Misalignment of the real exchange rate in the African Financial Community (CFA zone) and its policy implications

Abstract

The purpose of this study is to see whether the same pre-devaluation overvaluation (1980 to 1993) of the Communauté Financière Africaine (CFA) franc exists for the post-devaluation period (1995 to 2004). As overvaluation can have significant negative impacts on exports, it is imperative to investigate whether this has continued. First, we found that the same overvaluation exists for the post-devaluation period. Second, we observe that post-devaluation, the overvaluation trend has been reverting and recurring downwards until 1999 and 2000 and steadily increasing since 2001. In fact, the average overvaluation for the zone is estimated at 25% in 2004 which is about 85% of the pre-devaluation level. Third, the economies where agriculture dominates are more overvalued than the economies where agriculture does not dominate. This stylized fact is similar to pre-devaluation observations by Devarajan (1997). Fourth, we also found that the oil producing countries are less overvalued than the nonoil producing countries; and the middle income economies are less overvalued than lower income economies. The latter two stylized facts are contrary to the pre-devaluation dynamics observed by Devarajan (1997). Finally, the empirical analysis suggests that the anticipated contribution of the parity relationship to the overvaluation of the currency is minimal. Instead, the macroeconomic variables that have the most negative influences on the Real Exchange Rate (RER) of the CFA franc are terms of trade shocks, increases in investment and aid inflow.

Notes

¹ The countries in the CFA zone include: Benin, Burkina Faso, Cameroon, Chad, Republic of the Congo, Cote d’Ivoire, Central African Republic, Equatorial Guinea, Gabon, Mali, Niger, Senegal and Togo. Mali and Equatorial Guinea are not included in the sample because they were not in the zone during the duration of the period under study.

² Although the nominal exchange rate among the CFA zone countries is fixed (one-to-one) against each other, we use the RER, reflecting differences in the macroeconomic variables across countries to measure the magnitude of misalignment. We define the RER as the price of tradeables to nontradeables.

³ The income level subgroups follow the World Bank's (1995) classification scheme which uses Gross National Product (GNP) per capita of approximately ≤US$700 for lower income countries and ≥US$700 for middle income countries.

⁴ Misalignment was also estimated using the PPP method. The results are available upon request.

⁵ See Greene (2003, Chapter 13).

⁶ The following tests were performed: Ramsey's reset test (model specification/functional form), Wald's test (coefficient restrictions), Durbin's h-test (serial correlation) and redundant variable test. The model passes all of these diagnostic tests. The F-statistics are greater than their critical values at the 1 and 5% significance levels.

⁷ Chad started exporting oil in 2004. So, it has been included as a nonoil producing country from 1980 to 2003 and as an oil producing country in 2004.

⁸ The diversification index measures the extent to which exports are spread amongst many commodities. It has been calculated as the inverse of the Herfindahl index using disaggregated exports at four digits of the Standard International Trade Classification, Revision 3 (SITC3). The higher the index, the more the export diversification.

⁹ See Yoshino (2004).