nep-for New Economics Papers
on Forecasting
Issue of 2015‒04‒19
ten papers chosen by
Rob J Hyndman
Monash University

  1. Beating a Random Walk: “Hard Times” for Forecasting Inflation in Post-Oil Boom Years? By Huseynov, Salman; Ahmadov, Vugar; Adigozalov, Shaig
  2. Evaluation of the Forecasting Quality By Kurennoy, Alexey
  3. Forecasting in nonstationary environments: What works and what doesn't in reduced-form and structural models By Raffaella Giacomini; Barbara Rossi
  4. Business Tendency Surveys and Macroeconomic Fluctuations By Daniel Kaufmann; Rolf Scheufele
  5. The Forecasting of Spot Exchange Rates Based on the Forward Exchange Rates By Radim Gottwald
  6. EuroMInd-D: A Density Estimate of Monthly Gross Domestic Product for the Euro Area By Tommaso Proietti; Martyna Marczak; Gianluigi Mazzi
  7. Bringing an elementary agent-based model to the data: Estimation via GMM and an application to forecasting of asset price volatility By Ghonghadze, Jaba; Lux, Thomas
  8. A Simple Multivariate Filter for Estimating Potential Output By Patrick Blagrave; Roberto Garcia-Saltos; Douglas Laxton; Fan Zhang
  9. l1-Regularization of High-Dimensional Time-Series Models with Flexible Innovations By Marcelo C. Medeiros; Eduardo F. Mendes
  10. Theoretical Aspects of Modeling of the SVAR By Skrobotov, Anton; Turuntseva, Marina

  1. By: Huseynov, Salman; Ahmadov, Vugar; Adigozalov, Shaig
    Abstract: In this study, we investigate forecasting performance of various univariate and multivariate models in predicting inflation for different horizons. We design our forecast experiment for the post-oil boom years of 2010-2014 and compare forecasting ability of the different models with that of naïve ones. We find that for all forecast horizons simple naïve models have equal forecasting ability with relatively sophisticated models which allow for richer economic dynamics. To check whether forecasting ability of naïve models has not been inferior to relatively sophisticated ones in boom and pre-boom years as well, we repeat our forecast experiment and estimate the models for the period 2003-2006 and keep the years 2006-2010 for undertaking pseudo out-of-sample exercise. Our experiment reveals that surprising forecasting performance of naïve models in post-oil boom years is a new phenomenon and in fact, the employed models have exhibited significant forecasting advantage over naïve ones in boom and pre-boom years. We find that despite declining volatility in inflation over the post-oil boom years, it has become considerably difficult for our models to beat naïve ones due to recently unpredictable behavior of inflation.
    Keywords: Inflation; Forecasting; Time Series methods; Bayesian methods
    JEL: C11 C13 C32 C53
    Date: 2014–10–27
  2. By: Kurennoy, Alexey (Russian Presidential Academy of National Economy and Public Administration (RANEPA))
    Abstract: This preprint describes a number of statistical tests for (unselective) assess the quality of forecasting. For each of these assumptions are presented and discussed to be executed for the corresponding test can be used. In addition, preprint extends the scope of applicability of the Giacomini and White tests, spreading them in the event of forecasts prepared according to the recursive scheme, but almost entirely dependent on the short-term observations.
    Keywords: forecasting, forecasting quality, Giacomini test, White test
    Date: 2015–03
  3. By: Raffaella Giacomini; Barbara Rossi
    Abstract: This review provides an overview of forecasting methods that can help researchers forecast in the presence of non-stationarities caused by instabilities. The emphasis of the review is both theoretical and applied, and provides several examples of interest to economists. We show that modeling instabilities can help, but it depends on how they are modeled. We also show how to robustify a model against instabilities.
    Keywords: Forecasting, instabilities, structural breaks.
    Date: 2014–12
  4. By: Daniel Kaufmann (KOF Swiss Economic Institute, ETH Zurich, Switzerland); Rolf Scheufele (Swiss National Bank, Switzerland)
    Abstract: We investigate the information content of business tendency surveys for key macroeconomic variables in Switzerland. To summarise the information of a large data set of sectoral business tendency surveys we extract a small number of common factors by a principal components estimator. The estimator is able to deal with mixed-frequency data and missing observations at the beginning and end of the sample period. We show that these survey-based factors explain a relevant share of the movements of key macroeconomic variables, i.e., CPI inflation, GDP, employment, and an output gap. In particular, questions about the current and future expected situation are informative. However, backward-looking questions, for example questions about the situation compared to the previous year, do not contain additional information. We then examine the economic dimension of the data set. Questions about prices, real activity and capacity constraints contain important information for the corresponding macroeconomic variables. Finally, we estimate a dynamic relationship to produce forecasts for our factors and these key macroeconomic variables. It turns out that the predictive ability of our survey-based factor approach is quite encouraging. In a pseudo out-of-sample exercise, our approach beats relevant benchmarks for forecasting CPI inflation and an output gap and adds information to the benchmark forecasts for GDP and employment.
    Keywords: Business tendency surveys, dynamic factor models, mixed frequencies, missing observations, nowcasting, forecasting
    JEL: E32 E37 C53
    Date: 2015–04
  5. By: Radim Gottwald (Department of Finance, Faculty of Business and Economics, Mendel University in Brno)
    Abstract: The forecasting power of forward exchange rates for future spot exchange rates has been investigated by many researchers. In this paper, the author focuses on this topical economic theme too, and investigates the extent, to which the future spot exchange rates could be forecasted based on the current forward exchange rates. The paper aims at an assessment of the forecasting of spot USD/EUR exchange rates based on the forward exchange rates in the period from 2005 to 2013. Graphical and regression analyses are used to investigate the relationship between daily closing spot and forward rates, namely between 3 month rates and 6 month rates. The ordinary least squares method is used in order to forecast the chosen parameters. Hypotheses related to these parameters are tested at a significance level of 5%. By means of the augmented Dickey-Fuller test for a unit root in a time series sample, the author investigates whether the time series of the parameters is stationary. Afterwards, the time series is detrended in order to guarantee stationarity. Transformation into a non-linear econometric model with integrated autoregressive process AR(1) is used in order to reduce high positive autocorrelation in the residuals of the model. Thereafter, forecasts of the detrended model are made. Results revealed the following findings. According to the graphical analysis, the current forward exchange rates probably cannot be considered sufficiently reliable forecasters of the future spot exchange rates. According to the regression analysis, the forward forecasts even systematically undervalue the future spot exchange rates. Summarized, the current forward exchange rates cannot be considered sufficiently reliable forecasters of the future spot exchange rates. The above-mentioned findings are important for financial analysts working in financial companies or enterprises, which import or export some products, thus trading with foreign business partners using foreign currencies. Speculators on foreign exchange markets could make use of the presented findings as well.
    Keywords: Forward exchange rate, spot exchange rate, rational expectations theory, currency pair, FOREX
    JEL: C53 F31 O24
    Date: 2015–04
  6. By: Tommaso Proietti (DEF and CEIS, Università di Roma "Tor Vergata"); Martyna Marczak (Department of Economics, University of Hohenheim); Gianluigi Mazzi (Statistical Office of the European Communities, Eurostat)
    Abstract: EuroMInd-D is a density estimate of monthly gross domestic product (GDP) constructed according to a bottom–up approach, pooling the density estimates of eleven GDP components, by output and expenditure type. The components density estimates are obtained from a medium-size dynamic factor model of a set of coincident time series handling mixed frequencies of observation and ragged–edged data structures. They reflect both parameter and filtering uncertainty and are obtained by implementing a bootstrap algorithm for simulating from the distribution of the maximum likelihood estimators of the model parameters, and conditional simulation filters for simulating from the predictive distribution of GDP. Both algorithms process sequentially the data as they become available in real time. The GDP density estimates for the output and expenditure approach are combined using alternative weighting schemes and evaluated with different tests based on the probability integral transform and by applying scoring rules.
    Keywords: Density Forecast Combination and Evaluation; Mixed–Frequency Data; Dynamic Factor Models; State Space Models
    JEL: C32 C52 C53 E37
    Date: 2015–04–10
  7. By: Ghonghadze, Jaba; Lux, Thomas
    Abstract: We explore the issue of estimating a simple agent-based model of price formation in an asset market using the approach of Alfarano et al. (2008) as an example. Since we are able to derive various moment conditions for this model, we can apply generalized method of moments (GMM) estimation. We find that we can get relatively accurate parameter estimates with an appropriate choice of moment conditions and initialization of the iterative GMM estimates that reduce the biases arising from strong autocorrelations of the estimates of certain parameters. We apply our estimator to a sample of long records of returns of various stock and foreign exchange markets as well the price of gold. Using the estimated parameters to form the best linear forecasts for future volatility we find that the behavioral model generates sensible forecasts that get close to those of a standard GARCH (1,1) model in their overall performance, and often provide useful information on top of the information incorporated in the GARCH forecasts.
    Keywords: sentiment dynamics,GMM estimation,volatility forecasting
    JEL: G12 C22 C53
    Date: 2015
  8. By: Patrick Blagrave; Roberto Garcia-Saltos; Douglas Laxton; Fan Zhang
    Abstract: Estimates of potential output are an important ingredient of structured forecasting and policy analysis. Using information on consensus forecasts, this paper extends the multivariate filter developed by Benes and others (2010). Although the estimates in real time are more robust relative to those of naïve statistical filters, there is still significant uncertainty surrounding the estimates. The paper presents estimates for 16 countries and provides an example of how the filtered estimates at the end of the sample period can be improved with additional information.
    Date: 2015–04–07
  9. By: Marcelo C. Medeiros (Department of Economics PUC-Rio); Eduardo F. Mendes (Department of Economics Australian School of Business)
    Abstract: We study the asymptotic properties of the Adaptive LASSO (adaLASSO) in sparse, high-dimensional, linear time-series models. We assume that both the number of covariates in the model and the number of candidate variables can increase with the sample size (polynomially orgeometrically). In other words, we let the number of candidate variables to be larger than the number of observations. We show the adaLASSO consistently chooses the relevant variables as the number of observations increases (model selection consistency) and has the oracle property, even when the errors are non-Gaussian and conditionally heteroskedastic. This allows the adaLASSO to be applied to a myriad of applications in empirical finance and macroeconomics. A simulation study shows that the method performs well in very general settings with t-distributed and heteroskedastic errors as well with highly correlated regressors. Finally, we consider an application to forecast monthly US inflation with many predictors. The model estimated by the adaLASSO delivers superior forecasts than traditional benchmark competitors such as autoregressive and factor models.
    Date: 2015–04
  10. By: Skrobotov, Anton (Russian Presidential Academy of National Economy and Public Administration (RANEPA)); Turuntseva, Marina (Russian Presidential Academy of National Economy and Public Administration (RANEPA))
    Abstract: In this paper an overview of methods for the analysis of structural VAR models is provided. The fundamental properties of SVAR models, the estimated parameters, as well as various methods of identifying shocks and pritsnipe construct confidence intervals for impulse responses, are discussed. The paper also discusses the problems associated with non-stationary variables.
    Keywords: structural VAR models (SVAR), structural VECM (SVECM), impulse responses, decomposition of the forecast error variances, the identification of shocks
    Date: 2015–03

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.