The performance of the Italian mutual funds: Does the metric matter?

This paper discusses the differences among performance metrics in the Italian mutual fund industry. This industry is worthy of interest because it presents two characteristics (representative of other Continental Europe countries, less analyzed than Anglo-Saxon ones) that weaken the importance of the time-weighted approach: a dominant role of the sellers and a significant vertical integration between production and distribution. Based on an original dataset, never used before by any scholar, we simulate (by using a Monte Carlo simulation model) the dynamics of returns and cash flows in the 2003-2010 period, analyzing the metric spreads and their sensitivity to scenarios' characteristics (volatility and timing of returns, entity and volatility of subscriptions and withdrawals). The empirical findings suggest that metrics matter. In fact, spreads between time-weighted and money-weighted returns are significant at level of individual funds in the simulated scenarios (consistent with the dynamics of the Italian industry in the considered period), while are not significant when we consider aggregated data, since aggregation smooths the volatility of flows and returns. The analysis suggests that it would be useful: (i) to rethink asset managers' choices in terms of performance measurement; (ii) to provide all the measures of return that could satisfy the broad spectrum of interested parties and assessment purposes.