Standard testing approaches designed to identify funds with non-zero alphas do not account for the presence of lucky funds. Lucky funds have a significant estimated alpha, while their true alpha is equal to zero. This paper quantifies the impact of luck with new measures built on the False Discovery Rate (FDR). These FDR measures provide a simple way to compute the number and the proportion of funds with truly positive and negative performance in any portion of the tails of the cross-sectional alpha distribution. Using a large cross-section of U.S. domestic-equity funds, we find that 76.6% of them have zero alphas. 21.3% yield negative performance and are dispersed in the left tail of the alpha distribution. The remaining 2.1% with positive alphas are located at the extreme right tail. The same analysis is run on three investment categories (growth, aggressive growth, growth and income funds), as well as groups formed according to lagged fund characteristics (turnover, expense ratio, total net asset value).

Ž. The aim of this paper is to analyze the sensitivity of Value at Risk VaR with respect to portfolio allocation. We derive analytical expressions for the first and second derivatives of the VaR, and explain how they can be used to simplify statistical inference and to perform a local analysis of the VaR. An empirical illustration of such an analysis is given for a portfolio of French stocks. q

We revisit the apparent historical success of technical trading rules on daily prices of the DJIA from 1897 to 2008. We use the False Discovery Rate as a new approach to data snooping. The advantage of the FDR over existing methods is that it selects more outperforming rules and diversifies against model uncertainty. Persistence tests show that an investor would never have been able to select ex ante the future best-performing rules. Moreover, even the in-sample performance is completely offset by the introduction of transaction costs. Overall, our results seriously call into question the economic value of technical trading rules.2

Abstract. We develop a test of equality between two dependence structures estimated through empirical copulas. We provide inference for independent or paired samples. The multiplier central limit theorem is used for calculating p-values of the Cramér-von Mises test statistic. Finite sample properties are assessed with Monte Carlo experiments. We apply the testing procedure on empirical examples in finance, psychology, insurance and medicine.