15 Comments
User's avatar
James's avatar
Jan 4Edited

Also didn’t someone say it leaked to the NYT and WaPo anyway (I think my source for this is a Hasan Piker tweet so unclear if true). Doesn’t seem like this is worse.

Prediction markets are meant to have insider trading! I am begging people to understand that prediction markets encourage insider trading and this is _THE_ point!

Celeste 🌱's avatar

I have heard this argument alluded to before, could you steelman it please?

James's avatar
Jan 4Edited

Happy to! :)

Insider trading is not ontologically wrong. Different jurisdictions have very different insider trading rules. ~This was partially influential on the Jane Street India market manipulation thing being technically legal afaik~ (EDIT: this Jane Street claim seems wrong to me now, having looked into it more, but it doesn't materially affect the point)

Prediction markets are meant to predict reality. Under insider trading rules, if you meet a CEO at a private event and he tells you he’s doing a gram of ketamine a day, it’s unclear whether you can short the stock. In a prediction market, you totally can! That’s good for predictions! It’s a prediction market.

David Spies's avatar

But why would I want to bet in a market where anyone might have inside information? I'm continually at risk of being exploited.

The reason insider trading is illegal in the first place is because it discourages people from entering the market. I don't see how prediction markets are any different

James's avatar

Yeah so I think in that case then you should not bet on prediction markets.

They’re not primarily intended to be a tool for you to make money on (a la publicly traded companies, to some extent). They’re intended to be an information aggregator, and that includes inside information.

David Spies's avatar

"I" meaning the general I. Not me specifically. If everyone reasons that way, then no one bets, then the prediction market doesn't work

James's avatar

But like, insider trading is _currently_ allowed, and they _currently_ work. Somewhat, at least (the bigger issues are with gambling behaviour imo).

Shankar Sivarajan's avatar

If you think of it as a game in which people with "inside information" are cheating, you shouldn't. If you instead think of it as a way to purchase information, then it might still make sense to.

David Spies's avatar

I see.

If I bet on something I know a lot and care about and someone with inside information takes my money, then I'm effectively paying them for the valuable info

Shankar Sivarajan's avatar

Yes, that's how I'd think about it, though it's muddied by the possibility of some markets becoming kind of self-fulfilling, where someone might cause an otherwise unlikely event to happen.

James's avatar
Jan 4Edited

Like what you actually need is just _consistent_ insider trading rules is so that everyone knows they're playing under the same game. And the reason the UK and US has instituted them for publicly traded stocks is basically a "think of the little guy"-argument.

EDIT for more info: This doesn't apply for e.g. swaps, where insider-trading rules _are_ much more lax on the regular market because "think of the little guy" doesn't apply so much. If you're trading on prediction markets, you need to consider that some people may have more information than you, and this information may be inaccessible to you. That seems fine. That means we don't need insider-trading rules on prediction markets.

Benjamin's avatar

It's probably not just people who were super close to Trump who knew about it ahead of time? You'd have to assume some people in the intelligence/military had more info (even if it wasn't super precise per se), maybe whoever the CIA's contacts inside the Maduro regime were, maybe even some foreign intelligence agencies that were good at predicting this kind of thing? Otherwise I agree with your conclusion.

Shankar Sivarajan's avatar

I am reminded of this story (more of a parable) about prediction markets posted on LessWrong: https://www.lesswrong.com/posts/6h9p6NZ5RRFvAqWq5/the-summoned-heroine-s-prediction-markets-keep-providing

Ha Tran Nguyen Phuong's avatar

this is a really fun exploration, thanks for writing! i really like how you detail your thinking w the Bayesian estimate haha.

Mick Bransfield's avatar

There was already public information about the raid in advance.

1) Semafor reports that the NYT & WaPo knew in advance (but did not report)

2) Pentagon Pizza Watch shows a 769% spike that night: https://x.com/nyprepper1/status/2007327615070392726