Financial asset risk premia are widely agreed to vary over time. This paper decomposes these risk premia into expected excess returns earned in short windows around the times of macroeconomic news announcements (which mostly come out at 8:30am) and the expected excess returns that are earned at other times. Using intradaily data, we find that some, but not all, of the time-varying expected excess returns accrue right around macroeconomic announcements. In forecasting six-month cumulative bond returns, there is more predictability in announcement windows than at other times.

This chapter discusses recent developments in inflation forecasting. We perform a horse-race among a large set of traditional and recently developed forecasting methods, and discuss a number of principles that emerge from this exercise. We find that judgmental survey forecasts outperform model-based ones, often by a wide margin. A very simple forecast that is just a glide path between the survey assessment of inflation in the current-quarter and the long-run survey forecast value turns out to be competitive with the actual survey forecast and thereby does about as well or better than model-based forecasts. We explore the strengths and weaknesses of some specific prediction methods, including forecasts based on the Phillips curve and based on dynamic stochastic general equilibrium models, in greater detail. We also consider measures of inflation expectations taken from financial markets and the tradeoff between forecasting aggregates and disaggregates.

Employing a large number of financial indicators, we use Bayesian Model Averaging (BMA) to forecast real-time measures of economic activity. The indicators include credit spreads based on portfolios--constructed directly from the secondary market prices of outstanding bonds--sorted by maturity and credit risk. Relative to an autoregressive benchmark, BMA yields consistent improvements in the prediction of the cyclically-sensitive measures of economic activity at horizons from the current quarter out to four quarters hence. The gains in forecast accuracy are statistically significant and economically important and owe almost exclusively to the inclusion of credit spreads in the set of predictors.

We analyze retail prices and at-the-dock (import) prices of specific items in the Bureau of Labor Statistics' (BLS) CPI and IPP databases, using both databases simultaneously to identify items that are identical in description at the dock and when sold at retail. This identification allows us to measure the distribution wedge associated with bringing traded goods from the point of entry into the United States to their retail outlet. We find that overall U.S. distribution wedges are 50-70%, around 10 to 20 percentage points higher than that reported in the literature. We discuss the implications of this for measuring the size of the "pure" tradeables sector, exchange rate pass-through, and real exchange rate determination. We find that distribution wedges are very stable over time but there is considerable variation across items. There is some variation across the country of origin for the imported item, for our major trading partners, but not as much as the cross-item variation. We also investigate the determinants of distribution wedges, finding that wedges do not vary systematically with exchange rates, but are related to other features of the micro data

DSGE models have now reached a point where they can and do serve an important role in the monetary policy process. From the standpoint of real-world policymaking, however, there remain important areas of omission and coarse approximation in these models. I argue that macroeconomics should follow other fields such as toxicology in having a formal literature on how best to use models that are far from perfect as a basis for public policy.

While dynamic stochastic general equilibrium (DSGE) models for monetary policy analysis have come a long way, there is considerable difference of opinion over the role these models should play in the policy process. The paper develops three main points about assessing the value of these models. First, we document that DSGE models continue to have aspects of crude approximation and omission. This motivates the need for tools to reveal the strengths and weaknesses of the models--both to direct development efforts and to inform how best to use the current flawed models. Second, posterior predictive analysis provides a useful and economical tool for finding and communicating strengths and weaknesses. In particular, we adapt a form of discrepancy analysis as proposed by Gelman, et al. (1996). Third, we provide a nonstandard defense of posterior predictive analysis in the DSGE context against long-standing objections. We use the iconic Smets-Wouters model for illustrative purposes, showing a number of heretofore unrecognized properties that may be important from a policymaking perspective.

It is well known that augmenting a standard linear regression model with variables that are correlated with the error term but uncorrelated with the original regressors will increase the asymptotic efficiency of the original coefficients. We argue that in the context of predicting excess returns, valid augmenting variables exist and are likely to yield substantial gains in estimation efficiency and, hence, predictive accuracy. The proposed augmenting variables are ex post measures of an unforecastable component of excess returns: ex post errors from macroeconomic survey forecasts, the surprise components of asset price movements around macroeconomic news announcements, or even the weather. These "surprises" cannot be used directly in forecasting they are not observed at the time that the forecast is made but can nonetheless improve forecasting accuracy by reducing parameter estimation uncertainty. We derive formal results about the benefits and limits of this approach and apply it to standard examples of forecasting excess bond and equity returns. We find substantial improvements in out-of-sample forecast accuracy for standard excess bond return regressions; gains for forecasting excess stock returns are much smaller.

It is well known that augmenting a standard linear regression model with variables that are correlated with the error term but uncorrelated with the original regressors will increase asymptotic efficiency of the original coefficients. We argue that in the context of predicting excess returns, valid augmenting variables exist and are likely to yield substantial gains in estimation efficiency and, hence, predictive accuracy. The proposed augmenting variables are ex post measures of an unforecastable component of excess returns: ex post errors from macroeconomic survey forecasts and the surprise components of asset price movements around macroeconomic news announcements. These "surprises" cannot be used directly in forecasting--they are not observed at the time that the forecast is made--but can nonetheless improve forecasting accuracy by reducing parameter estimation uncertainty. We derive formal results about the benefits and limits of this approach and apply it to standard examples of forecasting excess bond and equity returns. We find substantial improvements in out-of-sample forecast accuracy for standard excess bond return regressions; gains for forecasting excess stock returns are much smaller. Jon Faust Johns Hopkins University Department of Economics Mergenthaler Hall 456 3400 N. Charles Street Baltimore, MD 25218 and NBER faustj@jhu.edu Jonathan H. Wright Division of International Finance The Federal Reserve Board 20th Street & Constitution Avenue, NW Washington, DC 20551 jonathan.h.wright@frb.gov 1. Introduction Many empirical papers in …nance explore the predictability of excess returns using a simple regression-based approach: estimate a regression for future excess returns based on current predictors and then measure the degree of predictive power of the estimated model. We derive a method to increase e¢ ciency in the estimation step and show that this method can lead to substantial gains in measured forecastability. The key idea is familiar from …rst-year econometrics. If we take any regression and augment it with regressors that are correlated with the error term, but are known to be uncorrelated with the original regressors in population, we increase asymptotic e¢ ciency of the estimates of the original coe¢ cients without compromising consistency. The augmenting variables are not of direct interest, but soak up some residual variance, increasing precision of the estimates of the coe¢ cients that are of interest. This idea is an example of the familiar principle that system estimation imposing correct cross-equation restrictions is more e¢ cient than single equation estimation,1 but we are not aware systematic treatment of the sort we are proposing in forecasting context. We argue that forecasting excess returns provides an excellent opportunity for gains from this approach. The standard predictive regression is of the form,

for their very helpful comments on earlier drafts, and to Lu Xu for research assistance. All errors are our sole responsibility.

Many recent articles have found that atheoretical forecasting methods using many predictors give better predictions for key macroeconomic variables than various small-model methods. The practical relevance of these results is open to question, however, because these articles generally use ex post revised data not available to forecasters and because no comparison is made to best actual practice. We provide some evidence on both of these points using a new large dataset of vintage data synchronized with the Fed’s Greenbook forecast. This dataset consist of a large number of variables as observed at the time of each Greenbook forecast since 1979. We compare realtime, large dataset predictions to both simple univariate methods and to the Greenbook forecast. For inflation we find that univariate methods are dominated by the best atheoretical large dataset methods and that these, in turn, are dominated by Greenbook. For GDP growth, in contrast, we find that once one takes account of Greenbook’s advantage in evaluating the current state of the economy, neither large dataset methods, nor the Greenbook process offers much advantage over a univariate autoregressive forecast.

This article aims to better understand the factors driving fluctuations in potential output measured by the production function approach (PFA.)To do so, the authors integrate a production function definition of potential output into a large-scale dynamic stochastic general equilibrium (DSGE) model in a fully consistent manner and give two estimated versions based on U.S. and euro-area data.The main contribution of this article is to provide a quantitative and comparative assessment of two approaches to potential output measurement, namely DSGE and PFA, in an integrated framework.The authors find that medium-term fluctuations in potential output measured by the PFA are likely to result from a large variety of shocks, real or nominal.These results suggest that international comparisons of potential growth using the PFA could lead to overstating the role of structural factors in explaining cross-country differences in potential output, while neglecting the fact that different economies are exposed to different shocks over time.

The 1960s were an exciting time – at least for macroeconomic modelers. An impressive new kind of macroeconometric model was entering central banking, and cutting-edge central banks were beginning to analyze policy as a problem of optimal control. The December 1965 edition of Time, the popular U.S. news magazine, has Keynes on the cover, quotes the experts of the day extensively, and is almost giddy in tone regarding the successes of countercyclical policy. Indeed, one gets the impression that the future of the business cycle might be rather dull: ‘[U.S. businessmen] have begun to take for granted that the Government will intervene to head off recession or choke off inflation.’ By the revealed practice of central bankers, the new econometric models of the 1960s were a long-term success. The original models and their direct descendents remained workhorses of policy analysis at central banks for the next forty years or so. Were it not for the role the models played in the tragic economic events of the 1970s, this would be a very happy tale of scientific advance. We are once again in exciting times for macro modelers: a new breed of policy analysis model is entering central banking. Cutting-edge central banks are again beginning to analyze monetary policy as an optimal control problem within those models. For the first time since the mistakes of the 1970s, science is gaining ground in discussions of the art and science of monetary policymaking (e.g., Mishkin, 2007). At a central banking conference in 2007, I heard a senior central banker lament that the modern strategy of model-based flexible inflation targeting might render central banking rather dull.

It is well known that augmenting a standard linear regression model with variables that are correlated with the error term but uncorrelated with the original regressors will increase asymptotic e¢ ciency of the original coe¢ cients. We argue that in the context of predicting excess returns, valid augmenting variables exist and are likely to yield substantial gains in estimation e¢ ciency and, hence, predictive accuracy. The proposed augmenting variables are ex post measures of an unforecastable component of excess returns: ex post errors from macroeconomic survey forecasts, the surprise components of asset price movements around macroeconomic news announcements, or even the weather. These “surprises” cannot be used directly in forecasting— they are not observed at the time that the forecast is made— but can nonetheless improve forecasting accuracy by reducing parameter estimation uncertainty. We derive formal results about the bene…ts and limits of this approach and apply it to standard examples of forecasting excess bond and equity returns. We …nd substantial improvements in out-of-sample forecast accuracy for standard excess bond return regressions; gains for forecasting excess stock returns are much smaller. KEYWORDS: Excess returns, e¢ ciency, predictive regression, term premiums, seemingly unrelated regression. JEL Classi…cations: C22, C53, E17, E43. 1. Introduction Many empirical papers in …nance explore the predictability of excess returns using a simple regression-based approach: estimate a regression for future excess returns based on current predictors and then measure the degree of predictive power of the estimated model. We derive a method to increase e¢ ciency in the estimation step and show that this method can lead to substantial gains in measured forecastability. The key idea is familiar from …rst-year econometrics. If we take any regression and augment it with regressors that are correlated with the error term, but are known to be uncorrelated with the original regressors in population, we increase asymptotic e¢ ciency of the estimates of the original coe¢ cients without compromising consistency. The augmenting variables are not of direct interest, but soak up some residual variance, increasing precision of the estimates of the coe¢ cients that are of interest. This idea is an example of the familiar principle that system estimation imposing correct cross-equation restrictions is more e¢ cient than single equation estimation,1 but we are not aware systematic treatment of the sort we are proposing in forecasting context. We argue that forecasting excess returns provides an excellent opportunity for gains from this approach. The standard predictive regression is of the form,

Central Banks regularly make forecasts, such as the Fed’s Greenbook forecast, that are conditioned on hypothetical paths for the policy interest rate. While there are good public policy reasons to evaluate the quality of such forecasts, up until now, the most common approach has been to ignore their conditional nature and apply standard forecast efficiency tests. In this paper we derive tests for the efficiency of conditional forecasts. Intuitively, these tests involve implicit estimates of the degree to which the conditioning path is counterfactual and the magnitude of the policy feedback over the forecast horizon. We apply the tests to the Greenbook forecast and the Bank of England’s inflation report forecast, finding some evidence of forecast inefficiency. Nonetheless, we argue that the conditional nature of the forecasts made by central banks represents a substantial impediment to the analysis of their quality—stronger assumptions are needed and forecast inefficiency may go undetected for longer than would be the case if central banks were instead to report unconditional forecasts.

This working paper comments on Monika Piazzesi and Martin Schneider's "Bond Positions, Expectations, and the Yield Curve," delivered at the Fiscal Policy and Monetary/Fiscal Policy Interactions conference held at the Atlanta Fed on April 19-20, 2007.

Many recent papers have studied movements in stock, bond, and currency prices over short windows of time around macro announcements. This paper adds to the announcement effects literature in two ways. First, we study the joint announcement effects across a broad range of assets--exchange rates and U.S. and foreign term structures. In order to evaluate whether the joint effects can be reconciled with conventional theory, we interpret the joint movements in light of uncovered interest rate parity or changes in risk premia. For several real macro announcements, we find that a stronger than expected release appreciates the dollar today, but that it must either (i) lower the relative risk premium for holding foreign currency rather than dollars, or (ii) imply considerable future expected dollar depreciation. The latter implies an overshooting behavior akin to that described by Dornbusch (1976). Second, we use a longer span of high frequency data than has been common in announcement work. A longer span of high frequency data contributes to the precision of our estimates and allows us to explore the possibility that the effects of macro surprises on asset prices have varied over time. We find evidence, for example, that PPI releases had a larger effect on U.S. interest rates before about 1992 than subsequently.