Market Manipulation

Nate Silver at FiveThirtyEight thinks there is something fishy going on with the presidential contracts at Intrade. He first points out that there is a big discrepancy between the probability that Obama will win on Intrade -- about 54% -- and his chances on another prediction market, Betfair -- about 62%:

This is the equivalent of the Giants being 3-point favorites at the Bellagio Sportsbook, and 7-point favorites at the Mirage down the block. Those things just don't happen in efficient, sufficiently liquid markets, because they create arbitrage opportunities

(I gave a possible explanation for this here.)

Silver also finds a suspicious pattern in the Obama and Hillary Clinton contracts:

Presumably the spikes you see are trades made by the same entity with some extra money to spend. The individual trade data show that on Sep 23, both contracts (as well as the one for McCain) saw significant action between 5:30 a.m. and 7:00 a.m. EST.

Felix was nice enough to send me an email exchange with Wharton's Justin Wolfers on this and here is what Wolfers thinks:

This is an issue that I've been tracking for the past few days, and the thought that the market is being manipulated is certainly credible. We've been looking at similar evidence, and Nate Silver's argument seems reasonably well constructed. The difficulty with detecting manipulation is that it involves people buying and selling securities, while informationally-based trading involves buying and selling securities.

Silver also thinks the rogue trader(s) may know something about why Obama won't be the next President. (If this were the case, the patterns we're seeing wouldn't be market manipulation since the trader is betting with information).

But it's more likely that if this is manipulation, then the manipulator is actually betting that Obama will win. Here's how it would work: first the manipulator sells the Obama contract to drive down the price, then when it reaches a sufficiently low enough price, the manipulator buys back in. The trick to doing this is that, in total, you keep your upside bigger than your downside. (This is sketchy evidence, but the second peak in the Obama contract is lower than the first.) This reasoning would also explain the surge in buying of Clinton contracts.

If this is manipulation, it's certainly not a great development for prediction market proponents. But let me try and find a silver-lining: This type of manipulation most likely couldn't happen in a super-low liquidity market. It looks like Intrade has gotten popular enough so that this sort of thing can go on. Meanwhile, this study found that manipulation is more likely to happen in a market with less liquidity than in a comparable one with more funds slushing around. That would suggest that if more investors were trading (volume is up about at least 400 percent over the 2004 elections), manipulating opportunities would be diminished. All that's left is for that to actually happen.