nep-for New Economics Papers
on Forecasting
Issue of 2019‒02‒11
thirteen papers chosen by
Rob J Hyndman
Monash University

  1. Evaluating the Conditionality of Judgmental Forecasts By Travis J. Berge; Andrew C. Chang; Nitish R. Sinha
  2. Time-Varying Risk Aversion and the Predictability of Bond Premia By Oğuzhan Çepni; Riza Demirer; Rangan Gupta; Christian Pierdzioch
  3. Forecasting Volatility with Time-Varying Leverage and Volatility of Volatility Effects By Leopoldo Catania; Tommaso Proietti
  4. The Macroeconomic Projections of the German Government: A Comparison to an Independent Forecasting Institution By Robert Lehmann; Timo Wolllmershäuser
  5. A Parametric Factor Model of the Term Structure of Mortality By Niels Haldrup; Carsten P. T. Rosenskjold
  6. Forecasting dynamically asymmetric fluctuations of the U.S. business cycle By Emilio Zanetti Chini
  7. Assessing the uncertainty in central banks' inflation outlooks By Knüppel, Malte; Schultefrankenfeld, Guido
  8. Forecaster’s utility and forecasts coherence By Emilio Zanetti Chini
  9. Forecasting French GDP with Dynamic Factor Models : a pseudo-real time experiment using Factor-augmented Error Correction Models By Stéphanie Combes; Catherine Doz
  10. Nowcasting Peruvian GDP using Leading Indicators and Bayesian Variable Selection By Pérez, Fernando
  11. CBO’s Economic Forecasting Record: 2017 Update By Congressional Budget Office
  12. Factor Investing: Hierarchical Ensemble Learning By Guanhao Feng; Jingyu He
  13. Forecasting Foreign Economic Growth Using Cross-Country Data By Hakkio, Craig S.; Nie, Jun

  1. By: Travis J. Berge; Andrew C. Chang; Nitish R. Sinha
    Abstract: We propose a framework to evaluate the conditionality of forecasts. The crux of our framework is the observation that a forecast is conditional if revisions to the conditioning factor are faithfully incorporated into the remainder of the forecast. We consider whether the Greenbook, Blue Chip, and the Survey of Professional Forecasters exhibit systematic biases in the manner in which they incorporate interest rate projections into the forecasts of other macroeconomic variables. We do not find strong evidence of systematic biases in the three economic forecasts that we consider, as the interest rate projections in these forecasts appear to be efficiently incorporated into forecasts of other economic variables.
    Keywords: Conditional forecast ; Forecast efficiency ; Macroeconomic forecasting
    JEL: C53 C22 E17
    Date: 2019–02–01
  2. By: Oğuzhan Çepni (Central Bank of the Republic of Turkey, Anafartalar Mah. Istiklal Cad. No:10 06050, Ankara, Turkey); Riza Demirer (Department of Economics and Finance, Southern Illinois University Edwardsville, Edwardsville, IL 62026-1102, USA); Rangan Gupta (Department of Economics, University of Pretoria, Pretoria, South Africa); Christian Pierdzioch (Department of Economics, Helmut Schmidt University, Holstenhofweg 85, P.O.B.700822, 22008 Hamburg, Germany)
    Abstract: We show that time-varying risk aversion captures significant predictive information over excess returns on U.S. government bonds even after controlling for a large number of financial and macro factors. Including risk aversion improves the predictive accuracy at all horizons (one- to twelve-months ahead) for shorter maturity bonds and at shorter forecast horizons (one- to three-months ahead) for longer maturity bonds. Given the role of Treasury securities in economic forecasting models and portfolio allocation decisions, our findings have significant implications for investors, policy makers and researchers interested in accurately forecasting return dynamics for these assets.
    Keywords: Bond premia, Predictability, Risk aversion, Out-of-sample forecasts
    JEL: C22 C53 G12 G17
    Date: 2019–01
  3. By: Leopoldo Catania (Aarhus University and CREATES); Tommaso Proietti (CEIS & DEF, University of Rome "Tor Vergata")
    Abstract: The prediction of volatility is of primary importance for business applications in risk management, asset allocation and pricing of derivative instruments. This paper proposes a novel measurement model which takes into consideration the possibly time-varying interaction of realized volatility and asset returns, according to a bivariate model aiming at capturing the main stylised facts: (i) the long memory of the volatility process, (ii) the heavy-tailedness of the returns distribution, and (iii) the negative dependence of volatility and daily market returns. We assess the relevance of "volatility in volatility"and time-varying "leverage" effects in the out-of-sample forecasting performance of the model, and evaluate the density forecasts of the future level of market volatility. The empirical results illustrate that our specification can outperform the benchmark HAR-RV, both in terms of point and density forecasts.
    Keywords: realized volatility, forecasting, leverage effect, volatility in volatility
    Date: 2019–02–06
  4. By: Robert Lehmann; Timo Wolllmershäuser
    Abstract: This paper investigates the macroeconomic projections of the German government since the 1970s and compares it those of the Joint Economic Forecast, which is an in-dependent forecasting institution in Germany. Our results indicate that nominal GDP projections are upward biased for longer forecast horizons, which seems to be driven by a false assessment of the decline in Germany’s trend growth and a systematic failure to correctly anticipate recessions. We show that the German government also deviates from the projections of the Joint Economic Forecast, which in fact worsened the forecast accuracy. Finally, we find evidence that these deviations are driven by political motives.
    Keywords: macroeconomic forecasting, forecast accuracy, independent forecasting, political economic biases
    JEL: E30 E37 E39
    Date: 2019
  5. By: Niels Haldrup (Aarhus University and CREATES); Carsten P. T. Rosenskjold (Aarhus University and CREATES)
    Abstract: The prototypical Lee-Carter mortality model is characterized by a single common time factor that loads differently across age groups. In this paper we propose a factor model for the term structure of mortality where multiple factors are designed to influence the age groups differently via parametric loading functions. We identify four different factors: a factor common for all age groups, factors for infant and adult mortality, and a factor for the "accident hump" that primarily affects mortality of relatively young adults and late teenagers. Since the factors are identified via restrictions on the loading functions, the factors are not designed to be orthogonal but can be dependent and can possibly cointegrate when the factors have unit roots. We suggest two estimation procedures similar to the estimation of the dynamic Nelson-Siegel term structure model. First, a two-step nonlinear least squares procedure based on cross-section regressions together with a separate model to estimate the dynamics of the factors. Second, we suggest a fully specified model estimated by maximum likelihood via the Kalman filter recursions after the model is put on state space form. We demonstrate the methodology for US and French mortality data. We find that the model provides a good fitt of the relevant factors and in a forecast comparison with a range of benchmark models it is found that, especially for longer horizons, variants of the parametric factor model have excellent forecast performance.
    Keywords: Mortality Forecasting, Term Structure of Mortality, Factor Modelling, Cointegration
    JEL: C1 C22 J10 J11 G22
    Date: 2018–01–12
  6. By: Emilio Zanetti Chini (University of Pavia and CREATES)
    Abstract: The Generalized Smooth Transition Auto-Regression (GSTAR) parametrizes the joint asymmetry in the duration and length of cycles in macroeconomic time series by using particular generalizations of the logistic function. The symmetric smooth transition and linear auto-regressions are peculiar cases of the new parametrization. A test for the null hypothesis of dynamic symmetry is discussed. Two case studies indicate that dynamic asymmetry is a key feature of the U.S. economy. Our model beats its competitors in point forecasting, but this superiority becomes less evident in density forecasting and in uncertain forecasting environments.
    Keywords: Density forecasts, Econometric modelling, Evaluating forecasts, Generalized logistic, Industrial production, Nonlinear time series, Point forecasts, Statistical tests, Unemployment
    JEL: C22 C51 C52
    Date: 2018–03–28
  7. By: Knüppel, Malte; Schultefrankenfeld, Guido
    Abstract: Recent research has found that macroeconomic survey forecasts of uncertainty exhibit several deficiencies, such as horizon-dependent biases and lower accuracy than simple unconditional uncertainty forecasts. We examine the inflation uncertainty forecasts from the Bank of England, the Banco Central do Brasil, the Magyar Nemzeti Bank and the Sveriges Riksbank to assess whether central banks' uncertainty forecasts might be subject to similar problems. We find that, while most central banks' uncertainty forecasts also tend to be underconfident at short horizons and overconfident at longer horizons, they are mostly not significantly biased. Moreover, they tend to be at least as precise as unconditional uncertainty forecasts from two different approaches.
    Keywords: Density Forecasts,Fan Charts,Forecast Optimality,Forecast Accuracy
    JEL: C13 C32 C53
    Date: 2018
  8. By: Emilio Zanetti Chini (University of Pavia and CREATES)
    Abstract: I provide general frequentist framework to elicit the forecaster’s expected utility based on a Lagrange Multiplier-type test for the null of locality of the scoring rules associated to the probabilistic forecast. These are assumed to be observed transition variables in a nonlinear autoregressive model to ease the statistical inference. A simulation study reveals that the test behaves consistently with the requirements of the theoretical literature. The locality of the scoring rule is fundamental to set dating algorithms to measure and forecast probability of recession in US business cycle. An investigation of Bank of Norway’s forecasts on output growth leads us to conclude that forecasts are often suboptimal with respect to some simplistic benchmark if forecaster’s reward is not properly evaluated.
    Keywords: Business Cycle, Evaluation, Locality Testing, Nonlinear Time Series, Predictive Density, Scoring Rules, Scoring Structures
    JEL: C12 C22 C44 C53
    Date: 2018–01–02
  9. By: Stéphanie Combes (INSEE Paris - INSEE Paris); Catherine Doz (PJSE - Paris Jourdan Sciences Economiques - UP1 - Université Panthéon-Sorbonne - ENS Paris - École normale supérieure - Paris - INRA - Institut National de la Recherche Agronomique - EHESS - École des hautes études en sciences sociales - ENPC - École des Ponts ParisTech - CNRS - Centre National de la Recherche Scientifique, PSE - Paris School of Economics)
    Abstract: Dynamic Factor Models (DFMs) allow to take advantage of the information provided by a large dataset, which is summarized by a small set of unobservable latent variables, and they have proved to be very useful for short-term forecasting. Since most of their properties rely on the stationarity of the series, these models have been mainly used on data which have been di_erenciated to achieve stationarity. However estimation procedures for DFMs with I(1) common factors have been proposed by Bai (2004) and Bai and Ng(2004). Further, Banerjee and Marcellino (2008) and Banerjee, Marcellino and Masten (2014) have proposed to extend stationary Factor Augmented VAR models to the non-stationary case, and introduced Factor augmented Error Correction Models (FECM). We rely on this approach and conduct a pseudoreal time forecasting experiment, in which we compare short term forecasts of French GDP based on stationary and non-stationary DFMs. We mimic the timeliness of data, and use in the non-stationary framework the 2-step estimator proposed by Doz, Giannone and Reichlin(2011). In our study, forecasts based on stationary or non-stationary DFMs have a similar precision.
    Date: 2018–06
  10. By: Pérez, Fernando (Banco Central de Reserva del Perú)
    Abstract: There exists a large set of leading indicators that are directly related with GDP growth. However, it is often very difficult to select which of these indicators can be used in order to choose the best shortterm forecasting (nowcasting) model. In addition, it may be the case that more than one model can do this job accurately. Therefore, it would be convenient to average these potentially non-nested models. Following Scott and Varian (2015), we estimate a Structural State Space model through Gibbs Sampling and a spike-slab prior in order to perform the Stochastic Search Variable Selection (SSVS) method. Posterior simulations can be used to then compute the inclusion probability of each variable for the whole set of models considered. In-sample GDP estimates are very precise, taking into account the large set of regressors considered for the estimation. Data comes from the BCRPs database plus other additional sources.
    Keywords: Nowcasting, Gibbs Sampling, Variable Selection, Model Averaging
    JEL: E43 E51 E52 E52 E58
    Date: 2018–12
  11. By: Congressional Budget Office
    Abstract: CBO regularly evaluates the quality of its economic forecasts by comparing them with the economy’s actual performance and with the Administration’s forecasts and the Blue Chip consensus, an average of about 50 private-sector forecasts. CBO’s forecasts have been comparable in quality to those of the Administration and the Blue Chip consensus. When CBO’s projections were inaccurate by large margins, the other two forecasters’ projections tended to have similar errors because all forecasters faced the same challenges.
    JEL: C53 H20
    Date: 2017–10–02
  12. By: Guanhao Feng; Jingyu He
    Abstract: We present a Bayesian hierarchical framework for both cross-sectional and time-series return prediction. Our approach builds on a market-timing predictive system that jointly allows for time-varying coefficients driven by fundamental characteristics. With a Bayesian formulation for ensemble learning, we examine the joint predictability as well as portfolio efficiency via predictive distribution. In the empirical analysis of asset-sector allocation, our hierarchical ensemble learning portfolio achieves 500% cumulative returns in the period 1998-2017, and outperforms most workhorse benchmarks as well as the passive investing index. Our Bayesian inference for model selection identifies useful macro predictors (long-term yield, inflation, and stock market variance) and asset characteristics (dividend yield, accrual, and gross profit). Using the selected model for predicting sector evolution, an equally weighted long-short portfolio on winners over losers achieves a 46% Sharpe ratio with a significant Jensen's alpha. Finally, we explore an underexploited connection between classical Bayesian forecasting and modern ensemble learning.
    Date: 2019–02
  13. By: Hakkio, Craig S. (Federal Reserve Bank of Kansas City); Nie, Jun (Federal Reserve Bank of Kansas City)
    Abstract: We construct a monthly measure of foreign economic growth based on a wide range of cross-county indicators. Unlike GDP data, which are normally released with a delay of one to two quarters in most countries, our monthly measure incorporates monthly information up to the current month. As new information arrives, this measure of foreign growth can be updated as frequently as daily. This monthly measure of foreign growth not only helps gauge the economic conditions in other countries but also provides a timely measure of foreign demand to help forecast U.S. export growth.
    JEL: E17 F17
    Date: 2019–12–01

This nep-for issue is ©2019 by Rob J Hyndman. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.