Who Made Money on the Presidential Prediction Markets and Who Lost

BY Paul Solman  October 20, 2013 at 12:25 PM EDT

By Paul Solman

President Barack Obama, seen here debating former Massachusetts Gov. Mitt Romney, wasn’t the only winner of the 2012 election. Thanks to offshore presidential prediction markets, one Romney-backer’s $4 million loss bankrolled a grad student’s winter break in Portugal. Photo courtesy of Jewel Samad/AFP.

Back in October 2012, we reported here on the Mitt Romney Arbitrage, an almost surefire way to make money on the offshore presidential prediction markets, assuming you weren’t an American citizen or, if you were, you were blasé about U.S. gambling laws.

(An arbitrage, as I explained here, refers to making the difference between prices. If something’s more expensive in one place, cheaper in the other, you buy one — sell the other. You make the difference. You lock it in.)

We speculated, with the help of MIT MBA student André Júdice Glória, that one or several Republican supporters were making Romney-will-win bets on the Irish gambling website Intrade, thereby pushing up Romney’s odds relative to other online betting markets and, among other outcomes, financing André’s prospective winter vacation in Portugal.

Now comes the news that one individual turns out to have been André’s sugar daddy.

“Romney backer lost $4 million plus on Intrade in possible manipulation attempt,” reports Tom Bemis on the Wall Street Journal blog MarketWatch.

“Researchers from Microsoft and Columbia University have found that a single trader on Intrade accounted for a third of bets Mitt Romney would win in the two weeks leading up to last year’s presidential election,” writes Bemis, “losing millions of dollars in a possible attempt to make the Republican’s chances look better.”

One of the researchers, David Rothschild of Microsoft, had discussed online markets with us during the presidential campaign. He and co-author Rajiv Sethi, of Barnard College, Columbia University, studied Intrade records and discovered that just one trader “accounted for one-third of all bets placed on Romney during our observational window, and lost almost four million dollars in the process,” a person who “could have been attempting to manipulate beliefs about the odds of victory in an attempt to boost fundraising, campaign morale, and turnout.”

So what does our MIT MBA arbitrageur have to say? Here’s André Júdice Glória’s reaction:

Intrade grew to become a reference in prediction markets, making the predictions themselves news worth reporting. As such, it is plausible that someone would support the price of a given prediction, like Romney winning the election, to make the candidate look better than polls would suggest. I find such a thing to be of dubious value to the campaign, but as an investor in predictions markets, I thank the person behind the idea and stand ready to take the other side of his trades anytime. I made some money last November thanks to this persistent arbitrage opportunity on Intrade and would make even more should the opportunity come up again.

Also, on a side note: I tried to replicate the results of my strategy for the Oscars and interestingly the arbitrage opportunity was not there. All the factors that would explain a persistent price difference that are mentioned in the Marketwatch report you link to are also present for the Oscars. I.e., Intrade was restricted to non-U.S. persons and had fee structures that differed from the typical betting site. Yet I did not find the same persistent price differences between contracts for a specific movie to win. Intriguing.

In other words, there’s no more plausible explanation for what happened last fall than that an earnest Romney backer inadvertently sent a grad student to Portugal — and bankrolled many more such astute market watchers — in a vain effort to make President Barack Obama’s opponent seem more likely to win than he ever was.


This entry is cross-posted on the Rundown — NewsHour’s blog of news and insight.