On the systemic nature of weather risk
Authors:
- 3 August 2010
- journal article
- Published by Emerald Publishing in Agricultural Finance Review
- Vol. 70 (2) , 267-284
Abstract
Purpose – The purpose of this paper is to assess the losses of weather-related insurance at different regional levels. The possibility of spatial diversification of insurance is explored by estimating the joint occurrence on unfavorable weather conditions in different locations, looking particularly at the tail behavior of the loss distribution. Design/methodology/approach – Joint weather-related losses are estimated using copulas. Copulas avoid the direct estimation of multivariate distributions but allow for much greater flexibility in modeling the dependence structure of weather risks compared with simple correlation coefficients. Findings – Results indicate that indemnity payments based on temperature as well as on cumulative rainfall show strong stochastic dependence even at a large regional scale. Thus the possibility to reduce risk exposure by increasing the trading area of insurance is limited. Research limitations/implications – The empirical findings are limited by a rather weak database. In that case the estimation of high-dimensional copulas leads to large estimation errors. Practical implications – The paper includes implications for the quantification of systemic weather risk which is important for the rate making of crop insurance and reinsurance. Originality/value – This paper's results highlight how important the choice of the statistical approach is when modeling the dependence structure of weather risks.Keywords
All Related Versions
This publication has 27 references indexed in Scilit:
- Biography: Agricultural Finance ReviewAgricultural Finance Review, 2009
- Basis risk and weather hedging effectivenessAgricultural Finance Review, 2008
- Using Index‐Based Risk Transfer Products to Facilitate Micro Lending in Peru and VietnamAmerican Journal of Agricultural Economics, 2007
- Copula Modeling: An Introduction for PractitionersFoundations and Trends® in Econometrics, 2007
- Measurement of aggregate risk with copulasThe Econometrics Journal, 2005
- Trends in indices for extremes in daily temperature and precipitation in central and western Europe, 1901-99International Journal of Climatology, 2005
- On the Possibility of a Private Crop Insurance Market: A Spatial Statistics ApproachJournal of Risk and Insurance, 2003
- Correlation and Dependence in Risk Management: Properties and Pitfalls2002
- An Introduction to CopulasTechnometrics, 2000
- A semiparametric estimation procedure of dependence parameters in multivariate families of distributionsBiometrika, 1995