Style analysis of portfolios and mutual funds is a subject of great interest both for investors and managers, as style management is related to funds’ risk and return characteristics. Two different methods make it possible to perform this analysis. The first one is the return-based style analysis (RBSA) and the second is called holding-based style analysis (HBSA).

Return-based style analysis and portfolio-based style analysis: two methods for analysing portfolio style

Return-based style analysis draws from Sharpe’s style analysis model, which stipulates that a manager’s investment style can be determined by comparing the returns on his portfolio with those of a certain number of selected indices. The oft-quoted words of Sharpe to justify his methodology are: “If it acts like a duck, assume it’s a duck.” As managers rarely have a pure style, Sharpe proposes a method whereby one can find the combination of style indices which gives the highest R-squared with the returns on the portfolio being studied. We recall that R-squared measures the proportion of variance explained by the model, and therefore gives the goodness of fit between the portfolio returns and the returns on the indices. The success of this technique relies heavily on the correct specification of the style benchmark indices used as regressors. They must correspond to the fund’s investment universe and must allow a complete description of the style of the fund. The major advantage of this method is that it is not necessary to know the securities that make up the portfolio or their proportions. It is therefore the only method that can be used when there is no data available on the composition of the portfolio, or if we are not sure that the data available are reliable.

The holding-based analysis, by contrast, consists in analysing each of the securities that make up the portfolio. The securities are studied and ranked according to the different characteristics that allow their style to be described. The results are then aggregated at the portfolio level to obtain the style of the portfolio as a whole. This method therefore requires the present and historical composition of the portfolio, together with the weightings of the different securities that it contains, to be known with precision. The analysis has to be carried out regularly, in order to take account of the evolution of the portfolio composition, as well as the evolution of the characteristics of the securities that make up the portfolio. The holding-based method requires more information on the portfolio than the return-based style analysis and will provide more precise information, as long as the data are exhaustive and reliable, so the problem of data availability appears to be the key point for this methodology. Moreover, the main weakness of this approach is the frequently subjective character of the classifications. Since the style analyses performed within this approach are specific to each manager, it is difficult for them to be reproduced by an external third party.

Holding-based style analysis: the difficulties

From a theoretical point of view, the holding-based method may appear to be better. The analysis results correspond to the characteristics of the portfolio currently held by the manager, and thus liable to influence his future performance. Consequently, this approach is assumed to integrate the evolution of the portfolio style over time to a greater degree than return-based style analysis. The return-based method, on the other hand, relies on an average of the past characteristics of the portfolio during the latest years. Confidence in results will depend on whether or not the fund style extensively changed during the period.

In practice, meanwhile, some difficulties occur in implementing holding-based style analysis. First, it is appropriate to have access to information concerning the fund composition that is reliable and representative of the fund’s investment policy over the period. This information includes the list of assets the fund is made up of, as well as their respective weights. The portfolio composition should be known not only at the beginning of the period, but throughout the whole analysis period. Moreover, this information must be updated each month, in order to get the latest data for the portfolio. Indeed, when external agencies perform fund analysis, the fund managers usually do not provide them with complete and accurate information about the fund’s composition and weightings. Portfolio managers are most often reluctant to disclose the details of their portfolios at regular dates, either for practical reasons or because they want to keep this information confidential. For the time being, no database is available for all portfolios with all information on their holdings at monthly frequency. In most cases, the information available is only partly updated and is provided only at an annual frequency. If up-to-date mutual fund holdings are not available, HBSA will lead to poor information.

Secondly, even if complete information about the fund holdings is available, the methodology requires that the style characteristics of each security be identified. This is necessary so as to determine the fund style as the weighted average of the style of the fund’s various assets. This constitutes the second pitfall of the HBSA model. Numerous studies (see for example Lucas and Riepe (1996)) have highlighted the difficulty in classifying securities according to their characteristics. Although it is relatively easy to obtain a consensus on the segmentation of classes, sectors or countries, style analysis, on the other hand, relies on more subjective classification. Commonly used attributes such as the book/price or earnings/price ratios are unstable as much with regard to market conditions as company-specific qualities. Moreover, the characteristics of a large number of securities do not allow satisfactory discriminatory analysis to be carried out. For example, the growth or value classification of a security, which is based more often than not on microeconomic attributes whose values — and therefore significance for the ranking — vary according to market conditions, is neither stable nor objective. This style classification will be tightly related to a risk analysis methodology, or to style index construction. Consequently, HBSA also involves certain issues, such as style indices choice and co-linearity between those indices, which are supposed to be relevant only in RBSA. As a matter of fact, considering the difficulty encountered with asset style, it is often necessary to turn to an RBSA approach, whether implicit or explicit, so that it no longer classifies the portfolio, but the assets it is made up of.

Finally, whether style analysis is performed using HBSA or RBSA, the important point is how the results are used. Most of the time, after meticulously collecting and analysing fund holdings, the rating agency goes on to discount the results of its own work by introducing style categories. Indeed, while HBSA makes it possible to know the risk of each position in the funds, the funds are then gathered into broad categories for comparison purposes. This leads to the loss of the potential accuracy of the risk approach previously adopted. While a portfolio is not always completely growth or value, or large cap or small cap, or well balanced in each of its two dimensions, the rating agency has to put the fund in a specific category, in which it is not possible to consider the specific risks taken by the fund. As a result, the comparisons between the funds that belong to one of these categories are not reliable, as these funds do not share many characteristics. Considering the problems introduced in style analysis by this final categorisation, the criticisms concerning the R-squared value or the regressors used in the RBSA approach appear to be minor details.

How best to deal with return-based style analysis

In light of this analysis of the difficulties related to sound use of HBSA, it appears that RBSA, by dealing with the problem of unreliable or missing information, offers the best solution in the analysis of a fund’s style. From a practical point of view, it is better to run the risk of statistical error than to rely on a manager’s stated objectives or the investment style he declares, or that which is inferred by the fund’s name. This problem has been illustrated by diBartolomeo and Witkowski’s 1997 study, which found that 40% of the funds studied were in a category other than the one declared. Moreover, the style of a fund may not be stable over time, so the category in which the fund is classified may differ from its current style category. A study by Kim, Shukla and Thomas (2000) shows that only 46% of the 1043 funds they considered had investment attributes that were consistent with the fund’s stated objectives, while 54% of funds were misclassified. Over one third of funds were severely misrepresented. Over the three-year period covered by the study, 57% of the funds that survived changed their investment style at some point. Only 27% of funds held their investment attributes throughout the period.

If well conducted, RBSA will make it possible to get a good representation of a fund’s style. This will require the analysis to be conducted regularly and statistical indicators to be monitored carefully. RBSA makes it possible to select the group of style indices that best describe a fund’s style, whereby both their types and their numbers are considered, in order to cover the basic investment styles of the fund considered, and whereby they are chosen as being mutually exclusive. The presence of multi-colinearity between indices can be checked by a cross-correlation indicator. This will make sure that the relative influence of each style index will be reliably determined and will confer robustness to the results, by providing coefficient estimates that are not sensitive to the block of data used.

Numerous statistical tests are available to improve confidence in RBSA results. Lobosco and diBartolomeo (1997) have developed a test that makes it possible to check whether the regression coefficients, i.e. the portfolio weights for style indices, are significant. The adequacy between the fund’s returns and the benchmark’s returns can be controlled with the adjusted R-squared. The higher the adjusted R-squared coefficient, the greater the ability of the passive style portfolio, i.e. the customised benchmark, to explain the fund’s performance. If this coefficient of determination is lower than an acceptable threshold, it means that the analysis has to be performed with another set of style indices that is likely to provide a better representation of its investment style.

Finally, the main drawback of return-based style analysis, namely that the fund style is assumed to remain constant during the analysis period, can be circumvented in several ways. One method consists in performing rolling regression over successive sub-periods to get the evolution of fund style throughout the whole period. Another, as proposed by Swinkels and Van Der Sluis (2002), consists in explicitly incorporating style changes in the model to get dynamic style exposure throughout the whole period. This will considerably improve the accuracy of style exposures for funds which tend to change their style over time with relatively high frequency.

If you do decide to use HBSA, it is important to be sure about the data used — specifically the classification models for asset styles. Otherwise, it is better to trust RBSA.

This article first appeared in Funds Europe magazine.

References

Bienstock, S. and E. Sorensen, ‘Segregating Growth from Value: It’s not Always Either/Or’, Salomon Brothers, Quantitative Equities Strategy, July 1992.

diBartolomeo, D. and E. Witkowski, ‘Mutual Fund Misclassification: Evidence Based on Style Analysis’, Financial Analysts Journal, September-October 1997.

Kim, R., R. Shukla and M. Thomas, ‘Mutual Fund Objective Misclassification’, Journal of Economics and Business, July-August 2000.

Lobosco, A. and D. diBartolomeo, ‘Approximating the Confidence Intervals for Sharpe Style Weights’, Financial Analysts Journal, July-August 1997, pp. 80-85.

Lucas, L. and M. W. Riepe, ‘The Role of Returns-Based Style Analysis: Understanding, Implementing and Interpreting the Technique’, Ibbotson Associates, May 1996.

Swinkels L. and P. J. Van Der Sluis, ‘Return-Based Style Analysis with Time-Varying Exposures’, Working Paper, October 2002.