Board of Governors of the Federal Reserve System

We study banks' use of the Fed's discount window during a time of relative calm in financial markets. We merge transaction‐level data on discount window borrowings with quarterly bank balance sheet information between 2010 and 2019. We develop a model of the borrowing decisions of banks and contrast its implications with the data. In line with the model, we find that discount window borrowing is tightly linked to the composition of banks' balance sheets. Importantly, banks holding less reserves are more likely to borrow from the discount window, as are banks with more expensive and fragile liabilities, and less marketable collateral.

Risk was incorporated into monetary aggregation over thirty-five years ago, using a stochastic version of the workhorse money-in-the-utility-function model. Nevertheless, the mathematical foundations of this stochastic model remain shaky. To firm the foundations, this paper employs richer probability concepts than Borel-measurability, enabling me to prove the existence of a well-behaved solution and to derive stochastic Euler equations. This measurability approach is less common in economics, possibly because the derivation of stochastic Euler equations is new. Importantly, the problem’s economics are not restricted by the approach. The results provide firm footing for the growing monetary aggregation under risk literature, which integrates monetary and finance theory. As crypto-currencies and stable coins garner attention, solidifying the foundations of risky money becomes more critical. The method also supports deriving stochastic Euler equations for any dynamic economics problem that features contemporaneous uncertainty about prices, including asset pricing models like capital asset pricing models and stochastic consumer choice models.

Using firm‐level transaction data for the United States, this paper presents a series of stylised facts on exporters in services industries, contrasting these findings with those of exporters in manufacturing and for domestic transactions. My results show that most of the basic facts on manufacturing exporters extend to the services sectors with three important differences. First, the export participation rate and export value share of firms in services are much lower than those of manufacturing firms. Second, the assortative patterns across services exporters and their buyers appear stronger than those for manufacturing firms. Third, the survival rate of services exporters tends to be lower than that of manufacturing exporters. Furthermore, I document that domestic relationships between buyers and services firms experience much higher churning than those in the foreign market. These facts are compatible with the hypothesis that firms in services sectors face larger trade costs.

Can automation complement economic incentives? We explore this question by randomly encouraging households to activate a feature on their existing smart thermostat that automates responsiveness to time-of-use electricity pricing. The feature reduces air conditioning use during the highest-priced afternoon period, raising indoor temperatures above a household's preferred temperature, primarily for customers who are typically home during the day. Customers infrequently override the feature when they experience discomfort, suggesting that they are willing to trade off monetary savings for small increases in discomfort. Automation thus enables low-cost changes in household energy use, with potentially large electricity supply-cost reductions at scale. (JEL D12, D91, L94, L98, Q48)

In this paper we identify demand shocks that can have a permanent effect on output through hysteresis effects. We call these shocks permanent demand shocks. They are found to be quantitatively important in the United States, in particular in samples starting in the 1980s. Recessions driven by permanent demand shocks lead to a permanent decline in employment and investment, while output per worker is largely unaffected. We find strong evidence that hysteresis transmits through a rise in long-term unemployment and a decline in labor force participation and disproportionately affects the least productive workers. (JEL C51, E22, E23, E24, E32, J22, J24)

Replacing faulty measurements with missing values can suppress outlier-induced distortions in state-space inference. We therefore put forward two complementary methods for enhanced outlier-robust filtering and forecasting: supervised missing data substitution (MD) upon exceeding a Huber threshold, and unsupervised missing data substitution via exogenous randomization (RMDX).Our supervised method, MD, is designed to improve performance of existing Huber-based linear filters known to lose optimality when outliers of the same sign are clustered in time rather than arriving independently. The unsupervised method, RMDX, further aims to suppress smaller outliers whose size may fall below the Huber detection threshold. To this end, RMDX averages filtered or forecasted targets based on measurement series with randomly induced subsets of missing data at an exogenously set randomization rate. This gives rise to regularization and bias-variance trade-off as a function of the missing data randomization rate, which can be set optimally using standard cross-validation techniques.We validate through Monte Carlo simulations that both methods for missing data substitution can significantly improve robust filtering, especially when combined together. As further empirical validation, we document consistently attractive performance in linear models for forecasting inflation trends prone to clustering of measurement outliers.

This paper uses data on bank connections with service providers to construct a representation of an operational network used to facilitate the sending of Fedwire transactions. Our data contains 227 connections between 215 banks (mostly community banks, but also some large banks) and four unique payment products used by the firms to send and receive Fedwire transactions. By constructing such an operational network between banks and payment providers, we can perform multiple analyses that are useful in operational resilience considerations. First, we use the mean daily Fedwire volume for each bank to create a dollar estimate of the "operational risk exposure" associated with each service platform based on its bank clients. Second, we examine how these bank payment risk exposure estimates compare with other, publicly available benchmarks, since payment data are usually confidential. Last, we use the network model to conduct analysis on network concentration, which provides an example of how such networks could be used in analyzing the likely impact of operational outages. Our results indicate that data on service provider connections such as that we analyze can provide important insights into the extent to which payment network resilience mitigates risk to the financial sector. Our results also indicate that several publicly available benchmarks can serve as substitutes (with certain caveats) for payments data in estimating payment risk exposure.

We develop a novel indicator of aggregate and sectoral wage inflation by leveraging multiple sources of wage data with detailed sectoral information, using a hierarchical dynamic factor model. Our empirical approach controls for data-specific measurement errors and industry-specific developments; this feature makes it particularly effective in assessing wage inflation during the Covid era, which saw increased dispersion in wage inflation across industries and larger divergences between measures of the level and trajectory of wage inflation. Our findings indicate that as the labor market has cooled, aggregate underlying wage inflation has returned to pre-pandemic levels, with wage growth in some sectors even falling below its pre-pandemic rate.

Since the early 2000s, there has been a growing divergence between the index of industrial production (IP) and the goods component of GDP—hereafter, goods GDP—breaking the close correlation the two series had maintained in earlier decades (figure 1). This note revisits the factors behind this divergence and provides a novel quantification of their role.

Disability can be devastating financially for households, as it can severely limit earnings potential (Meyer and Mok, 2019; Benito, Glassman,`and Hiedemann, 2016; Jolly, 2013). The earnings penalty for households with disabilities is estimated to range from 15 to 70 percent of earnings, depending on the nature of the disability (Meyer and Mok, 2019). Households receiving disability insurance (either employer-based or via Social Security) receive payments based on their prior earnings to help account for this earnings penalty.

The decomposition of GDP into potential output—the level of output consistent with current technologies and "normal" use of capital and labor—and the output gap—the percentage deviation of GDP from its potential—is a fundamental task for policymakers. Potential output tells us how fast an economy can grow in the long run; the output gap helps assess the cyclical position of the economy and, thus, potential inflationary pressures (Jarociński and Lenza, 2018).

In this paper, we assemble the most comprehensive dataset to date on the characteristics of colleges and universities, including dates of operation, institutional setting, student body, staff, and finance data from 2002 to 2023. We provide an extensive description of what is known and unknown about closed colleges compared with institutions that did not close. Using this data, we first develop a series of predictive models of financial distress, utilizing factors like operational revenue/expense patterns, sources of revenue, metrics of liquidity and leverage, enrollment/staff patterns, and prior signs of significant financial strain. We benchmark these models against existing federal government screening mechanisms such as financial responsibility scores and heightened cash monitoring. We document a high degree of missing data among colleges that eventually close and show that this is a key impediment to identifying at risk institutions. We then show that modern machine learning techniques, combined with richer data, are far more effective at predicting college closures than linear probability models, and considerably more effective than existing accountability metrics. Our preferred model, which combines an off-the-shelf machine learning algorithm with the richest set of explanatory variables, can significantly improve predictive accuracy even for institutions with complete data, but is particularly helpful for predicting instances of financial distress for institutions with spotty data. Finally, we conduct simulations using our estimates to contemplate likely increases in future closures, showing that enrollment challenges resulting from an impending demographic cliff are likely to significantly increase annual college closures for reasonable scenarios.

Using inflation swap prices, we study how changes in expected inflation affect firm-level credit spreads and equity returns, and uncover evidence of a time-varying inflation sensitivity. In times of “good inflation,” when inflation news is perceived by investors to be more positively correlated with real economic growth, movements in expected inflation substantially reduce corporate credit spreads and raise equity valuations. Meanwhile in times of “bad inflation,” these effects are attenuated and the opposite can take place. These dynamics naturally arise in an equilibrium asset pricing model with a time-varying inflation-growth relationship and persistent macroeconomic expectations.

This paper studies the estimation and inference of time-varying impulse response functions in structural vector autoregressions (SVARs) identified with external instruments. Building on kernel estimators that allow for nonparametric time variation, we derive the asymptotic distributions of the relevant quantities. Our estimators are simple and computationally trivial and allow for potentially weak instruments. Simulations suggest satisfactory empirical coverage even in relatively small samples as long as the underlying parameter instabilities are sufficiently smooth. We illustrate the methods by studying the time-varying effects of global oil supply news shocks on US industrial production.

Owners of real assets often have informational advantages over other investors about asset- specific cash flows and local market conditions. When local property values suddenly decline, these investors may have to sell their assets quickly and have less access to credit due to the decline in the value of their collateral. This opens up the market to outside investors who lack the same informational advantages but have better access to capital. We find that, in times of stress, the volume of transactions in hard-hit markets falls, and entry from out-of-market buyers increases. Out-of-market buyers consistently pay less for properties, by about 20 basis points in terms of a cap rate, during both boom and bust markets, which is consistent with them reducing their bids when they are at an informational disadvantage. Out-of-market investors also tend to generate higher holding period returns for the previous owners, suggesting that they may still be overbidding for properties. Out-of-market buyers are less likely to purchase properties where they have a significant information disadvantage, such as troubled properties during a market decline.

This article documents that a single‐factor model based on shocks to the residential investment share, or the ratio of residential‐to‐nonresidential investment, exhibits strong explanatory power for expected returns across various characteristic‐sorted portfolios in equity and other asset classes. The residential investment share captures time‐varying demand for housing services and is a state variable of the economy. Consequently, innovations to the share emerge as a risk factor in asset prices in the cross section. The empirical results are robust to controlling for other factor models based on durable consumption, financial intermediaries, household heterogeneity, and return‐based multifactor models designed to price these assets.

The ability to conduct reproducible research in Stata is often limited by the lack of version control for community-contributed packages. In this article, we introduce the require command, a tool designed to ensure package dependencies are compatible across users and computer systems. Given a list of packages, require verifies that each package is installed, checks for a minimum or exact version or package release date, and optionally installs the package if prompted by the researcher.

Empirical research involves multiple, seemingly‐minor choices that can substantially impact a study's findings. While acknowledged, the importance of these “degrees of flexibility” on published estimates is not well understood. We examine the considerable literature focused on the impacts of early COVID‐19 policies on social distancing to assess the role of researchers' degrees of flexibility on the estimated effects of mobility‐reducing policies. We find that estimates reported in previous studies are not robust to minor changes in typically‐unexplored dimensions of the degree of flexibility space, and usual robustness tests systematically fail to detect these issues.

Plain English Summary Japanese government loan program boosts employment and long-term investment and helps firms optimize production inputs. Due to their inherent riskiness, small businesses frequently have difficulty borrowing funds from banks and other lenders. Therefore, many governments have created programs to facilitate lending to these businesses. We study the effects of the Small Business Managerial Improvement Loan (MIL) program in Japan. We find that small businesses that receive an MIL loan have higher employment growth and invest more than other, similar firms that do not receive a loan. Recipients appear to move toward a more optimal production process and sales growth is consistently higher over the long term. Differences in debt levels are persistent, which suggests that the MIL program successfully makes loans to businesses to whom private lenders do not lend.

This paper investigates the impact of vaccine lotteries in the United States on spending, mobility, and employment, three dimensions of activity representing novel outcomes relative to the existing literature. This analysis combines hand-collected data on the introduction of vaccine lotteries across states over the second and third quarters of 2021 with daily state-level data from Fiserv (spending), Apple and Google (mobility), and Homebase (employment). In a dynamic event design setting, lotteries significantly boosted retail spending but did not meaningfully affect other measures of activity. As the introduction of the lotteries was largely unexpected and there were no significant pre-trends, we argue that these effects can be interpreted causally, differently from the contributions strictly relating vaccinations and economic outcomes. Finally, these effects were translated into implications for economic growth. Under the assumption that there were no meaningful substitution patterns, the findings imply that the introduction of lotteries added 0.4 percentage points to gross domestic product growth in 2021.