Put Your Money Where Your Mouth is: Polls versus Prediction Markets in Predicting Election Results


Did you hear? There’s an election coming. Just kidding. I’m sure that, by now, most readers feel as if they’ve been in a pinball machine, being bombarded from side to side with rhetoric from both political aisles as well as their PACs and the media.

Today, we look at the current odds as they stand in two markets. The first is the traditional polling markets such as Rasmussen, Reuters, Gallup and USA Today. In these endeavors, a somewhat statistically significant sample group of citizens (maybe 500 to 2000) are asked about various issues. The results are compiled and tabulated and are reported in the press. These polls are intended to represent the opinions of the broader population.

In the second market, known as prediction markets, people actually make bets with real money on an issue. The idea is that when people are backing their prediction of the future by putting real money on the line, the results will be more reliable. Maybe they are, and maybe they aren’t, but it’s basically a cash market on predictions for the future, much like the stock market.

Starting with the polling results in Chart 1, we can see that, while the Presidential race has been close at times, President Obama has generally led in the polls throughout the race. While Mitt Romney got a small boost during the Republican National Convention, President Obama erased the Governor’s gains and then some after the Democratic National Convention. According to Real Clear Politics, who compiles and averages out the results of the major national polling organizations, the President now holds a lead of about 3.2%, based on polls taken during the seven day period of 9/4 – 9/11.

Polling Forecast of Obama vs. Romney Win in November
Poll Forecast, Obama vs Romney

The prediction market in Chart 2 generally confirms the results of the polls, but the gap is much larger. Intrade.com data as of 9/12/2012 shows President Obama with a 61% chance of re-election, while Mitt Romney has a 38% chance. This is a much wider margin of victory for the President than is being forecast by polls!

Prediction Market Forecast of Obama vs. Romney Win in November
Prediction Market Forecast of Obama vs. Romney Win in November

Both methods show President Obama with a lead over Governor Romney, but when correcting for margin of error, the polling data is much closer to a dead heat than the prediction market data, which has no margin of error. Much like the stock market, there is no margin for error on a transaction. The price is what you get.

I would personally be shocked if President Obama wins by the 24% margin being forecast in the prediction markets. After all, don’t forget that the Electoral College system and the voting districts setup by the two party system typically causes inherently close elections. The counter point to that argument would be that market prices reflect all available data. Clearly, the makeup of the Electoral College and voting districts is known information. At the same time, the biggest popular vote defeat in our nation’s history came when Lyndon Johnson defeated Barry Goldwater by 22.6%.

Barring some serious traction from Governor Romney or a bombshell negative issue for President Obama, it looks like we will see four more years of an Obama Presidency. Of course, the polls on Congress are very solidly on the side of a Republican controlled House and Senate, so even if President Obama wins, he will likely be faced with continued political gridlock and an uphill battle in pushing his agenda in Washington.

The opinions expressed are my own and are for informational and educational purposes only. These opinions are not a recommendation or solicitation to buy, sell or hold any security. The information in this article has been researched and is believed to be accurate, but readers should not make investment decisions solely on this information. Talk to an advisor before pursuing any investment strategy. Past performance is no guarantee of future results. Investment returns and principal fluctuate through time so shares may be worth more or less than their original cost when redeemed. Current performance may be lower or higher. Indexes are unmanaged baskets of securities that are not available for direct investment; they do not reflect the deduction of advisory fees or other investment expenses such as taxes and transaction costs.