Working Paper: NBER ID: w30187
Authors: Ralph S. J. Koijen; Hae Kang Lee; Stijn Van Nieuwerburgh
Abstract: We study aggregate lapsation risk in the life insurance sector. We construct two lapsation risk factors that explain a large fraction of the common variation in lapse rates of the 30 largest life insurance companies. The first is a cyclical factor that is positively correlated with credit spreads and unemployment, while the second factor is a trend factor that correlates with the level of interest rates. Using a novel policy-level database from a large life insurer, we examine the heterogeneity in risk factor exposures based on policy and policyholder characteristics. Young policyholders with higher health risk in low-income areas are more likely to lapse their policies during economic downturns. We explore the implications for hedging and valuation of life insurance contracts. Ignoring aggregate lapsation risk results in mispricing of life insurance policies. The calibrated model points to overpricing on average. In the cross-section, young, low-income, and high-health risk households face higher effective mark-ups than the old, high-income, and healthy.
Keywords: No keywords provided
JEL Codes: E32; E44; G12; G22; G52
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| macroeconomic conditions (E66) | lapsation rates (C41) |
| economic downturns (F44) | lapsation rates (C41) |
| credit spreads (G12) | lapsation rates (C41) |
| unemployment (J64) | lapsation rates (C41) |
| higher health-risk households (I14) | lapsation rates (C41) |
| individual policyholder characteristics (G52) | lapsation cycle (C41) |
| aggregate lapsation risk (G33) | mispricing of life insurance policies (G52) |
| mortality effect (J17) | mispricing of life insurance policies (G52) |