An Agnostic Analysis of Exchange Rate Movement in Ghana

Abstract

Based on quarterly data for the period 2006:3-2018:4, the effect of exchange movement on a set of price indices in Ghana is examined via a Bayesian Vector Autoregressive model. Using normal inverted-Wishart priors, the posterior estimates are generated by Markov Chain Monte Carlo draws via a sign restriction algorithm. Findings showed that the response of consumer prices (CPI), producer prices (PPI) and non-food prices (NFP) to exchange rate shocks is low and incomplete. Furthermore, the forecast error variance decomposition (FEVD) indicated that CPI is most responsive to exchange rate impulses than NFP and PPI. In addition, inflationary pressures in Ghana emanated from exchange rate sources other than monetary sources. The paper recommends "pricing in local currency" as a deliberate policy to insulate domestic prices from volatilities in the exchange rate. Contribution/Originality: This study used a new estimation methodology to examine the effect of exchange rate dynamics on prices in Ghana.