Hedging late frost risk in viticulture with exotic options

Purpose – The purpose of this paper is to model and to value a temperature derivative to hedge late frost risk in viticulture. Design/methodology/approach – Starting from 11 years of historical temperature data collected in Mendoza, Argentina, the authors reconstruct the missing data using principal component analysis. The frequency content of time series is examined by the periodogram method; ordinary least squares are used to estimate the trends of minimum, maximum and average temperatures, and hypothesis tests of univariate and bivariate normality are performed on deseasonalized and filtered temperature returns. The authors express the temperature dynamics by correlated Ornstein‐Uhlenbeck processes and historical data were fitted into the model to obtain parameters estimates. An Asian‐type option on a temperature index is constructed and its price and sensitivities are computed by Monte Carlo method. Findings – The authors define an index in terms of minimum and average temperatures that, under some simplifying hypotheses, quantifies the damage produced by a late frost. To hedge the late frost risk, an Asian‐type option on the index is constructed. Together with the results concerning the design and pricing of the option, the analysis of historical data reveals non‐negligible linear trends, negative in minimum temperature and positive in maximum and average temperatures. These findings may be consistent with the hypothesis of global warming or with the presence of out‐of‐phase very low frequency components. Originality/value – The authors have not found in the literature a similar option to hedge the risk of spring frosts faced by fruit producers.