Prediction Markets. Kalshi & Polymarket
Growing more than 100x in just over two years
Original Business Case Weekly articleMost industries grow by solving problems. Prediction markets grew by solving a problem nobody in power wanted solved: proving that crowds with money on the line are better forecasters than credentialed experts.
From three economists drinking beer in Iowa City to a $44 billion industry with the Trump family on its boards, this is the story of how a segment nobody could legally define kept getting killed and kept coming back.
Three Economists Walk Into a Bar, 1988–1999
In March 1988, Jesse Jackson won the Michigan Democratic caucuses. Every major poll had predicted a comfortable victory for Michael Dukakis. The polls were wrong, and three economics professors at the University of Iowa were irritated.
Robert Forsythe, George Neumann, and Forrest Nelson were sitting in a bar in Iowa City, complaining about exactly this kind of failure. Their frustration had a specific intellectual edge. If financial markets already did a reasonable job of aggregating dispersed information about the future price of commodities, why couldn’t a similar mechanism aggregate dispersed information about elections? People had skin in the game when trading corn futures. Why not give them skin in the game about political outcomes?
The three built what they called the Iowa Political Stock Market, launching it in time for the 1988 presidential election with 192 traders drawn from the University of Iowa community. University lawyers were nervous. They found a loophole in the state code allowing betting pools within employee groups. The stakes were small, a maximum investment of $500 per trader, but the results were immediate and striking. The 192 traders predicted George H.W. Bush’s vote share within 0.2 percentage points of the actual result, outperforming every major poll.
By 1992, the operation had grown into the Iowa Electronic Markets (IEM), a permanent internet-based exchange. The CFTC granted it a no-action letter, a regulatory term of art meaning the agency would not pursue enforcement as long as the operation stayed small and academic. Between 1988 and 2004, IEM prices outperformed traditional polls roughly 74% of the time across five presidential elections.
Your call
If you sat in that Iowa City bar in 1988 and could invest $10,000 in the concept of “prediction markets as a business,” would you? The academic model clearly worked. But the path to legal, commercial operation was completely uncharted. What would your biggest concern have been?