How Big are the Ambiguity-Based Premiums on Mortgage Insurances?

This paper studies how ambiguity aversion affects the pricing of mortgage insurance (MI). We consider pricing-kernel ambiguity arising from market incompleteness. This ambiguity model is applied to a standard framework of MI-ML (mortgage loan) structural pricing. Our quantitative results show that insurers' ambiguity aversion generates substantial positive effects on MI premium. Ambiguity impacts are highly sensitive to loan-to-value ratio, ambiguity magnitude, and the tightness of information constraints. By using the U.S. city-level housing and mortgage data, we estimate that, on average, ambiguity aversion increases MI premium rate by 77% (46bps), and explains about 60-90% of pricing errors.