Using a clustering procedure, we classify Italian funds ex-post on the basis of the composition of their portfolios and find that the optimal number of clusters is equal to 4. The four groups which result from the statistical classification closely match the 4-level aggregation of the 20 ex-ante categories used by the Italian mutual funds association.
We then estimate the risk-adjusted performance of Italian equity funds, using both net and gross returns and employing both one-factor CAPM benchmarks and multi-factor benchmarks. In addition to the standard Jensen's a, we measure risk-adjusted performance using the Positive Period Weighting measure (PPW), which is not influenced by managers' market-timing strategy. Using net returns (calculated after management fees and taxes but before load fees) the Italian equity funds' performance is not significantly different from zero. However, when the funds' performance is evaluated on the basis of gross returns (i.e. returns computed adding back management fees paid each year by the funds), the performance of the Italian equity funds is always positive. In particular, when both a 2-index benchmark that takes account of the funds' investments in government bonds and a 5-factor APT benchmark are considered, performance is positive and significant using both Jensen's a and the PPW. This result supports Grossman and Stiglitz's (1980) view of market efficiency, suggesting that informed investors (investment funds) are compensated for their information gathering.
Published in 2002 in: Journal of Banking and Finance, v. 26, 1, pp. 99-126.
