Prediction markets solve this problem by allowing each participant to exploit their own self-knowledge. Each participant knows their own strengths and weaknesses, and for which subset of events they have confidence in their own forecast relative to other experts across the rest of the market. Some participants, for example, will have expertise in the extremes of a distribution – accordingly they will concentrate their trading to correcting what they perceive as price anomalies there. Conventional survey-based mechanisms can elicit team-specific forecast distributions, but not this additional layer of differentiation and relative weighting. And in a prediction market, these individual forecasts adjust continually in real time in reaction to events and updates. Whenever new relevant information arrives, it is built in. This continual updating is not possible with surveys or other conventional forecasts. OBTAINING THE BEST RESULTS We want the best results, so we need to incentivise. To do this, we give participants an endowment, an amount of on-platform credit. When we launched our first Atlantic Hurricane Market, for instance, we endowed it with £20,000. Participants start with a certain balance of on-platform credits. They begin trading, and the value of their credits increases in accordance with the amount of new information they bring. When the market closes, participants can receive in pounds sterling the difference between their balance of credits and their initial endowment; or they take those on-platform credits and roll them over into future markets, say for cyclones in the Indian Ocean. WHAT HAPPENS NOW? We have received funding from the SCOR Foundation for Science that will allow us to expand beyond our initial Atlantic hurricanes market. We want to establish a long-horizon two-dimensional market where you have atmospheric CO2 concentration on one axis, and temperatures above pre-industrial levels on the other. This allows us to determine the distribution of temperature anomaly conditional upon specific intervals of cumulative atmospheric CO2. This conditional forecast information is essential for informing public and corporate planning and decision-making. The price distribution in this market structure contains information that can be used to determine a range of derived quantities that have value to planners and decision makers. For instance, the Intergovernmental Panel on Climate Change (IPCC) and other organisations produce CO2 emissions scenarios, but how likely is any one of those scenarios? With information contained in the twodimensional market, we can derive the likelihood of one scenario relative to another, and thereby allow planners and decision makers to infer which scenarios are more likely. What types of expertise do we seek? For the temperature anomaly dimension, most participants will come from the physical sciences and cognate disciplines. But for the CO2 concentration dimension, social scientists, country specialists, political scientists, regulation specialists, economists, and international relations specialists are needed. You need experts who understand not only atmospheric natural processes, but also the innovation side – what is possible in terms of energy efficiency, carbon capture and storage. Since four countries account for 55% of annual CO2 emissions (China, US, India, and Russia), expertise in these countries is needed, but equally for the emitters responsible for the remaining 45%. You need people who understand political economies. Will and can countries change their policies and stances? What needs to happen economically, socially and politically for that to take place? HOW CAN WE HELP? The insurance industry is very interested in our work. They are highly skilled in analysing data and in drawing implications from historical data. But particularly for weather and climate, they have found increasingly in recent years that climate change has started to limit the value of historical data and traditional forecasting methods for extreme weather-linked losses. They can see the value in our data, which comes from all these assorted experts and methodologies When we generate information, it is of use for policymakers, regulators, financial markets, and insurance companies – not necessarily in that order! As we develop more markets and bring more teams of experts on board, we hope the amount and quality of that information will only increase, creating a virtuous cycle of benefits and involvement. FIFTY FOUR DEGREES | 41 Dr Kim Kaivanto is a Senior Lecturer in the Department of Economics, and Principal Investigator of the Climate Risk and Uncertainty Collective Intelligence Aggregation Laboratory (CRUCIAL) initiative. CRUCIAL is a joint initiative between Lancaster University and the University of Exeter, funded by the SCOR Foundation for Science. It employs expert prediction to elicit, aggregate, and summarise knowledge on future climate risks, drawing on expertise from physical, social, and policy science. k.kaivanto@lancaster.ac.uk CRUCIAL
RkJQdWJsaXNoZXIy NTI5NzM=