We develop tests for detecting possibly episodic predictability induced by a persistent pre-dictor. Our framework is that of a predictive regression model with threshold effects and our goal is to develop operational and easily implementable inferences when one does not wish to impose a ̀ priori restrictions on the parameters of the model other than the slopes corresponding to the persistent predictor. Differently put our tests for the null hypothesis of no predictability against threshold predictability remain valid without the need to know whether the remain-ing parameters of the model are characterised by threshold effects or not (e.g. shifting versus non-shifting intercepts). One interesting feature of our setting is that our test statistics remain unaffected by whether some nuisance parameters are identified or not. We subsequently apply our methodology to the predictability of aggregate stock returns with valuation ratios and doc-ument a robust countercyclicality in the ability of some valuation ratios to predict returns.

In this paper we study empirically the implications of macroeconomic disagreement for the time variation in bond market risk premia. If there is a source of heterogeneity in the belief structure of the economy then differences in beliefs can affect equilibrium asset prices, and the dynamics of disagreement may generate a source of predictable variation in excess bond returns. Using survey data on macroeconomic forecasts of fundamentals spanning interest rates, real aggregates and inflation variables at different horizons we propose a new empirically observable proxy to aggregate macroeconomic disagreement and find a number of novel results. First, consistent with a general equilibrium model, heterogeneity in beliefs affects the price of risk so that disagreement regarding the real economy, inflation, short, and long ends of the yield curve explain excess bond returns with R 2 between 22%- 29%. Second, we show that macroeconomic disagreement is exogenous to time t price innovations. Third, disagreement not only contains information on expected bond returns, but also contains significant information on expected changes in future interest rates: the combination of disagreement and the forward slope predict 1-year changes in the spot rate with an R 2

We decompose long-term yields into a persistent component and maturity-related cycles to study the predictability of bond excess returns. Predictive regressions of one-year excess bond returns on a common factor constructed from the cycles give R 2 ’s up to 60 % across maturities. The result holds true in different data sets, passes a range of out-of-sample tests, and is not sensitive to the inclusion of the monetary experiment (1979/83), or the recent crisis (2007/09). We identify a simple economic mechanism that underlies this robust feature of the data: Cycles represent deviations from the long-run relationship between yields and the slow-moving component of inflation and savings. This single observation extends to a number of insights. First, we show that a key element for return predictability is contained in the first principal component of yields—the level. Once we account for this information, there is surprisingly little we can learn about term premia from other principal components. Second, we interpret the standard predictive regression using forward rates as a constrained special case of a more general return forecasting factor that could have been exploited by bond investors in real time. Third, using a simple dynamic term structure model, we quantify the cross-sectional impact of that encompassing factor on yields. We find that the factor is spanned, i.e. it has a nontrivial effect on yields which increases with the maturity of the bond. Finally, conditional on those findings, we revisit the additional predictive content of macroeconomic fundamentals for bond returns. By rendering most popular predictors insignificant,

This paper presents preliminary findings and is being distributed to economists and other interested readers solely to stimulate discussion and elicit comments. The views expressed in this paper are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the authors.

ce pte d M an us cri pt

We study risk premium in U.S. Treasury bonds. We decompose Treasury yields into inflation expectations and maturity-specific interest-rate cycles, which we define as variation in yields orthogonal to expected inflation. The short-maturity cycle captures the real short-rate dynamics. Jointly with expected inflation, it comprises the expectations hypothesis (EH) term in the yield curve. Controlling for the EH term, we extract a measure of risk-premium variation from yields. The risk-premium factor forecasts excess bond returns in and out of sample and subsumes the common bond return predictor obtained as a linear combination of forward rates. (JEL E43, G12) We study the time variation in the risk premium that investors require for holding Treasury bonds. We propose a novel way of decomposing the nominal yield curve into the risk-premium component and the expectations hypothesis (EH) term (i.e., the average expected short-term interest rate that investor expect to prevail during the life of a bond). Specifically, we decompose Treasury yields into inflation expectations and maturity-specific interest-rate cycles, which we define as the variation in yields that is orthogonal to expected inflation. The