Hacker News new | ask | show | jobs
by gmiller123456 3201 days ago
That's a pretty weird story. His "expert" ability at picking funds is what was being tested. He then goes on to give his "expert" opinion on why his expert opinion failed.

Yes, it's totally possible that Buffet just got lucky. But of the five stocks, over the 9 year period (45 data points), there are only 9 data points where a fund beat the S&P for the year. They all beat the S&P the first year, then for the next 8 years there were only four instances of a fund beating the S&P, and the S&P won for the other 36 data points.

If we were to assume that the funds vs the S&P was an equal playing field (each side had a 50% chance of winning), then there's just a 20% chance that Buffet just got lucky. But the point of actually paying a fund for advice isn't to just have equal odds, the funds are supposed to beat amateurs, and by a significant margin.

What we consider "significant" is obviously open to interpretation. But the most significant win a fund had over the S&P was in 2015 where fund C beat the S&P 5.4% to 1.4%, almost a 4x higher return. But the S&P had larger than 4x returns than the funds 12 times. I'm not exactly sure how to work all of that into a probability, but it's looking like the odds that Buffet just got lucky is pretty close to zero.

1 comments

Typo on the 20% chance (that's actually the percentage it lost). The actual probability is about 1/(2^13).