Multivariate Long-Memory Modeling of Daily Surface Air Temperatures and the Valuation of Weather Derivative Portfolios
25 Pages Posted: 25 Jul 2003
Date Written: July 2002
It has been shown in earlier work that long memory can be detected in multidecadal time series of daily surface air temperature measured at individual weather stations. Here, we present a multivariate stochastic model capable of capturing both the long memory at each station, and the lagged cross-correlations between the stations. It is shown that this model can be useful for the management of portfolios of weather derivatives.
Keywords: weather derivatives, weather risk, daily temperatures, temperature time-series, arfima, varfima, multivariate
JEL Classification: G12, G13
Suggested Citation: Suggested Citation