Knowledge that Transforms

There is causal evidence that mortgage credit expansions increase house prices. Does an expansion of margin lending increase stock prices? Because unconstrained arbitrageurs are more important for pricing stocks than homes, the impact is not obvious. Tests are limited because sizable shocks to margin lending are rare. We examine a major Chinese margin-lending expansion between 2010 and 2015. Institutional holding, regression discontinuity, and event study evidence—exploiting the rollout of margin lending across stocks—shows that arbitrageurs anticipated and bought in advance of a significant causal effect of credit. We develop a model to rationalize our findings. Our estimates suggest that margin debt contributes to stock market fluctuations.

We present evidence that loan allocations and loan terms are closely linked to the strength of social connections between bank and borrower regions. Lending increases with social connectedness, particularly in the presence of strong screening incentives. If connectedness is high, banks discriminate between borrowers more, the terms of originated loans are more borrower friendly, and loan performance is better. Furthermore, social connectedness is associated with higher bank profitability and with lending-related increases in borrower-region GDP growth and employment. Our results suggest that lending barriers decrease with social connectedness, which helps our understanding of geographic lending patterns and regional economic disparities.

We develop a tractable dynamic contracting framework to study bank bail-in regimes. In the presence of a repeated monitoring problem, the optimal bank capital structure combines standard debt, which induces liquidation and provides strong incentives, and bail-in debt, which restores solvency but provides weaker incentives. Given fire sales, an optimal policy response entails joint regulation: a bail-in regime reduces standard debt while leverage regulation reduces total debt. Bail-ins replace bailouts as a recapitalization tool.

When a macroeconomic shock arrives, variation in household balance sheet health (captured by the presence of financial distress, or “FD”) leads to differential access to credit and hence a distribution in consumption responses. As we document, though, over the past two recessions, households in prior FD also experienced macroeconomic shocks more intensely than others, leading to a distribution of shock severity. Quantifying the importance of each dimension of heterogeneity (FD or shock severity) for consumption requires a structural model. We find that heterogeneity in FD matters more for shaping the responses of individual and aggregate consumption to shocks.

The second special issue on big data in finance showcases advancements in research related to data of large size, high dimension, and complex structure since the first NBER/RFS big data conference. The papers published in this next chapter address some questions that were proposed in the initial special issue in 2021. Other papers are more directly connected to recent developments in the markets. We discuss some new research directions, following on the papers published here. They include evaluating market microstructure reforms, understanding medium-frequency trading, improving missing data imputations, and deepening data valuation. We look forward to more developments to follow.

I show that investor composition affects bond price dynamics and capital allocation during crises. Using large-scale holdings data and within-firm ownership variation across near-identical bonds, I causally identify bond returns’ investor composition elasticities. Corporate bonds held predominantly by insurers rather than mutual funds suffer milder losses in downturns: increasing insurer holdings by half a bond’s size causes 20% shallower drawdowns. A shift-share instrument isolates variation from large insurers’ idiosyncratic primary-market allocations. Differences in intermediaries’ liability structures drive these results, which hold across countries. During crises, firms with more stable bondholders maintain higher borrowing at lower cost and invest more.

We model firms’ allocation of resources across surplus-creating (ie, productive) and surplus-appropriating (ie, rent-seeking) activities. Our model predicts that industry-wide technological advancements, such as recent progress in data collection and processing, induce a disproportionate and socially inefficient reallocation of resources toward surplus-appropriating activities. As technology improves, firms rely more on appropriation to obtain their profits, thereby endogenously reducing the impact of technological progress on economic progress and inflating the price of the resources used for both types of activities. Our theoretical insights shed light on the rise of high-frequency trading.

Using a semisupervised topic model on 7 million New York Times articles spanning 160 years, we test whether topics of media discourse predict future stock market excess returns to test rational and behavioral hypotheses about market valuation of disaster risk. Media discourse data address the challenge of sample size even when disasters are rare. Our methodology avoids look-ahead bias and addresses semantic shifts. Our discourse topics positively predicts market excess returns, with War having an out-of-sample R^2 of 1.35%. We call this effect the war return premium. The war return premium has increased in more recent time periods.

We show a negative effect of opioid prescriptions on subsequent individual employment among employers in our sample using doctor-opioid-prescribing propensity as our instrument. This finding has implications for firms that must now contend with lower local labor supply. We find a negative relationship between opioid prescriptions and subsequent establishment growth. However, firms respond to labor shortages by investing more in technology, replacing the relatively scarcer labor with capital, especially when they are not financially constrained. We find positive abnormal returns, upon the passage of state laws intended to limit opioid prescriptions, that are driven by firms more reliant on labor.

We document political sentiment effects on U.S. inventors. Democratic inventors are more likely to patent (relative to Republicans) after the 2008 election of Obama but less likely after the 2016 election of Trump. These effects are at least twice as strong among politically active Democrats and are present even within firms and within firm$ × $technology. We also show that partisans tend to cluster in technologies (e.g., Democrats in Biotechnology and Republicans in Weapons), so that sentiment effects aggregate to more patents in the technologies dominated by the winning party.