The content and tone of 10-K risk disclosures each independently predict firms' future exposures to aggregate risk, even after controlling for standard firm characteristics.

We extract information from 10-K risk disclosures: topic models measure what is discussed, and context models measure how it is discussed. We find that both contain significant predictive information about future aggregate risk exposures, even controlling for (structured) firm characteristics, and that this information is both economically valuable and statistically distinct. We present evidence that management's discussion helps to predict some future corporate actions, which in turn can alter the firm's exposure to aggregate risk.

Aggregate dividend-price ratios robustly forecast both cash flows and returns, implying that both cash-flow and discount-rate expectations significantly drive stock prices.

The dividend-price present-value identity includes buybacks and issuance, from an aggregate perspective. Aggregate dividend-price ratios forecast buybacks and issuance, as well as returns, in the data. An alternative aggregate ratio, combining dividends and buybacks, also forecasts cash flows and returns. The long-run variance decomposition of either value ratio says that both cash-flow and discount-rate expectations significantly drive stock prices.

ESG characteristics do not earn alpha or define new risk factors, but do explain variation in firms' conditional exposures to existing aggregate risks.

We evaluate ESG investing through the lens of conditional asset pricing. We find that ESG characteristics do not define a new priced factor or deliver alpha, but they do explain variation in firms' exposure to existing aggregate risks—particularly with respect to value. These conditional beta effects differ across ESG data providers, revealing an important economic dimension of ESG disagreement—even among positively correlated scores—helping to reconcile conflicting findings on the "greenium" in the literature. Finally, we show that negative ESG screening can improve a portfolio's ESG profile with little impact on risk-adjusted returns.

IPCA is a latent factor model that uses observable characteristics as instruments for time-varying loadings, consistently estimating factors and their economic determinants from large panels.

We propose a new approach of latent factor analysis that, in addition to the main panel of interest, introduces other relevant data that serve as instruments for dynamic factor loadings. The method, called IPCA, provides a parsimonious means of incorporating vast conditioning information into factor model estimates. This improves the efficiency of estimates for the latent factors and their loadings, and helps to ascertain the economic relationships among factors and individuals via the observable instruments. The estimation is fast to calculate and accommodates unbalanced panels. We show consistency and asymptotic normality under general panel data generating processes. We demonstrate the advantages of IPCA in simulated data and in applications to equity asset pricing and international macroeconomics.

A conditional five-factor model for corporate bond returns dramatically outperforms prior models and recommends systematic bond portfolios that beat leading credit strategies.

We propose a conditional factor model for corporate bond returns with five factors and time-varying factor loadings. We have three main empirical findings. First, our factor model excels in describing the risks and returns of corporate bonds, improving over previously proposed models in the literature by a large margin. Second, our benchmark model recommends a systematic bond investment portfolio that significantly outperforms leading corporate credit investment strategies. Third, we find closer integration between debt and equity markets than found in prior literature.

Momentum and long-term reversal reflect time-varying risk compensation, as captured by a conditional factor model in which loadings depend on observable firm characteristics.

Stock momentum, long-term reversal, and other past return characteristics that predict future returns also predict future realized betas, suggesting these characteristics capture time-varying risk compensation. We formalize this argument with a conditional factor pricing model. Using instrumented principal components analysis, we estimate latent factors with time-varying factor loadings that depend on observable firm characteristics. We show that factor loadings vary significantly over time, even at short horizons over which the momentum phenomenon operates (one year), and that this variation captures reliable conditional risk premia missed by other factor models commonly used in the literature. Our estimates of conditional risk exposure can explain a sizeable fraction of momentum and long-term reversal returns and can be used to generate even stronger return predictions.

Households effectively insure against much of the earnings risk facing primary earners, facing roughly half the countercyclical risk increase experienced by males alone.

Using detailed IRS administrative data on millions of households, we find that households effectively insure against much of the risk facing primary earners. We show that households face less risk than males alone, and households face roughly half the countercyclical risk increase. As a result of these risk differences, household certainty equivalent earnings are 19% higher than for males alone, and household certainty equivalent earnings fall by about half as much during recessions. To facilitate related research, we make available the aggregated data used in our analysis.

IPCA explains the cross section of returns with five latent factors, finding that only 10 characteristics drive nearly all of the model's accuracy with anomaly intercepts statistically indistinguishable from zero.

We propose a new modeling approach for the cross section of returns. Our method, Instrumented Principal Component Analysis (IPCA), allows for latent factors and time-varying loadings by introducing observable characteristics that instrument for the unobservable dynamic loadings. If the characteristics/expected return relationship is driven by compensation for exposure to latent risk factors, IPCA will identify the corresponding latent factors. If no such factors exist, IPCA infers that the characteristic effect is compensation without risk and allocates it to an "anomaly" intercept. Studying returns and characteristics at the stock level, we find that five IPCA factors explain the cross section of average returns significantly more accurately than existing factor models and produce characteristic-associated anomaly intercepts that are small and statistically insignificant. Furthermore, among a large collection of characteristics explored in the literature, only 10 are statistically significant at the 1% level in the IPCA specification and are responsible for nearly 100% of the model's accuracy.

ECB purchases under the Securities Markets Programme caused robust and lasting reductions in sovereign bond liquidity premia, consistent with a search-based asset pricing model.

To "ensure depth and liquidity," the European Central Bank intervened in sovereign debt markets through its Securities Markets Programme (SMP), providing a unique opportunity to estimate the effects of large-scale asset purchases on sovereign bond liquidity premia. From reduced-form estimates, we find robust, economically significant impact and lasting reductions in sovereign bonds' liquidity premia in response to official purchases. We develop a search-based asset-pricing model to understand our empirical results. The theory implies that bond liquidity premia fall in response to both official purchases and rising sovereign default probabilities, as seen in the data.

Using forecaster surveys to sidestep the zero lower bound censoring problem, we find that after the Global Financial Crisis the Fed's perceived inflation response weakened while its unemployment response strengthened.

Did the Federal Reserve's response to economic fundamentals change with the onset of the Global Financial Crisis? Estimation of a monetary policy rule to answer this question faces a censoring problem since the interest rate target has been set at the zero lower bound since late 2008. Surveys by forecasters allow us to sidestep the problem and to use conventional regressions and break tests. We find that, in the opinion of forecasters, the Fed's inflation response has decreased and the unemployment response has increased, which suggests that the Federal Reserve's commitment to stable inflation has become weaker in the eyes of the professional forecasters.

Changes in systemic risk measures skew the distribution of subsequent macroeconomic shocks, and dimension-reduction indexes constructed from many measures predict macroeconomic outcomes out of sample.

This article studies how systemic risk and financial market distress affect the distribution of shocks to real economic activity. We analyze how changes in 19 different measures of systemic risk skew the distribution of subsequent shocks to industrial production and other macroeconomic variables in the US and Europe over several decades. We also propose dimension reduction estimators for constructing systemic risk indexes from the cross section of measures and demonstrate their success in predicting future macroeconomic shocks out of sample.

The 3PRF forecasts a target series using many predictors by requiring only knowledge of the number of factors relevant to the target, regardless of the full factor space.

We forecast a single time series using many predictor variables with a new estimator called the three-pass regression filter (3PRF). It is calculated in closed form and conveniently represented as a set of ordinary least squares regressions. 3PRF forecasts are consistent for the infeasible best forecast when both the time dimension and cross section dimension become large. This requires specifying only the number of relevant factors driving the forecast target, regardless of the total number of common factors driving the cross section of predictors. The 3PRF is a constrained least squares estimator and reduces to partial least squares as a special case. Simulation evidence confirms the 3PRF's forecasting performance relative to alternatives. We explore two empirical applications: forecasting macroeconomic aggregates with a large panel of economic indices, and forecasting stock market returns with price-dividend ratios of stock portfolios.

Young workers face greater cyclical volatility in both hours and wages than prime-aged workers, which a model of labor demand differences—not supply differences—can explain.

Over the business cycle young workers experience much greater volatility of hours worked than prime-aged workers. This can arise from age differences in labor supply or labor demand characteristics. To distinguish between these, we document that, for young workers, both the cyclical volatilities of hours and wages are greater than those of the prime-aged. We argue that a general class of models featuring only age-specific labor supply differences cannot reconcile these facts. We then show that a simple model featuring labor demand differences can.

A single factor extracted from the cross section of book-to-market ratios forecasts aggregate stock market returns and cash flows out of sample with an annual R-squared as high as 13%.

Returns and cash flow growth for the aggregate U.S. stock market are highly and robustly predictable. Using a single factor extracted from the cross section of book-to-market ratios, we find an out-of-sample return forecasting R-squared as high as 13% at the annual frequency (0.9% monthly). We document similar out-of-sample predictability for returns on value, size, momentum, and industry-sorted portfolios. We present a model linking aggregate market expectations to disaggregated valuation ratios in a dynamic latent factor system. We find that spreads in growth and value portfolios' exposures to economic shocks are key to identifying predictability and are consistent with duration-based theories of the value premium. Our findings suggest that discount rates are far less persistent, and their shocks far more volatile, than implied by leading asset pricing models.

Introducing data uncertainty into a model of a learning Federal Reserve makes the learning process more sluggish, explaining the rise and fall of U.S. inflation through a perceived Phillips curve trade-off.

Economic agents who are uncertain of their economic model learn, and this learning is sensitive to the presence of data uncertainty. I investigate this idea in a framework that successfully describes inflation as a learning Federal Reserve's optimal policy but fails to satisfactorily motivate these policy shifts. I modify the framework to account for data uncertainty: the learning process is made more sluggish by its presence. Consequently, the estimated model provides an explanation for the rise and fall in inflation: the concurrent rise and fall in the perceived Phillips curve trade-off between inflation and unemployment.

Using macroeconomic news to identify the market-perceived Fed policy rule, we find that between 1994 and 2007 the output response vanished while the inflation response became more gradual but larger in long-run magnitude.

We introduce a novel method for estimating a monetary policy rule using macroeconomic news. We estimate directly the policy rule agents use to form their expectations by linking news' effects on forecasts of both economic conditions and monetary policy. Evidence between 1994 and 2007 indicates that the market-perceived Federal Reserve policy rule changed: the output response vanished, and the inflation response path became more gradual but larger in long-run magnitude. These response coefficient estimates are robust to measurement and theoretical issues with both potential output and the inflation target.

TRACE bond returns differ meaningfully from ICE data used by practitioners, though the core results of Kelly et al. (2023) are robust to using WRDS/TRACE data instead.

TRACE bond returns are meaningfully different from ICE bond return data used by banks, asset managers, and hedge funds. The core results of Kelly et al. (2023) are robust to instead using WRDS/TRACE data.

Investment-good relative prices can fluctuate due to endogenous markups, not only technological shocks

The two sector model presented in this note suggests a simple structural decomposition of movements in the price of investment goods into exogenous and endogenous sources. The endogenous fluctuations arise in the presence of countercyclical markups which vary differently across the consumption and investment sectors. In turn, the movements in the markups are due to endogenous procyclical net business formation. The model, while being consistent with the countercyclicality of the price of investment goods, suggests that about a quarter of the movement in the price series can be attributed to this endogenous mechanism.

The pseudo-information filter solves linear state-space models with significant computational advantages over the Kalman filter when data are high-dimensional.

The pseudo-information filter is defined in order to solve a linear state-space model. The filter offers significant computational advantages over the Kalman filter when the data are high-dimensional.