European natural gas seasonal effects on futures hedging

This paper is the first to discuss the design of futures hedging strategies in European natural gas markets (NBP, TTF and Zeebrugge). A common feature of energy prices is that conditional mean and volatility are driven by seasonal trends due to weather, demand, and storage level seasonalities. This paper follows and extends the Ederington and Salas (2008) framework and considers seasonalities in mean and volatility when minimum variance hedge ratios are computed. Our results show that hedging effectiveness is much higher when the seasonal pattern in spot price changes is approximated with lagged values of the basis (futures price minus spot price). This fact remains true for short (a week) and long (one, three and six months) hedging periods. Furthermore, volatility of weekly price changes also has a seasonal pattern and is higher in winter than in summer. A simple volatility seasonal model that is based on sinusoidal functions on the basis improves the risk reduction obtained by strategies in which hedging ratios are estimated with linear regressions. Seasonal hedging strategies, linear regression based strategies, or even a naïve position, perform better than more sophisticated statistical methods.