Some Lessons From CDO Markets on Mathematical Models

Abstract

Bad mathematical models or over reliance on mathematical models is frequently cited as one of the culprits of the financial crisis. While models surely played a role, putting such a responsibility on models is giving too much credit to the influence that models have in the larger political decisions in financial institutions and in political decision making. Decisions to run banks with extreme leverage, to systematically try to loosen up capital constraints using off-balance sheet vehicles, searching-for-yield behavior by institutional and private investors, political desires to increase house ownership even at the cost of increasing riskiness in lending — these decisions are not deeply rooted in models. When they are, one is suspicious that the model selection has often been such that the decisions were justified in the selected models.