The Dynamics of Insurance Underwriting Regimes in the UK: Evidence from Panel VAR

Using the unique dataset of the five major UK insurance sectors, we adopt a novel approach in the insurance literature and model the evolution of underwriting returns as Regime Switching processes. This produces estimates of time-varying conditional regime probabilities and captures non-normality characteristics present in the data. Using Dynamic Panel and Panel Vector Auto-Regressions we study the joint dynamics of underwriting regime probabilities and their attribution to economic factors. Our evidence uncovers high/low volatility switching for all sectors, where their joint evolution is mainly attributed to industry-specific factors. High volatility is linked with low profitability while impulse response functions and variance decompositions identify a negative association of changes in premiums and a positive association of changes in claims and interest rates with the likelihood of the low-profitability regime.