Housing Tenure and Mortgage Choice

This paper examines the relationship between housing tenure and mortgage contract. We present a model showing that, given expected mobility, borrowers will have incentive to self-select into the appropriate mortgage product such that their fixed-rate period is directly related to their probability of moving. We empirically test this hypothesis using housing tenure data derived from a large national database of repeat mortgage transactions. After controlling for borrower characteristics, the mobility hazards of 3/1, 5/1 and 7/1 ARMs, compared to a 30-year fixed rate, are estimated to be 28%, 14% and 11% higher, respectively