Modeling the Daily Electricity Price Volatility with Realized Measures

Abstract

We propose using the Realized GARCH model to estimate the daily price volatility in the EPEX power markets. The model specification extracts the volatility-related information from realized measures, which substantially improves the in-sample fit of the data compared to the standard EGARCH model. More importantly, evidence on the out-of-sample forecasts reinforces the value of the specifications as the forecast quality is improved over the benchmark model under eight conventional criteria. The increased forecast accuracy is robust under both the rolling-window and recursive estimation scheme. Finally, we show that intra-day range is an effective volatility indicator in the power market as the benefit of including intra-day range is substantial as compared to realized variance.

Keywords: volatility forecasting, intra-day range, Realized GARCH, electricity

JEL Classification: C51, C53, G10