Global liquidity flows are largely channeled through banks and nonbank financial institutions. The common drivers of global liquidity flows include monetary policy in advanced economies and risk conditions. At the same time, the sensitivities of liquidity flows to changes in these drivers differ across institutions and have been evolving over time. Microprudential regulation of banks plays a role, influencing leverage and capitalization, changing sensitivities to shocks, and also driving risk migration from banks to nonbank financial institutions. Risk sensitivities and flightiness of global liquidity are now strongest in more leveraged nonbank financial institutions, raising challenges in stress episodes. Current policy initiatives target linkages across different types of financial institutions and associated risks. Meanwhile, significant gaps remain. This paper concludes by discussing policy options for addressing systemic risk in banks and nonbanks.
This paper studies large sample classical and Bayesian inference in a prototypical linear DSGE model and demonstrates that inference on the structural parameters based on a Gaussian likelihood is unaffected by departures from Gaussianity of the structural shocks. This surprising result is due to a cancellation in the asymptotic variance resulting into a generalized information equality for the block corresponding to the structural parameters. The underlying reason for the cancellation is the certainty equivalence property of the linear rational expectation model. The main implication of this result is that classical and Bayesian Gaussian inference achieve a semi-parametric efficiency bound and there is no need for a “sandwich-form” correction of the asymptotic variance of the structural parameters. Consequently, MLE-based confidence intervals and Bayesian credible sets of the deep parameters based on a Gaussian likelihood have correct asymptotic coverage even when the structural shocks are non-Gaussian. On the other hand, inference on the reduced-form parameters characterizing the volatility of the shocks is invalid whenever the structural shocks have a non-Gaussian density and the paper proposes a simple Metropolis-within-Gibbs algorithm that achieves correct large sample inference for the volatility parameters.
Does the effect of monetary policy depend on the prevailing level of inflation? In order to answer this question, we construct a parsimonious nonlinear time series model that allows for inflation regimes. We find that the effects of monetary policy are markedly different when year-over-year inflation exceeds 5.5 percent. Below this threshold, changes in monetary policy have a short-lived effect on prices, but no effect on the unemployment rate, giving a potential explanation for the recent “soft landing” in the United States. Above this threshold, the effects of monetary policy surprises on both inflation and unemployment can be larger and longer lasting.
Increasing personal bankruptcy protection raises consumers’ desire to borrow and lenders’ cost of extending credit; the impact on equilibrium borrowing is ambiguous. Using bankruptcy protection changes between 1999 and 2005 across U.S. states, we find that borrowers respond to greater protection by increasing their unsecured debt. Border county estimates suggest that local economic conditions do not drive these results. Borrowers pay more for protection through higher interest rates, yet delinquency is unaffected. Remarkably, our results indicate that rising borrower demand outstripped decreasing supply. Increased protections did not reduce the aggregate level of household debt but affected the composition of borrowing.
We examine the effect of undiagnosed memory disorders on credit outcomes using individually-matched nationally representative credit reporting and Medicare data. We find effects of early stage disease, years before diagnosis, on a wide range of financial outcomes, including credit card account payment delinquency and amount of delinquent balance, credit utilization among credit card account holders, mortgage delinquency and delinquent balance amount, and credit scores. Effects are pervasive, affecting seniors in single and coupled households, racial/ethnic minorities and non-minorities, and older adults living in areas with higher and lower education levels. Early stage effects are greater among singles and Black individuals.
We use high-frequency interbank payments data to trace deposit flows in March 2023 and identify twenty-two banks that suffered a run—significantly more than the two that failed but fewer than the number that experienced large negative stock returns. The runs were driven by large (institutional) depositors, rather than many small (retail) depositors. While the runs were related to weak fundamentals, we find evidence for the importance of coordination because run banks were disproportionately publicly traded and many banks with similarly bad fundamentals did not suffer a run. Banks that survived a run did so by borrowing new funds and then raising deposit rates—not by selling liquid securities.
I study the labor market risks associated with being self-employed. I document that the self-employed are subject to larger earnings fluctuations than employees and that they frequently transition into unemployment. Given that the self-employed are not eligible to unemployment insurance, I analyze the provision of benefits targeted at these risks using a calibrated search model with (i) precautionary savings, (ii) work opportunities in paid and self-employment, and (iii) skill heterogeneity. This exercise suggests that extending the current U.S. unemployment insurance scheme to the self-employed comes with a clear increase in the transition rate from self-employment to unemployment and an unequal benefits-to-contributions ratio across skill groups. At the calibrated parameters, the self-employed in the middle of the skill distribution lose welfare.
We use twenty-five years of tax records for the Norwegian population to study the mobility of wealth over people’s lifetimes. We find considerable wealth mobility over the life cycle. To understand the underlying mobility patterns, we group individuals with similar wealth rank histories using agglomerative hierarchical clustering, a tool from statistical learning. The mobility patterns we elicit provide evidence of segmented mobility. Over 60 percent of the population remains at the top or bottom of the wealth distribution throughout their lives. Mobility is driven by the remaining 40 percent, who move only within the middle of the distribution. Movements are tied to differential income trajectories and business activities across groups. We show parental wealth is the key predictor of who is persistently rich or poor, while human capital is the main predictor of those who rise and fall through the middle of the distribution.
Most macroeconomic models impose a tight link between expected future short rates and the term structure of interest rates via the expectations hypothesis (EH). While the EH has been systematically rejected in the data, existing work evaluating the EH generally assumes either full-information rational expectations or stationarity of beliefs, or both. As such, these analyses are ill-equipped to refute the EH when these assumptions fail to hold, fueling hopes for a “resurrection” of the EH. We introduce a model of expectations formation which features time-varying means and accommodates deviations from rationality. This model tightly matches the entire joint term structure of expectations for output growth, inflation, and the short-term interest rate from all surveys of professional forecasters in the U.S. We show that deviations from rationality and drifting long-run beliefs consistent with observed measures of expectations, while sizable, do not come close to bridging the gap between the term structure of expectations and the term structure of interest rates. Not only is the EH decisively rejected in the data, but model-implied short-rate expectations generally display, at best, only a weak co-movement with the forward rates of corresponding maturities.
We examine how investors’ perception of bank balance sheet risk evolved before and during the March-April 2023 bank run. To do so, we estimate the covariance (“beta”) of bank excess stock returns with returns on factors constructed from long-short portfolios sorted on shares of uninsured deposits and unrealized losses on securities. We find that the market’s perception of bank risk shifted in both the time series and the cross-section. From January 2022 to February 2023, both factor betas were mostly insignificant, but after the bank run started, they became positive and significant for all banks on average. However, in the cross-section, only the factor betas of banks put on downgrade watch on March 13 were significant, consistent with our finding that this announcement was informative. When additional banks were downgraded in April, their factor betas also became significant, even though we find the April announcements to be noninformative for these banks. We suggest that investors with limited attention focused on the banks included in the April announcements to update their priors on balance sheet risk.
Equilibrium assumptions posit relations between different people's beliefs and behavior without describing a process that causes these relations to hold. I show that because equilibrium models do not describe a causal process whereby one endogenous variable affects another, attempts to decompose the effects of shocks into “direct” and “indirect” effects can suggest misleading predictions about how these models work. Equilibrium assumptions also imply absurd paradoxes: history can determine future behavior without affecting any intervening state variables today; individuals can learn information that no one originally possesses by observing each other’s actions. This makes equilibrium models unreliable tools to study how economic systems coordinate activity and aggregate dispersed information. I describe how to construct non-equilibrium models that avoid these paradoxes and can be interpreted causally.
I analyze the value workers ascribe to the gender composition of their workplace and the consequences of these valuations for occupational segregation, tipping, and welfare. To elicit these valuations, I survey 9,000 U.S. adults using a hypothetical job choice experiment. This reveals that on average women and men value gender diversity, but these average preferences mask substantial heterogeneity. Older female workers are more likely to value gender homophily. This suggests that gender norms and discrimination, which have declined over time, may help explain some women’s desire for homophily. Using these results, I estimate a structural model of occupation choice to assess the influence of gender composition preferences on gender sorting and welfare. I find that workers’ composition valuations are not large enough to create tipping points, but they do reduce female employment in male-dominated occupations substantially. Reducing segregation could improve welfare: making all occupations evenly gender balanced improves utility as much as a 0.4 percent wage increase for women and a 1 percent wage increase for men, on average.
Do global credit conditions affect local credit and business cycles? Using a large cross-section of equity and corporate bond market returns around the world, we construct a novel global credit factor and a global risk factor that jointly price the international equity and bond cross-section. We uncover a global credit cycle in risky asset returns, which is distinct from the global risk cycle. We document that the global credit cycle in asset returns translates into a global credit cycle in credit quantities, with a tightening in global credit conditions predicting extreme capital flow episodes and declines in the stock of country-level private debt. Furthermore, global credit conditions predict the mean and left tail of real GDP growth outcomes at the country level. Thus, the global pricing of corporate credit is a fundamental factor in driving local credit conditions and real outcomes.
Using a New Keynesian Phillips curve, we document the rapid and persistent increase in the natural rate of unemployment, ut∗, in the aftermath of the pandemic and characterize its implications for inflation dynamics. While the bulk of the inflation surge is attributed to temporary supply factors, we also find an important role for current and expected negative unemployment gaps. Through the lens of the model, the 2022–2023 disinflation was driven by the expectation that the unemployment gap will close through a progressive decline in ut∗ and a rise in the unemployment rate. This implies that convergence to long-run price stability depends critically on expectations about labor market tightness. Using a variety of cross-sectional data sources we provide corroborating evidence of unusually tight labor market conditions, consistent with our estimated rise in ut∗.
We study international monetary policy spillovers and spillbacks in a tractable two-country Heterogeneous Agent New Keynesian model. Relative to Representative Agent (RANK) models, our framework introduces a precautionary-savings channel, as households in both countries face uninsurable income risk, and a real-income channel, as households have heterogeneous marginal propensities to consume (MPC). While both channels amplify the size of spillovers/spillbacks, only precautionary savings can change their sign relative to RANK. Spillovers are likely to be larger in economies with higher fractions of high MPC households and more countercyclical income risk. Quantitatively, both channels amplify spillovers by 30-60% relative to RANK.
We study estimation and inference in panel data regression models when the regressors of interest are macro shocks, which speaks to a large empirical literature that targets impulse responses via local projections. Our results hold under general dynamics and are uniformly valid over the degree of signal-to-noise of aggregate shocks. We show that the regression scores feature strong cross-sectional dependence and a known autocorrelation structure induced only by leads of the regressor. In general, including lags as controls and then clustering over the cross-section leads to simple, robust inference.
The share of U.S. dollar assets in the official foreign exchange reserve portfolios of central banks is sometimes taken as an indicator of dollar status. We show that the observed decline in the aggregate share of U.S. dollar assets does not stem from a systematic shift in currency preferences away from holding dollar assets. Instead, a small group of countries with large foreign exchange reserve balances drive the dollar share decline observed in aggregate statistics. This arises either due to countries conducting monetary policy vis-à-vis the euro or due to preference shifts away from dollars. Regression analysis shows that interest rate differentials between traditional and nontraditional reserve currencies can tilt portfolio composition, particularly in relation to the scale of investment tranches within overall central bank portfolios. Geopolitical distance from the United States and financial sanctions are associated with lower U.S. dollar shares, especially if the primary foreign currency liquidity needs of the central bank are already satisfied.
This paper studies how quantitative easing (QE) affects household welfare across the wealth distribution. I build a Heterogeneous Agent New Keynesian (HANK) model with household portfolio choice, wage and price rigidities, endogenous unemployment, frictional financial intermediation, an effective lower bound (ELB) on the policy rate, forward guidance, and QE. To quantify the contribution of the various channels through which monetary policy affects inequality, I estimate the model using Bayesian methods, explicitly taking into account the occasionally binding ELB constraint and the QE operations undertaken by the Federal Reserve during the 2009-15 period. I find that the QE program unambiguously benefited all households by stimulating economic activity. However, it had nonlinear distributional effects. On the one hand, it widened the income and consumption gap between the top 10 percent and the rest of the wealth distribution by boosting profits and equity prices. On the other hand, QE shrank inequality within the lower 90 percent of the wealth distribution, primarily by lowering unemployment. On net, it reduced overall wealth and income inequality, as measured by the Gini index. Surprisingly, QE has weaker distributional consequences compared with conventional monetary policy. Lastly, forward guidance and an extended period of zero policy rates amplified both the aggregate and the distributional effects of QE.
We show a significant loss in U.S. Treasury market functionality when intensive use of dealer balance sheets is needed to intermediate bond markets, as in March 2020. Although yield volatility explains most of the variation in Treasury market liquidity over time, when dealer balance sheet utilization reaches sufficiently high levels, liquidity is much worse than predicted by yield volatility alone. This is consistent with the existence of occasionally binding constraints on the intermediation capacity of bond markets.
