Working papers

We examine how banks’ depositor composition shapes lending behavior, using granular supervisory data on deposits, loans, and securities for the largest U.S. banks. Classifying banks by depositor specialization, we find persistent differences in funding that translate to differences in asset allocations. Retail-depositor oriented banks hold longer-maturity loans and conduct more real estate lending, while corporate- and NBFI-oriented banks, whose funding is more volatile, hold shorter loans and liquid securities. Loan-level analyses show that stable funding is associated with lower rates, longer maturities, and larger loans. Growth in deposits is allocated differently depending on the depositor specialization of the bank, something we can explore using exogenous deposit growth during COVID.

This paper develops probability pricing, extending cash flow pricing to quantify the willingness-to-pay for changes in probabilities. We show that the value of any marginal change in probabilities can be expressed as a standard asset-pricing formula with hypothetical cash flows derived from changes in the survival function. This equivalence between probability and cash flow valuation allows us to construct hedging strategies and systematically decompose individual and aggregate willingness-to-pay. Four applications examine the valuation of changes in the distribution of aggregate consumption, the efficiency effects of varying performance precision in principal-agent problems, and the welfare implications of public and private information.

Recent technological change in lending converts previously subjective assessments into structured, easily accessible data. We study this transformation in a credit market competition model that distinguishes between information span (breadth) and signal precision (quality). Borrower quality depends on multidimensional fundamentals, assessed through hard or soft signals. Two banks observe private hard signals, but only the specialized bank receives a soft signal. Expanding the span of hard information allows the non-specialized bank to evaluate characteristics previously only available to the specialist, and reducing its winner’s curse. By contrast, greater precision of hard signals strengthens the specialized bank’s informational advantage.

This paper shows how to identify and estimate price informativeness, a necessary step to test theories of information aggregation. Starting from a pricing equation and a stochastic process for payoffs, we show how to recover relative price informativeness from regressions of asset price changes on changes in payoffs. Applying our identification results, we estimate a panel of stock-specific measures of informativeness for U.S. stocks. In the cross-section, large stocks with high turnover, idiosyncratic volatility, institutional-ownership, and analyst-coverage have higher informativeness. In the time-series, the median, mean, and standard deviation of the distribution of informativeness have steadily increased since the mid-1980s.

Using detailed supervisory data on the loan portfolios of large US banks, we document that these banks “specialize” by concentrating their lending disproportionately into a few industries. This specialization improves a bank's industry-specific knowledge, reflected in a reduced risk of loan defaults, lower aggregate charge-offs, and higher propensity to lend to opaque firms in the preferred industry. Banks attract high-quality borrowers by offering generous loan terms in their specialized industry, especially to borrowers with alternative options. Banks focus on their preferred industry in times of instability and relatively lower tier 1 capital as well as after sudden surges in deposits.

We study how competition between asymmetrically informed banks, one specialized and one non-specialized, affects loan prices. Both banks possess “general” signals regarding the borrower’s quality, which they use to screen loans. The specialized bank also has access to a “specialized” signal on which it bases its loan pricing. This private-information-based pricing makes the specialized bank bid more aggressively, mitigating the informational rent effect that gives it monopolistic power. Our findings explain why loans from specialized lenders feature lower interest rates and better ex-post performance. Supporting empirical evidence emphasizes the role of specialized information in shaping credit market outcomes.

We examine the optimal financing of infrastructure when governments can expropriate rents from private sector firms that manage infrastructure. While private firms need incentives to implement projects well, governments need incentives to limit expropriation. This double moral hazard limits the willingness of outside investors to fund infrastructure projects. Optimal financing contracts involve government guarantees to investors against project failure to incentivize the government to agree not to expropriate which improves private sector incentives and project quality. The contract also reflects several other features prevalent in infrastructure financing in practice such as government co-investment, tax subsidies, and development rights.

We study the optimal design of stress scenarios. A principal manages the unknown risk exposures of agents by asking them to report losses under hypothetical scenarios before taking remedial actions. We apply a Kalman filter to solve the learning problem and we relate the optimal design to the risk environment, the principal’s preferences and available interventions. Applied to banking stress tests, optimal capital requirements cover losses under an adverse scenario while targeted interventions depend on covariances among residual exposures and systematic risks. Our calibration reveals that information is particularly valuable for targeted interventions as opposed to broad capital requirements.

In a banking model with imperfect information, I find that more precise information increases the economy's vulnerability to bank runs. For low transparency levels, depositors cannot distinguish bad from good states based on their private signals and, absent liquidity shocks, have no incentives to withdraw early. As transparency increases, and private signals become more informative, depositors' incentives to withdraw strengthen and run-proof contracts become costlier in risk-sharing terms: the bank must offer less to early withdrawers to prevent runs. When transparency is high enough, the bank would rather forgo return and hold excess liquidity than choose a run-proof deposit contract.

This paper studies the social value of closing price differentials in financial markets. We show that arbitrage gaps exactly correspond to the marginal social value of executing an arbitrage trade. Moreover, arbitrage gaps and price impact measures are sufficient to compute the total social value from closing an arbitrage gap, which may emerge for different reasons, including non-pecuniary benefits of holding particular assets. Theoretically, we show that, for a given arbitrage gap, the total social value of arbitrage is higher in more liquid markets. We compute the welfare gains from closing arbitrage gaps for covered interest parity violations.

This paper studies the relation between volatility and informativeness in financial markets. We identify two channels (noise-reduction and equilibrium-learning) that determine the volatility-informativeness relation. When informativeness is sufficiently high (low), volatility and informativeness positively (negatively) comove in equilibrium. We identify conditions on primitives that guarantee that volatility and informativeness comove positively or negatively. We introduce the comovement score, a statistic that measures the distance of a given asset to the positive/negative comovement regions. Empirically, comovement scores i) have trended downwards over the last decades, ii) are positively related to value and idiosyncratic volatility and negatively to size and institutional ownership.

We study the determinants of asset market fragmentation in a model with strategic investors that disagree about the value of an asset. Investors choices determine the market structure. Fragmented markets are supported in equilibrium when disagreement between investors is low. In this case, investors take the same side of the market and are willing to trade in smaller markets with a higher price impact to face less competition when trading against a dealer. The maximum degree of market fragmentation increases as investors' priors are more correlated. Dealers can benefit from fragmentation, but investors are always better off in centralized markets.

We study the effect of trading costs on information aggregation and acquisition in financial markets. For a given precision of investors' private information, an irrelevance result emerges when investors are ex-ante identical: price informativeness is independent of the level of trading costs. When investors are ex-ante heterogeneous, a change in trading costs can increase or decrease price informativeness, depending on the source of heterogeneity. Our results are valid under quadratic, linear, and fixed costs. Through a reduction in information acquisition, trading costs reduce price informativeness. We discuss how our results inform the policy debate on financial transaction taxes/Tobin taxes.

I develop a dynamic model of optimal funding to understand why financial assets are used as collateral instead of being sold to raise funds. Firms need funds to invest in risky projects with non-observable returns. Since holding these assets allows firms to raise these funds, investing firms value the asset more than non-investing ones. When assets are less than perfectly liquid and investment opportunities are persistent, collateralized debt minimizes asset transfers from investing to non-investing firms and, thus, is optimal. Frictions in asset markets lead to an illiquidity discount and a collateral premium, which increase with the asset's illiquidity.

Money Market Funds (MMFs), which are crucial to short-term funding markets, rely on voluntary support of fund sponsors to maintain stable share values. I develop a general equilibrium model of MMFs to study how sponsor support affects the industry’s fragility and regulation. Adverse asset-quality shocks lead MMFs to liquidate assets. When liquidity in asset markets is limited, asset prices are lower if more funds liquidate. Lower asset prices, in turn, make sponsor support costlier and even more liquidations occur. This feedback leads to complementarities in sponsors’ support decisions. Based on the model’s insights, I derive implications for the regulation of MMFs.

Since Sandmo (1981), many articles have analyzed optimal fiscal policies in economies with tax evasion. All share a feature: they assume that the cost of enforcing the tax law is exogenous. However, governments often invest resources to reduce these enforcement costs. In a very simple model, we incorporate such investments in the analysis of an optimal fiscal policy. We characterize their optimal level and we show numerically how they interact with the other dimensions of the optimal fiscal policy. Finally, we highlight the differences between our results and those obtained in a model without investment in the tax administration.