Monday, January 27, 2020

"the betting on Klobuchar getting the NYT endorsement started really soaring just a few hours before it was officially revealed "



From Matt Levine's Newsletter:
Not a lot of people bet on Klobuchar, and she was trading at under 10 cents on the dollar, before Sunday. But as Bloomberg’s Joe Weisenthal pointed out on Twitter, “the betting on Klobuchar getting the NYT endorsement started really soaring just a few hours before it was officially revealed,” and she traded up on heavy volume, hitting 68 cents before the contract was settled (and paid out $1.00). Hmm!

I don’t know what public information you would have used, on Sunday evening, to conclude that Amy Klobuchar was much more likely to receive the Times endorsement than she had been a few hours earlier. Like maybe you could have read some position papers on her website and said “hmm this sounds like stuff the Times would like,” or you could have read the transcript of her interview with the Times, but you could have done that last week too; none of that was new on Sunday evening. If you assume that the market had efficiently incorporated all public information about Klobuchar's chances of getting the Times’s endorsement as of Sunday morning, there was not much reason for those chances to rise dramatically over the course of Sunday evening, unless the rise was based on non-public information.

But part of the standard justification for prediction markets is exactly that they will be informed by insider trading. The idea of political prediction markets is that campaign aides and party insiders and other informed people will place bets and move the prices in line with real informed probabilities and thus provide information to the public. The point is not just to allow you to have fun making uninformed bets on the candidates you like; the point is to provide the public with information about politics.