Giant Pool of Money
In this radio story about the events leading up to the subprime mortgage crisis, it’s clearly demonstrated that a few psychological biases and heuristics were present and played important roles in forming the crisis. The most critical ones I’ve identified are the confirmation bias and the social proof phenomenon in the development of the crisis.
Confirmation Bias
Mainly two types of confirmation bias were observed in the subprime crisis: the confirmation trap as well as anchoring heuristics. Anchoring caused the banks to miscalculate real risks of the mortgages and related MBS/CDO, while the confirmation trap caused most people in the industry to ignore potential risks and keep playing the dangerous game.
Bazerman and Moore defined the anchoring heuristic as “Individuals make estimates for values based on an initial value (derived from past events…), and typically make insufficient adjustments from that anchor when establishing a final value.” And the Wall Street banks made exactly this mistake leading up to the mortgage crisis.
When estimating the default rate for subprime mortgage products, Wall Street banks used historical data which indicated an average default rate of 2%, then extrapolated to estimate the absolute worst-case scenario to be at worst 10% to 12%. However, such anchoring is misleading because most of the data were years old were based on loans with strict income, asset and credit requirements. On the contrary, the subprime loans on which the MBS and CDO were based were mostly issued with very low, if any, income/asset requirements. For instance, many of the loans were NINA loans, which were “no income no asset” and did not require any income or asset proof as prerequisite for the loans. Subsequently, since the Wall Street banks only anchored their new estimates on the old data and did not make sufficient adjustments from