The Impact of Forecasting Jumps on Forecasting Electricity Prices

The paper is devoted to forecasting hourly day-ahead electricity prices from the perspective of existence jumps. We compare results different jump detection techniques and identify common features price apply jump-diffusion model with a double exponential distribution sizes explanatory variables. In order improve accuracy forecasts, we take into account time-varying intensity occurrences. forecast moments occurrences depending on several factors, including seasonality weather conditions, by means generalised ordered logit model. study conducted basis data Nord Pool power market. empirical indicate that jumps mechanism prediction useful in for peak hours, i.e., probabilities downward, no or upward improves forecasts prices.