A highlight of this year's Annual Meeting was a presentation by Olivier Scaillet (pictured at the Meeting) on the application of false discovery rate (FDR) statistics to the estimation of mutual fund performance.
Scaillet uses FDR to reveal when good fund performance can be attributed to chance alone, and is therefore likely to deteriorate in the long term.
The Geneva-based faculty member of the SFI become interested in the application of FDR statistics to finance after attending a presentation by Berkeley statistician John Rice at a statistics conference in Brazil.
Though Rice applies the method to the detection of stellar occultations, Scaillet recognized the application to mutual fund selection when Rice stated:
We face the problem of searching for needles in a haystack when we do not know what the needles look like. How do we automate serendipity?
We face the problem of searching for needles in a haystack when we do not know what the needles look like. How do we automate serendipity?
At this point on a Friday afternoon I am glad to entrust Scaillet to select the correct statistical test for this particular application.
The end result is that the private banking industry will benefit from an objective and relatively simple method to select wisely among the estimated 2100 US domestic mutual funds available to today's investor.
