State Space Modeling of Mortgage Default Rates under Natural Hazard Shocks

TL;DR

This paper develops a state-space model to analyze how natural hazard-induced economic losses influence mortgage default rates, revealing significant relationships especially during high-loss events.

Contribution

It introduces a sliced state-space modeling approach to isolate the impact of natural hazards on mortgage defaults, accounting for latent economic factors.

Findings

Natural hazard losses significantly increase mortgage default rates.

The sliced model captures effects only evident during high-loss periods.

Insights aid insurers in risk assessment and management.

Abstract

Mortgage default rates, on the one hand, serve as a measure of economic health to support decision-making by insurance companies, and on the other hand, is a key risk factor in the asset-liability management (ALM) practice, as mortgage related assets constitute a significant proportion of insurers' investment portfolios. This paper studies the relationship between economic losses due to natural hazards and mortgage default rates. The topic is greatly relevant to the insurance industry, as excessive insurance losses from natural hazards can lead to a surge in mortgage defaults, creating compounded challenges for insurers. To this end, we apply a state-space modeling approach to decouple the effect of natural hazard losses on mortgage default rates after controlling for other economic determinants through the inclusion of latent variables. Moreover, we consider a sliced variant of the classical SSM to capture the subtle relationship that only emerges when natural hazard losses are sufficiently high. Our model verifies the significance of this relationship and provides insights into how natural hazard losses manifest as increased mortgage default rates.