nep-for New Economics Papers
on Forecasting
Issue of 2011‒08‒29
twelve papers chosen by
Rob J Hyndman
Monash University

  1. Out-of-Sample Forecast Tests Robust to the Choice of Window Size By Inoue, Atsushi; Rossi, Barbara
  2. Forecasting the Yield Curve for the Euro Region By Benjamin M. Tabak; Daniel O. Cajueiro; Alexandre B. Sollaci
  3. Evaluating the forecasting performance of commodity futures prices By Trevor A. Reeve; Robert J. Vigfusson
  4. Nowcasting Irish GDP By D'Agostino, Antonello; McQuinn, Kieran; O'Brien, Derry
  5. Stock market firm-level information and real economic activity By Filippo di Mauro; Fabio Fornari; Dario Mannucci
  6. Are some forecasters really better than others? By D'Agostino, Antonello; McQuinn, Kieran; Whelan, Karl
  7. Mapping the State of Financial Stability By Sarlin, Peter; Peltonen, Tuomas A.
  8. A Dynamic Factor Model for World Trade Growth By Stéphanie Guichard; Elena Rusticelli
  9. The "CAPS" Prediction System and Stock Market Returns By Christopher Avery; Judith A. Chevalier; Richard J. Zeckhauser
  11. Risky Curves: From Unobservable Utility to Observable Opportunity Sets By Daniel Friedman; Shyam Sunder
  12. Predicting Peaks and Troughs in Real House Prices By Linda Rousová; Paul van den Noord

  1. By: Inoue, Atsushi; Rossi, Barbara
    Abstract: This paper proposes new methodologies for evaluating out-of-sample forecasting performance that are robust to the choice of the estimation window size. The methodologies involve evaluating the predictive ability of forecasting models over a wide range of window sizes. We show that the tests proposed in the literature may lack the power to detect predictive ability and might be subject to data snooping across different window sizes if used repeatedly. An empirical application shows the usefulness of the methodologies for evaluating exchange rate models' forecasting ability.
    Keywords: estimation window; forecast evaluation; predictive ability testing
    JEL: C22 C52 C53
    Date: 2011–08
  2. By: Benjamin M. Tabak; Daniel O. Cajueiro; Alexandre B. Sollaci
    Abstract: This paper compares the forecast precision of the Functional Signal plus Noise (FSN), the Dynamic Nelson-Siegel (DL), and a random walk model. The empirical results suggest that both outperform the random walk at short horizons (one-month) and that the the FSN model outperforms the DL at the one-month forecasting horizon. The conclusions provided in this paper are important for policy makers, fixed income portfolio managers, financial institutions and academics.
    Date: 2011–08
  3. By: Trevor A. Reeve; Robert J. Vigfusson
    Abstract: Commodity futures prices are frequently criticized as being uninformative for forecasting purposes because (1) they seem to do no better than a random walk or an extrapolation of recent trends and (2) futures prices for commodities often trace out a relatively flat trajectory even though global demand is steadily increasing. In this paper, we attempt to shed light on these concerns by discussing the theoretical relationship between spot and futures prices for commodities and by evaluating the empirical forecasting performance of futures prices relative to some alternative benchmarks. The key results of our analysis are that futures prices have generally outperformed a random walk forecast, but not by a large margin, while both futures and a random walk noticeably outperform a simple extrapolation of recent trends (a random walk with drift). Importantly, however, futures prices, on average, outperform a random walk by a considerable margin when there is a sizeable difference between spot and futures prices.
    Keywords: Commodity futures ; Futures market ; Prices ; Economic forecasting
    Date: 2011
  4. By: D'Agostino, Antonello; McQuinn, Kieran; O'Brien, Derry
    Abstract: In this paper we present a dynamic factor model that produces nowcasts and backcasts of Irish quarterly GDP using timely data from a panel dataset of 35 indicators. We apply a recently developed methodology, whereby numerous potentially useful indicator series for Irish GDP can be availed of in a parsimonious manner and the unsynchronized nature of the release calendar for a wide range of higher frequency indicators can be handled. The nowcasts in this paper are generated by using dynamic factor analysis to extract common factors from the panel dataset. Bridge equations are then used to relate these factors to quarterly GDP estimates. We conduct an out-of-sample forecasting simulation exercise, where the performance of the factor model is compared with that of a standard benchmark model.
    Keywords: GDP; Forecasting; Factors
    JEL: C53 E52 C33
    Date: 2011
  5. By: Filippo di Mauro (European Central Bank, Kaiserstrasse 29, D-60311 Frankfurt am Main, Germany.); Fabio Fornari (European Central Bank, Kaiserstrasse 29, D-60311 Frankfurt am Main, Germany.); Dario Mannucci (Prometeia, Via G. Marconi 43, 40122 Bologna, Italy.)
    Abstract: We provide evidence that changes in the equity price and volatility of individual firms (measures that approximate the definition of 'granular shock' given in Gabaix, 2010) are key to improve the predictability of aggregate business cycle fluctuations in a number of countries. Specifically, adding the return and the volatility of firm-level equity prices to aggregate financial information leads to a significant improvement in forecasting business cycle developments in four economic areas, at various horizons. Importantly, not only domestic firms but also foreign firms improve business cycle predictability for a given economic area. This is not immediately visible when one takes an unconditional standpoint (i.e. an average across the sample). However, conditioning on the business cycle position of the domestic economy, the relative importance of the two sets of firms - foreign and domestic - exhibits noticeable swings across time. Analogously, the sectoral classification of the firms that in a given month retain the highest predictive power for future IP changes also varies significantly over time as a function of the business cycle position of the domestic economy. Limited to the United States, predictive ability is found to be related to selected balance sheet items, suggesting that structural features differentiate the firms that can anticipate aggregate fluctuations from those that do not help to this aim. Beyond the purely forecasting application, this finding may enhance our understanding of the underlying origins of aggregate fluctuations. We also propose to use the cross sectional stock market information to macro-prudential aims through an economic Value at Risk. JEL Classification: C53, C58, F37, G15.
    Keywords: Business cycle forecasting, granular shock, international linkages.
    Date: 2011–08
  6. By: D'Agostino, Antonello; McQuinn, Kieran; Whelan, Karl
    Abstract: In any dataset with individual forecasts of economic variables, some forecasters will perform better than others. However, it is possible that these ex post differences reflect sampling variation and thus overstate the ex ante differences between forecasters. In this paper, we present a simple test of the null hypothesis that all forecasters in the US Survey of Professional Forecasters have equal ability. We construct a test statistic that reflects both the relative and absolute performance of the forecaster and use bootstrap techniques to compare the empirical results with the equivalents obtained under the null hypothesis of equal forecaster ability. Results suggest little support for the idea that the best forecasters are actually innately better than others, though there is evidence that a relatively small group of forecasters perform very poorly.
    Keywords: Forecasting; Bootstrap
    JEL: C53 E27 E37
    Date: 2011
  7. By: Sarlin, Peter (BOFIT); Peltonen, Tuomas A. (BOFIT)
    Abstract: The paper uses the Self-Organizing Map for mapping the state of financial stability and visualizing the sources of systemic risks on a two-dimensional plane as well as for predicting systemic financial crises. The Self-Organizing Financial Stability Map (SOFSM) enables a two-dimensional representation of a multidimensional financial stability space and thus allows disentangling the individual sources impacting on systemic risks. The SOFSM can be used to monitor macro-financial vulnerabilities by locating a country in the financial stability cycle: being it either in the pre-crisis, crisis, post-crisis or tranquil state. In addition, the SOFSM performs better than or equally well as a logit model in classifying in-sample data and predicting out-of-sample the global financial crisis that started in 2007. Model robustness is tested by varying the thresholds of the models, the policymaker’s preferences, and the forecasting horizon.
    Keywords: systemic financial crisis; systemic risk; self-organizing maps; visualisation; prediction; macroprudential supervision
    JEL: E44 E58 F01 F37
    Date: 2011–08–22
  8. By: Stéphanie Guichard; Elena Rusticelli
    Abstract: This paper reviews the main monthly indicators that could help forecasting world trade and compares different type of forecasting models using these indicators. In particular it develops dynamic factor models (DFM) which have the advantage of handling larger datasets of information than bridge models and allowing for the inclusion of numerous monthly indicators on a national and world-wide level such as financial indicators, transportation and shipping indices, supply and orders variables and information technology indices. The comparison of the forecasting performance of the DFMs with more traditional bridge equation models as well as autoregressive benchmarking models shows that, the dynamic factor approach seems to perform better, especially when a large set of indicators is used, but also that the marginal gains in adding indicators seems to diminish after a certain stage.<P>Un modèle à facteurs dynamiques pour prévoir la croissance du commerce mondial<BR>Ce document passe en revue les principaux indicateurs mensuels pouvant aider á prévoir le commerce mondial et compare différents types de modèles de prévision utilisant ces indicateurs. En particulier, il développe des modèles á facteurs dynamiques (DFM) qui ont l'avantage de permettre l’utilisation de plus de séries que les modèles d’étalonnage et donc d’inclure des indicateurs mensuels au niveau national et mondial tels que les indicateurs financiers, de transport et d’expédition, d’approvisionnement et de carnets de commandes ou encore et de technologie de l’information. La comparaison de la performance de prévision des DFM avec des modèles d’étalonnage plus traditionnels ou des modèles autoregressifs montre que l'approche en facteurs dynamiques semble plus performante, surtout quand un vaste ensemble d'indicateurs est utilisé ; les gains marginaux en ajoutant des indicateurs semblent toutefois diminuer après un certain stade.
    Keywords: forecasting, world trade, dynamic factor models, bridge models, prévisions, Commerce mondial, modèles á facteurs dynamiques, modèle d’étalonnage
    JEL: C53 E37 F17 F47
    Date: 2011–05–31
  9. By: Christopher Avery; Judith A. Chevalier; Richard J. Zeckhauser
    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.
    JEL: G12 G14
    Date: 2011–08
  10. By: Albu, Lucian-Liviu (Romanian Academy, INSTITUTE FOR ECONOMIC FORECASTING)
    Abstract: There are evidences that the actual global crisis affected the convergence process in EU. Generally, just new adhered countries were more affected by the actual crisis. Today all forecasts are suffering by uncertainty. Last time, economists, with their methods and models, are invoked for actual crisis, but real causes could be found in policymaker’s actions and in public opinion’s influence, as a rule focussed only on short or very short-term. However, the methods and predictive models that economists are using seem to be not adequate for the new situation, especially due to the large extension of the actual crisis, at its turn caused by the globalization phenomenon. For the post-crisis period, we are presenting some likely challenges both in the real economy and in economics. There are different opinions regarding how deep and how long the convergence process will be affected. Synthetically, the pessimistic authors are viewing the future economic dynamics as one of so-called L type or U type or W type. Coming from lessons supplied by standard economic growth theories (Ramsey model, Solow-Swan model, Mankiw, Romer, and Weil model, etc.) and by empirical evidences, we are considering the convergence in the level of income per capita as a result of structural changes in economy.
    Keywords: convergence, structural changes, spatial distribution, inflation - interest rate - growth rate correlation
    JEL: A11 B41 C13 C31 E27 O11 O47 O52
    Date: 2011–08
  11. By: Daniel Friedman (Dept. Economics, UC Santa Cruz; CESifo); Shyam Sunder (Yale School of Management)
    Abstract: Most theories of risky choice postulate that a decision maker maximizes the expectation of a Bernoulli (or utility or similar) function. We tour 60 years of empirical search and conclude that no such functions have yet been found that are useful for out-of-sample prediction. Nor do we find practical applications of Bernoulli functions in major risk-based industries such as finance, insurance and gambling. We sketch an alternative approach to modeling risky choice that focuses on potentially observable opportunities rather than on unobservable Bernoulli functions.
    Keywords: Expected utility, Risk aversion, St. Petersburg Paradox, Decisions under uncertainty, Option theory
    JEL: C91 C93 D11 D81 G11 G12 G22 L83
    Date: 2011–08
  12. By: Linda Rousová; Paul van den Noord
    Abstract: OECD work prior to the financial crisis suggested that real prices in several housing markets had become vulnerable to a change in financial and economic conditions, with the risk of a subsequent downturn becoming increasingly possible, as proved to be the case. With corrections in many, but not all, housing markets having now occurred, and, in some countries, prices having rebounded rapidly in the low interest rate environment, the issue of whether prices are now close to another turning point is again of considerable policy interest. As a means of addressing this issue, probit models have been estimated to provide an indication of possible peaks and troughs in real house prices in 2011 and 2012, using data for 20 OECD countries. Predictions based on these models have been reported in OECD Economic Outlook, No. 89 and this paper provides information on the methodology underpinning these predictions.<P>Comment prévoir les fluctuations des prix réels des logements ?<BR>Les travaux menés par l’OCDE avant la crise financière laissaient entendre que les prix réels sur plusieurs marchés du logement ne sauraient résister à une modification des conditions financières et économiques, alors que le risque que leur vulnérabilité n’entraîne une crise économique devenait de plus en plus probable, ainsi que les événements ultérieurs l’ont d’ailleurs démontré. Les prix s’étant désormais rétablis sur un grand nombre, bien que pas sur la totalité, des marchés du logement, et ayant même, dans certains pays, enregistré une remontée rapide favorisée par la faiblesse des taux d’intérêt, la question de savoir s’ils font aujourd’hui face à un nouveau changement de cap mobilise fortement l’intérêt des gouvernements. Pour tenter de répondre à cette question, des modèles probits, dont on estime qu’ils pourraient fournir une indication des fluctuations possibles des prix réels des logements en 2011 et 2012, ont été sollicités et utilisés avec des données concernant 20 pays de l’OCDE. Les prévisions établies sur la base de ces modèles ont été reproduites dans les Perspectives économiques de l'OCDE, n°89, et le présent document donne des informations sur la méthodologie employée à cet effet.
    Keywords: house prices, business cycles, housing bubbles, cycle économique, prix des logements, bulles immobilière
    JEL: E32 F42 R31
    Date: 2011–07–13

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.