Research

Published and Forthcoming Papers:

• "Deposit Competition and Financial Fragility: Evidence from the US Banking Sector" [with Ali Hortacsu and Gregor Matvos, American Economic Review, 2017]

• "The Market for Financial Adviser Misconduct" [with Gregor Matvos and Amit Seru, Journal of Political Economy, 2019]

• "Brokers vs. Retail Investors: Conflicting Interests and Dominated Products"  [The Journal of Finance, 2019]

• "When Harry Fired Sally: The Double Standard in Punishing Misconduct"  [with Gregor Matvos and Amit Seru; Journal of Political Economy, 2022]

• "Recovering Investor Expectations from Demand for Index Funds" [with Alex MacKay and Hanbin Yang; Review of Economic Studies, 2022]

• "The Cross Section of Bank Value" [with Stefan Lewellen and Adi Sunderam; Review of Financial Studies, 2022; Winner of the Douglas D. Evanoff Best Paper Award at the Chicago Financial Institutions Conference]

• "The Value of Intermediation in the Stock Market"  [with Marco Di Maggio and Francesco Franzoni, Journal of Financial Economics; 2021]

• "Conflicting Interests and the Effect of Fiduciary Duty — Evidence from Variable Annuities" [with Shan Ge and Johnny Tang; Review of Financial Studies, 2022]

• "Arbitration with Uninformed Consumers"  [with Gregor Matvos and Amit Seru; Review of Economic Studies, 2025]

• "What Drives Variation in Investor Portfolios? Evidence from Retirement Plans" [with Alex MacKay and Hanbin Yang; Review of Financial Studies, 2026]

Working Papers:

• "Competition in Generative AI: Evidence from Consumer Demand" [with Dominic Russel, Claire Shi, and Vincent Yao, NBER WP 35953]

Abstract: How competitive will the market for generative AI be as it matures? Using novel transaction-level credit and debit card data covering 174 million consumers and 114 billion transactions, we study the adoption of paid services and the demand-side foundations of competition in this market. As of April 2026, approximately 2.9% of U.S. consumers pay for at least one major generative AI service, with adoption highest among younger and higher-income consumers. Nearly all of these paid users subscribe to basic plans that cost $30 or less per month. Three patterns suggest limited differentiation and low switching costs. First, although multihoming is common, it has declined as providers' capabilities have expanded. Second, annual customer retention rates range from 20% to 45% across major AI providers, lower than those for many other subscription products. Third, switching rises around major model releases. Motivated by these findings, we develop and estimate a structural model of AI service demand and supply that reveals strong competitive pressure. Consumers are highly price sensitive, have much weaker preferences for particular providers than for AI overall, and view AI services as increasingly close substitutes. Our results suggest that, absent a substantial increase in AI model quality differentiation, growing substitutability among services will constrain providers' pricing power and allow consumers to capture a large share of the surplus generated by these services.

• "Who Pays for Payments" [with Gregor Matvos, Amit Seru, Lulu Wang, and Vincent Yao 35067]

Abstract: We use novel data on the composition and cost of payments across U.S. merchants to quantify consumer redistribution in the payment system. Cards charge interchange fees to merchants to fund consumer rewards. When merchants raise prices for all consumers in response to these costs, users of low-cost payment methods (e.g., cash and debit) cross-subsidize high-reward credit card users who shop at the same merchant. This standard mechanism implicitly assumes that consumers using different payment methods shop at the same merchants and that merchants face similar fees. We show instead that incidence depends on the joint distribution of payment choices across merchants. We document two key forces that shape redistribution. First, consumer sorting—where consumers who use different payment methods shop at different merchants—limits the exposure of cash and debit users to the effects of high interchange fees. Second, interchange fees vary across merchants; where users of different payment methods overlap, such as at large grocery stores, fees are lower due to sector discounts and private negotiations. We embed these forces in a sufficient-statistics framework that maps observed heterogeneity directly into redistribution. We estimate that interchange fees transfer approximately $30 billion every year from cash and debit users to credit card users. Consumer sorting and merchant fee heterogeneity reduce the magnitude of this regressive transfer by 25%, but do not eliminate it. Finally, we show that both the Durbin Amendment and the rise of premium credit cards have been regressive, highlighting how policy and innovation can reshape the incidence of platform fees.

• "Dynamic Competition for Sleepy Deposits" [with Ali Hortacsu, Nathan Kaplan, Adi Sunderam, and Vincent Yao; Revise and Resubmit at the American Economic Review]

Abstract: We examine the implications of sleepy deposits and their impact on competition, bank value, and financial stability in the US banking sector. We first document the shopping behavior of depositors using novel data on account openings and closures. Depositors infrequently shop for deposits, with 5-15% of depositors opening a new account each year. Shopping behavior is idiosyncratic: deposit accounts are more likely to be closed due to the depositor either moving or dying than because the depositor switched to a new account offering higher rates or better services. Building on these facts, we develop an empirical model of the supply and demand for "sleepy deposits." In the model, banks face dynamic "invest-versus-harvest" incentives in competing for depositors who shop infrequently. We estimate the model and find that depositor sleepiness accounts for 58% of the average bank's deposit franchise value. Sleepiness softens competition, particularly raising markups and franchise value for banks in low-concentration areas, as well as for banks with either low-quality deposit services or high marginal costs. Sleepiness also creates stability in the banking sector. For two main money center banks in the US, the probability of default after the Federal Reserve's 2022-2023 hiking cycle would have increased to more than 20% in a counterfactual without sleepy depositors.

• "Why Do Index Funds Have Market Power? Quantifying Frictions in the Index Fund Market" [with Zach Brown, Jihye Jeon, Chuqing Jin, and Alex Wu; NBER WP 31778]

Abstract: Index funds are one of the most common ways investors access financial markets and are perceived to be a transparent and low-cost alternative to active investment management. Despite these purported virtues of index fund investing and the introduction of new products and competitors, many funds remain expensive and fund managers appear to exercise substantial market power. Why do index funds have market power? We develop a novel quantitative dynamic model of demand for and supply of index funds. In the model, investors are subject to inertia, search frictions, and have heterogeneous preferences. These frictions on the demand side create market power for index fund managers, which fund managers can further exploit by price discriminating and charging higher expense ratios to retail investors. Our results suggest that the average expense ratios paid by retail investors are roughly 45% higher as a result of search frictions and are 40% higher as a result of inertia compared to the friction-less baseline. In our counterfactuals, we find an interaction between search frictions and inertia---inertia imposes higher (lower) costs on investors when search frictions are low (high).

• "How Do Investors Value ESG?" [with Malcolm Baker and Suproteem Sarkar; NBER WP 30708 ]

Abstract: Environmental, social, and governance (ESG) objectives have risen to near the top of the agenda for corporate executives and boards, driven in large part by their perceptions of shareholder interest. We quantify the value that shareholders place on ESG using a revealed preference approach, where shareholders pay higher fees for ESG-oriented index funds in exchange for their financial and non-financial benefits. We find that investors are willing, on average, to pay 20 basis points more per annum for an investment in a fund with an ESG mandate as compared to an otherwise identical mutual fund without an ESG mandate, suggesting that investors as a group expect commensurately higher pre-fee, gross returns, either financial or non-financial, from an ESG mandate. Our point estimate has risen from 9 basis points in 2019 when our sample begins to as much as 28 basis points in 2022. When we incorporate the possibility that investors are willing to accept lower financial returns in exchange for the psychic and societal benefits of ESG, when we consider that the holdings of ESG and non-ESG index funds overlap, when we measure the ESG ratings of these holdings, and when we focus on 401(k) participants who report being concerned about climate change or who work in industries with lower levels of emissions, we find that the implicit value that investors place on ESG stocks is higher still. A simple model of supply suggests that the large majority of these benefits accrue to investors and firms, with intermediaries capturing 5.9 basis points in fees, half of which reflect higher markups.

• "The Effects of Regulating Greenwashing: Evidence from Europe’s Sustainable Finance Disclosure Regulation (SFDR)" [with Hunt Allcott, Paul Smeets, and Hanbin Hang; NBER WP 34624]

Abstract: We examine the impact of the European Union’s Sustainable Finance Disclosure Regulation (SFDR) on mutual fund flows and investment sustainability. The SFDR classifies funds into three categories to promote transparency and curb greenwashing: those with a sustainable investment objective (Article 9 or “dark green”), those that promote environmental characteristics (Article 8 or “light green”), and others (Article 6). Using a difference-in-differences design, we find that the SFDR had little effect on fund flows or portfolio sustainability. The disclosures were ineffective in part because they offered little new or clear information beyond what investors could already infer from fund names and mandates. In an experimental setting, we show that the current disclosures have minimal impact on investor decisions, but making the information more intuitive could improve the regulation’s effectiveness.

Other Work:

• "Adjusting Economic Measures for Health: Is the Business Cycle Countercyclical?" [with Casey Mulligan and Tomas Philipson, NBER WP 19058]

Abstract: Many national accounts of economic output and prosperity, such as gross domestic product (GDP) or net domestic product (NDP), offer an incomplete picture by ignoring, for example, the value of leisure, home production, and the value of health. Previous discussed shortcomings of such accounts have focused on how unobserved dimensions affect GDP levels but not their cyclicality, which affects the measurement of the business cycle. This paper proposes a new methodology to measure economic fluctuations that incorporates monetized changes in health of the population in the United States and globally during the past 50 years. In particular, we incorporate in GDP the dollar value of mortality, treating it as depreciation in human capital analogous to how net domestic product (NDP) treats depreciation of physical capital. Because mortality tends to be pro?cyclical, we find that adjusting for mortality reduces the measured deviations of GDP from trend during the past 50 years by about 30% both in the United States and internationally. 

• "Non-Adherence in Health Care: A Positive and Normative Analysis" [with Tomas Philipson, NBER WP 20330]

Abstract: Non-adherence in health care results when a patient does not initiate or continue care that has been recommended by a provider. Previous researchers have identified non-adherence as a major source of waste in US healthcare, totaling approximately 2.3% of GDP, and have proposed a plethora of interventions to improve adherence. However, little explicit analysis exists in health economics of the dynamic demand behavior that drives non-adherence. We argue that while providers may be more informed about the population-wide effects of treatments, patients are more informed about their individual treatment effect. We interpret a patient's adherence decision as an optimal stopping problem where patients learn the value of a treatment through experience. Our positive analysis derives an "adherence survival function" and shows how various observable factors affect adherence. Our normative analysis derives the efficiency effects of non-adherence, the conditions under which adherence is too high or too low, and why many common interventions aimed at raising adherence produce indeterminate welfare effects. We calibrate these welfare effects for one of the largest US drug categories, cholesterol reducing drugs. Contrary to frequent normative claims of under-adherence, our estimates suggest that the ex-post efficiency loss from over-adherence is over 80% larger than from under-adherence. 

• "International Health Economics" [with Tomas Philipson, NBER WP 19280]

Abstract: Economists have paid little attention to interactions between domestic health care economies. However, the growth in domestic health care sectors is often attributed to medical innovations whose returns are earned worldwide. Due to the public-goods nature of providing world returns, domestic health care reimbursement policies will underprovide returns to innovation and a given country's profit provision will depend negatively on those of others. We estimate the degree to which national reimbursements are strategic substitutes in this manner and the implied indeterminate effects on world returns and innovation from growth in emerging markets.