nep-for New Economics Papers
on Forecasting
Issue of 2015‒11‒07
fourteen papers chosen by
Rob J Hyndman
Monash University

  1. Forecasting Unemployment across Countries: the Ins and Outs By Barnichon, Régis; Garda, Paula
  2. Forecasting employment in Europe: Are survey results helpful? By Lehmann, Robert; Weyh, Antje
  3. Financialization of Agricultural Commodity Markets: Do Financial Data Help to Forecast Agricultural Prices By Etienne, Xiaoli
  4. Predicting Recessions in Germany With Boosted Regression Trees By Jörg Döpke; Ulrich Fritsche; Christian Pierdzioch
  5. Using low frequency information for predicting high frequency variables By Claudia Foroni; Pierre Guérin; Massimiliano Marcellino
  6. With string model to time series forecasting By Richard Pin\v{c}\'ak; Erik Barto\v{s}
  7. Robust stress testing By Bidder, Rhys; McKenna, Andrew
  8. The Robustness of Tests for Consumer Choice Inconsistencies By Jason Abaluck; Jonathan Gruber
  9. An Alternative Reference Scenario for Global CO2Emissions from Fuel Consumption: An ARFIMA Approach By José Belbute; Alfredo M. Pereira
  10. Unemployment Hysteresis and Structural Change in Europe By Kurmaş Akdoğan
  11. Looking into the Black Box of Boosting: The Case of Germany By Lehmann, Robert; Wohlrabe, Klaus
  12. Aggregation level in stress testing models By Hale, Galina; Krainer, John; McCarthy, Erin
  13. Nowcasting BRIC+M in Real Time By Tatjana Dahlhaus; Justin-Damien Guénette; Garima Vasishtha
  14. Identifying GM crops for cultivation in the EU through a Delphi forecasting By McFarlane, Ian; Jones, Philip; Park, Julian; Tranter, Richard

  1. By: Barnichon, Régis; Garda, Paula
    Abstract: This paper evaluates the flow approach to unemployment forecasting proposed by Barnichon and Nekarda (2012) for a set of OECD countries characterized by very different labor markets. We find that the flow approach yields substantial improvements in forecast accuracy over professional forecasts for all countries, with especially large improvements at longer horizons (one-year ahead forecasts) for European countries. Moreover, the flow approach has the highest predictive ability during recessions and turning points, when unemployment forecasts are most valuable.
    Keywords: steady-state unemployment; stock-flow model
    JEL: E24 E27 J6
    Date: 2015–11
  2. By: Lehmann, Robert; Weyh, Antje (Institut für Arbeitsmarkt- und Berufsforschung (IAB), Nürnberg [Institute for Employment Research, Nuremberg, Germany])
    Abstract: "In this paper we evaluate the forecasting performance of employment expectations for employment growth in 15 European states. Our data cover the period from the first quarter 1998 to the fourth quarter 2014. With in-sample analyses and pseudo out-ofsample exercises, we find that for most of the European states considered, the survey-based indicator model outperforms common benchmark models. It is therefore a powerful tool for generating more accurate employment forecasts. We observe the best results for one quarter ahead predictions that are primarily the aim of the survey question. However, employment expectations also work well for longer forecast horizons in some countries." (Author's abstract, IAB-Doku) ((en))
    Keywords: Arbeitsmarktprognose, Prognoseverfahren, Arbeitsmarktforschung, Meinungsforschung, Befragung, Datengewinnung, Prognosegenauigkeit, Erwartung, Arbeitsplatzsicherheit, Arbeitsmarktrisiko, Arbeitsmarktindikatoren, Europa, Österreich, Belgien, Bulgarien, Tschechische Republik, Estland, Finnland, Frankreich, Bundesrepublik Deutschland, Ungarn, Italien, Niederlande, Portugal, Slowakei, Schweden, Großbritannien
    JEL: E27 J00 J49
    Date: 2015–10–29
  3. By: Etienne, Xiaoli
    Abstract: The dramatic rise in commodity index investment have made many market analysts and researchers believe that commodity markets have undergone a financialization process that forged a closer link between commodity and financial markets. I empirically test whether this hypothesis is true in a forecasting context by using high-frequency financial data to forecast monthly US corn prices. Specific financial series examined include the Baltic Dry Index, the US exchange rate, the Standard and Poor’s 500 market index, the 3-month US Treasury bill interest rate, and crude oil futures prices. Using a recently developed statistical model that deals with mixed-frequency data, I find that while some improvements may be made when including high-frequency financial data in the forecasting model, the improvements in mean-squared prediction error and directional accuracy using such models are minimal, and that models generated from random walk and autoregressive models perform satisfactory well compared to more complicated models.
    Keywords: Agribusiness, Agricultural Finance,
    Date: 2015
  4. By: Jörg Döpke (Hochschule Merseburg (University of Applied Sciences Merseburg)); Ulrich Fritsche (Universität Hamburg (University of Hamburg)); Christian Pierdzioch (Helmut-Schmidt-Universität (Helmut-Schmidt-University))
    Abstract: We use a machine-learning approach known as Boosted Regression Trees (BRT) to reexamine the usefulness of selected leading indicators for predicting recessions. We estimate the BRT approach on German data and study the relative importance of the indicators and their marginal effects on the probability of a recession. We then use receiver operating characteristic (ROC) curves to study the accuracy of forecasts. Results show that the short-term interest rate and the term spread are important leading indicators, but also that the stock market has some predictive value. The recession probability is a nonlinear function of these leading indicators. The BRT approach also helps to recover how the recession probability depends on the interactions of the leading indicators. While the predictive power of the short-term interest rates has declined over time, the term spread and the stock market have gained in importance. We also study how the shape of a forecaster’s utility function affects the optimal choice of a cutoff value above which the estimated recession prob- ability should be interpreted as a signal of a recession. The BRT approach shows a competitive out-of-sample performance compared to popular Pro- bit approaches.
    Keywords: Recession forecasting, Boosting, Regression trees, ROC curves
    JEL: C52 C53 E32 E37
    Date: 2015–10
  5. By: Claudia Foroni (Norges Bank); Pierre Guérin (Bank of Canada); Massimiliano Marcellino (Bocconi University, IGIER and CEPR)
    Abstract: We analyze how to incorporate low frequency information in models for predicting high frequency variables. In doing so, we introduce a new model, the reverse unrestricted MIDAS (RU-MIDAS), which has a periodic structure but can be estimated by simple least squares methods and used to produce forecasts of high frequency variables that also incorporate low frequency information. We compare this model with two versions of the mixed frequency VAR, which so far had been only applied to study the reverse problem, that is, using the high frequency information for predicting low frequency variables. We then implement a simulation study to evaluate the relative forecasting ability of the alternative models in finite samples. Finally, we conduct several empirical applications to assess the relevance of quarterly survey data for forecasting a set of monthly macroeconomic indicators. Overall, it turns out that low frequency information is important, particularly so when it is just released.
    Keywords: Mixed-Frequency VAR models, temporal aggregation, MIDAS models
    JEL: E37 C53
    Date: 2015–10–29
  6. By: Richard Pin\v{c}\'ak; Erik Barto\v{s}
    Abstract: Overwhelming majority of econometric models applied on a long term basis in the financial forex market do not work sufficiently well. The reason is that transaction costs and arbitrage opportunity are not included, as this does not simulate the real financial markets. Analyses are not conducted on the non equidistant date but rather on the aggregate date, which is also not a real financial case. In this paper, we would like to show a new way how to analyze and, moreover, forecast financial market. We utilize the projections of the real exchange rate dynamics onto the string-like topology in the OANDA market. The latter approach allows us to build the stable prediction models in trading in the financial forex market. The real application of the multi-string structures is provided to demonstrate our ideas for the solution of the problem of the robust portfolio selection. The comparison with the trend following strategies was performed, the stability of the algorithm on the transaction costs for long trade periods was confirmed.
    Date: 2015–11
  7. By: Bidder, Rhys (Federal Reserve Bank of San Francisco); McKenna, Andrew (Federal Reserve Board of Governors)
    Abstract: In recent years, stress testing has become an important component of financial and macro-prudential regulation. Despite the general consensus that such testing has been useful in many dimensions, the techniques of stress testing are still being honed and debated. This paper contributes to this debate in proposing the use of robust forecasting analysis to identify and construct adverse scenarios that are naturally interpretable as stress tests. These scenarios emerge from a particular pessimistic twist to a benchmark forecasting model, referred to as a ‘worst case distribution’. This offers regulators a method of identifying vulnerabilities, even while acknowledging that their models are misspecified in possibly unknown ways. We first carry out our analysis in the familiar Linear-Quadratic framework of Hansen and Sargent (2008), based on an estimated VAR for the economy and linear regressions of bank performance on the state of the economy. We note, however, that the worst case so constructed features undesirable properties for our purpose in that it distorts moments that we would prefer were left undistorted. In response, we formulate a finite horizon robust forecasting problem in which the worst case distribution is required to respect certain moment conditions. In this framework, we are able to allow for rich nonlinearities in the benchmark process and more general loss functions than in the L-Q setup, thereby bringing our approach closer to applied use.
    Date: 2015–09–22
  8. By: Jason Abaluck; Jonathan Gruber
    Abstract: We explore the in- and out- of sample robustness of tests for consumer choice inconsistencies based on parameter restrictions in parametric models, with a focus on tests proposed by Ketcham, Kuminoff and Powers (2015). We start by arguing that non-parametric alternatives are inherently conservative with respect to detecting mistakes (and one specific test proposed by KKP is incorrect). We then consider several proposed robustness checks of parametric models and argue that they do not separately identify misspecification and choice inconsistencies. We also show that, when implemented using a comprehensive goodness of fit measure, the Keane and Wolpin (2007) test of out of sample forecasting demonstrates that a model allowing for choice inconsistencies forecasts substantially better than one that does not. Finally, we explore the robustness of our 2011 results to alternative normative assumptions.
    JEL: D12 I11 J14
    Date: 2015–10
  9. By: José Belbute (Department of Economics, University of Évora, Portugal and CEFAGE-UE, Portugal); Alfredo M. Pereira (Department of Economics, College of William and Mary, Williamsburg)
    Abstract: We provide alternative reference forecasts for global CO2 emissions based on an ARFIMA model estimated with annual data from 1750 to 2013. These forecasts are free from additional assumptions on demographic and economic variables that are commonly used in reference forecasts, as they only rely on the properties of the underlying stochastic process for CO2emissions, as well as on all the observed information it incorporates. In this sense, these forecasts are more based on fundamentals. Our reference forecast suggests that in 2030, 2040 and 2050, in the absence of any structural changes of any type, CO2 would likely be at about 25%, 34% and 39.9% above 2010 emission levels, respectively. These values are clearly below the levels proposed by other reference scenarios available in the literature. This is important, as it suggests that the ongoing policy goals are actually within much closer reach than what is implied by the standard CO2reference emission scenarios. Having lower and more realistic reference emissions projections not only gives a truer assessment of the policy efforts that are needed, but also highlights the lower costs involved in mitigation efforts, thereby maximizing the likelihood of more widespread energy and environmental policy efforts.
    Keywords: Forecasting, reference scenario, CO2 emissions, long memory, ARFIMA.
    JEL: C22 C53 O13 Q47 Q54
    Date: 2015
  10. By: Kurmaş Akdoğan (Central Bank of Turkey)
    Abstract: We examine the unemployment hysteresis hypothesis for 31 European countries, US and Japan, using linear and nonlinear unit root tests. Two types of smooth transition models - Exponential Smooth Transition Autoregressive (ESTAR) and Asymmetric Exponential Smooth Transition Autoregressive (AESTAR) - are employed to account for the mean-reverting behaviour in unemployment due to heterogeneity in hiring and firing costs across firms. Four main results emerge: First, the hysteresis hypothesis is rejected for 60 percent of the countries in our sample. Second, nonlinear models capture the asymmetries in unemployment dynamics over the business cycle for some countries. Third, many of the series display multiple structural breaks which might point out shifts in mean level of unemployment. Fourth, forecasting powers of our nonlinear models are moderately better than the random walk model in the longer term. The results have policy implications for the debate on the benefits of demand or supply side policies for tackling the current unemployment problem in Europe.
    Keywords: Unemployment, Hysteresis, Nonlinear Adjustment, Structural Breaks, Forecasting
    JEL: E24 C22 E27
    Date: 2015
  11. By: Lehmann, Robert; Wohlrabe, Klaus
    Abstract: This paper looks into the 'fine print' of boosting for economic forecasting. By using German industrial production for the period from 1996 to 2014 and a data set consisting of 175 monthly indicators, we evaluate which indicators get selected by the boosting algorithm over time and four different forecasting horizons. It turns out that a number of hard indicators like turnovers, as well as a small number of survey results, get selected frequently by the algorithm and are therefore important to forecasting the performance of the German economy. However, there are indicators such as money supply that never get chosen by the boosting approach at all.
    Keywords: boosting; economic forecasting; industrial production
    JEL: C53 E17 E37
    Date: 2015–11–03
  12. By: Hale, Galina (Federal Reserve Bank of San Francisco); Krainer, John (Federal Reserve Bank of San Francisco); McCarthy, Erin
    Abstract: We explore the question of optimal aggregation level for stress testing models when the stress test is specified in terms of aggregate macroeconomic variables, but the underlying performance data are available at a loan level. Using standard model performance measures, we ask whether it is better to formulate models at a disaggregated level (“bottom up”) and then aggregate the predictions in order to obtain portfolio loss values or is it better to work directly with aggregated models (“top down”) for portfolio loss forecasts. We study this question for a large portfolio of home equity lines of credit. We conduct model comparisons of loan-level default probability models, county-level models, aggregate portfolio-level models, and hybrid approaches based on portfolio segments such as debt-to-income (DTI) ratios, loan-to-value (LTV) ratios, and FICO risk scores. For each of these aggregation levels we choose the model that fits the data best in terms of in-sample and out-of-sample performance. We then compare winning models across all approaches. We document two main results. First, all the models considered here are capable of fitting our data when given the benefit of using the whole sample period for estimation. Second, in out-of-sample exercises, loan-level models have large forecast errors and underpredict default probability. Average out-of-sample performance is best for portfolio and county-level models. However, for portfolio level, small perturbations in model specification may result in large forecast errors, while county-level models tend to be very robust. We conclude that aggregation level is an important factor to be considered in the stress-testing model design.
    JEL: C18 C52 G21 G28
    Date: 2015–09–28
  13. By: Tatjana Dahlhaus; Justin-Damien Guénette; Garima Vasishtha
    Abstract: Emerging-market economies have become increasingly important in driving global GDP growth over the past 10 to 15 years. This has made timely and accurate assessment of current and future economic activity in emerging markets important for policy-makers not only in these countries but also in advanced economies. This paper uses state-of-theart dynamic factor models (DFMs) to nowcast real GDP growth in five major emerging markets—Brazil, Russia, India, China and Mexico (“BRIC+M”). The DFM framework allows us to efficiently handle data series characterized by different publication lags, frequencies and sample lengths. This framework is particularly suitable for emerging markets for which many indicators are subject to significant publication lags and/or have been compiled only recently. The methodology also allows us to extract model-based “news” from a data release and assess the impact of this news on nowcast revisions. Results show that the DFMs generally outperform simple univariate benchmark models for the BRIC+M. Overall, our results suggest that the DFM framework provides reliable nowcasts for GDP growth for the emerging markets under consideration.
    Keywords: Econometric and statistical methods, International topics
    JEL: C33 C53 E37
    Date: 2015
  14. By: McFarlane, Ian; Jones, Philip; Park, Julian; Tranter, Richard
    Abstract: This paper reports the design and implementation of a Delphi forecasting exercise carried out to identify crop traits that could feasibly be introduced to the advantage of European arable farmers, and for the general benefit of members of the public in EU member states. An expert stakeholder panel was recruited, and in the first round of the consultation, asked for opinions regarding a number of scenarios concerning the availability of GM events, and also scenarios that envisage novel crops developed using advanced technology that is not classified as GM. In a second round of consultation, panel members were asked to comment anonymously on opinions elicited in the first phase. Preliminary results indicate that crops with input traits most likely to become available in EU before 2025 are HTIR maize, HT sugarbeet and HT soybean; these crops are already widely adopted outside Europe. The crops with output traits most likely to become available are winter-sown varieties of rape with reduced saturated fats, spring varieties of which are already available outside EU (notably Canadian Canola).
    Keywords: Forecasting, genetic modification, herbicide tolerance, insect resistance, Crop Production/Industries, Food Consumption/Nutrition/Food Safety, International Development, B4, O3, Q1,
    Date: 2015–11

This nep-for issue is ©2015 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.