underwritings

Lowering of underwriting standards may have contributed much to the unprecedented recent rise and subsequent fall of mortgage volumes and house prices.

Keywords: underwriting, standards, mortgages, house prices, LTV

Underwriting standards for residential mortgages (henceforth: underwriting) are now generally regarded as having been unusually lax during the mid-2000s.

Conventional data do not satisfactorily measure aggregate underwriting over the past decade: Easing of underwriting, inside and especially outside of banks, was likely much more extensive than these data indicate.

Although we have several indicators of underwriting, none seems sufficient alone.

We argue that, for this time period, the method of principal components (PC) can usefully summarize the information in those variables about underwriting. We used the PC method to generate a single, new data series and show that the resulting indicator variable exhibits correlations with the input variables to the PC method that are consistent with its measurement of underwriting.

For the practicing economist, having a single variable that both summarizes multiple measures and more accurately measures a phenomenon, like underwriting, can be very valuable.

To assess the caliber of the derived underwriting variable, we used the variable in a vector auto-regression (VAR).

Further, the estimated VAR suggests how, in turn, the other variables, such as house prices and the volume of and interest rates on mortgages, were affected by changes in underwriting. Again, to the extent that the estimated responses to underwriting laxity and tightening fit our understanding, they further raise our confidence in the indicator of underwriting that we derived.

Underwriting, mortgage lending, and house prices: 1996-2008