nep-for New Economics Papers
on Forecasting
Issue of 2011‒09‒16
twelve papers chosen by
Rob J Hyndman
Monash University

  1. Economic Forecasting in the Great Recession By Herman O. Stekler; Raj M. Talwar
  2. Forecasting Volatility with Copula-Based Time Series Models By Oleg Sokolinskiy; Dick van Dijk
  3. Forecasting Macroeconomic Variables using Neural Network Models and Three Automated Model Selection Techniques By Anders Bredahl Kock; Timo Teräsvirta
  4. Forecasting performance of three automated modelling techniques during the economic crisis 2007-2009 By Anders Bredahl Kock; Timo Teräsvirta
  5. Can oil prices forecast exchange rates? By Domenico Ferraro; Ken Rogoff; Barbara Rossi
  6. Improving the reliability of real-time Hodrick-Prescott filtering using survey forecasts By Jaqueson K. Galimberti; Marcelo L. Moura
  7. Professional Forecasters: How to Understand and Exploit Them Through a DSGE Model By Luis E. Rojas
  8. Differences in Early GDP Component Estimates Between Recession and Expansions By Tara M. Sinclair; H.O. Stekler
  9. The string prediction models as application to financial forex market By Marian Repasan; Richard Pincak
  10. Improving GDP Measurement: A Forecast Combination Perspective By Boragan Aruoba; Francis X. Diebold; Jeremy Nalewaik; Frank Schorfheide; Dongho Song
  11. Reward Prediction Error in Online Game Trades By Christoph Safferling
  12. Heterogeneous sunspots solutions under learning and replicator dynamics By Michele Berardi

  1. By: Herman O. Stekler (George Washington University); Raj M. Talwar (George Washington University)
    Date: 2011–08
  2. By: Oleg Sokolinskiy (Erasmus University Rotterdam); Dick van Dijk (Erasmus University Rotterdam)
    Abstract: This paper develops a novel approach to modeling and forecasting realized volatility (RV) measures based on copula functions. Copula-based time series models can capture relevant characteristics of volatility such as nonlinear dynamics and long-memory type behavior in a flexible yet parsimonious way. In an empirical application to daily volatility for S&P500 index futures, we find that the copula-based RV (C-RV) model outperforms conventional forecasting approaches for one-day ahead volatility forecasts in terms of accuracy and efficiency. Among the copula specifications considered, the Gumbel C-RV model achieves the best forecast performance, which highlights the importance of asymmetry and upper tail dependence for modeling volatility dynamics. Although we find substantial variation in the copula parameter estimates over time, conditional copulas do not improve the accuracy of volatility forecasts.
    Keywords: Nonlinear dependence; long memory; copulas; volatility forecasting
    JEL: C22 C53
    Date: 2011–09–05
  3. By: Anders Bredahl Kock (Aarhus University and CREATES); Timo Teräsvirta (Aarhus University and CREATES)
    Abstract: In this paper we consider the forecasting performance of a well-defined class of flexible models, the so-called single hidden-layer feedforward neural network models. A major aim of our study is to find out whether they, due to their flexibility, are as useful tools in economic forecasting as some previous studies have indicated. When forecasting with neural network models one faces several problems, all of which influence the accuracy of the forecasts. First, neural networks are often hard to estimate due to their highly nonlinear structure. In fact, their parameters are not even globally identified. Recently, White (2006) presented a solution that amounts to converting the specification and nonlinear estimation problem into a linear model selection and estimation problem. He called this procedure the QuickNet and we shall compare its performance to two other procedures which are built on the linearisation idea: the Marginal Bridge Estimator and Autometrics. Second, one must decide whether forecasting should be carried out recursively or directly. Comparisons of these two methodss exist for linear models and here these comparisons are extended to neural networks. Finally, a nonlinear model such as the neural network model is not appropriate if the data is generated by a linear mechanism. Hence, it might be appropriate to test the null of linearity prior to building a nonlinear model. We investigate whether this kind of pretesting improves the forecast accuracy compared to the case where this is not done.
    Keywords: artificial neural network, forecast comparison, model selection, nonlinear autoregressive model, nonlinear time series, root mean square forecast error, Wilcoxon’s signed-rank test
    JEL: C22 C45 C52 C53
    Date: 2011–08–26
  4. By: Anders Bredahl Kock (Aarhus University and CREATES); Timo Teräsvirta (Aarhus University and CREATES)
    Abstract: In this work we consider forecasting macroeconomic variables dur- ing an economic crisis. The focus is on a specific class of models, the so-called single hidden-layer feedforward autoregressive neural net- work models. What makes these models interesting in the present context is that they form a class of universal approximators and may be expected to work well during exceptional periods such as major economic crises. These models are often difficult to estimate, and we follow the idea of White (2006) to transform the speci?fication and non- linear estimation problem into a linear model selection and estimation problem. To this end we employ three automatic modelling devices. One of them is White's QuickNet, but we also consider Autometrics, well known to time series econometricians, and the Marginal Bridge Estimator, better known to statisticians and microeconometricians. The performance of these three model selectors is compared by look- ing at the accuracy of the forecasts of the estimated neural network models. We apply the neural network model and the three modelling techniques to monthly industrial production and unemployment se- ries of the G7 countries and the four Scandinavian ones, and focus on forecasting during the economic crisis 2007-2009. Forecast accuracy is measured by the root mean square forecast error. Hypothesis testing is also used to compare the performance of the different techniques with each other.
    Keywords: Autometrics, economic forecasting, Marginal Bridge estimator, neural network, nonlinear time series model, Wilcoxon's signed-rank test
    JEL: C22 C45 C52 C53
    Date: 2011–08–26
  5. By: Domenico Ferraro; Ken Rogoff; Barbara Rossi
    Abstract: This paper investigates whether oil prices have a reliable and stable out-of-sample relationship with the Canadian/U.S. dollar nominal exchange rate. Despite state-of-the-art methodologies, the authors find little systematic relation between oil prices and the exchange rate at the monthly and quarterly frequencies. In contrast, the main contribution is to show the existence of a very short-term relationship at the daily frequency, which is rather robust and holds no matter whether the authors use contemporaneous (realized) or lagged oil prices in their regression. However, in the latter case the predictive ability is ephemeral, mostly appearing after instabilities have been appropriately taken into account.
    Keywords: Foreign exchange rates ; Economic forecasting
    Date: 2011
  6. By: Jaqueson K. Galimberti; Marcelo L. Moura
    Abstract: Incorporating survey forecasts to a forecast-augmented Hodrick-Prescott filter, we evidence a considerable improvement to the reliability of US output-gap estimation in realtime. Odds of extracting wrong signals of output-gap estimates are found to reduce by almost a half, and the magnitude of revisions to these estimates accounts to only three fifths of the output-gap average size, usually an one-by-one ratio. We further analyze how this end-of-sample uncertainty evolves as time goes on and observations accumulate, showing that a 90% rate of correct assessments of the output-gap sign can be attained with five quarters of delay using survey forecasts.
    Date: 2011
  7. By: Luis E. Rojas
    Abstract: This paper derives a link between the forecasts of professional forecasters and a DSGE model. I show that the forecasts of a professional forecaster can be incorporated to the state space representation of the model by allowing the measurement error of the forecast and the structural shocks to be correlated. The parameters capturing this correlation are reduced form parameters that allow to address two issues i) How the forecasts of the professional forecaster can be exploited as a source of information for the estimation of the model and ii) How to characterize the deviations of the professional forecaster from an ideal complete information forecaster in terms of the shocks and the structure of the economy.
    Date: 2011–08–15
  8. By: Tara M. Sinclair (George Washington University); H.O. Stekler (George Washington University)
    Abstract: In this paper we examine the quality of the initial estimates of the components of both real and nominal U.S. GDP. We introduce a number of new statistics for measuring the magnitude of changes in the components from the initial estimates available one month after the end of the quarter to the estimates available 3 months after the end of the quarter. We further specifically investigate the potential role of changes in the state of the economy for these changes. Our analysis shows that the early data generally reflected the composition of the changes in GDP that was observed in the later data. Thus, under most circumstances, an analyst could use the early data to obtain a realistic picture of what had happened in the economy in the previous quarter. However, the differences in the composition of the vectors of the two vintages were larger during recessions than in expansions. Unfortunately, it is in those periods when accurate information is most vital for forecasting.
    Keywords: Flash Estimates, Data Revisions, GDP Components, Statistical Tests, Business Cycles
    JEL: C82 E32 C53
    Date: 2011–02
  9. By: Marian Repasan; Richard Pincak
    Abstract: In this paper we apply a new approach of the string theory to the real financial market. The strings are defined here by the boundary conditions, characteristic length, real values and the method of redistribution of information. The map represents the detrending and data standardization procedure. We used 1-end-point, 2-end-point open string and partially compactified strings that satisfy the Dirichlet and Neumann boundary conditions. We established two different models to predict the behavior of financial forex market. The first model is based on the correlation function as invariant and the second one is an application based on the deviations from the closed string/pattern form. We found the difference between these two approaches. The first model cannot predict the behavior of the forex market with good efficiency in comparison with the second one which is, in addition, able to make relevant profit per year.
    Date: 2011–09
  10. By: Boragan Aruoba (Department of Economics, University of Maryland); Francis X. Diebold (Department of Economics, University of Maryland); Jeremy Nalewaik (Federal Reserve Board, Washington D.C.); Frank Schorfheide (Department of Economics, University of Pennsylvania); Dongho Song (Department of Economics, University of Pennsylvania)
    Abstract: Two often-divergent U.S. GDP estimates are available, a widely-used expenditure side version, GDPE, and a much less widely-used income-side version, GDPI . We propose and explore a "forecast combination" approach to combining them. We then put the theory to work, producing a superior combined estimate of GDP growth for the U.S., GDPC. We compare GDPC to GDPE and GDPI, with particular attention to behavior over the business cycle. We discuss several variations and extensions.
    Keywords: National Income and Product Accounts, Output, Expenditure, Economic Activity, Business Cycle, Recession
    JEL: E01 E32
    Date: 2011–09–06
  11. By: Christoph Safferling (Institut für Betriebswirtschaftslehre, Universität Wien, Austria)
    Abstract: We use trade data from an online game economy to test the 'dopaminergic reward prediction error' (DRPE) hypothesis: upon buying a game item at a price which is obviously too low, a player should become more active in the trading market. We find that players are more willing to buy goods in the in-game market after such an trade incident. Hence, the effect predicted by the DRPE model is visible. Yet, contrary to the prediction of DRPE, the magnitude of the prediction error does not have any effect on the post-error trading activity.
    JEL: C99 D01 D12 D87
    Date: 2011–09–02
  12. By: Michele Berardi
    Abstract: In a linear stochastic forward-looking univariate model with predetermined variables, we consider the possibility of heterogeneous equilibria with sunspots emerging endogenously through adaptive learning and replicator dynamics. In particular, we investigate equilibria where only a fraction of agents in the economy condition their forecasts on a sunspot, and equilibria where di¤erent groups of agents use di¤erent sunspots. We comclude that, although such heterogeneous equilibria exist and can be stable under adaptive learning, they do no survive under endogenous replicator dynamics. Moreover, we show that even homogeneous sunspot equilibria require some degree of coordinations among agents for them to emerge in an economy. We conclude that heterogeneous equilibria with sunspots are fragile under endogenous selection of predictors by agents, and that even the relevance of homogeneous sunspot equilibria is questioned once agents are allowed to doubt about the importance of sunspots in their forecasts.
    Date: 2011

This nep-for issue is ©2011 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.