This paper contributes to two strands of business cycle literature — news shocks and bounded rationality — by assessing the empirical importance of TFP news shocks while relaxing the rational expectations assumption. We estimate a medium-scale DSGE model, incorporating financial frictions and TFP news shocks, under two different expectation formation mechanisms: rational expectations (RE) and adaptive learning (AL). The results suggest that AL amplifies the effects of financial market frictions, leading to three key findings. First, AL improves the model’s fit, as shown in the related literature, and better replicates the volatility of several aggregate variables. Second, the AL amplification results in a deflationary response and a more persistent reaction of lending spreads to TFP news shocks. Third, AL increases the importance of pure news shocks (i.e. purely anticipated shocks), amplifying their effects through both expectation and credit channels. Finally, we show that the dynamics generated by the DSGE model under AL align more closely with empirical VAR evidence than those produced by the RE version of the DSGE model.

This paper shows that inflation expectations and those embedded in short-term interest rate expectations as reported in the Survey of Professional Forecasters show evidence of misaligned expectations. This misalignment seems to have been substantial in recent times, featuring a low correlation between inflation and the policy rate. This empirical evidence motivates an alternative explanation, based on uncertainty rather than risk, of the bond term premium measures found in the literature. This paper estimates an expectational term premium driven by misaligned short-term interest rate expectations from a behavioral DSGE model that introduces model uncertainty by assuming adaptive learning with discretionary beliefs. The estimated 10-year expectational term premium shares important features with the corresponding term premium measures obtained using no-arbitrage affine term structure models. Thus, the expectational term premium is sizable, highly persistent, mildly countercyclical, and highly correlated with those term premium measures in the most recent period studied. In short, a potential misalignment of short-term interest expectations with inflation expectations provides an important channel for explaining the bond premium lately.

This paper builds on the Euler-equation approach to adaptive learning by introducing the term structure of interest rates into a medium-scale DSGE model, where bond yields are priced with separate Euler equations. Term structure information enables us to characterize agents' forecasting models using only term spread information available at the time when expectations are formed. Our estimated DSGE model under adaptive learning substantially improves the model fit to the data, including both macroeconomic and yield curve data, compared to the rational expectations version. The out-of-sample forecasting performance also increases under our learning approach. Despite large modeling differences, the estimated non-standard term-risk premium from our adaptive learning model resembles those term premia estimated using no-arbitrage affine term structure models. The model fit gain obtained assuming adaptive learning with term structure information instead of rational expectations largely increases when disciplining model expectations with forecasts from the Survey of Professional Forecasters.

This paper provides a solution to the equity premium puzzle. We modify the standard constant relative risk aversion utility function by assuming that the representative consumer also has a preference for consumption predictability. While keeping the conditional mean of the stochastic discount factor close to one, this feature not only reinforces consumption smoothing, but it also results in large increases in the variability of the stochastic discount factor which is crucial for this solution to the puzzle. The large increase in variability for the stochastic discount factor in our modified model is primarily determined by large, realized consumption forecast errors. Although these oversized forecast errors arise infrequently, when they do arise, they result in very high aversion to risk and enhanced interest in smoothing consumption.

This paper assesses the significance of quality-of-capital (QoC) news shocks and their transmission through the credit channel in explaining aggregate fluctuations. Our framework is an estimated medium-scale DSGE model augmented with a financial sector where two alternative sources of news shocks are considered. One is a (standard) total-factor-productivity (TFP) news shock; the other is a QoC news shock. The latter has a clear meaning that enables a close link to be built up between financial markets and the macroeconomy through the credit and expectation channels, which greatly improves model fit and largely displaces TFP news shocks as a source of the business cycle. The significance of pure (rather than realized) news underscores the role of expectations.

Abstract We revisit three major US recessions through the lens of a standard medium-scale DSGE model (Smets, F., and R. Wouters. 2007. “Shocks and Frictions in US Business Cycles: A Bayesian DSGE Approach.” The American Economic Review 97: 586–606) augmented with financial frictions. We first estimate the DSGE model using a Bayesian approach for three alternative periods, each containing a major US recession: the Great Depression, the Stagflation and the Great Recession. Then, we assess the stability of structural parameters, and analyze what frictions were particularly important and what shocks were the main drivers of aggregate fluctuations in each historical period. This exercise can be understood as a test of the standard New-Keynesian DSGE model with financial accelerator in closed economies. We find that the estimated DSGE model is able to provide a sound explanation of all three recessions by closely relating both estimated structural shocks and frictions with well known economic events.

This paper estimates a term premium from an estimated DSGE model under adaptive learning (AL) with arbitrary beliefs. The estimated AL term premium shares important features with the corresponding measures obtained in the literature using no-arbitrage affine term structure models. Thus, the AL term premium is sizable, highly persistent, and countercyclical. The estimation results suggest that AL, by introducing much more economic uncertainty, presents an additional channel for explaining the bond premium. This feature can be viewed as an alternative to other features needed to generate a sizable term premium under rational expectations, such as a huge risk aversion coefficient. Shocks associated with wage markup, investment, and monetary policy explain roughly 75% of AL term premium fluctuations, but their relative contributions change quite dramatically with the forecast horizon used in the variance decomposition.

This paper builds on the Euler-equation approach to adaptive learning (AL) by considering term structure information in addition to macroeconomic data. We consider a medium-scale DSGE model where the term structure expectations hypothesis is imposed. The model estimated under AL using term structure information has a better fit. Estimation results show that term structure provides useful real-time information in forecasting macroeconomic variables above and beyond that provided by revised macroeconomic data. In particular, term structure information greatly improves the matching of AL expectations to the corresponding forecasts reported in the Survey of Professional Forecasters, which further contributes to the empirical validity of AL.

Recent studies show that the estimated parameters of rational expectations dynamic stochastic general equilibrium models of the business cycle are largely time-varying. This paper shows that assuming adaptive learning (rather than rational expectations) strongly reduces the estimated parameter variability of standard models (by around 75%). Moreover, the reduction in parameter variability induced by adaptive learning is much stronger for the subsets of parameters that control nominal price and wage rigidity and the subset of policy rule parameters (at 98% and 83%, respectively). Furthermore, our estimation results suggest that adaptive learning helps to explain the recent swings in the comovements between real and nominal US macroeconomic variables, but the swing in the relative weight of supply and demand shocks seems to be the most important driving force.

Agents can learn from financial markets to predict macroeconomic outcomes, and learning dynamics can feed back into both the macroeconomy and financial markets. This paper builds on the adaptive learning (AL) model of [Slobodyan, S. and R. Wouters (2012a) American Economic Journal: Macroeconomics 4, 65–101.] by introducing the term structure of interest rates. This extension enables term structure information to fully characterize agents’ expectations in real time. This feature addresses an imperfect information issue neglected in the related AL literature. The term structure of interest rates results in a strong channel of persistence driven by multi-period forecasting. Including the term structure in the AL model results in a model fit similar to that obtained in the rational expectation (RE) version of the model, but it greatly reduces the importance of other endogenous sources of aggregate persistence such as price and wage stickiness and the elasticity of the cost of adjusting capital. The model estimated also shows that term premium innovations are a major source of persistent fluctuations in nominal variables under AL. This stands in sharp contrast to the lack of transmission of term premium shocks to the macroeconomy under REs.

The volatility of unemployment fluctuations has been about 3 times higher in Spain than in Germany over the recent business cycles (1996–2013). Besides, the rates of unemployment of these two countries have moved on opposite directions featuring negative cross correlation. We estimate a DSGE model with unemployment and find these explanatory factors: (i) wage rigidity has been higher in Spain, (ii) the elasticity of hours per worker has been lower in Spain, (iii) labor force shocks have been stronger and more persistent in Spain, (iv) risk-premium shocks have deteriorated labor demand in Spain while fiscal/net exports shocks have stimulated labor demand in Germany, and (v) the idiosyncratic shocks from the ECB single monetary policy have switched from reducing Spanish unemployment (before 2007) to increasing it (after 2007).

The primary goal of this article is to investigate whether properly modelling real-time data and optimal real-time decision-making of a monetary planner provides new insights into monetary policy behaviour and outcomes. This article extends a variant of the asymmetric preference model suggested by Ruge-Murcia to investigate the use of real-time data available to policymakers when making their decisions and revised data which more accurately measure economic performance, but is only available much later. In our extended model, the central banker targets a weighted average of revised and real-time inflation together with a weighted average of revised and real-time output. Moreover, we allow for an asymmetric central bank response to real-time data depending on whether the unemployment rate is high or low. Our model identifies several new potential sources of inflation bias due to data revisions. Our empirical results suggest that the Federal Reserve Bank focuses on targeting revised inflation during low unemployment periods, but it weighs heavily real-time inflation during high unemployment periods. The inflation bias due to data revisions is comparable in magnitude to the bias from asymmetric central banker preferences with the bias being somewhat larger during high unemployment.

This paper extends the asymmetric preference model suggested by Ruge-Murcia (2003a) to investigate the use of real-time data which roughly measure the type of data available to policy makers when making their decisions and revised data which more accurately measure economic performance (Croushore, 2011). In our extended model, the central banker monitors a weighted average of revised and real-time inflation. Moreover, we allow for an asymmetric central bank focus on real-time inflation depending on whether the unemployment rate is high or low. Our model identifies a source of inflation bias due to inflation revisions. Our empirical results suggest that the Federal Reserve Bank focuses on monitoring revised inflation during low unemployment periods, but it weights real-time inflation heavily during high unemployment periods. In contrast, the Bank of England seems to focus on an equally-weighted average of real-time and revised inflation when monitoring inflation which is fairly robust over the business cycle.

In recent business cycles, U.S. inflation has experienced a reduction of volatility and a severe weakening in the correlation to the nominal interest rate (Gibson paradox). We examine these facts in an estimated dynamic stochastic general equilibrium model with money. Our findings point at a flatter New Keynesian Phillips Curve (higher price stickiness) and a lower persistence of markup shocks as the main explanatory factors. In addition, a higher interest-rate elasticity of money demand, an increasing role of demand-side shocks, and a less systematic behavior of Fed's monetary policy also account for the recent patterns of U.S. inflation dynamics. (JEL E32, E47)

This paper introduces the term structure of interest rates into a medium-scale DSGE model. This extension results in a multi-period forecasting model that is estimated under both adaptive learning and rational expectations. Term structure information enables us to characterize agents’ expectations in real time, which addresses an imperfect information issue mostly neglected in the adaptive learning literature. Relative to the rational expectations version, our estimated DSGE model under adaptive learning largely improves the model fit to the data, which include not just macroeconomic data but also the yield curve and the consumption growth and inflation forecasts reported in the Survey of Professional Forecasters. Moreover, the estimation results show that most endogenous sources of aggregate persistence are dramatically undercut when adaptive learning based on multi-period forecasting is incorporated through the term structure of interest rates.

This paper empirically investigates US fiscal policy sustainability and cyclicality in an empirical structure that allows fiscal policy responses to exhibit asymmetric behavior. We investigate this over two quarterly intervals, both of which begin in 1955:1. The short sample ends in 1995:2 and is most similar to the one used by Bohn (Q J Econ 113:949–963, 1998 ), whereas the full sample ends in 2013:3. Our estimation results show that the full sample period is sufficiently different from the short sample period, that the asymmetric (nonlinear) empirical models used in this paper are important and that the sustainability of US government debt topic needed to be revisited. Indeed, the short sample provides evidence of fiscal policy sustainability in line with Bohn’s ( 1998 ) findings. However, when considering the full sample, US fiscal policy is found sustainable during good economic times only according to the best fitting nonlinear model, but unsustainable for all specifications studied during times of distress. With regard to cyclicality, both samples show policy is asymmetric. Moreover, both samples show countercyclical policy during times of distress and the full sample results show some evidence that policy may be procyclical during good economic times.

Revisions of U.S. macroeconomic data are persistent, correlated with real-time data, and with high variability (around 80% of U.S. real-time data volatility). This paper adapts a DSGE-style model to accommodate both real-time and revised data from the U.S. economy. The results show a lesser role of both habit formation and price indexation than in the standard model. In the simulations, revision shocks to both output and inflation are expansionary because the Fed reacts by cutting interest rates. Consumption revisions, in contrast, are countercyclical, consumption mirrors the observed reduction in real-time consumption. In the variance decomposition, data revisions explain 9.3% of output changes.

In this paper we show that there are striking differences in impulse response results when using revised and real-time data under opportunistic threshold structures. Notably, the impulse responses using revised data show almost no threshold behavior differences between the two sides of an opportunistic threshold structure while real-time data show significant differences between the two sides. This result is important to recognize since policy makers do not have ex-post revised data during their decision deliberations. As a result, economists using revised data may misread policy making behavior that can only be revealed using real-time data. JEL Classification: C32, E32, E52

The price–dividend (PD) ratio must be stationary for the present value model to be valid. However, several market episodes show stock prices drifting apart from dividends. This paper investigates PD ratio stationarity by considering a Markov-switching model featuring an asymmetric adjustment speed toward a unique attractor. A three-regime model displays the best regime identification. Within this specification, the post-war period is mainly characterized by a stationary state featuring slow reversion to a high attractor, the growing PD ratio period of 1996–2000 features a high-reversion stationary regime, and the subprime crisis episode is classified into a temporary nonstationary regime.