Electoral hedging gains votes

Academic experiments have shown that “electoral markets” – where players make cash bets on the outcome of a ballot – are better predictors of the result than traditional opinion polls. Money talks.

In a provocative technical paper in this month’s Risk (see “Hedging Electoral Risk”, pages 95–98), assistant professor Steve Kou and professor Michael Sobel of New York’s Columbia University suggest a way in which these electoral forecasting markets (EFM) can be developed into an effective market for hedging political risk – an electoral hedging market (EHM).

Such a market could provide companies with a better alternative to the established ways that they use to hedge their political risks, by spending money on lobbying and party political donations for example.

Scroll back to November 7 last year – the day the US went to the polls to elect a new president. Nobody anticipated the tortuous weeks of confusion that would follow and the almost impossibly narrow result by which George W Bush finally won. During those weeks, financial markets were no better informed than any individual US citizen about the likelihood of the eventual outcome.

That experience might provide the motivation for establishing an EHM. Bush versus Gore was not the most market-sensitive of contests, but consider the following scenario.
Candidate Smith states that if he becomes president, he will severely restrict the right to bear arms. The ammunition makers believe the implementation of the policy will force them out of business. Short of diversifying their product range, how could they hedge the financial risk inherent in the election? They could make a donation to Smith’s strongest rival, candidate Brown.

If the donation helps Brown win, then this is money well spent. But, if Smith wins, it has been wasted and – worse still – could encourage Smith to pursue his hard-line policy further. There is a large downside risk.

In 1988, the University of Iowa created an alternative to the polls: an EFM called the Iowa Electronic Market (IEM). In total, 192 traders played the market on that year’s Bush-Dukakis presidential contest.

The IEM is a real-money market, where participants trade shares in the presidential candidate they believe will win. In the 1988 IEM, traders received a payout proportional to the candidates’ final share of the popular vote in the real election.

Over the years, the final prices in the IEM have proven to be accurate indicators of the presidential results, although the precise reasons for this remain a matter of debate.
Kou and Sobel have analysed the results of the IEM and other electoral markets using options theory. In terms of probability theory, candidates’ market prices are expectations taken with respect to a real-world probability measure.

To create an EHM, Kou and Sobel propose developing this precisely defined forecasting tool into something akin to a futures market. For this to happen, investment caps need to be removed and short-selling and buying on margin should be allowed. “We hope that individuals and corporate actors will appreciate the potential of our work and that such markets will be established in the near future,” says Sobel.