Exchange rate movements and oil price expectation shocks in selected African countries: Evidence from a recursive methodology

Many studies investigating the exchange rate-oil price nexus do not impose realistic gains on time-varying parameters, which more accurately depict how market participants react to shocks on exchange rates, especially those emanating from large movements in oil prices. The advantage of estimating time-varying parameters over the conventional ordinary least squares (OLS) estimator is that imposing realistic gains on time-varying parameters truly captures the behavior of foreign exchange market participants following an oil-price expectation shock. The main objective of this paper is to investigate how oil price expectation shocks affect exchange rate movements in selected African countries, which include a mix of net oil-exporting and net oil-importing countries. We deploy a recursive methodology to estimate time-varying parameters from an exchange rate–oil price regression. Since OLS weighs all observations equally and assigns an average parameter estimate, we find that time-varying parameter estimates tend to be larger and more accurately represent the behavior of market participants, especially profit-maximizing speculators. Moreover, the evolution of our time-varying parameter estimates mimics the exchange rate data generation process. Our results provide new insights into how exchange rates behave following an oil-price expectation shock at different quarterly time horizons and have important implications for monetary policy.