Improving Weather Derivative Trading with Probabilistic S2S Forecasts

Salient Predictions uses machine learning to provide differentiated subseasonal-to-seasonal (S2S) weather forecasts. Traders can apply these forecasts directly to weather-linked derivatives such as Heating Degree Days (HDDs) / Cooling Degree Days (CDDs) and to weather-influenced commodities like power or gas. Salient's probabilistic S2S forecast leverages deep weather connections 2 to 52 weeks in advance, which provides non-consensus insights not available to traditional 14-day deterministic forecasts.

This study compares actual settlement values to forecasts at HDD/CDD locations traded on the Chicago Mercantile Exchange. We propose and test a decision making framework that quantifies the value of an accurate, reliable probabilistic forecast. While no individual forecast is guaranteed to be perfect, over time a quality forecast creates a measurable edge. Salient's forecasts enable more informed decisions that can lead to increased profits and reduced risk.