Short-Run Forecasting of Argentine Gross Domestic Product Growth

Abstract

We propose a small-scale dynamic factor model for monitoring Argentine gross domestic product (GDP) in real time using economic data at mixed frequencies (monthly and quarterly), which are published with different time lags. Our model not only produces a coincident index of the Argentine business cycle in striking accordance with professional consensus and the history of the Argentine business cycle, but also generates accurate short-run forecasts of the highly volatile Argentine GDP growth. By using a pseudo real-time empirical evaluation, we show that our model produces reliable backcasts, nowcasts, and forecasts well before the official data are released.

KEY WORDS:

• Argentine GDP
• business cycles
• mixed frequencies
• real-time forecasting
• state-space models

Acknowledgments

The authors are very grateful to Gustavo Martin and to seminar participants in the XVI Applied Economics Meeting (Granada, Spain, June 2013), the BBVA Research Seminar (Madrid, Spain, July 2013), and the XLVIII Meeting of the Argentine Association of Political Economy (Rosario, Argentina, November 2013) for their very useful comments and suggestions. All remaining errors are the authors’ responsibility. The views expressed in this article are those of the authors and do not necessarily represent those of BBVA Research. M. Camacho thanks CICYT and Fundación Seneca for their support through grants ECO2010-19830, ECO2013-45698 and 11998/PHCS/09.

Notes

1. For the sake of brevity, we focus this review on recent advances on small-scale dynamic factor models and on the articles that have recently tried to forecast Argentine GDP growth.

2. Aruoba and Diebold (2010) also build on Stock and Watson (1989, 1991) and Mariano and Murasawa (2003) and examine the real-time performance of the common factor as a business-cycle indicator, but their focus is on the assessment of current economic activity and not on forecasting.

3. This econometric model is, however, quite flexible. For an application to a very different field, foreign exchange (FX) rate forecasting, see Dal Bianco et al. (2012).

4. Despite its high volatility, the Argentine GDP growth does not follow an autoregressive conditional heteroskedasticity (ARCH) process. We carried out ARCH tests for the residual from an AR(1) specification of Argentine real GDP growth and were not able to reject the null of no presence of ARCH in the residuals at the usual significance levels. These results are available from the authors upon request.

5. D’Amato et al. (2011a, 2011b) do not present “backcasting” results, which are the estimation on a given quarter of the previous quarter rate of growth before they are published by the statistical agency; that is, within the ten weeks of delay.

6. Aruoba et al. (2009) extend this analysis to include high-frequency data using an exact algorithm, as opposed to the approximate algorithm of Mariano and Murasawa (2003). However, Aruoba et al. (2009) face the cost of assuming deterministic trends in the series.

7. A description of how these equations look for an illustrative simplified model is set out in the Appendix of the working paper version of this work (available at http://www.um.es/econometria/Maximo/).

8. In particular, if the log of a variable appears as nonstationary according to Ng and Perron (2001) unit root tests, then the data are used in growth rates. To save space, the results are not presented, but they are available from the authors upon request.

9. The lag lengths used in the empirical exercise were always set to two since the AR(2) specification is able to model very rich dynamics in the time series. However, we performed several exercises to check that our results were robust to other reasonable choices of the lag lengths.

10. Camacho et al. (2013) show that although the fully Markov-switching dynamic factor model is generally preferred to the shortcut of computing inferences from the common factor obtained from a linear factor model, its marginal gains rapidly diminish as the quality of the indicators used in the analysis increases. This is precisely our case.

11. Following Camacho and Perez-Quiros (2007), we did not include any lags in the factor. We checked that the resulting model is dynamically complete in the sense that the errors are white noise.

12. That is why the last shaded month in the graph is March 2012, which corresponds with the last data analyzed by Jorrat (2012), but this does not imply that this apparent recession ended on that month, as it could have lasted longer.

13. This high commonality in switch times of probabilities with Argentine business-cycle phases validates the interpretation of state s_t = 1 as recession and the probabilities plotted in Figure 2 as probabilities of being in recession.