Timing the Recession

Don Fishback and Barry Ritholtz are shocked — shocked! — that the InTrade recession contract is based on hard GDP numbers rather than, um, something else, maybe an NBER pronouncement the timing of which is entirely unknown.

The fact is that prediction markets have to be based on something hard and fast like GDP numbers: they need an expiry date, and the results need to be completely unambiguous. This isn’t a "flaw in the betting process," as Ritholtz would have it: it’s a feature, not a bug.

Ritholtz even complains that InTrade is using the final GDP figures, and that they’re not waiting indefinitely to see whether those final GDP figures get revised again. Does he ever want the winner of this bet to get paid out?

And neither Fishback nor Ritholtz points out that the Recession 09 contract is now trading at twice the level of the Recession 08 contract. This is par for the course, as far as the current slowdown is concerned: the people predicting recession are likely to be proved right eventually, but it’s taking much longer than anybody expected. Consider this, from Nouriel Roubini:

Given the recent flow of dismal U.S. economic indicators (Q2 GDP report, July payrolls, service ISM, etc.) I am now taking the view that the odds of a U.S. recession by year end have increased from my previous 50% to 70% now.

The year in question was 2006. No matter what the NBER finally pronounces, I’m pretty sure they’re not going to declare a recession which began in 2006.

For all the chaos in financial markets, the supertanker US economy seems to take a really, really long time to slow down. That’s the real reason why the people betting on a 2008 recession are now out of the money — it’s got nothing to do with the entirely sensible ground rules laid down by InTrade.

This entry was posted in economics, prediction markets. Bookmark the permalink.