Dongkeun Choi
Thank you for visiting my website! I am a Ph.D. Candidate in Economics at UC San Diego.
My research focuses on macroeconomics, monetary economics, and financial intermediation.
I visited the Federal Reserve Board as a Dissertation Fellow this summer.
I am on the job market in 2026–2027.
Curriculum Vitae: CV
Email: dochoi@ucsd.edu
Job Market Paper
Unintended Risks of Bank Capital Regulation
[Abstract] I show that tighter bank capital regulation can increase aggregate credit risk by shifting credit supply toward nonbank financial intermediaries (NBFIs) that operate outside the scope of regulation. Using U.S. mortgage data, I document that NBFIs approve observably similar applicants at higher rates than banks and that NBFI loans have higher default rates. I show that tighter bank capital regulation contributed to a shift in credit toward riskier borrowers with higher default rates. I develop a general equilibrium model in which banks and NBFIs choose loan rates and underwriting standards. In the model, tighter regulation leads banks to raise loan rates and tighten underwriting standards, shifting applicants toward NBFIs. NBFIs' looser underwriting standards extend credit to applicants who would otherwise remain unfunded, increasing aggregate credit risk. The model further shows that tighter regulation can amplify credit losses from adverse financial shocks.
Working Papers
Why Is Monetary Policy Less Effective at Low Interest Rates: The Role of the Banking Sector
[Abstract] This paper investigates the effectiveness of monetary policy in a low-interest-rate environment, focusing on how loan and deposit rates respond to changes in nominal interest rates. Using U.S. bank data, I show that deposit rates reach zero when the policy rate falls to about 2 percent, with both deposit and loan rates becoming unresponsive to further cuts below the threshold. Net interest margins do not shrink, meaning that banks’ equity and lending capacity do not respond to cuts in the policy rate. For these reasons, monetary policy loses its effectiveness well before the policy rate reaches zero. Stabilizing demand-driven recessions near this threshold requires earlier fiscal coordination.
Equipment, Structures, and the Limits of Investment-Specific Technological Change (with Munseob Lee)
[Abstract] Models of investment-specific technological change treat the falling relative price of equipment as a source of growth. The relative price of structures has moved the other way. It has risen in rich countries, and in the United States it is 80 percent higher than in 1970. We document this divergence across countries, show that the U.S. rise covers nearly every type of structure, and attribute about half of the post-1996 rise in construction prices to falling construction productivity. To quantify its cost, we build an endogenous growth model in which equipment and structures enter both final production and R&D. In the U.S. calibration, rising structures prices subtract 0.63 percentage points a year from per-capita growth, half of the 1.24 points that cheaper equipment adds, with 30 percent of the drag operating through R&D. A nested CES extension finds that structures substitute for unskilled labor and that the two capital-labor bundles substitute for each other. These margins lower the drag to about 0.4 points. Removing the substitution between structures and unskilled labor raises the model's 1963--2019 rise in the skill premium by 30 percent.
Selected Work in Progress
Redistribution through the Secondary Mortgage Market
[Abstract] This paper investigates how a relaxation in GSE purchase and underwriting rules can redistribute credit from safer to riskier borrowers through the secondary mortgage market. Exploiting the 2017 removal of additional documentation requirements for high-DTI mortgages, I show that ex post default rates increase among affected borrowers, while interest-rate spreads respond only modestly. This muted price adjustment suggests that lenders do not fully incorporate the DTI-induced deterioration in credit risk into loan pricing. The combination of expanded high-DTI acquisitions and incomplete risk-based pricing implies a reallocation-subsidy mechanism. When secondary-market pricing is imperfectly risk-adjusted, expanded purchase eligibility reallocates GSE credit toward riskier borrowers without commensurate increases in fees, shifting a larger share of expected losses onto safer segments within the pooled GSE system. I quantify the resulting increase in expected credit losses borne by the GSEs, decompose it into composition and performance components, and highlight the importance of a secondary-market pricing framework that more fully reflects borrower risk.