Conversation

2) The whole point of golden geese is that sometimes they don't work. Maybe they rarely do. But they're valuable enough that they're positive expected value anyway.

10:30 AM · Aug 1, 2021Twitter Web App

3) So, if you find an example of a time when a lot of 'golden geese' falter, the reaction shouldn't be "whelp guess golden geese are fucked". The _default_ state of potential golden geese is that they falter; adding on another few examples doesn't change much.

Replying to

4) And, anyway, the article is referencing a 50% downturn since the peak, for a set of bets that had doubled many times already.

5) If you had told early investors in BABA that it would be a $530B company on 2021-08-01, they would have been ecstatic. The fact that it was briefly hit $800B doesn't diminish the net growth to $500B.

6) Which is one thing I never quite got about some of the blowback on Softbank from the WeWork shitshow. Yeah, they lost on it. If you're really going for the biggest wins, you'll lose sometimes, maybe frequently. And sure, you should think about mismanagement.

7) But if the goal of a strategy is to make 100 bets, lose on 50 of them, and 5x on the other 50, then there's only so much you'll learn from an individual loss. That's expected. You have to look at the overall performance.

8) IDK how to think about the *real* trajectory--its noisy! But, FWIW, here's the % performance of Softbank (red) vs the S&P500 (blue) in the last 10 years: