Exchange Rate Pass-Through to Prices: Bayesian VAR Evidence for Ghana

Abstract

Using quarterly data from 2006q3 to 2017q4, this paper employed sign restrictions with rejection method in a Vector Autoregression to estimate the pass-through of exchange rate dynamics to domestic prices in Ghana. The priors of the model belongs to the flat Normal inverted-Wishart family. Markov Chain Monte Carlo (MCMC) is used to collect 1000 draws from the posterior distribution of the SVAR parameters that satisfy the sign restrictions. The model specification included some idiosyncratic features of the Ghanaian economy such as the dependence on primary export commodities for foreign exchange revenue and the dependence on foreign aid. Impulse response functions was used to analyze exchange rate pass-through whilst variance decomposition was used to explain the most dominant source of inflation in the study sample. The impulse response showed a fairly large but not unitary pass-through of exchange rate dynamics to domestic prices. The implication herein is that exchange rate depreciation led to upsurge in prices in Ghana albeit, the impact is incomplete. Results from the variance decomposition indicated a monetary expansion was most dominant in explaining inflationary pressures in Ghana. For inflation to be lowered, policy directives should be geared towards exchange rate stability as well as ensuring a stable interest rate environment.