The results of a correlation analysis are outlined in table 3.
The evaluation of the correlation as a measure of goodness of fit shows impressively that the indices not only improved their correlation in the last quarters, but have also moved in a smaller range. This result does not depend on the considered universe either.
We therefore conclude that the European benchmark discussion is not settled yet. From an investors’ perspective, at some point in the future one might expect some consolidation among Pan-European indices.
Andreas Grünbichler is professor of finance at the University of St Gallen and a member of the board of directors of the Swiss Institute of Banking and Finance. Steffen Graf is a PhD student in finance at the University of St Gallen and a research scholar at the Swiss Institute of Banking and Finance