A spatial Bayesian approach to weather derivatives

Nicholas D. Paulson, Chad E. Hart, Dermot J. Hayes

Research output: Contribution to journalArticle


Purpose – While the demand for weather-based agricultural insurance in developed regions is limited, there exists significant potential for the use of weather indexes in developing areas. The purpose of this paper is to address the issue of historical data availability in designing actuarially sound weather-based instruments. Design/methodology/approach – A Bayesian rainfall model utilizing spatial kriging and Markov chain Monte Carlo techniques is proposed to estimate rainfall histories from observed historical data. An example drought insurance policy is presented where the fair rates are calculated using Monte Carlo methods and a historical analysis is carried out to assess potential policy performance. Findings – The applicability of the estimation method is validated using a rich data set from Iowa. Results from the historical analysis indicate that the systemic nature of weather risk can vary greatly over time, even in the relatively homogenous region of Iowa. Originality/value – The paper shows that while the kriging method may be more complex than competing models, it also provides a richer set of results. Furthermore, while the application is specific to forage production in Iowa, the rainfall model could be generalized to other regions by incorporating additional climatic factors.

Original languageEnglish (US)
Pages (from-to)79-96
Number of pages18
JournalAgricultural Finance Review
Issue number1
StatePublished - May 11 2010


  • Agriculture
  • Climatic loading
  • Insurance
  • Modelling
  • Rainfall
  • United States of America

ASJC Scopus subject areas

  • Agricultural and Biological Sciences (miscellaneous)
  • Economics, Econometrics and Finance (miscellaneous)

Fingerprint Dive into the research topics of 'A spatial Bayesian approach to weather derivatives'. Together they form a unique fingerprint.

  • Cite this