Volatility Models for Electricity Prices with Intra-daily information

In this work, we explore the impact that intra-daily information could have on explaining and forecasting the conditional volatility of daily electricity returns. Returns are computed on Italian spot prices. The basic model considers an autoregressive structure on the conditional mean, daily dummies for capturing weekly seasonality and lagged total daily volumes . The conditional variance equation is modeled with ARCH(1) and GARCH(1,1) models with intra-daily regressors given by total traded volumes and variance of hourly returns at time t-1. The inclusion of intra-daily information reduces the volatility persistence, hence inducing better volatility forecasts of standardized returns.