Underwriting Cycles in UK Motor Insurance

The purpose of this paper is to research adjustment process of the insurance premium and examine its cointegrated relationship with economic loss ratio, claim, interest rate and retail price index in UK motor insurance market. First, the article introduces methods of integration, cointegration, unit root tests and error correction models. Second, three hypotheses based on previous literature are proposed in this dissertation, so the first hypothesis is a cointegrated relationship between economic loss ratio, retail price index and interest rate. Then, the second hypothesis is a cointegrated relationship between premium, retail price index and interest rate. The third is the main hypothesis of this paper that is a cointegrated relationship between premium, retail price index, interest rate and claim, which is not tested by previous researchers. The Engle and Granger two steps error correction model is adapted to prove cointegration, which is a linear error correct mechanism to represent dynamics of the insurance premium. Third, the empirical result of this paper indicates adjustment dynamics of the insurance premium is continuous, symmetrical and linear in UK motor insurance market. The result of the first hypothesis presents a statistically significant cointegrated relationship between economic loss ratio and interest rate, but retail price do not share a long run relationship with economic loss ratio. The second hypothesis result shows premium has a long-term cointegrated relationship with interest rate and retail price index, but short-term shocks to both economic variables do not affect the premium. The third hypothesis result proves cointegrated relationship between claim and premium, but retail price index and interest rate do not share a long run relationship in this linear regression. This result indicates claim is a more important factor to premium than retail price index and interest rate in UK motor insurance so that previous research may be a misspecification in linear regression to underwriting cycles