Essays on Securitization and the Resolution of Financial Distress

This thesis consists of three empirical essays on securitization and the resolution of financial distress, examining the information content of prices of mortgage-backed securities, the role of securitization in preventing mortgage renegotiation and the effect of minimum bids in ascending price auctions of repossessed assets. In the first chapter, I present evidence that the prices at origination of residential mortgage backed securities contain information on the quality of the underlying asset pools above what is reflected in the ratings. Yield spreads at issuance predict both future downgrades and defaults even after the information contained in ratings is taken into account. This holds for all rating classes except triple-A. Yield spreads of the highest rated securities have no predictive power for future performance. This suggests that investors in triple-A were less informed about the quality of the securitized assets than investors in riskier, more information sensitive securities. In the second chapter (co-authored with Kris Gerardi and Paul Willen), we show that securitization does not explain the reluctance among lenders to renegotiate home mortgages. We focus on seriously delinquent borrowers from 2005 through the third quarter of 2008 and show that servicers renegotiate similarly small fractions of securitized and portfolio loans. The results are robust to several different definitions of renegotiation and hold in subsamples where unobserved heterogeneity is likely to be small. We argue that information issues endemic to home mortgages where lenders negotiate with large numbers of borrowers lead to barriers to renegotiation fundamentally different from those present with other types of debt. In the third chapter (co-authored with Antoinette Schoar), we study the effect of changing the reservation prices in online English auctions of repossessed motorcycles. The main intervention involves lowering the minimum bid initially set by the auctioneer by up to 10 percent in randomly selected vehicles. We find that lowering the minimum bids increases the probability of sale and reduces the final price conditional on a motorcycle being sold. This is in line with the predictions of standard auction theory and is evidence against auction fever being induced by lower initial minimum bids. The fact that bidders are regular participants in auctions and experienced buyers of the items for sale may help explain why they weren't subject to auction fever. Also, in our setting the number of bidders increased only slightly for discounted items, which is the opposite of what is necessary to induce an auction fever effect. For a subset of motorcycles that were bundled together for the auctions we a find lower probability of sale and lower prices conditional on sale. Thesis Supervisor: Antoinette Schoar Title: Michael Koerner '49 Professor of Entrepreneurial Finance