Cross-city hedging with weather derivatives using bivariate DCC GARCH models

As monopolies gave their way to competitive wholesale electricity markets, volumetric risk came into play. Electricity supplier can buy weather derivatives to protect from volumetric risk due to unexpected weather conditions. However, contracts can only be negotiated for weather variables measured at few selected locations. To hedge their specific risk, electricity supplier have to correlate their risk with the risk at tradeable locations. In this paper, we concentrate on temperature derivatives. More precisely, we examine if and how bivariate GARCH models with dynamic conditional correlations can help in modelling correlation between two distinct temperature time series. The knowledge of correlation dynamics between the temperature time series enables an electricity supplier to correlate his risk with the risk of a traded city and to construct a sensible hedge. It turns out that the application of bivariate DCC GARCH models to three German temperature time series provides encouraging results.