AGORA is a platform for hosting sophisticated prediction markets that can aggregate the output of quantitative predictive models alongside more tacit forms of expert knowledge. AGORA was originally created by the investment manager Winton Group before being further developed by its technology spinout, Hivemind.
AGORA supports markets that generate high-resolution joint probability distributions from the trading activity of participants, who can use the AGORA UI or trade programmatically using an API.
AGORA has been tested over the past three years and used to run prediction markets with expert participants for forecasting seasonal temperatures and rainfall, Atlantic hurricane activity, and El Niño events. Hivemind transferred the IP for AGORA to Lancaster University in 2022 for use in CRUCIAL.
Figure: The evolution of prices in a joint-outcome market for UK monthly temperature and rainfall for June 2018. The participants were 24 teams from British universities. Winton Group sponsored the real-money incentive pot (£55,000). The black dots show the values of temperature and rainfall for previous Junes and the blue cross shows the actual value for June 2018.
CRUCIAL’s inaugural market is live
This paper reports viability tests of prediction markets with highly granular, monthly UK rainfall and temperature joint outcome spaces. The experiments demonstrate these markets can aggregate the judgments of experts with relevant expertise, and suggest similarly structured markets, with longer horizons, could provide a mechanism to produce credible forecasts of climate-related risks for policy making, planning, and risk disclosure.
This article examines the performance of 24 prediction markets for climate-related variables that have been run over the past five years. The markets had horizons of 2 to 12 months. The predictions of the markets were consistent with good reliability, given the resolving power afforded by the sample size.
Forward-looking information about climate risks is critical for decision makers, but the provision and accuracy of such information is limited. Innovative prediction-market designs could provide a mechanism to enhance applied climate research in an incentive-compatible way.
Send us your CRUCIAL message.