By Werner Antweiler, The Globe And Mail, Thursday 16 May 2013
Tuesday’s resounding election win by the B.C. Liberals has caught most people by surprise as polls, pundits and prediction markets were all forecasting a large win by the NDP. How could they all have been so wrong?
Traders in a prediction market aggregate news and information from a variety of sources, but misleading polls led them down the same path as the public. Only the few contrarian minds in UBC’s prediction markets foresaw the B.C. Liberals’ win – and made a more than handsome profit.
The prediction markets for the B.C. election had low participation compared to past campaigns. Without sufficient numbers of participants and liquidity, prediction markets fail to do their magic “crowd sourcing.”
Prediction markets have an advantage, though. If polls continue to have an anti-incumbency bias, traders will take this into consideration and adjust their positions. If polls influence the outcome, traders will be aware of this possibility from now on. Prediction markets have the potential to outperform polls through traders’ ability to learn and adapt and correct for the biases inherent in polls.