Working paper Weather, Storage, and Natural Gas Price Dynamics: Fundamentals and Volatility

This paper assesses how market fundamentals affect asset return volatility by drawing on evidence from the U.S. natural gas futures market. One of the novel features of this paper is the use of the deviation of temperatures from normal (weather surprise) as a proxy for demand shocks and a determinant of the conditional volatility of natural gas futures returns. I estimate a GARCH model using daily natural gas futures data from January 1997 to December 2000. The empirical result shows that the weather surprise variable has a significant effect on the conditional volatility of natural gas prices and the inclusion of the weather surprise variable in the conditional variance equation reduces volatility persistence. Combined with the evidence that volatility is considerably higher on Monday and the day when natural gas storage report is released, these results show that information about market fundamentals are important determinants of natural gas price volatility. Aside from these findings, I also document that returns of the first month futures are more volatile than those of the second month futures, which is consistent with Samuelson’s (1965) hypothesis that commodity futures price volatility declines with contract horizon. Acknowledgement This paper is based on a chapter of my dissertation. I am grateful to Timothy Dunne, Aaron Smallwood and Daniel Sutter for their guidance. Also I thank Dennis O’Brien and the Institute for Energy Economics and Policy for financial support to acquire natural gas futures trading data, Peter Lamb, Mark Richmond and Reed Timmer for help on weather data. All errors remain mine.