Abstract
Several factors considered significant in determining the level of construction lending risk were tested for association with samples of troubled and good residential construction loans. Generally, a priori expectations were confirmed. The strongest associations with risk were found for unavailability of materials, followed by inflationary cost overruns and borrower experience.
Interest rates on the sample of troubled loans did not significantly differ from the rates on the sample of good loans, suggesting the lender did not adequately recognize elements of lending risk, or that those risks were ignored.