[ub]

I guess this is a way for YC to participate in the upside of the most successful companies without creating signaling risk. But from a pure investment perspective, there's a possibility it might not end up being that prudent. It will all depend on the home runs. If YC can create a few multi-billion dollar companies, this will work out well.

[briholt]

Looks like there's a useful built-in selection bias. YC commits itself to investing in future rounds, but only good companies will be able to raise future rounds, so they probably won't get stuck doubling down on too many failing companies.

[_delirium]

> only good companies will be able to raise future rounds

Given the rather poor returns of the VC sector overall, I'm not sure you can make this assumption without more qualification.

[briholt]

Fair enough:

* ...only not-completely-screwed-up-trainwrecks will raise future rounds...

[namenotrequired]

Maybe the inverted statement is more appropriate?

... companies that aren't able to raise further rounds will fail...

[vasilipupkin]

it's actually hard to generate really high returns without doing prorata. If you don't do it and you invest into 1,000 companies, your pot of money is spread mostly over companies not doing well. If you keep pro-rata it will be more concentrated among companies that are raising subsequent rounds, which is correlated with doing well. Thus, your money will lean naturally towards better companies