A dynamic insurance market model with competitive insurers, financial investment, and liquidity management typically fails to produce an equilibrium. The combination of underwriting, investment, recapitalization, and dividend decisions creates feedback loops that prevent prices from settling. The model is too flexible for its own good.
Introduce model uncertainty — make insurers act under worst-case probability beliefs — and equilibrium existence is restored. Doubt stabilizes the market. Insurers who assume the worst are more conservative, and this conservatism breaks the feedback loops that prevented equilibrium under precise beliefs.
The mechanism inverts the usual relationship between uncertainty and instability. In most economic models, uncertainty is a source of volatility, excess risk premia, market failures. Here, it is the cure. The worst-case behavior that model uncertainty induces acts as an implicit coordination device: all insurers becoming simultaneously cautious produces the regularity that precise optimization destroys.
A second counterintuitive finding: when insurance gains and financial returns are positively correlated, insurance prices can carry negative loadings — prices decrease despite the added correlation risk. Conventional pricing says positively correlated risks should cost more to insure. The model shows this need not hold in dynamic equilibrium, where the interplay between investment returns and underwriting cycles creates pricing patterns that contradict static intuition.
Certainty breaks the market. Doubt restores it. Correlated risk reduces prices rather than raising them. The system's behavior under precise knowledge is worse than its behavior under acknowledged ignorance.