1. Introduction2. Subprime MortgagesB. Subprime Mortgage Design3. The Design and Complexity of Subprime RMBS Bonds4. Collateralized Debt Obligations (CDOs)5. The Panic7. DiscussionElectronic copy available at: http://ssrn.com/abstract=1276047 The Subprime Panic+ Gary Gorton Yale School of Management and NBER Prepared for the journal European Financial Management This version: September 30, 2008 Abstract Understanding the ongoing credit crisis or panic requires understanding the designs of a number of interlinked securities, special purpose vehicles, and derivatives, all related to subprime mortgages. I describe the relevant securities, derivatives, and vehicles to show: (1) how the chain of interlinked securities was sensitive to house prices; (2) how asymmetric information was created via complexity; (3) how the risk was spread in an opaque way; and (4) how trade in the ABX indices (linked to subprime bonds) allowed information to be aggregated and revealed. These details are at the heart of the origin of the Panic of 2007. The events of the panic are described. + This paper is a much shorter, somewhat revised, version of a paper entitled “The Panic of 2007,” which was prepared for the Proceedings of a Symposium on “Maintaining Stability in a Changing Financial System,” Federal Reserve Bank of Kansas City, Jackson Hole Conference, August 2008. See http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1255362 . For comments, suggestions, and assistance with data and examples on the Jackson Hole paper, I thank Geetesh Bhardwaj, Omer Brav, Adam Budnick, Jared Champion, Kristan Blake Gochee, Itay Goldstein, Ping He, Bengt Holmström, Lixin Huang, Matt Jacobs, Arvind Krishnamurthy, Tom Kushner, Bob McDonald, Hui Ou-Yang, Ashraf Rizvi, Geert Rouwenhorst, Hyun Shin, Marty Wayne, Axel Weber, and to those who wished to remain anonymous.Electronic copy available at: http://ssrn.com/abstract=12760471 1. Introduction Subprime mortgages are a financial innovation designed to provide home ownership opportunities to riskier borrowers in the U.S. Such borrowers are indeed riskier (also poor and disproportionately minority), and lending to this group involved a particular mortgage design feature, that resulted in linking the outcome to house price appreciation. Subprime mortgages were then financed via securitization, which in turn has a unique design, reflecting the subprime mortgage design. Subprime securitization tranches were then often sold into CDOs. Tranches of CDOs were, in turn, often purchased by market value off-balance sheet vehicles, and money market mutual funds. Additional subprime risk was created (though not on net) with derivatives. This nexus of off-balance sheet vehicles, derivatives, securitization, and, in addition, the growth of the repo (repurchase agreement) market constitute what has come to be known as the “shadow banking system.” When the U.S. housing prices did not rise as expected, this chain of securities, derivatives, and off-balance sheet vehicles could not be penetrated by most investors or counterparties in the financial system to determine the location and size of the risks. Faced with this lack of information, financial intermediaries refused to deal with each other and began to hoard cash. The panic began. An important part of the information story is the introduction, in 2006, of new synthetic indices of subprime risk, the ABX.HE (“ABX”) indices. These indices trade over-the-counter. For the first time information about subprime values and risks was aggregated and revealed. While the location of the risks was unknown, market participants could, for the first time, express views about the value of subprime bonds, by buying or selling protection. In 2007 the ABX prices plummeted. The common knowledge created, in a volatile way, ended up with the demand for protection pushing ABX prices down. At the root of the information story are the details of the chain of securities and vehicles through which the risk was distributed. In this paper I describe the relevant securities, derivatives, and vehicles and to show: (1) how the chain of interlinked securities was sensitive to house prices; (2) how asymmetric information was created via complexity; (3) how the risk was spread in an opaque way; and (4) how the ABX indices allowed information to be aggregated and revealed. I argue that these details are at the heart of the answer to the question of the origin of the Panic of 2007. The panic poses challenges for economists as well as regulators and policymakers. None of the various layers of intertwined securities, off-balance sheet vehicles (and their liabilities), or derivatives are traded in markets that resemble those that economists tend to focus on, namely, the secondary market for equities. Nor does the banking system that I will describe look very much like what is taught in courses on “banking.” Further, there is some empirical work on crises, but little if it weighted by the importance of the event. (I mention some of this work later.) The panic should be a momentous event for economic research. Section 2 is devoted to explaining how subprime mortgages work. The focus is on implicit contract features, which link the functioning of these mortgages to home price appreciation. Subprime mortgage originators financed their businesses via securitization, but the securitization of subprime mortgages is very different from the securitization of other types of assets (e.g. prime mortgages, credit cards, auto loans). Subprime securitization has dynamic tranching as a function of excess spread and prepayment and is sensitive to house prices as a result. This is explained in Section 3. That is not the end of the story, because tranches of subprime residential mortgage-backed securities (RMBS) were often sold to2 collateralized debt obligations (CDOs). Section 4 briefly explains the link to CDOs and the inner workings of these vehicles, the issuance of CDOs, links to subprime, and the synthetic creation of subprime RMBS risk. Section 5 is about the panic itself, the falling house prices, the role of the ABX indices, the runs on the SIVs. I also try to summarize the information argument of the paper. In Section 6 I briefly discuss the “originate-to-distribute” hypothesis. Some final discussion is contained in Section 7. A final section, Section 8, discusses the U.S. Treasury Secretary’s proposed plan for addressing