Rational overconfidence in the tens of billions: recent example

post by banev · 2022-11-13T22:48:31.073Z · LW · GW · 3 comments

Contents

  There is a difference between one die roll of 1,000 and 1,000 rolls of 1 and the corresponding difference in the probability of survival. 
None
3 comments

Still, when SBF analyzed the bright future that lay before him, something wasn’t right. He was, he realized, too secure. SBF’s mind had been trained almost from birth to calculate. As a schoolboy the hedonic calculous of utilitarianism had him trying to maximize the utility function (measured in “utils,” of course) for abortion. During his teenage gaming years, his mathematical abilities allowed him to sharpen his tactics—and win. And, of course, every trade SBF ever made at Jane was the subject of a risk/reward calculation. All of it boiled down to expected value. The formula is fairly simple. If the amount won multiplied by the probability of winning a bet is greater than the amount lost multiplied by the probability of losing a bet, then you go for it—irrespective of units. Utils, euros, dollars were all subject to the same reckoning.

But, at Jane, SBF osmosed another trading principle. He learned to be “risk-neutral”: In simple terms, a trader, given a choice between $50 and a 50 percent chance at $100, must be agnostic if they want to maximize the expected value of earnings over a lifetime. Those who prefer the sure win are “risk-averse,” and those who would rather gamble are “risk-lovers.” But both risk-lovers and the risk-averse are suckers, equally. Because, over the long run, they lose out to the risk-neutral, who take both deals without prejudice. 

Here, SBF realized, was the rub: When he applied this principle to his own life, he came up short.

From the promptly deleted webpage at the Sequoia who proud to be "partners for the long term" with all backed by them companies, which page has been saved by all-remembering web.archive.org. 

Read the story, it's interesting, no wonder they deleted it. 

There is a difference between one die roll of 1,000 and 1,000 rolls of 1 and the corresponding difference in the probability of survival. 

I'm even not sure there is a fraud. It could be just stupidity and overconfidence of person who see himself as very rational actor. Just read this: 

There are some modest smiles in the Zoom windows, including one from Singh, FTX’s director of engineering. He knows firsthand, from his trading days at Alameda, how frustrating a rickety exchange can be. “As always, we have some changes to make going forward,” SBF says, “but nothing really huge.” The next order of business is a round of belt-tightening. “In general, money is going to be tighter for everyone in our industry and other industries too—it’s not super-crypto-specific,” SBF says. “So, if you are thinking about expenses that are, say, above $100 million, we should have a chat about that,” he continues. “That’s our new set of takeaways,” he says, wrapping up his portion of the meeting. “Oh, and yeah: Ten-second updates!”

With that, the floor is open. And indeed, various FTX execs take over the presenters’ mic, updating the company, in ten seconds or less, about what is going on.

• FTX Stocks is now open for a private beta.
• An independent FTX fan club has emerged: the FTT DAO.
• Time-weighted average price functionality is being tested.
• The sponsorship deals of the Met Gala and the amfAR Gala have been successful—so far.
• The NBA Eastern Conference final starts tonight at the FTX Arena: the Miami Heat versus the Boston Celtics.

That last item is the only one that provokes any kind of comment from SBF: “Go, Heat!” he shouts.

All-hands adjourned. Total meeting time? Ten minutes.

I remark to the person sitting next to me that that was rather quick for an all-hands. “That was actually an unusually long one,” she says. “They’re usually five minutes.”

Very rational indeed. So much time is saved, and time is money, right? 

Be just a little bit less rational and more human, really. Listen to your grandma. Go play football. 

It will pay off. 

3 comments

Comments sorted by top scores.