ABSTRACT
This article investigates the asymmetric relationship between real oil prices and real GDP for selected ASEAN countries (ASEAN-5), using the nonlinear autoregressive distributed lag (NARDL) model. The sample data consist of annual frequencies from 1970 to 2015. Furthermore, the article pays attention to structural breaks in testing the asymmetric nexus. The empirical findings demonstrate long-run asymmetry for the ASEAN-5 countries, however, short-run asymmetry is found for Malaysia and Singapore only. Moreover, the empirical findings show that the increase in oil prices leads to positive effect on GDP for all countries (highly significant for all countries except for the Philippines), yet, such outcome does not go in line with the vast majority of the studies conducted. As for the direction of causality between oil prices and GDP, the article finds mixed results by using the Toda and Yamamoto noncausality test.
KEY WORDS:
Acknowledgments
Thanks are due to three anonymous reviewers and Ali Kutan-Editor for their useful comments and helpful suggestions that allowed me to improve the article. As usual, all remaining errors are my own.
Supplementary material
Supplemental data for this article can be accessed on the publisher’s website.
Notes
1. While it is more useful to use higher frequency data, such as quarterly data, as it offers more observations, this article employed empirically annual data due to the fact that the quarterly data for the ASEAN-5 countries were available for different time periods.
2. For more details, see Bai and Perron (Citation1998).
3. As mentioned previously in the early footnote that the empirical analysis used annual frequencies, nonetheless, we applied the NARDL test for the available quarterly data, which were not available for the same time span. The results are reported in the Supplementary Material (available online) as Tables S2 and S3, for both the long- and short-run dynamics (i.e., see Tables S2 and S3, available online). The results reported did not change substantially compared to the results of the annual frequencies that was employed.
4. For more details, kindly see: Silvey (Citation1959).
5. For more details, kindly see: Ramsey (Citation1969).
6. However, the quarterly data did not support the long-run asymmetric effect in case of Thailand (1993: Q1-2016: Q1). This could be due to the short time span.
7. Although in the case of Indonesia, the CUSUMSQ shows instability, however, since one of the two tests; CUSUM and CUSUMSQ, shows stability, then this is not an issue.
8. However, the quarterly data show short-run asymmetric effect in the case of Philippines and Singapore only.