Working Papers

Global vs. Local Banking: A Double Adverse Selection Problem


This paper provides a new theory of credit allocation in financial systems with both global and local banks, and tests it using cross-country loan-level data. I first point out that the traditional theory in banking and corporate finance of firm-bank sorting based on hard versus soft information does not explain the sorting patterns between firms and global versus local banks. In light of this puzzle, I propose a new perspective: global banks have a comparative advantage in extracting global information, and local banks have a comparative advantage in extracting local information. I formalize this view in a model in which firms have returns dependent on global and local risk factors, and each bank type can observe only one component of the firms’ returns. This double information asymmetry creates a segmented credit market with a double adverse selection problem: in equilibrium, each bank lends to the worst type of firms in terms of the unobserved risk factors. Moreover, I show that the adverse selection problem has important macroeconomic implications. When one of the bank types faces a funding shock, the adverse selection affects credit allocation at both the extensive and intensive margins, generating spillover and amplification effects through adverse interest rates. I formally test the model using empirical strategies that tightly map to the model set-up. I find firm-bank sorting patterns, and effects of US and Euro area monetary policy shocks on credit allocation, that support the model predictions. This evidence reveals a novel adverse selection channel of international monetary policy transmission.

Scarred Consumption

with Ulrike Malmendier


We show that personal lifetime experiences can “scar” consumers: Having lived through times of high unemployment, consumers have persistently pessimistic beliefs about their future financial situation, even though their actual future income is uncorrelated with those past experiences, after controlling for income, demographics, and time effects, and future wealth build-up is positively related. Nevertheless worse lifetime experiences predict significantly reduced consumption spending, after including the same set of controls. The results hold both between and within households, and are robust to a battery of variations in liquid- and illiquid wealth controls, in income controls, and further robustness checks. We use the stochastic life-cycle model of Low, Meghir, and Pistaferri (2010) to show that financial constraints, income scarring, and unemployment scarring fail to generate the estimated negative relationship between experiences and consumption. It is instead consistent with experience-based learning, i. e., the notion that consumers per- sistently overweight past experiences in their belief formation process. We replicate the PSID results in the Nielsen Homescan Panel, and the Consumer Expenditure Survey (CEX). The Nielsen data also reveals that households who have lived through times of high unemployment are particularly likely to use coupons and to purchase sale items or lower-end products. Moreover, as predicted by experience-based learning models, the effects of a macro shock are stronger for younger than for older cohorts. Our results sug- gest a novel micro-foundation of fluctuations in aggregate demand, and explain long-run effects of macroeconomic shocks.

Capital Flows and the Real Economy: A “China Shock” in the US Real Estate Market

with Zhimin Li and Calvin Zhang


This paper documents an unprecedented surge in housing purchases by foreign Chinese in the US over the past decade and analyzes the effects of the purchases on US local economies. Using detailed transaction-level housing purchase data, we find that the share of purchases in the California real estate market by foreign Chinese increased more than tenfold over the period of 2007-2013 relative to earlier years. In particular, the purchases have been concentrated in zip codes that are historically populated by ethnic Chinese, making up more than 10% of the total real estate transactions in these neighborhoods by 2013. We exploit the cross-sectional variation in the concentration of Chinese population stemming from pre-sample period differences in Chinese population settlement across zip codes to instrument for the volume of housing purchases by foreign Chinese. Our results show that housing purchases by foreign Chinese significantly increase local housing prices as well as local employment in the non-tradable sector. In our main specification, housing demand by foreign Chinese explains 13% of the housing price differences between zip codes in the top quartile and those in the bottom quartile based on foreign Chinese real estate purchases. To rationalize the results, we develop a simple model that shows the impact of housing demand from foreigners on local housing prices and labor markets through the housing wealth channel. Our results indicate that foreign investments and capital inflow can have significant effects on domestic local economies, especially during economic downturns.

A New Perspective on the Welfare Implications of Business Cycle Fluctuations: Evidence from Consumption Quality

with Casper Nordal Jørgensen


This paper introduces a new approach for estimating the welfare costs of business cycle fluctuations. We quantify and evaluate a new channel that consumers use to smooth macroeconomic shocks—the quality channel of consumption reallocation. Using detailed micro-level panel data on household expenditures, we show that there exists significant heterogeneity in the degree of reallocation across the quality vs. quantity margins of consumption reallocation across income groups: high-income households tend to adjust their consumption at the quality channel when facing negative shocks, while the low-income households are more likely to adjust at the quantity margin. Our results suggest the poor may be rationed in their margins of consumption reallocations when hit by a negative shock and thereby bear a disproportionately greater share of the cost of business cycle fluctuations. We develop a model in which households have non-homothetic preferences and value both the quality and quantity of purchased products. Using the model, we estimate the structural parameters that are consistent with the patterns of consumption behavior observed in the data and analyze the welfare implications of business cycles fluctuations.

Valuing Bank Complexity

with Linda Goldberg

(Draft available upon request)

Banks have progressively evolved from mainly standalone entities toward financial conglomerates that are complex in organizational structure, business scope, and geographical reach. We investigate the role of complexity in market valuation of US banking organizations, using detailed information on the structure of all publicly listed bank holding companies over the period 1990-2017. We find that more organizationally complex banking organizations have significantly lower risk-adjusted equity returns, conditional on standard risk factors and bank size. The results point to a complexity factor in bank returns. This factor is sensitive to bank regulatory changes, suggesting that the market incorporates potential government guarantees for too-complex-to-fail banks in its pricing.