Monetary policy in low income countries: the case of Uganda

This thesis addresses interrelated issues that influence the implementation of monetary policy in low income countries (LICs). These include the role of inflation persistence, financial frictions and the potential impact of regime-changes or large shocks. The analysis is applied to data for the Ugandan economy. Chapter 3 extends the quantile regression approach to investigate inflation persistence in LICs. The results suggest mean-reversion for the whole sample, however, there is evidence of asymmetric mean-reversion within specific quantiles. In addition, it is noted that the level of persistence increased after 2006 and during the inflation-targeting period. The study also suggests that a measure of core inflation that is derived from wavelet techniques appears to provide a useful measure of this variable. Chapter 4 considers the role of financial frictions in Uganda. It makes use of a dynamic stochastic general equilibrium (DSGE) model that incorporates several small open-economy features. The model parameters are estimated with the aid of Bayesian techniques using quarterly macroeconomic data. The results suggest that the central bank currently responds to changes in the interest rate spread and that it may be possible to derive a more favourable sacrifice ratio by making use of a slightly more aggressive response to macroeconomic developments. Chapter 5 employs a Markov-switching DSGE model to consider the possibility of regime-switching behaviour. Two variants of regime-switching models are considered: One that incorporates regime-switching features in the monetary policy rule (only) and another that incorporates regime-switching features in both the monetary policy rule and in the volatility of the shock processes. Most of the parameters are again estimated with the aid of Bayesian techniques. The results suggest that the model parameters do not remain constant over the two regimes and the transition probabilities appear to capture important economic events. In addition, the out-of-sample evaluation suggests that the regime-switching models may provide a more accurate description of the data generating processes