nep-for New Economics Papers
on Forecasting
Issue of 2013‒04‒06
thirteen papers chosen by
Rob J Hyndman
Monash University

  1. The Predictive Role of Stock Market Return for Real Activity in Thailand By Jiranyakul, Komain
  2. Measuring Inaccuracy in Travel Demand Forecasting: Methodological Considerations Regarding Ramp Up and Sampling By Bent Flyvbjerg
  3. Food versus Fuel: Causality and Predictability in Distribution By Andrea Bastianin; Marzio Galeotti; Matteo Manera
  4. Forecasting with Non-spurious Factors in U.S. Macroeconomic Time Series By Yohei Yamamoto
  5. Federal reserve forecasts: asymmetry and state-dependence By Julieta Caunedo; Riccardo DiCecio; Ivana Komunjer; Michael T. Owyang
  6. Analyst Forecast Revisions and Overconfidence By Jean-Sébastien Michel; J. Ari Pandes
  7. A discrete-choice econometrician's tale of monetary policy identification and predictability. By SIRCHENKO, Andrei
  8. Inefficiency in Survey Exchange Rates Forecasts By Francesca Pancotto; Filippo Maria Pericoli; Marco Pistagnesi
  9. Biofuels and Food Prices: Searching for the Causal Link By Andrea Bastianin; Marzio Galeotti; Matteo Manera
  10. Econometric Models for Mixed-Frequency Data. By FORONI, Claudia
  11. Do real-time Okun's law errors predict GDP data revisions? By Michelle L. Barnes; Fabià Gumbau-Brisa; Giovanni P. Olivei
  12. What cost for photovoltaic modules in 2020? Lessons from experience curve models By Arnaud De La Tour; Matthieu Glachant; Yann Ménière
  13. Predicting Bank of England’s Asset Purchase Decisions with MPC Voting Records By Matthias Neuenkirch

  1. By: Jiranyakul, Komain
    Abstract: Stock market return is one of financial variables that contain information to forecast real activity such as industrial production and real GDP growth. However, it is still controversial that stock market return can have a predictive content on real activity. This paper attempts to investigate the ability of stock market return to predict industrial production growth (or real activity) in Thailand, which is an emerging market economy. The standard causality test and the equal forecast evaluation of nested models are employed. For the purpose of forecasting, the data are divided into two periods: the data for the in-sample and the out-of-sample periods. The test of equal forecasting ability is also used. Using monthly data from January 1993 to December 2011, it is found that the model augmented with stock return variable outperforms the benchmark model in the forecast horizon of two months. The results seem to support the notion that stock market return is a predictor of industrial output growth in the short run. Moreover, the standard Granger causality test using the in-sample data also supports this notion. The findings offers a useful insight to investors, financial managers and policymakers on the role of stock market return in forecasting real economic activity. Specifically, a change in stock market return is a signal for revising investment decision by investors and portfolio managers.
    Keywords: Stock return, real activity, emerging market, forecasting, causality
    JEL: C14 C22 E44
    Date: 2012–12
  2. By: Bent Flyvbjerg
    Abstract: Project promoters, forecasters, and managers sometimes object to two things in measuring inaccuracy in travel demand forecasting: (1) using the forecast made at the time of making the decision to build as the basis for measuring inaccuracy and (2) using traffic during the first year of operations as the basis for measurement. This paper presents the case against both objections. First, if one is interested in learning whether decisions about building transport infrastructure are based on reliable information, then it is exactly the traffic forecasted at the time of making the decision to build that is of interest. Second, although ideally studies should take into account so-called demand "ramp up" over a period of years, the empirical evidence and practical considerations do not support this ideal requirement, at least not for large-N studies. Finally, the paper argues that large samples of inaccuracy in travel demand forecasts are likely to be conservatively biased, i.e., accuracy in travel demand forecasts estimated from such samples would likely be higher than accuracy in travel demand forecasts in the project population. This bias must be taken into account when interpreting the results from statistical analyses of inaccuracy in travel demand forecasting.
    Date: 2013–03
  3. By: Andrea Bastianin (University of Milan-Bicocca and FEEM); Marzio Galeotti (University of Milan and IEFE-Bocconi); Matteo Manera (University of Milan-Bicocca and FEEM)
    Abstract: This paper examines the relationship between biofuels and commodity food prices in the U.S. from a new perspective. While a large body of literature has tried to explain the linkages between sample means and volatilities associated with ethanol and agricultural price returns, little is known about their whole distributions. We focus on predictability in distribution by asking whether ethanol returns can be used to forecast different parts of field crops returns distribution, or vice versa. Density forecasts are constructed using Conditional Autoregressive Expectile models estimated with Asymmetric Least Squares. Forecast evaluation relies on quantile-weighed scoring rules, which identify regions of the distribution of interest to the analyst. Results show that both the centre and the left tail of the ethanol returns distribution can be predicted by using field crops returns. On the contrary, there is no evidence that ethanol can be used to forecast any region of the field crops distribution.
    Keywords: Biofuels, Ethanol, Field Crops, Density Forecasting, Granger Causality, Quantiles
    JEL: C22 C53 Q13 Q42 Q47
    Date: 2013–03
  4. By: Yohei Yamamoto
    Abstract: Time instability in factor loadings can induce an overfitting problem in forecasting analyses since the structural change in factor loadings inflates the number of principal components and thus produces spurious factors. This paper proposes an algorithm to estimate non-spurious factors by identifying the set of observations with stable factor loadings based on the recursive procedure suggested by Inoue and Rossi (2011). I found that 51 out of 132 U.S. macroeconomic time series of Stock and Watson (2005) have stable factor loadings. Although crude principal components provide eight or more factors, there are only one or two non-spurious factors. The forecasts using non-spurious factors significantly improve out-of-sample performance.
    Keywords: dynamic factor model, principal components, structural change, spurious factors, out-of-sample forecasts, overfitting
    JEL: C12 C38 E17
    Date: 2013–02
  5. By: Julieta Caunedo; Riccardo DiCecio; Ivana Komunjer; Michael T. Owyang
    Abstract: We jointly test the rationality of the Federal Reserve’s Greenbook forecasts of infiation, unemployment, and output growth using a multivariate nonseparable asymmetric loss function. We find that the forecasts are rationalizable and exhibit directional asymmetry. The degree of asymmetry depends on the phase of the business cycle: The Greenbook forecasts of output growth are too pessimistic in recessions and too optimistic in expansions. The change in monetary policy that occured in the late 1970s has been attributed in the literature to the Fed coming to terms with the difficulties in predicting real variables. Our results offer an alternative explanation: A combination of different preferences over expansions and recessions and less frequent recessions in the latter part of the sample.
    Keywords: Forecasting ; Rational expectations (Economic theory)
    Date: 2013
  6. By: Jean-Sébastien Michel; J. Ari Pandes
    Abstract: We find evidence that supports the notion that analysts who provide extreme forecast revisions are overconfident, especially in assessing the earnings prospects of high information uncertainty firms. We further examine whether analyst overconfidence is associated with stock market performance, and find that a portfolio of extreme forecast revisions underperforms a portfolio of modest forecast revisions in high information uncertainty firms, but not in low information uncertainty firms. Finally, we find that experienced analysts are more overconfident than inexperienced analysts, particularly for high information uncertainty firms, suggesting that analysts do not reduce their overconfidence bias by learning from experience.
    Keywords: Analyst forecast revisions, overconfidence, information uncertainty, forecast accuracy, stock returns
    JEL: G12 G14 G24
    Date: 2013
  7. By: SIRCHENKO, Andrei
    Abstract: This thesis studies the econometric identification and predictability of monetary policy. It addresses the discrete and collective nature of policy decisions, and the use of the real-time versus currently available revised data. The first chapter combines the ordered probit model, novel real-time data set and policy-making meetings as a unit of observation to estimate highly systematic reaction patterns between policy rate decisions and incoming economic data. The paper measures the empirical significance of the rate discreteness and demonstrates that both the discrete-choice approach and real-time "policy-meeting" data do matter in the econometric identification of monetary policy. The estimated rules surpass the market anticipation made one day prior to a policy meeting, both in and out of sample. The second chapter provides empirical evidence that a prompter release of policy- makers’ votes could improve the predictability of policy decisions. The voting patterns reveal strong and robust predictive content even after controlling for policy bias and responses to inflation, real activity, exchange rates and financial market indicators. They contain information not embedded in the spreads and moves in the market interest rates, nor in the explicit forecasts of the next policy decision made by market analysts. Moreover, the direction of policymakers’ dissent explains the direction of analysts’ forecast bias. The third chapter develops a two-stage model for ordinal outcomes (such as discrete changes to the policy interest rates) that are characterized by abundant observations, potentially generated by different processes, in the middle neutral category (no change to the rate). In the context of policy rate setting, the first stage, a policy inclination decision, determines policy stance (loose, neutral or tight) as a reaction to economic conditions, whereas two amount decisions at the second stage are driven mostly by the institutional features. There are three types of zeros: "neutral" zeros, generated directly by the neutral policy stance, and two kinds of "offset" zeros, "loose" and "tight" zeros, generated by the loose or tight stance, offset at the second stage. The model is applied to the individual policymakers’ votes for the interest rate. Both the empirical applications and simulations demonstrate superiority with respect to the conventional models.
    Date: 2012
  8. By: Francesca Pancotto; Filippo Maria Pericoli; Marco Pistagnesi
    Abstract: We use a novel database of a panel of quarterly survey of exchange rates forecasts available on the Bloomberg platform, for the following .ve bilateral exchange rates: EUR/GBP, EUR/JPY, EUR/USD, GBP/USD and USD/JPY, for the timespan ranging from the third quarter 2006 up to the fourth quarter of 2011. We .nd that forecasters are on average irrational, failing to identify the true data generating process of bilateral exchange rates and generally overreacting to past observed information. Moreover, exploring individual performance, we can state that .nancial analysts irrationally do not look at their past forecast errors to improve the quality of their later forecasts
    Keywords: survey forecasts, exchange rates, overreaction;
    JEL: F31
    Date: 2013–03
  9. By: Andrea Bastianin (University of Milan-Bicocca and FEEM); Marzio Galeotti (University of Milan and IEFE-Bocconi); Matteo Manera (University of Milan-Bicocca and FEEM)
    Abstract: We analyze the relationship between the prices of ethanol, agricultural commodities and livestock in Nebraska, the U.S. second largest ethanol producer. The paper focuses on long-run relations and Granger causality linkages between ethanol and the other commodities.The analysis takes possible structural breaks into account and uses a set of techniques that allow to draw inferences about the existence of long-run relations and of short-run in-sample Granger causality and out-of-sample predictive ability. Even after taking breaks into account, evidence that the price of ethanol drives the price dynamics of the other commodities is extremely weak. It is concluded that, on the basis of a formal, comprehensive and rigorous causality analysis we do not find evidence in favour of the Food versus Fuel debate.
    Keywords: Ethanol, Field Crops, Granger Causality, Forecasting, Structural Breaks
    JEL: C22 C53 Q13 Q42 Q47
    Date: 2013–03
  10. By: FORONI, Claudia
    Abstract: This thesis addresses different issues related to the use of mixed-frequency data. In the first chapter, I review, discuss and compare the main approaches proposed so far in the literature to deal with mixed-frequency data, with ragged edges due to publication delays: aggregation, bridge-equations, mixed-data sampling (MIDAS) approach, mixed-frequency VAR and factor models. The second chapter, a joint work with Massimiliano Marcellino, compares the different approaches analyzed in the first chapter, in a detailed empirical application. We focus on now- and forecasting the quarterly growth rate of Euro Area GDP and its components, using a very large set of monthly indicators, with a wide number of forecasting methods, in a pseudo real-time framework. The results highlight the importance of monthly information, especially during the crisis periods. The third chapter, a joint work with Massimiliano Marcellino and Christian Schumacher, studies the performance of a variant of the MIDAS model, which does not resort to functional distributed lag polynomials. We call this approach unrestricted MIDAS (U-MIDAS). We discuss the pros and cons of unrestricted lag polynomials in MIDAS regressions. In Monte Carlo experiments and empirical applications, we compare U-MIDAS to MIDAS and show that U-MIDAS performs better than MIDAS for small differences in sampling frequencies. The fourth chapter, a joint work with Massimiliano Marcellino, focuses on the issues related to mixed-frequency data in structural models. We show analytically, with simulation experiments and with actual data that a mismatch between the time scale of a DSGE or structural VAR model and that of the time series data used for its estimation generally creates identification problems, introduces estimation bias and distorts the results of policy analysis. On the constructive side, we prove that the use of mixed-frequency data can alleviate the temporal aggregation bias, mitigate the identification issues, and yield more reliable policy conclusions.
    Date: 2012
  11. By: Michelle L. Barnes; Fabià Gumbau-Brisa; Giovanni P. Olivei
    Abstract: Using U.S. real-time data, we show that changes in the unemployment rate unexplained by Okun's Law have significant predictive power for GDP data revisions. A positive (negative) error in Okun's Law in real time implies that GDP will be later revised to show less (more) growth than initially estimated by the statistical agency. The information in Okun's Law errors about the true state of real economic activity also helps to improve GDP forecasts in the near term. Our findings add a new dimension to the interpretation of real-time Okun's Law errors, as they show that these errors can convey information other than a change in potential GDP, the equilibrium unemployment rate, or the use of labor's intensive margin.
    Keywords: Unemployment ; Gross domestic product
    Date: 2013
  12. By: Arnaud De La Tour (CERNA - Centre d'économie industrielle - MINES ParisTech - École nationale supérieure des mines de Paris); Matthieu Glachant (CERNA - Centre d'économie industrielle - MINES ParisTech - École nationale supérieure des mines de Paris); Yann Ménière (CERNA - Centre d'économie industrielle - MINES ParisTech - École nationale supérieure des mines de Paris)
    Abstract: Except in few locations, photovoltaic generated electricity remains considerably more expensive than conventional sources. It is however expected that innovation and learning-bydoing will lead to drastic cuts in production cost in the near future. The goal of this paper is to predict the cost of PV modules out to 2020 using experience curve models, and to draw implications about the cost of PV electricity. Using annual data on photovoltaic module prices, cumulative production, R&D knowledge stock and input prices for silicon and silver over the period 1990 - 2011, we identify a experience curve model which minimizes the difference between predicted and actual module prices. This model predicts a 67% decrease of module price from 2011 to 2020. This rate implies that the cost of PV generated electricity will reach that of conventional electricity by 2020 in the sunniest countries with annual solar irradiation of 2000 kWh/year or more, such as California, Italy, and Spain.
    Keywords: Learning curve; solar photovoltaic energy; cost prediction
    Date: 2013–03
  13. By: Matthias Neuenkirch (University of Aachen)
    Abstract: We use MPC voting records to predict changes in the volume of asset purchases. We find, first, that minority voting favoring an increase in the volume of asset purchases raises the probability of an actual increase at the next meeting. Second, minority voting supporting a higher Bank Rate decreases the likelihood of further asset purchases.
    Keywords: Asset Purchases, Bank of England, Monetary Policy, Monetary Policy Committee, Predictability, Voting Records
    JEL: E43 E52 E58
    Date: 2013

This nep-for issue is ©2013 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.