EDHEC-Risk Indices in institutional investment management: results of a European survey
Noël Amenc, Professor of Finance, EDHEC Business School, Director, EDHEC-Risk Institute
Felix Goltz, Head of Applied Research, EDHEC-Risk Institute
Lin Tang, Research Assistant, EDHEC-Risk Institute
As the choice of an index is a crucial step in both asset allocation and performance measurements, it is useful to investigate index use and perceptions about indices. The recent EDHEC-Risk European Index Survey 2011, which elicited responses from 104 European institutional investment professionals, aims to contribute to this awareness by analysing the current uses of and opinions on equity indices and fixed income indices. It is our hope that this survey will provide unique insight into the users' perspective in the index industry.
What makes a good-quality index? Investors' evaluation criteria
We began the questionnaire by asking our respondents about how to evaluate the quality of an index. We find that liquidity, objectivity and transparency are the most important criteria investors have for indices. The importance of these criteria is quite intuitive. Liquidity of an index means that it can be used by many investors at the same time - trading large positions in vehicles tracking the index does not have an impact on the market prices of securities in the index. The importance of objectivity and transparency suggests that investors want to be able to understand and follow what exactly drives the index risk and return. On the other hand, respondents to our survey suggest that a buy-and-hold character is not a requirement for an index. This finding is interesting as their buy-and-hold nature is often cited as one of the key reasons for the dominance of cap-weighted indices in various asset classes. If objectivity, transparency and liquidity are key, this opens many possibilities to construct indices as in fact any set of transparent and liquid portfolio construction rules could be used as an index by investors.
Furthermore, the results also reveal that there is an obvious confusion linked to equating indexing with "passive investing". The majority of respondents actually does not think that indices should only reflect passive strategies (58%). Together with the low importance of the buy-and-hold property of an index, this suggests that respondents widely accept the notion that their indices can move away from traditional cap-weighted approaches. Respondents do however indicate that indices should not be based on alpha (75.2%). Overall, the responses thus suggest that investors indeed ask for "normal returns" from "systematic" portfolios when investing in an index. Of course, finance theory and concepts leave a wide range of choice for what these systematic portfolios could look like.
Slicing and dicing the market: broad indices versus sub-segment indices
Within equity indices or bond indices, investors have a wide set of categorisations to choose from. In equity investing, some investors use broad worldwide or regional indices, while others prefer to separate indices by country, sector or style. When comparing the results for the importance of sub-segment indices across different asset classes, we find that sub-segment indices are of relatively little importance to equity investors - where broad market indices dominate - but are of prime importance to bond index users (figure 1). This finding may be explained by the fact that equity investors mainly seek exposure to the overall equity risk premium without necessarily having a clear view on drivers of differences of expected returns within equity markets. Of course, the main motivation behind equity investing is often to seek performance in return for taking on equity risk, rather than hedging of clearly specified risk factors.
On the contrary, hedging of clearly defined risk factors is a much more important issue for bond investors. As investors are concerned with their exposure to bond market risk factors, such as interest rate risk, inflation risk, and credit risk, they need fixed-income instruments with specific characteristics to match their desired risk exposures. Hence, sub-segment indices could offer more flexible solutions for different needs of fixed-income investors.
Another important background to the different findings for the use of equity sub-segment indices versus fixed income sub-segment indices is that the relevant underlying risk factors are quite consensual in the fixed income arena but much more debatable for equity investments. While, at least since Ross's work in the 1970s on Arbitrage Pricing Theory and its subsequent application in investment practice, there is a consensus that equity returns are driven by multiple risk factors, there is no such consensus when it comes to specifying the risk factors or the portfolios that should represent them. In fact, researchers often prefer to use implicit factors to let the data decide which factors matter most, while index providers have largely concentrated on explicit factors due to the more straightforward use of explicit stock characteristics in index construction.
While index providers have recently made forays into equity indices that replicate explicit financial factors, such indices are not widely used. Likewise, sector and style indices are relatively little used despite the evidence that sectors and styles are important drivers of equity correlations, thus justifying the use of such indices as building blocks in portfolio construction.
In the end, due perhaps to the lack of consensus on relevant explicit equity risk factors, and due to the lack of precise hedging demands of investors for categories like styles and sectors, sub-segment indices play a rather minor role in equity indexation. Whether or not categories could be defined that are more relevant for investors and would lead to more widespread usage is a question beyond the scope of what we can answer from the survey results.
Generic indices versus investors' objectives
The take-up rate of sub-segment indices is not the only marked difference between equity and bond index use. When comparing the problems investors see with the existing indices across different asset classes, we find that there are pronounced differences. For example, equity investors are mainly concerned that standard cap-weighted indices overinvest in overpriced stocks and provide poor diversification within the constituent universe. In contrast, fixed-income index users pay more attention to reliable duration exposure and are concerned with liquidity issues.
This difference in perceived problems with indices in different asset classes is also consistent with the results above on the importance of broad and sub-segment indices: equity indices are used for seeking performance while bond indices are used to hedge risk exposures.
Despite these significant differences on objectives and in perceived issues, index providers typically apply the standard cap-weighting approach indifferently to both equity and fixed income indices. It seems surprising that a weighting methodology is applied uniformly to different asset classes, which investors use to pursue different objectives and for which they see different sets of issues. Investors' answers across the survey suggest that, even if the cap-weighting approach were the best approach in one asset class, it is unlikely that it is also necessarily the best answer to different issues in another asset class.
Even within each asset class, it should be noted that investors see a variety of issues with index investments and the most appropriate weighting scheme for a given investor may depend on which problem is the most relevant for him. It seems unreasonable to assume that a single weighting scheme could be applied uniformly to all investors, given the differences in their objectives.
Rather, there may be a case for the design of more "objective oriented" as opposed to generic indices. Designing indices that explicitly relate to objectives may be more challenging than designing somewhat ad-hoc indices, but would potentially provide more useful tools for performance measurement and asset allocation of investors.
Of course, standard cap-weighted indices do not take into account such investor-specific objectives, as they are constructed on the notion of providing an aggregate snapshot of the market average. But even many of the more recently introduced alternative weighting schemes show little difference with cap-weighting when it comes to taking into account differences between investors or different requirements for different asset classes. We now turn to a discussion of the use of various non-cap-weighted indices.
Adoption of alternative weighting schemes
Figure 3 shows that the adoption of alternative weighting schemes is reasonably high among respondents concerning equity indices, but in the bond index area, adoption of non-standard indices remains much lower.
Interestingly though, for both stock and bond indices only a minority of respondents does not see the existing standard indices as problematic. For bond indices, this implies that although investors are not satisfied with the current weighting schemes, they have not made a move to alternatives.
The last row of figure 3 offers a reasonable explanation for this observation - about one-fifth of respondents are not familiar with alternative approaches in bond indices, while only about 10% of equity investors are unfamiliar with alternative indices in this area. Hence, unfamiliarity with alternative approaches for bond indices is one of the most plausible barriers to the adoption of alternative weighting schemes at the current stage.
Overall, our survey results have shown that European institutional investors are well aware of a range of important problems with existing standard indices in the equity and fixed income arena. As a consequence, they have started to adopt alternative weighting schemes. As long as indices remain transparent and objective and focused on beta management rather than alpha generation, respondents to our survey are open to exploring new ways of designing indices that are suitable to their investment objectives, thus opening the room for further innovation and research in the field of index construction.