3 ms·
We see a lot of stories on HN about how so many medical studies are false positives, because when you set up your analysis to yield 95% confidence, and most of
by mikeash 11y ago
We see a lot of stories on HN about how so many medical studies are false positives, because when you set up your analysis to yield 95% confidence, and most of your studies are on things that have no effect, the 5% false positive chance translates to a much higher false positive rate.
Active management is subject to the same thing. Everyone hammers on the idea that "past results do not guarantee future returns" but that's all anyone ever looks at, what with it being very difficult to observe the future. Some active managers will be successful simply by luck. They'll tout their results and get more customers from it. And then eventually their luck stops and they'll revert to the mean, minus their fees. Because of the proportions involved, lucky managers will heavily outnumber those who are actually good at it (if there are any).