Asymmetric Risk

Energy market participants are exposed to risk. A significant risk factor is the “busted” temperature forecast. Various methods are used to manage this risk. These methods are often based on an incorrect assumption that the risk envelope around the temperature forecast takes a Gaussian form. The risk is then expressed as a confidence interval displaying a one sigma or two sigma plus/minus range around the mean temperature forecast. In reality, the risk of any given temperature forecast is often non-Gaussian, producing a directional signal of the risk profile. In this presentation, we describe several techniques to ascertain asymmetric risk. We then show historical examples of asymmetric risk temperature forecast events. We also show how employing asymmetric risk techniques benefit the decision making of the risk manager.