nep-for New Economics Papers
on Forecasting
Issue of 2012‒05‒29
thirteen papers chosen by
Rob J Hyndman
Monash University

  1. Do Oil Prices Help Forecast U.S. Real GDP? The Role of Nonlinearities and Asymmetries By Kilian, Lutz; Vigfusson, Robert J.
  2. Financial Markets Forecasts Revisited: Are they Rational, Herding or Bold? By Ippei Fujiwara; Hibiki Ichiue; Yoshiyuki Nakazono; Yosuke Shigemi
  3. Improving Bayesian VAR density forecasts through autoregressive Wishart Stochastic Volatility By Karapanagiotidis, Paul
  4. The "CAPS" Prediction System and Stock Market Returns By Avery, Christopher N.; Chevalier, Judith; Zeckhauser, Richard Jay
  5. Estimating and Forecasting APARCH-Skew-t Models by Wavelet Support Vector Machines By Li, Yushu
  6. A Solution to Overoptimistic Forecasts and Fiscal Procyclicality: The Structural Budget Institutions Pioneered by Chile By Frankel, Jeffrey A.
  7. Household optimism and borrowing By Hyytinen, Ari; Putkuri, Hanna
  8. History Repeating: Spain Beats Germany in the EURO 2012 Final By Achim Zeileis; Christoph Leitner; Kurt Hornik
  9. Euro Area Money Demand and International Portfolio Allocation: A Contribution to Assessing Risks to Price Stability By De Santis, Roberto A; Favero, Carlo A.; Roffia, Barbara
  10. Predicting rating changes for banks: How accurate are accounting and stock market indicators? By Distinguin, Isabelle; Hasan, Iftekhar; Tarazi , Amine
  11. Implications of Wealth Heterogeneity For Macroeconomics By Christopher D. Carroll
  12. Supply Function Prediction in Electricity Auctions By Matteo Pelagatti
  13. Local Distance-Based Generalized Linear Models using the dbstats package for R By Eva Boj; Pedro Delicado; Josep Fortiana; Anna Esteve; Adria Caballe

  1. By: Kilian, Lutz; Vigfusson, Robert J.
    Abstract: There is a long tradition of using oil prices to forecast U.S. real GDP. It has been suggested that the predictive relationship between the price of oil and one-quarter ahead U.S. real GDP is nonlinear in that (1) oil price increases matter only to the extent that they exceed the maximum oil price in recent years and that (2) oil price decreases do not matter at all. We examine, first, whether the evidence of in-sample predictability in support of this view extends to out-of-sample forecasts. Second, we discuss how to extend this forecasting approach to higher horizons. Third, we compare the resulting class of nonlinear models to alternative economically plausible nonlinear specifications and examine which aspect of the model is most useful for forecasting. We show that the asymmetry embodied in commonly used nonlinear transformations of the price of oil is not helpful for out-of-sample forecasting; more robust and more accurate real GDP forecasts are obtained from symmetric nonlinear models based on the three-year net oil price change. Finally, we quantify the extent to which the 2008 recession could have been forecast using the latter class of time-varying threshold models.
    Keywords: Asymmetry; Nonlinearity; Oil price; Out-of-sample forecast; Real GDP
    JEL: C32 C53 Q43
    Date: 2012–05
  2. By: Ippei Fujiwara (Associate Professor, Australian National University, CAMA, EABCN, and Globalization and Monetary Policy Institute (FRB Dallas) (E-mail:; Hibiki Ichiue (Director, Monetary Affairs Department, Bank of Japan (E-mail:; Yoshiyuki Nakazono (Waseda University and Research Fellow of the Japan Society for the Promotion of Science (Email:; Yosuke Shigemi (Director, Institute for Monetary and Economic Studies, (currently Information System Services Department), Bank of Japan (E-mail:
    Abstract: We test whether professional forecasters forecast rationally or behaviorally using a unique database, QSS Database, which is the monthly panel of forecasts on Japanese stock prices and bond yields. The estimation results show that (i) professional forecasts are behavioral, namely, significantly influenced by past forecasts, (ii) there exists a stock-bond dissonance: while forecasting behavior in the stock market seems to be herding, that in the bond market seems to be bold in the sense that their current forecasts tend to be negatively related to past forecasts, and (iii) the dissonance is due, at least partially, to the individual forecastersf behavior that is influenced by their own past forecasts rather than othersf. Even in the same country, forecasting behavior is quite different by market.
    Keywords: Anchoring, Bold, Herding, Survey Forecasts
    JEL: D03 G17
    Date: 2012–05
  3. By: Karapanagiotidis, Paul
    Abstract: Dramatic changes in macroeconomic time series volatility pose a challenge to contemporary vector autoregressive (VAR) forecasting models. Traditionally, the conditional volatility of such models had been assumed constant over time or allowed for breaks across long time periods. More recent work, however, has improved forecasts by allowing the conditional volatility to be completely time variant by specifying the VAR innovation variance as a distinct discrete time process. For example, Clark (2011) specifies the volatility process as an independent log random walk for each time series in the VAR. Unfortunately, there is no empirical reason to believe that the VAR innovation volatility process of macroeconomic growth series follow log random walks, nor that the volatility of each series is independent of the others. This suggests that a more robust specification on the volatility process—one that both accounts for co-persistence in conditional volatility across time series and exhibits mean reverting behaviour—should improve density forecasts, especially over the long run forecasting horizon. In this respect, I employ a latent Inverse-Wishart autoregressive stochastic volatility specification on the conditional variance equation of a Bayesian VAR, with U.S. macroeconomic time series data, in evaluating Bayesian forecast efficiency against a competing log random walk specification by Clark (2011).
    Keywords: InverseWishart distribution; stochastic volatility; predictive likelihoods; MCMC; macroeconomic time series; density forecasts; vector autoregression; steady state priors; Bayesian econometrics;
    JEL: C32 C53 E17 C11
    Date: 2012–03–10
  4. By: Avery, Christopher N.; Chevalier, Judith; Zeckhauser, Richard Jay
    Abstract: We study the predictive power of approximately 2.5 million stock picks submitted by individual users to the “CAPS†website run by the Motley Fool company ( These picks prove to be surprisingly informative about future stock prices. Indeed, a strategy of shorting stocks with a disproportionate number of negative picks on the site and buying stocks with a disproportionate number of positive picks produces a return of over nine percent per annum over the sample period. These results are mostly driven by the fact that negative picks on the site strongly predict future stock price declines; positive picks on the site produce returns that are statistically indistinguishable from the market. A Fama French decomposition suggests that these results are largely due to stock-picking rather than style factors or market timing.
    Date: 2011
  5. By: Li, Yushu (Department of Economics, Lund University)
    Abstract: This paper concentrates on comparing estimation and forecasting ability of Quasi-Maximum Likelihood (QML) and Support Vector Machines (SVM) for financial data. The financial series are fitted into a family of Asymmetric Power ARCH (APARCH) models. As the skewness and kurtosis are common characteristics of the financial series, a skew t distributed innovation is assumed to model the fat tail and asymmetry. Prior research indicates that the QML estimator for the APARCH model is inefficient when the data distribution shows departure from normality, so the current paper utilizes the nonparametric-based SVM method and shows that it is more efficient than the QML under the skewed Student’s t-distributed error. As the SVM is a kernel-based technique, we further investigate its performance by applying a Gaussian kernel and a wavelet kernel. The wavelet kernel is chosen due to its ability to capture the localized volatility clustering in the APGARCH model. The results are evaluated by a Monte Carlo experiment, with accuracy measured by Normalized Mean Square Error ( NMSE ). The results suggest that the SVM based method generally performs better than QML, with a consistently lower NMSE for both in sample and out of sample data. The outcomes also highlight the fact that the wavelet kernel outperforms the Gaussian kernel with a lower NMSE , is more computation efficient and has better generation capability.
    Keywords: SVM; APARCH; Wavelet Kernel; Monte Carlo Experiment
    JEL: C14 C53 C61
    Date: 2012–05–21
  6. By: Frankel, Jeffrey A.
    Abstract: Historically, many countries have suffered a pattern of procyclical fiscal policy: spending too much in booms and then forced to cut back in recessions, thereby exacerbating the business cycle. This problem has especially plagued Latin American commodity-producers. Since 2000, fiscal policy in Chile has been governed by a structural budget rule that has succeeded in implementing countercyclical fiscal policy. The key innovation is that the two most important estimates of the structural versus cyclical components of the budget – trend output and the 10-year price of copper – are made by expert panels and thus insulated from the political process. Chile’s fiscal institutions could usefully be emulated everywhere, but especially in other commodity-exporting countries. This paper finds statistical support for a series of hypotheses regarding forecasts by official agencies that have responsibility for formulating the budget. 1) Official forecasts of budgets and GDP in a 33-country sample are overly optimistic on average. 2) The bias toward over-optimism is stronger the longer the horizon 3) The bias is greater among European governments that are politically subject to the budget rules in the Stability and Growth Pact (SGP). 4) The bias is greater at the extremes of the business cycle, particularly in booms. 5) In most countries, the real growth rate is the key macroeconomic input for budget forecasting. In Chile it is the price of copper. 6) Real copper prices mean-revert in the long run, but this is not always readily perceived. 7) Chile’s official forecasts are not overly optimistic on average. 8) Chile has apparently avoided the problem of official forecasts that unrealistically extrapolate in boom times. The conclusion: official forecasts, if not insulated from politics, tend to be overly optimistic, and the problem can be worse when the government is formally subject to budget rules. The key innovation that has allowed Chile in general to achieve countercyclical fiscal policy, and in particular to run surpluses in booms, is not just a structural budget rule in itself, but a regime that entrusts to panels of independent experts the responsibility for estimating the extent to which contemporaneous copper prices and GDP have departed from their long-run trends.
    Date: 2011
  7. By: Hyytinen, Ari (University of Jyväskylä and Yrjö Jahnsson Foundation); Putkuri, Hanna (Bank of Finland)
    Abstract: A unique Finnish household-level data from 1994 to 2009 allow us to measure how households’ financial expectations are related to the subsequent outcomes. We use the difference between the two to measure forecast errors and household optimism and link the errors to households’ borrowing behaviour. We find that households making greatest optimistic forecast errors carry greater levels of debt and are most likely to suffer from excessive debt loads (overindebtedness). They also are less attentive to forecast errors than their pessimistic counterparts when forming their expectations for a subsequent period.
    Keywords: forecast errors; ex ante optimism; borrowing
    JEL: D21 L20
    Date: 2012–05–08
  8. By: Achim Zeileis; Christoph Leitner; Kurt Hornik
    Abstract: Four years after the last European football championship (EURO) in Austria and Switzerland, the two finalists of the EURO 2008 - Spain and Germany - are again the clear favorites for the EURO 2012 in Poland and the Ukraine. Using a bookmaker consensus rating - obtained by aggregating winning odds from 23 online bookmakers - the forecast winning probability for Spain is 25.8% followed by Germany with 22.2%, while all other competitors have much lower winning probabilities (The Netherlands are in third place with a predicted 11.3%). Furthermore, by complementing the bookmaker consensus results with simulations of the whole tournament, we can infer that the probability for a rematch between Spain and Germany in the final is 8.9% with the odds just slightly in favor of Spain for prevailing again in such a final (with a winning probability of 52.9%). Thus, one can conclude that - based on bookmakers' expectations - it seems most likely that history repeats itself and Spain defends its European championship title against Germany. However, this outcome is by no means certain and many other courses of the tournament are not unlikely as will be presented here. All forecasts are the result of an aggregation of quoted winning odds for each team in the EURO 2012: These are first adjusted for profit margins ("overrounds"), averaged on the log-odds scale, and then transformed back to winning probabilities. Moreover, team abilities (or strengths) are approximated by an "inverse" procedure of tournament simulations, yielding estimates of all pairwise probabilities (for matches between each pair of teams) as well as probabilities to proceed to the various stages of the tournament. This technique correctly predicted the EURO 2008 final (Leitner, Zeileis, Hornik 2008), with better results than other rating/forecast methods (Leitner, Zeileis, Hornik 2010a), and correctly predicted Spain as the 2010 FIFA World Champion (Leitner, Zeileis, Hornik 2010b). Compared to the EURO 2008 forecasts, there are many parallels but two notable differences: First, the gap between Spain/Germany and all remaining teams is much larger. Second, the odds for the predicted final were slightly in favor of Germany in 2008 whereas this year the situation is reversed.
    Keywords: consensus, agreement, bookmakers odds, sports tournaments, EURO 2012
    JEL: C53 C40 D84
    Date: 2012–05
  9. By: De Santis, Roberto A; Favero, Carlo A.; Roffia, Barbara
    Abstract: This paper argues that a stable broad money demand for the euro area over the period 1980-2011 can be obtained by modelling cross border international portfolio allocation. As a consequence, model-based excess liquidity measures, namely the difference between actual M3 growth (net of the inflation objective) and the expected money demand trend dynamics, can be useful to predict HICP inflation.
    Keywords: Euro area money demand; inflation forecasts; monetary policy; portfolio allocation
    JEL: E4 E44
    Date: 2012–05
  10. By: Distinguin, Isabelle (Université de Limoges, LAPE); Hasan, Iftekhar (Fordham University and Bank of Finland); Tarazi , Amine (Université de Limoges, LAPE)
    Abstract: We aim to assess how accurately accounting and stock market indicators predict rating changes for Asian banks. We conduct a stepwise process to determine the optimal set of early indicators by tracing upgrades and downgrades from rating agencies, as well as other relevant factors. Our results indicate that both accounting and market indicators are useful leading indicators but are more effective in predicting upgrades than downgrades, especially for large banks. Moreover, early indicators are only significant in predicting rating changes for banks that are more focused on traditional banking activities such as deposit and loan activities. Finally, a higher reliance of banks on subordinated debt is associated with better accuracy of early indicators.
    Keywords: bank failure; bank risk; ratings; emerging market
    JEL: G21 G28
    Date: 2012–04–12
  11. By: Christopher D. Carroll
    Abstract: Today’s dominant strain of macroeconomic models supposes that aggregate consumption can be understood by assuming the existence of a ‘representative agent’ whose behavior rationalizes observed outcomes. But representative agent models yield embarrassingly implausible (and empirically inaccurate) descriptions of consumption behavior. When push comes to shove, real-world forecasters (including those at the Fed) properly disregard these implications. As a result, consumption forecasting remains very much a seat-of-the-pants enterprise. I will argue that if the representative agent assumption is replaced with a model that generates wealth heterogeneity that matches the empirical data, the improved model can provide a sensible analysis of economic questions like "What might the consumption response be to economic stimulus payments?"
    Date: 2012–05
  12. By: Matteo Pelagatti (Dipartimento di Statistica, Università degli Studi di Milano-Bicocca)
    Abstract: In the fast growing literature that addresses the problem of the optimal bidding behaviour of power generation companies that sell energy in electricity auctions it is always assumed that every firm knows the aggregate supply function of its competitors. Since this information is generally not available, real data have to be substituted by predictions. In this paper we propose two alternative approaches to the problem and apply them to the hourly prediction of the aggregate supply function of the competitors of the main Italian generation company.
    Keywords: electricity auctions, functional prediction, reduced rank regression
    Date: 2012–03–01
  13. By: Eva Boj (Dept. de Matematica Economica, Financera i Actuarial, Univ. de Barcelona, Diagonal 690, 08034 Barcelona, Spain.); Pedro Delicado (Universitat Politecnica de Catalunya); Josep Fortiana (Universitat de Barcelona); Anna Esteve (CEEISCAT); Adria Caballe (Universitat Politecnica de Catalunya)
    Abstract: This paper introduces local distance-based generalized linear models. These models extend (weighted) distance-based linear models firstly with the generalized linear model concept, then by localizing. Distances between individuals are the only predictor information needed to fit these models. Therefore they are applicable to mixed (qualitative and quantitative) explanatory variables or when the regressor is of functional type. Models can be fitted and analysed with the R package dbstats, which implements several distancebased prediction methods.
    Keywords: Distance-based prediction, Generalized Linear Model, Local Likelihood, Iteratively Weighted Least Squares, R
    Date: 2012–05

This nep-for issue is ©2012 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.