Date of Award

Summer 2024

Degree Type

Restricted to Claremont Colleges Dissertation

Degree Name

Economics, PhD

Program

School of Social Science, Politics, and Evaluation

Advisor/Supervisor/Committee Chair

Josh Tasoff

Dissertation or Thesis Committee Member

Pierangelo De Pace

Dissertation or Thesis Committee Member

Michal Kowalik

Terms of Use & License Information

Terms of Use for work posted in Scholarship@Claremont.

Rights Information

© 2024 Ya Qing (Annie) Zhang

Subject Categories

Economics

Abstract

Regulatory policy plays an important role in the US financial banking system, which it is interesting to observe the interaction among policy change, credit demand and credit supply. In my first paper, I focus on studying the impact of a new FASB accounting rule called Current Expected Credit Losses (CECL). This is a new FASB accounting standard in 2016 to modify the way how financial institutions measure allowance. CECL requires a forward looking process to estimate reserve for lifetime credit losses for financial institutions’ credit portfolios. Before the date of CECL implementation in 2020, there are several concerns and discussions around whether CECL will impact banks to make major changes within credit risk modeling, risk tolerance and capital management. Therefore, the objective of this study is to assess CECL’s treatment effect on bank’s lending strategy and risk profile management. My main model is a difference-in-difference model to study the treatment effect on banks’ lending portfolio shares and I find significant changes in certain portfolio shares based on firm-level data. In addition, I use high performance loan-level data from the Federal Reserve, and observe banks’ risk distribution profile shifts toward lower-risk segment in both mortgages portfolio and credit card portfolio. Finally, event studies are used to study parallel trend and assess time-varying treatment effect overtime, and I find some variations in the treatment effect, possibly due to long horizon of post-treatment period.

In my second paper, I learn that mortgage forbearance provides consumers critical debt relief during the COVID-19 pandemic. We document an anomaly in forbearance take up – wealthier individuals were more likely to enroll in mortgage forbearance, even after controlling for income, leverage, and other borrower and loan characteristics. We test a number of hypotheses and rule out the possibilities that wealthier borrowers were more likely to take forbearance due to precaution or out of greater need. We find evidence that financial sophistication and strategic behavior contribute to the anomaly. Finally, we estimate the credit score benefits borrowers have extracted from forbearance. Our results demonstrate some wealth inequalities in debt relief during the pandemic period.

In my third paper, I focus on modeling bank’s automobile portfolio’s loss-given-default (LGD) variable. Loss given default modeling is a key task in credit risk modeling and crucial for a financial institution’s risk management practice. In this paper, we develop LGD models for retail auto loans using the restricted FR Y-14 data. With the high quality auto loan loss and recovery data from large bank holding companies in U.S., we show that a simple OLS model can produce strong model performance in both stress and benign periods. Through the inclusion of interest rate variables as macro variables as well as loan attributes, we quantify the effects of interest rate environment on LGD. Moreover, bank fixed effect significantly improve our LGD model’s prediction power since it captures institutional differences which may be omitted from the data. In addition, we find that conventional macro variables which perform well in Great Financial Crisis and pre-pandemic periods perform poorly in the COVID-19 Pandemic periods.

ISBN

9798384075998

Download

Share

COinS