Abstract
Traditional measures of mutual funds persistence are based on measures like Jensen's alpha, factor models and so on. As expected, their results on a broad data-set of European mutual funds show a small, but relevant, amount of persistent funds, depending on the method of estimation used. We go further in our analysis and, by using a multiple hypothesis testing methodology, find that those measures overstate the performance on the sample across countries, signalling persistence in situations where the positive results are due to luck.
Acknowledgements
We thank Mark Roomans from Morningstar Spain for the data. We also thank several managers for comments to this version and MEC Project MTM 2006-15433 for research support.