The Performance of Unemployment Rate Predictions in Romania. Strategies to Improve the Forecasts Accuracy

Open access

Abstract

The evaluation and improvement of forecasts accuracy generate growth in the quality of decisional process. In Romania, the most accurate predictions for the unemployment rate on the forecasting horizon 2001-2012 were provided by the Institute for Economic Forecasting (IEF) that is followed by European Commission and National Commission for Prognosis (NCP). The result is based on U1, but if more indicators are taken into consideration at the same time using the multi-criteria ranking, the conclusion remains the same. A suitable strategy for improving the degree of accuracy for these forecasts is represented by the combined forecasts. The accuracy of NCP predictions can be improved on the horizon 2001-2012, if the initial values are smoothed using Holt-Winters technique and Hodrick-Prescott filter. The use of Monte Carlo method to simulate the forecasted unemployment rate proved to be the best way to improve the predictions accuracy. Starting from an AR(1) model for the interest variable, the uncertainty analysis was included, the simulations being made for the parameters. Actually, the means of the forecasts distributions for unemployment are considered as point predictions which outperform the expectations of the three institutions. The strategy based on Monte Carlo method is an original contribution of the author introduced in this article regarding the empirical strategies of getting better predictions.

ABREU, I. (2011). International organizations’ vs. private analysts’ forecasts: an Evaluation, Banco de Portugal, Retrieved from http://www.bportugal.pt/en-US/BdP%20Publications%20Research/wp201120.pdf

ALLAN, G. (2012). Evaluating the usefulness of forecasts of relative growth, Strathclyde, Discussion Papers in Economics, No. 12-14.

ARMSTRONG, J. S., FILDES, R. (1995), On the selection of Error Measures for Comparisons Among Forecasting Methods, Journal of Forecasting, 14, p. 67-71. DOI: 10.1002/for.3980140106

BATES, J., GRANGER C. W. J. (1969). The Combination of Forecasts. Operations Research Quarterly, 20(4), 451-468.

BRATU M. (2012). Strategies to Improve the Accuracy of Macroeconomic Forecasts in USA. LAP LAMBERT Academic Publishing, ISBN-10: 3848403196, ISBN-13: 978-3848403196.

BRATU (SIMIONESCU), M. (2013). Filters or Holt Winters technique to improve the forecasts for USA inflation rate?. Acta Universitatis Danubius. OEconomica Vol 9, issue no. 1/2013.

DIEBOLD, F.X., MARIANO, R. (1995). Comparing Predictive Accuracy. Journal of Business and Economic Statistics, 13, pp. 253-265

DOBRESCU, E. (2003). Possible Evolutions of the Romanian Economy (Macromodel Estimations), Romanian Journal of Economic Forecasting, Supplement 2003, pp. 32-64

DOVERN, J., WEISSER J. (2011). Accuracy, unbiasedness and efficiency of professional macroeconomic forecasts: An empirical comparison for the G7, International Journal of Forecasting, 27 (2), pp. 452-465.

FRANSES, P. H., MCALEER, M., LEGERSTEE, R. (2012). Evaluating Macroeconomic Forecasts: A Concise Review of Some Recent Developments, Working paper/ Department of Economics and Finance, University of Canterbury.

GORR W. L. (2009). Forecast accuracy measures for exception reporting using receiver operating characteristic curves, International Journal of Forecasting, Volume 25, Issue 1, January-March 2009, pp. 48-61. DOI: 10.1016/j.ijforecast.2008.11.013

HEILEMANN, U., STEKLER, H. (2007), Introduction to The future of macroeconomic forecasting, International Journal of Forecasting, 23(2), pp. 159-165. DOI: 10.1016/j.ijforecast.2007.01.001

RUTH, K. (2008). Macroeconomic forecasting in the EMU: Does disaggregate modeling improve forecast accuracy?. Journal of Policy Modeling, Volume 30, Issue 3, May-June 2008, pp. 417-429. DOI: 10.1016/j.jpolmod.2007.12.002

Review of Economic Perspectives

Národohospodárský obzor; The Journal of Masaryk University

Journal Information


CiteScore 2017: 0.42

SCImago Journal Rank (SJR) 2017: 0.153
Source Normalized Impact per Paper (SNIP) 2017: 0.351

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 94 94 16
PDF Downloads 28 28 10