Wednesday, April 22, 2009

Why bankers assumed U.S. housing prices couldn't fall

From The Economist:
Alex Pollack reckons finance professionals did believe housing prices could fall, but just not on a national level. The data used to stress test mortgage assets was based on the experience of Texas and other oil-patch states in the 1970s and 80s. It provided an instance of a housing bubble that led to falling house prices. The problem was, since the Depression, house prices had never fallen on a national level. There existed no data that contained a large and positive correlation of home price across different regions and also had prices falling. This is the limitation of historical data; you use the past to predict the future. When you enter a new regime you are left with your own ad-hoc judgement. Rather than take on that sort of responsibility, most prefer to base their assumptions on historical data.
All the bankers had to do was look at Japan.


  1. well, and the other thing is that historically speaking, the mortgage industry had always been very conservative and careful. Starting late 90's til things started to fall apart first with subprime and then now the prime borrowers, we had unprecedented loose lending standards and very low interest rates for unprecedented lenght of time. They just should have put that together and said,"hmm, can we really apply data from the past to this loose & crazy lending environment?" I mean weren't these guys genius math geeks and MBA toting whiz kids? Duh!


  2. Another thing they forgot to consider is that lending was localized up until this bubble.

    If you put all your eggs in the same basket you can't rely on past results.