Goetzmann on Real Estate Collapse

The collapse of the housing market caught many real estate experts by surprise. Even after prices started to fall, mortgage lenders continued to approve loans to buyers who clearly lacked the ability to meet payments. The reasons were many: greed and sometimes outright fraud; compensation structures that valued quantity of loans provided over the quality of applicant; the fact many lenders immediately sold off mortgages to investment banks, making any defaults someone else’s problem. But one culprit that hasn’t received as much attention is the models banks commonly use to calculate worst-case scenarios for performance.

These models collect data going back years to figure out what a serious downturn in a market will look like. They are crucial in helping financial institutions decide whether an applicant is worth the risk. But a study by William Goetzmann, Edwin J. Beinecke Professor of Finance and Management Studies, and two co-authors, found a major flaw in the construction of such models. While they often went back ten years or more, the models didn’t include data from further back. As a result, they missed steeper dives in the markets, producing an overly rosy picture of a worst-case scenario and helping to lead to the greatest financial crisis since the Great Depression.

(6:48)