Developing and testing new measures of exchange market pressure in Sub-Saharan Africa

Abstract

Attempts by economists to identify exchange rate crisis have led to the development of the exchange market pressure (EMP) index, which has been modified to include the interest rate variable, to better capture pressure on domestic currencies. The index in this state (current) is more appropriate for countries who operate an inflation targeting (IT) monetary policy framework or a policy framework close to it. This observation among others makes the EMP index sensitive to some assumptions if sub-Saharan African countries (SSA) are involved. SSA countries employ different monetary policy frameworks making the usage of the EMP index across board misleading. Using different variations of the model specified by authors specialised in the subject, this study identified an EMP index which will be effective in identifying currency crisis in 32 sub-Saharan African (SSA) countries for which data is available. Using a ridge regression model and a dynamic OLS model as a control, the study showed that the original EMP index was good in explaining changes in the exchange rate variable in SSA if the changes in reserves are adjusted for by broad money. The explanatory power is further enhanced if the components are normalised by their standard deviations without adjusting for the money variable. It was discovered that the current version of the index which includes the interest rate variable is not good at explaining changes in the exchange rate variable in SSA even though it had some explanatory effect. Using a discrete threshold regression model this study resolves two major problem associated with the use of the exchange market pressure index in identifying exchange rate crisis. The results point to un-even implementation of the Inflation Targeting framework in the study countries.Dwelling on the assumption that a country will be able to wade of speculative attack on its currency if it has reserves that can cater for previous year's imports or short-term external debt or broad money, this study proposes two variations to the original index by Girton and Ropper (1975). Using a Generalised Method of Moments (GMM) regression model and a ridge regression model, this study found the index with the reserve variable adjusted for by imports to have the most explanatory effect on the exchange rate variable in SSA. This study recommends a combination of these indexes since the phase of economic development is not static.In testing the EMP indexes on the banking crisis in Ghana, the retail prices of premium gasoline in Ghana and Kenya and also testing the impact of the exchange rate variable in the debt sustainability of Ghana and Kenya. This study found that the exchange rate vii played a significant role in the banking sector and fluctuations in the retail price of premium gasoline in Ghana in the long run. It was also found to be putting increasing pressure on the gross debt of both Ghana and Kenya. The study recommends the creating of domestic buffers leading to the accumulation of the needed reserves to smoothen volatilities in the ...