Abdullah Yavas’s research while affiliated with University of Wisconsin–Madison and other places

We investigate behavior in long‐run projects and its relationship with experimentally elicited time preferences. Participants engage in a longitudinal project requiring sustained real effort, with their time preferences estimated through monetary outcomes. We observe a tendency to front‐load real effort, with choices reflecting both present and future bias, the former being more prevalent and severe. We also find evidence of naive choice reversals. However, there is no support for the quasi‐hyperbolic discounting model in monetary choices, and its predictions do not align with real effort allocation patterns. Nevertheless, discount rate and present bias parameters derived from monetary outcomes demonstrate predictive power over real effort allocation.

We examine matched rent-price ratios and rent transaction prices for single-family houses in Miami-Dade County between January 2009 and April 2014. The primary dataset consists of properties that are purchased and then rented within 240 days of the purchase. Each of the buyers in the sample are considered investors since each property included has a rental event indicating they are not owner occupied. We examine the relationship between housing and market characteristics and the impact active investors have on single-family rents and rent-price ratios. Entities that purchase the largest number of units pay more for properties, obtain marginally higher rent and obtain lower rent-to-price ratios.

Using a unique nationwide mortgage servicing dataset, this paper investigates the underreporting of second liens in portfolio mortgages and compares underreporting in portfolio versus privately securitized mortgages. Portfolio loans have more than 40% of the second liens underreported. Low documentation securitized loans have a 47% (in relative terms) lower misreporting rate than observably similar portfolio loans. The portfolio setting allows us to provide strong evidence that misreporting happens in the early stages of intermediation by lenders. The decreased occurrence of misreporting in sold loans subjected to more rigorous screening indicates the effectiveness of MBS issuers' screening, though its overall effect remains limited. Further, we show that the lender‐MBS issuer affiliation also plays a role in misreporting.

This study investigates whether flipping activities impose an externality on the transaction prices of the neighboring non‐flipped properties. Using a dataset of residential property transactions in Clark County, Nevada for the period 2003–2013, we find that flippers impose a significant positive impact on the price of neighboring non‐flipped properties in an up market, but a significant negative effect in a down market. This pro‐cyclical impact of flipping activity contributes to the volatility of housing prices, hence magnifying boom and bust cycles and increasing the likelihood of a mortgage crisis. This article is protected by copyright. All rights reserved

This study offers a simple theoretical model and empirical evidence to address the impact of social capital on mortgage delinquency. Social capital includes the norms, values, trust, and information common to a social network, which enable cooperative and shared actions. Using a new county-level dataset between 1999 and 2011 for the U.S, we find new evidence to show that social capital significantly affects the likelihood of mortgage delinquency. In particular, we find that a one-standard-deviation increase in social capital leads to a 0.13 standard deviation decrease in mortgage delinquency. The effect of social capital remains significant after controlling for location fixed effects and addressing endogeneity. The primary explanation is that social norms or trust could limit opportunistic behavior among homeowners and negatively affect strategic default activities. We also find that the impact of social capital on mortgage delinquency increased after the recent financial crisis. Furthermore, we show that the impact of social capital is more pronounced when the default is more likely to be strategic. Our findings have important implications for players in the mortgage industry and for policymakers in that cooperative and shared actions can play an important role in the mortgage default process. Thus, the assessment of default risk should consider social capital in addition to the factors already documented in the literature.

Loan performance varies by default probability and loss given default (LGD). While the relationship between securitization and default probability has been studied intensively, the effect of securitization on LGD remains unexplored. Using a unique dataset containing over forty thousand mortgage liquidations, this paper studies the effect of private securitization on LGD. We document that securitized loans incur higher LGD than observably similar portfolio loans. The results indicate that the effect comes mainly from the difference in loan quality between securitized and portfolio loans. Securitized opaque mortgages incur more than 18% (in relative terms) higher loan losses than observably similar portfolio loans. In addition, a difference‐in‐difference analysis provides evidence that pre‐crisis securitization standards contributed to higher loan losses. This article is protected by copyright. All rights reserved