Luck should have been mentioned but the sheer volume of trades and absurd average over 18 years makes it seem like they DO have some special sauce... and a lot of it.
Those are the 9 or 10 events I'm talking about. These people are probably rarely betting on something more than 1:10 odds. But for those huge cases, they were probably getting 1:100 or 1:1000. Those are the ones it is so important to be on the right side of, and dampen quite a bit of suboptimal whatever in the meanwhile.
He quite specifically bet against the housing and dotcom boom as mentioned in the article, therefore shows to be betting most importantly on those 1:1000 events.
I was more trying to get the point across that they weren't rolling one die for 18 years but rolling it thousands of times.
All this talk about special sauce is making me hungry...
It's really a failure of regulation.
Given enough series of die rolls by enough different rollers, there is inevitably going to be some with long strings of success.
It only takes looking at 1024 cases before you find someone who bet correctly on 10 binary decisions.
There are assuredly more than 1024 funds.
He might have some special sauce, or may have done some smart things, but it is just as possible he's just #1024