Use of Climatological Data in Weather Insurance

Abstract

There are three major types of crop-related weather insurance: hail, all perlis, and rain insurance. The development of rates is an exercise in applied climatology, and the importance of the historical data selected for assessing risk (and developing rates) is revealed by the problems encountered by firms insuring against deficient summer rainfall during the drought of 1988. Extensive purchase of premiums costing $9 million for coverage ($400 million) in the Midwest occurred with buyers (farmers) apparently aware during May and June 1988 that an unusually dry event was in progress. The levels of loss (50% or less of average June-August rainfall) offered by insurance firms were exceeded throughout the Midwest. The firms attempted to refund the record number of premiums accepted in June, and in turn, 8000 farmers filed a class-action suit against the firms for failure to accept premiums and to provide coverage. The insurance firms ultimately settled by agreeing to pay $48 million in claims. The coverage offered was based on the most recent 25 years of data, and this unusually wet period did not represent the longer-term likelihood of areally extensive dry summers. Uses of climatic data by the insurance industry include planning for the occurrence of such extreme event considerations, plus point vs area probabilities of these anomalous events; choosing the periods to select for routinely establishing new rates (rerating is typically done an 2- to 10-year cycles); and for determining the averages most appropriate to use for rate levels and sales considerations.