A Market for Trading Forecasts: A Wagering Mechanism

Abstract

The ever-increasing interest in the collection of data by advancing technical and social sectors makes it distributed in terms of ownership. Also, the diverse expertise of these owners results in the extraction of varying quality of predictive information. Thus, the platforms for pooling forecasts based on distributed data and heterogeneous predictive skills allow gaining a collective value for a prediction task. For this purpose, we design a wagering-based forecast elicitation market platform, where a buyer intending to improve their forecasts posts a prediction task, and sellers respond to it with their forecast reports and wagers. This market delivers an aggregated forecast to the buyer (pre-event) and allocates a payoff to the sellers (post-event) for their contribution. We propose a payoff mechanism and prove that it satisfies several desirable economic properties, including those specific to electronic platforms. Furthermore, we discuss the properties of the scoring rules that allow us to provide ex-ante theoretical support for showing that the quantile averaging produces higher-quality aggregate forecasts in terms of scores, compared to the widely used linear pooling methods. Finally, we provide numerical examples to illustrate the structure and properties of the proposed market platform.

Online coalitional games for real-time payoff distribution with applications to energy markets

“…Motivational example: Let us consider an online forecast valuation scheme, inspired by [22] and [23], for pooling the information and expertise held by different owners and generating a combined forecast. First, let us introduce the forecasting markets that are designed to predict an event e.g.…”