Asymmetric Cointegration, Nonlinear ARDL, and the J-Curve: A Bilateral Analysis of China and Its 21 Trading Partners

ABSTRACT

This article follows the nonlinear Autoregressive Distributed Lag (ARDL) error-correction methodology to explore nonlinearity in the relationship between the trade balances and the real exchange rates for China and its 21 partners. We find evidence for short-run asymmetric effects of exchange rate in cases of 18 partners, short-run adjustment asymmetry in cases of 11 partners, short-run cumulative asymmetry in cases of seven partners, and a significant long-run asymmetric effect cases of five partners. We find support for the “J-curve” that is only due to appreciation or depreciation of the Yuan in cases of five partners, including the U.S.

KEY WORDS:

• asymmetry cointegration
• bilateral J-curve
• China
• nonlinear ARDL

JEL CLASSIFICATION:

• F31

Notes

1. Please refer to Bahmani-Oskooee and Ratha (2004) and Bahmani-Oskooee and Hegerty (2010) for a detailed review of the literature.

2. Other studies which have only estimated a linear model for China by assuming symmetric effects are Brada, Kutan, and Zhou (1993), Zhang (1998, 1999), and Weixian (1999). Zhang (1999) also reviews China’s exchange rate policy.

3. This section closely follows Bahmani-Oskooee and Fariditavana (2016) who used such methods to estimate bilateral trade balance models between the U.S. and six of her major trading partners.

4. For the precise normalization procedure, see Bahmani-Oskooee and Fariditavana (2015).

5. Note that since critical values do account for integrating properties of variables, there is no need to perform unit-root tests and variables could be combination of I(0) and I(1) which is the case for almost all macro variables.

6. See Shin, Yu, and Greenwood-Nimmo (2014, 291) for details.

7. For some other application of these methods see Apergis and Miller (2006), Verheyen (2013), and McFarlane et al. (2014), Gregoriou, Healy, and Savvides (2014), Durmaz (2015), Gogas and Pragidis (2015), Baghestani and Kherfi (2015), Nusair (2017), Al-Shayeb and Hatemi-J. (2016), Lima et al. (2016), and Aftab et al. (2017).

8. Note that since in the ARDL approach of Pesaran, Shin, and Smith (2001, 303), variables are a combination of I(0) and I(1), they also tabulate an upper bound and a lower bound critical values for this t test. However, since their values are only for large samples, whereas Banerjee, Dolado, and Mestre’s (1998, 276) values are for small as well as large samples, we use values from this later source.

9. The results from the linear ARDL model are more or less consistent with Bahmani-Oskooee and Wang (2006) who tested the J-curve between China and each of her 13 partners.

10. A few other studies have considered bilateral trade balance of China with its partners. Wang, Lin, and Yang (2012) using a panel framework have focused on post-2005 period and used monthly data over the period, August 2005-September 2009, across 19 partners to show that a real appreciation of Yuan had adverse effects on China’s trade balance. As indicated in the text, an approach such as this is prone to the aggregation bias as the general conclusion may hold for one partner, but may not hold with the other. Indeed, this seems to be the case when Yang, Zhang, and Tokgoz (2013) assessed the impact of Yuan appreciation on different regions and found that appreciation improves China’s trade balances with the EU, Australia and New Zealand, ASEAN, Japan and Korea, but not with the U.S. In contrast, using the Chinese-U.S. trade flows over the period 1994–2012, Cheung, Chinn, and Qian (2016) suggest that “the value of China’s exports to the US responds negatively to real renminbi (RMB) appreciation, while imports respond positively”. None of these studies, however, allows for asymmetric effects which is the main focus of this article.

11. In the linear as well as in the nonlinear models with Hong Kong, the real Yuan-Hong Kong dollar does not seem to play a significant role. We suspect that this is due the lack of variation in the exchange rate between two regions. It is also worth pointing out that some portion of the Chinese export goes through Hong Kong. In these cases, the relatively stable exchange rate between Yuan and Hong Kong Dollar does not mask the variation and the importance of the exchange rate of Yuan with other currencies.

Additional information

Funding

This work was supported by the Shanghai Education Development Foundation and Shanghai Municipal Education Commission [Shanghai Pujiang Program and the “Shuguang Program”].