A Symbolic Data Approach

The best-known and widely used rating system is provided by Morningstar, which assigns stars to a mutual fund based on the historical performance of the fund in comparison with its peers. The Morningstar rating is based on the MRAR (Morningstar Risk Adjust Return) taking into account return/volatility. The heavy use of Morningstar ratings in mutual fund advertising suggests that fund companies assume that investors use Morningstar ratings for deciding on their investments. It is likely that investors will avoid funds with the lowest ratings and choose funds with the highest ratings in the expectation that they will increase the future returns received on their investments. Del Guercio and Tkac (2002) provide empirical evidence that the Morningstar rating itself has a significant effect on fund flows. A second methodology has been developed by the Europerformance – Edhec rating system. The Europerformance ratings are constructed by combining three criteria. Indeed, one measures extreme risks with a modified VaR (MVAR). A performance measure is taken into account by the Jensen’s alpha coefficient and a potential persistence of performance is estimated by the Hurst exponent. In this chapter, we describe a new rating indicator based on a symbolic data analysis (SDA) as developed by Diday (1987) and we use this new indicator to compare rating systems. There are two main reasons why we use this approach to construct a new rating of mutual funds. First, unlike previous research, which uses only single quantitative or categorical values,