Pricing Mortgage‐Backed Securities: Integrating Optimal Call and Empirical Models of Prepayment
构建了一个理性抵押贷款提前还款模型，同时包含最优赎回和非最优提前还款行为，通过区分交易成本和非利率因素，更精确地解释提前还款滞后和“倦怠”现象，并用实际证券价格验证了模型的无偏性。
Residential mortgage borrowers frequently appear to behave suboptimally with respect to their mortgage prepayment options. Many borrowers fail to exercise even well‐into‐the‐money options while others prepay when the call option is out‐of‐the‐money. To account for these apparently suboptimal prepayments, the recent trend in mortgage‐backed securities research has been away from optimal call valuation models, in which the decision to exercise is determined endoge‐nously, in favor of models in which prepayment behavior is exogenously specified based on empirical estimation. This paper develops a rational model of mortgage prepayment which incorporates both types of “non‐optimal” prepayment and retains endogenous call. This enables the model to disentangle and compare the separate effects of the interest rate call, impeded by transaction costs, and of non‐interest‐rate driven prepayment. In addition, by recognizing heterogenous borrower transaction costs, the model presents a way to account more precisely for the varying prepayment lags associated with well‐into‐the‐money call options and to account for the phenomenon of “burnout” within a mortgage pool. The paper includes an empirical test of the unbiasedness of the integrated pricing model by comparing simulated prices from our theoretical model to observed prices on traded Fannie Mae and GNMA securities.