An Examination of Human Needs Addressed in the Expository Preaching of David Crosby Following Hurricane Katrina

This dissertation presents studies of default behavior with endogenous mortgage pricing in two life cycle models with fixed rate mortgages (FRM) and adjustable rate mortgages (ARM) respectively.

A potential mortgage holder has the option; during the lifetime of the loan to remain a homeowner; default on the mortgage, or prepay the mortgage by selling the house over the life cycle.

A financial intermediary prices a potential mortgage by finding a suitable fixed rate or mortgage premium such that the expected present value of discounted cash flows equals the initial mortgage debt.

Both equilibrium FRM rates and ARM uremia offered to a potential new homeowner are highly correlated with the probability of future default.

This research finds that negative equity risk and interest rate risk are the primary drivers of default, but their effect is not uniform across the two types of mortgages.

In particular, negative equity has a, stronger impact on the default decision for ARM holders.

Additionally, FRM debtors are more likely to default when interest rates are low, while the opposite is true for ARM debtors.

I discuss two effects, the utility share effect and wealth share effect, which determine the level of importance negative equity has on the default decision of a mortgage debtor.

In response to the recent mortgage crisis, this dissertation supports the idea that biased beliefs in the evolution of house prices helped lead to a rise in the homeownership rate and proportion of debtor default.

In addition to the fundamental analysis of the models.

I develop an empirical counterpart to these theoretical models using a competing risk hazard model framework.

This allows for simulations of the model to be utilized in a comparison to real world mortgage performance data while allowing for the testing of the robustness of the theoretical models.

Finally, I describe several extensions and policy objectives for which the models may be utilized.