Robust Self Exciting Threshold AutoRegressive models for electricity prices

In this paper we suggest the use of robust GM-SETAR(Self Exciting Threshold AutoRegressive) processes tomodel and forecast electricity prices observed on deregulatedmarkets. The robustness of the model is achieved by extendingto time series the generalized M-type (GM) estimator first introducedfor independent multivariate data. As it has been shown ina very recent paper [1], the polynomial weighting function overperformsthe classical ordinary least squares method when extremeobservations are present. The main advantage of estimatingrobust SETAR models is the possibility to capture two very well knownstylized facts of electricity prices: nonlinearity producedby changes of regimes and the presence of sudden spikes due toinelasticity of demand. The forecasting performance of the modelapplied to the Italian electricity market (IPEX) is improved bythe introduction of predicted demand as an exogenous regressor.The availability of this regressor is a particular feature of theItalian market. By means of prediction performance indexes andtests, it will be shown that this regressor plays a crucial role andthat robust methods improve the overall forecasting performanceof the model.