2010 CDO Thesis - Harvard Kennedy School - Harvard University

1. Introduction. Collateralized debt obligations (CDOs), once a money making machine on Wall ... The basic principle behind a CDO involves the re-packaging of fixed income ...... While this finding is not surprising, it is a stark illustration of the.
3MB Sizes 43 Downloads 324 Views
The Story of the CDO Market Meltdown: An Empirical Analysis

Anna Katherine Barnett-Hart

Presented to the Department of Economics in partial fulfillment of the requirements for a Bachelor of Arts degree with Honors

Harvard College Cambridge, Massachusetts

March 19, 2009


The Story of the CDO Market Meltdown: An Empirical Analysis*

Abstract: Collateralized debt obligations (CDOs) have been responsible for $542 billion in write-downs at financial institutions since the beginning of the credit crisis. In this paper, I conduct an empirical investigation into the causes of this adverse performance, looking specifically at asset-backed CDO’s (ABS CDO’s). Using novel, hand-collected data from 735 ABS CDO’s, I document several main findings. First, poor CDO performance was primarily a result of the inclusion of low quality collateral originated in 2006 and 2007 with exposure to the U.S. residential housing market. Second, CDO underwriters played an important role in determining CDO performance. Lastly, the failure of the credit ratings agencies to accurately assess the risk of CDO securities stemmed from an overreliance on computer models with imprecise inputs. Overall, my findings suggest that the problems in the CDO market were caused by a combination of poorly constructed CDOs, irresponsible underwriting practices, and flawed credit rating procedures.


I would like to thank the following people and businesses who willingly gave of their time and expertise to help me tell the story of the CDO market meltdown: Michael Blum, Michael Blum Consulting, Ann Rutledge, Sylvain Raynes, R&R Consulting, Eliot Smith, Sam Jones, Mark Adelson, Mark McKenna, Thomas Giardi, Arturo Cifuentes, Douglas Lucas, Paul Muolo, Richard Baker, Eric Siegel, and Richard Gugliada. I am also grateful for the guidance and advising of the following Harvard Professors and doctoral students: Efraim Benmelech, Paul Healy, Erik Stafford, Allen Ferrell, Martin Feldstein, Erkko Etula, Laura Serban, Jenn Dlugosz, and David Seif. All remaining errors are my own.


1. Introduction Collateralized debt obligations (CDOs), once a money making machine on Wall Street, have been responsible for $542 billion of the nearly trillion dollars in losses suffered by financial institutions since 2007.1 Perhaps most disturbing about these losses is that most of the securities being marked down were initially given a rating of AAA by one or more of the three nationally recognized credit rating agencies,2 essentially marking them as “safe” investments.3 While the credit rating agencies have taken heavy criticism for their role in mis-rating billions of dollars in CDO tranches,4 they were not alone in their mistake. Indeed, almost all market participants, from investment banks to hedge funds, failed to question the validity of the models that were luring them into a false sense of security about the safety of these manufactured securities. How could so many brilliant financial minds have misjudged, or worse, simply ignored, the true risks associated with CDOs? In this paper, I use novel, hand-collected data from 735 ABS CDOs to shed light on this mystery, investigating the causes of adverse performance in CDOs backed by asset-backed securities (ABS CDOs).5 I characterize the relative importance of general CDO properties, underwriting banks, and credit rating agencies in contributing to the collapse of the CDO market and document several findings. First, the properties of the CDO collateral, including asset class and vintage, are the most important factor in explaining the variation in CDO performance. In particular,

According to CreditFlux Newsletter, as of January 8, 2008. Moody’s, S&P, and Fitch. 3 According to financial consultant Mike Blum, underwriters would often pay for all three agencies to rate their deals to “convey the impression that these bonds were rock-solid.” 4 See Roger Lowenstein’s article, “Triple-A Failure,” for