Exchange Rate and Sectoral Trade Balance Dynamics: Empirical Evidence from Eastern Africa Panel Data

ABSTRACT

This study examines the effect of the real exchange rate (RER) on the three sectors' trade balances in East Africa by using ARDL procedures. The linear ARDL results show that RER depreciation improves the manufacturing and mining trade balances, but worsens the agricultural trade balance in the long run. However, nonlinear ARDL results show an absence of the asymmetric effect of RER on trade balances, except the manufacturing sector. Heterogeneity effects can be obscured by the weak effect of RER on the aggregate trade balance. Our results suggest that sectoral investigation is crucial in analyzing the effects of RER on trade dynamics.

KEYWORDS:

• ARDL procedures
• Eastern Africa
• exchange rate
• sectoral trade balance

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ This may explain why some studies using aggregate data fail to support the J-curve hypothesis.

² Few existing studies investigated the relation between real exchange rate and aggregate trade balance in East Africa. For example, Caporale, Gil‐Alana, and Mudida (2015) for Kenya, Gebeyehu (2014) for Ethiopia, Bahmani-Oskooee and Gelan (2012) for Burundi and Tanzania, and Yol and Baharumshah (2007) for Uganda found a positive effect of foreign exchange rate on trade balance. On the other hand, Kennedy (2013) for Kenya, Yol and Baharumshah (2007) for Tanzania, and Asmamaw (2008) for Ethiopia found that the exchange rate has a negative effect on trade balance. Kwalingana et al. (2012) for Malawi found the effect of foreign exchange rate on trade balance is insignificant. Hunegnaw and Kim (2017) also investigated the relation between real exchange rate and aggregate trade balance in East African countries.

³ $\mathrm{Y}\mathrm{f}={\sum }_{k=1}^m(\theta \mathrm{k}\mathrm{Y}\mathrm{k})$.

⁴ The East African region is comprised of Ethiopia, Kenya, Tanzania, Zimbabwe, Zambia, Mozambique, Madagascar, Rwanda, Burundi, Malawi, Somalia, Mauritius, Djibouti, Eritrea, Seychelles, South Sudan, and Comoros, based on the 2010 UN classification. Among them, 12 countries where data are available were considered.

⁵ The appendix can be found online at www.tandfonline.com/uitj.

⁶ Past studies provided criticisms on the Johansen procedure when the number of observations is small. For example, Toda (1995) suggested that less than 100 observations can lead to misleading results.

⁷ The PMG estimator is an intermediate one because it involves both pooling and averaging. One advantage of the PMG over the panel FMOLS and DOLS approaches is that it can allow the short-run dynamic specification to differ from country to country while the long-run coefficients are constrained to be the same. If countries are expected to have a similar sectoral structure in the long run, the PMG estimator is the appropriate one. The MG estimator is the least restrictive one. It allows for heterogeneity of all the parameters. It consists of estimating separate regressions for each country and computing averages of the country-specific coefficients. The required assumptions are quite strong; the group-specific parameters are distributed independently of the regressors, and the regressors are strictly exogenous.

⁸ SBC is preferred because it provides more parsimonious specifications. See Acaravci and Ozturk (2012) and Pesaran and Shin (1995, 1998).

⁹ The panel DOLS model, developed by Kao and Chiang (2000), includes leads and lags of the independent variables. The panel DOLS estimation is fully parametric and offers a computational convenience to correct endogeneity problems. On the other hand, panel FMOLS correct for the endogeneity bias in a non-parametric way.

¹⁰ Refer to Asteriou and Hall (2007) for details.

¹¹ It is difficult to test the J-curve effect by using the PMG estimator for all the countries because the estimation procedure does not allow us to include enough lags in the panel.

Additional information

Funding

This work was supported by the Economics Department Foundation of Seoul National University.