Funny how so many people are skeptical, as if they they want to be convinced that it's not possible. The fact Renaissance Technologies has been so successful for so long shows it can be done. Easy? no, but possible, yes.
I've been in this industry for a while and folks making high returns solo either:
1) Got lucky for a very short amount of time, then their strategy stopped working. On avg those folks net lose money trying to make "the thing" work again.
2) Were running a super risky levareged strategy and went bust eventually
What about prop funds, which are trading for their own account? Seems like those wouldn’t keep it up very long if they were losing tons of money.
Not sure where to find a link to the statistics, but you hear it when talking to the community.
"this algo is levered far higher than it ought to be and while you've dodged landmines this month, you won't in the long run,"
vs.
"your 'minimum viable stat arb' and competent implementation outperforms the best-of-the-best outlier."
I think skepticism is warranted as most people at this point end up blowing up, but hey, I guess there's some slim possibility that it's the latter?
The Medallion Fund is also larger than a single quant's trading portfolio. This is important because a given trading strategy loses its effectiveness (in percentage terms) when scaled up. A strategy that yields 100% return on a 10k investment is unlikely to return 100% on a $10M investment, for instance.
I did originally buy into the explanation that they collected/digitized stock data that is no longer available.
But really it’s more likely to be some sort of fraud.