This article presents an augmented deep factor model that generates latent factors for cross-sectional asset pricing. The conventional security sorting on firm characteristics for constructing long–short factor portfolio weights is nonlinear modeling, while factors are treated as inputs in linear models. We provide a structural deep-learning framework to generalize the complete mechanism for fitting cross-sectional returns by firm characteristics through generating risk factors (hidden layers). Our model has an economic-guided objective function that minimizes aggregated realized pricing errors. Empirical results on high-dimensional characteristics demonstrate robust asset pricing performance and strong investment improvements by identifying important raw characteristic sources.

What are the implications of an aging population for financial stability? To examine this question, we exploit geographic variation in aging across U.S. counties. We establish that banks with higher exposure to aging counties increase loan-to-income ratios. Laxer lending standards lead to higher nonperforming loans during downturns, suggesting higher credit risk. Inspecting the mechanism shows that aging drives risk-taking through two contemporaneous channels: deposit inflows due to seniors’ propensity to save in deposits; and depressed local investment opportunities due to seniors’ lower credit demand. Banks thus look for riskier clients, especially in counties where they operate no branches.

We show that demand effects generated by institutional frictions can influence systematic return predictability patterns in stocks and mutual funds. Identification relies on a reform to the Morningstar rating system, which we show caused a structural break in style-level positive feedback trading by mutual funds. As a result, momentum-related factors in stocks, as well as performance persistence and the “dumb money effect” in mutual funds, experienced a sharp decline. Consistent with the proposed channel, return predictability declined right after the reform, was limited to the U.S. market, and was concentrated in factors and mutual funds most exposed to the mechanism.

We examine the role of peer (e.g., Lipper manager indices) versus pure (e.g., S&P 500) benchmarks in fund manager compensation. We model their impact on manager incentives and then test those predictions using novel data. We find that 71% of managers are compensated based on peer benchmarks. Consistent with the model, peer-benchmarked fund managers exhibit higher effort generating higher gross performance and collect higher fee income. Analyzing advisors’ choice between benchmark types, we show that peer-benchmarking advisors cater to more sophisticated and performance-sensitive investors, and are more likely to sell through direct channels, consistent with investor heterogeneity and market segmentation.

We propose a novel model-free approach to obtain the joint risk-neutral distribution among several assets that is consistent with options on these assets and their weighted index. We implement this approach for the nine industry sectors comprising the S&P 500 index and find that their option-implied dependence is highly asymmetric and time-varying. We then study two conditional correlations: when the market moves down or up. The risk premium is strongly negative for the down correlation but positive for the up correlation. Intuitively, investors dislike the loss of diversification when markets fall, but they actually prefer high correlation when markets rally.

We view innovation investments as real options and explore the implications of risk (volatility) as well as a newly defined outcome independent measure of ambiguity (Knightian uncertainty) for innovation decisions. The empirical analysis uses stock returns to compute an implementable measure of ambiguity. We also control for risk and other determinants of innovation. We find a consistently significant negative effect of ambiguity on R&D, patents, and citations, as predicted. The effect of risk on R&D is positive and significant, but the corresponding effect on patents and citations is negative and significant. Ambiguity matters more for high-tech firms, consistent with intuition.

We develop 50 indices of state-level economic policy uncertainty (SEPU) based on newspaper coverage frequency using 204 million newspaper articles from Mar. 1984 to Dec. 2019. We assess the validity of our measures. Our SEPU indices vary counter-cyclically with respect to state-specific economic conditions, rise before close gubernatorial elections, and exhibit a large cross-sectional variation. We demonstrate that SEPU indices are associated with the cross-sectional variation in state-level GDP, employment, income as well as industry investment decisions. Our findings highlight the importance of economic policy uncertainty at the state level in addition to the nationwide level.

This study examines how antitrust law adoptions affect horizontal merger and acquisition outcomes. Using the staggered introduction of competition laws in 20 countries, we find antitrust regulation decreases acquirers’ 5-day cumulative abnormal returns surrounding horizontal merger announcements. A decrease in deal value, target book assets, and industry peers’ announcement returns are consistent with the market power hypothesis. Exploiting antitrust law adoptions addresses a downward bias to an estimated effect of antitrust enforcement. The potential bias from heterogeneous treatment effects does not nullify our results. Overall, antitrust policies seem to deter post-merger monopolistic gains, potentially improving customer welfare.

We investigate how men and women fare in the managerial labor market in the plausibly exogenous circumstance of their firms being acquired when most target-firm managers (about 90%) are displaced. These career disruptions result in a larger drop in rank and compensation for female managers, despite similar job search attributes. Gender differences are mitigated when hiring firms have more women in upper-echelon positions. Rich managerial experience and external board service also reduce gender-related differences. Overall, results point to a (implicit) “gender penalty” in terms of managerial job mobility, but also indicate contexts in which penalty is alleviated, and even reversed.

We use New York City (NYC) taxi data to identify trips between mutual fund offices and local firm headquarters. NYC funds overweight the stocks of local firms they visit via taxi, and firm visits are associated with superior investment performance. Firm visits are elevated prior to earnings announcements, and mutual fund trades that are associated with firm taxi visits predict earnings surprises. The results are generally stronger when fund and firm executives share educational connections. Additional tests support the conclusion that funds’ local bias and investment performance are driven by portfolio managers’ efforts and ability to actively gather material information.

Using a large sample of leveraged loans, we provide evidence that, despite having fewer creditor control rights, covenant-lite (Cov-Lite) loans have similar recovery rates and significantly lower spreads than loans with maintenance covenants. We find that the propensity to borrow Cov-Lite is related to various proxies for the reputational capital of a borrowing firm’s private equity sponsor. We construct a simple model to illustrate the relationship between reputational capital, covenants, and loan spreads in the leveraged loan market. Our model illustrates how reputational capital can substitute for covenants in mitigating agency costs of debt, leading to lower loan spreads for Cov-Lite loans.

The cross section of options holds great promise for identifying return distributions and risk premia, but estimating dynamic option valuation models with latent state variables is challenging when using large option panels. We propose a particle Markov Chain Monte Carlo framework with a novel filtering approach and illustrate our method by estimating index option pricing models. Estimates of variance risk premiums, variance mean reversion, and higher moments differ from the literature. We show that these differences are due to the composition of the option sample. Restricting the option sample’s maturity dimension has the strongest impact on parameter inference and option fit in these models.

This article uses the enactment of China’s 2007 Property Law (the Law), which reduces the risk of expropriation by local governments, as the setting to investigate the importance of property rights protection for private firm investment. Using propensity score matching and a difference-in-differences design, we find that firms facing weaker property rights protection prior to the Law significantly increase their investment and investment efficiency after the Law. Cross-sectional analyses document evidence consistent with a decrease in firms’ perceived expropriation risk as the main mechanism underlying the Law’s effect. Finally, we show that the Law improves local economic outcomes and employment.

We show that bank competition diminishes banks’ incentives to produce information about prospective borrowers. We exploit the deregulation of U.S. interstate branching as a shock to competition and use borrowers’ stock returns after loan announcements to measure bank information production. Positive loan announcement returns are reduced in states that deregulate interstate branching, especially for opaque and bank-dependent firms and smaller banks that rely on soft information. Existing (i.e., inside) banks reduce information production more than new (i.e., outside) banks after deregulation, suggesting that they do so to deter borrower poaching. Furthermore, the probability of a covenant violation increases following deregulation.

We examine the structural stability of Gaussian shadow rate term structure models in a sample of Treasury yields that includes the “effective lower bound” (ELB) period from 2008 to 2015. After highlighting the challenges of testing for structural breaks in a latent-factor model, we proceed to document various pieces of empirical evidence for a structural break. As one of several practical implications, the expected policy rate paths during ELB years are notably shallower in our model that accommodates a structural break compared with a model that imposes structurally stability.