Modeling hourly Electricity Spot Market Prices as non stationary functional times series

The instantaneous nature of electricity distinguishes its spot prices from spot prices for equities and other commodities. Up to now electricity cannot be stored economically and therefore demand for electricity has an untempered effect on electricity prices. In particular, hourly electricity spot prices show a vast range of dynamics which can change rapidly. In this paper we introduce a robust version of functional principal component analysis for sparse data. The functional perspective interprets spot prices as functions of demand for electricity and allows to estimate a single price curve for each day. Variations in market fundamentals such as commodity prices are absorbed by the first principal components.