Plain Sense Economics

Prediction markets represent a fascinating development on the Internet. They’re a live, real-world example of the power of markets. True confession…I understand the basics of these markets, but there is much more for me to learn about them. Still I thought it important to introduce them here. Plus – there are some particularly relevant predictions featured on these sites.

An example, from the prediction market Intrade:

As of Monday February 11 Intrade markets give Sen. Obama a 70.0 percent chance of being the Democratic Presidential Nominee and a 47 percent chance of winning the election in November. Sen. Clinton has a 30 percent chance of getting the nomination and a 20.4% chance of winning in November.

So, where do these numbers come from? Not from polls, in the way we think of them. Instead people invest real money, betting on the outcomes. Let’s say that after this weekend you are very sure that Sen. Obama will win the nomination. The Intrade market for this prediction is at 70 percent. If you are so sure of the outcome you might rate the likelihood at 80 percent. For you the odds are better than the market indicates and you expect a profit.

So you buy a prediction contract for $7.00 (70 percent points times 10 cents a point.) You can hold onto these contracts until the Democratic nomination is set. If Obama gains the nomination you receive $10 (100 percent chance times ten cents a point) and make $3. If Sen. Clinton wins the nomination your contract receives nothing – you lose your investment. Your purchase, however, adds demand to this market and nudges the equilibrium price upwards. As more people decide to purchase a contract for Obama getting the nomination, the price rises. This continues until no one is willing to buy another contract. The market price reflects the market’s prediction for this event.

You can also sell your contract at any time. Let’s say the price for Sen. Obama rises to 85 and you’re still thinking 80 percent is the best guess. You can sell your contract for $8.50 and earn a $1.50 profit from your initial $7 purchase. The more people who start selling their contracts, the more the equilibrium price will drop.

There are thoughtful academic journal artices on prediction markets. As with any market the greater the number of participants, and the more available information is to everyone the better the market will work. On the other hand, from a political polling perspective, market participants are not a random sample of likely voters. They self-select to participate in the market.

We’ll keep an eye on this prediction market phenomenon and either edit mistakes in this post or add other posts along the way.

Historical side note… As I read more about these I learn that during and prior to WWII there were either no or few scientific polls, including for political purposes. There were, however, thriving markets on Wall Street and elsewhere that operated like today’s prediction markets – just with no help from computers or the Internet. Newspapers would routinely report on the odds reported in these markets, in lieu of the polling results stories we see today.