Forecasting the Spread Between HUPX and EEX DAM Prices the Case of Hungarian and German Wholesale Electricity Prices

摘要

Electricity price forecasting is a rapidly developing field. Using time series analysis to estimate future wholesale electricity prices is a popular approach, however forecasting price difference between markets is not typical. In this paper the spread between Hungarian (HUPX) and German (EEX) day-ahead power prices are forecasted, using combined ARIMA-GARCH type models. Models are built on 2011–2017 daily average price data, and forecasts are made for Q1 2018. All four applied model types (NGARCH, TGARCH, EGARCH, GJR GARCH) bring similar results, with an RMSE of 4.22-4.75 and a MAE of 2.87-2.99 while the average spread value in the forecasted period is 3.88. All models perform better than the simple ARIMA model.