Credible Interval Computation in Weather Derivatives Pricing

71 Pages Posted: 23 Nov 2012

See all articles by Shree Khare

Shree Khare

Risk Management Solutions

Stephen Jewson

Risk Management Solutions

Date Written: November 23, 2012

Abstract

This paper examines the quantification of uncertainty in weather derivatives pricing. The focus is on the propagation of a posterior distribution on uncertain model parameters through to relevant payoff statistics, summarizing the uncertainty using variance based credible intervals for any given payoff statistic. We also demonstrate the use of sensitivity analysis to determine the most influential parameters on the credible interval size. This paper serves a multitude of purposes. Firstly, we provide a basic review of the underlying theory behind uncertainty/sensitivity analysis, along with authors' novel ideas for numerical implementation. We also provide a novel and informative set of numerical results for credible interval computation and sensitivity analysis for a market relevant 5-station temperature problem. This paper serves as a foundation for exploring the use of Bayesian uncertainty/sensitivity quantification in weather derivatives portfolio management.

Keywords: Uncertainty, Sensitivity, Weather Derivatives, Emulators

Suggested Citation

Khare, Shree and Jewson, Stephen, Credible Interval Computation in Weather Derivatives Pricing (November 23, 2012). Available at SSRN: https://ssrn.com/abstract=2180041 or http://dx.doi.org/10.2139/ssrn.2180041

Shree Khare (Contact Author)

Risk Management Solutions ( email )

London EC3R 8NB
United Kingdom

Stephen Jewson

Risk Management Solutions ( email )

London EC3R 8NB
United Kingdom

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
111
Abstract Views
820
Rank
444,645
PlumX Metrics