Journées de l'optimisation 2007 Optimizations days

In this paper, we provide an in-depth analysis of the performance of newly launched mutual funds. In a simple framework of an agent utility maximization problem, we compare the optimal allocations between new funds and family funds to those observed in the market. In this latter case, we find that investors are reluctant to exit from funds they have already invested in. We then provide a new performance measure for newly launched funds. This measure decomposes the fund risk into two components: family risk and fund specific risk. We also improve the precision of betas and alphas estimation using two methods: a combined-sample and an empirical Bayesian estimator. These methods correct the bias resulting from the data shortage of new funds. Our efficiency tests based on the RMSE confirms the superiority of the Bayesian estimator. These results have many implications in assessing new fund performance.

Nous présenterons une étude basée sur l'évaluation et la réplication en temps discret d'un payoff mensuel. Une extension de l'algorithme de tarification d'option américaine en temps discret de Del Moral, Rémillard et Rubenthaler sera proposée. Nous développerons une stratégie de couverture optimale qui assurera une erreur de réplication nulle tout en minimisant le carré de cette erreur. Nous comparerons la performance de notre réplication avec celle proposée par Kat et Palaro (2005) dans un contexte de création de fond synthétique.

In this paper, we investigate the economic impact of new mutual fund introduction. We first demonstrate that on average new funds have higher performance, higher fees, higher turnover, and smaller size than existing funds. Moreover, new funds do not have a specific diversification and liquidity portfolios objectives. We then study the persistence of their performance within two-time windows following their entry. Our results suggest that top performers are more persistent than poorer funds. In addition, a high proportion of new funds belongs to extreme deciles and is therefore riskier. Finally, we analyze the impact of newly introduced funds on stock markets and fund industry performance. We find that stock prices increase after the introduction of new funds which is consistent with the information spreading theory. We also document a positive and significant relationship between the number of new funds and new stocks.