Financial Sector Reforms and Monetary Policy in Zambia

Abstract

The dissertation comprises four chapters focusing on issues concerning policy reforms and monetary policy in Zambia.

Chapter 1 briefly outlines the theoretical foundations for the reforms undertaken in Zambia since the mid 1980s and the process thereof. The main issues addressed were the removal of interest rate and credit controls, exchange rate devaluation and the use of indirect instruments in implementing monetary policy. The review indicates that although the control of inflation is still difficult and figures are still in double digit levels, annual inflation rates have reduced significantly compared to levels achieved in the early 1990s. The nominal exchange rate has been depreciating prompting increased intervention from the central bank. Despite the increase in nominal interest rates, real deposit rates have remained negative.

Chapter two analyses the monetary transmission mechanism in Zambia. Vector auto -regressions are estimated for the pre-reform and post-reform periods.Variance decompositions and impulse response functions are examined to see whether there are any changes observed in the monetary transmission mechanism after the reforms. Different systems are estimated in each period using alternate variables as measures of monetary policy shocks. The results show that contractionary monetary policy is followed by a fall in both output and prices. When compared, results from the two estimation periods show that both the responsiveness of prices and output to policy shocks and the magnitude of their forecast error variance decompositions explained by these variables have increased since the reforms. The results also show that the impact lags have reduced. There is evidence of the bank lending channel both before and after the reforms. Of the mechanisms estimated, the exchange rate mechanism seems to be the most important mechanism for transmission of policy shocks to both prices and output during the post-reform period.

Chapter three investigates whether monetary aggregates have useful information for predicting inflation other than that provided by inflation itself. Forecasting experiments are conducted to see whether monetary aggregates and selected financial sector variables are useful in predicting inflation.We perform forecasting experiments and compare the performance of different models. We also estimate an error correction model of inflation. Of the monetary aggregates considered, M2 contains the most information and its growth rate is significant in the inflation model. The external sector variables are also important. The results indicate that inflation exhibits a high level of inertia suggesting the presence of implicit indexation and significant inflationary expectations possibly due to past fiscal effects and low policy credibility. Overall, the foreign sector variables seem to be more important for movements in prices than monetary aggregates even in the long run.

The importance of the exchange rate to stabilisation policy in Zambia is underscored by the results obtained in chapters 2 and 3. In this paper, we pursue this idea by investigating the effect of central bank intervention on exchange rates in Zambia. Using a GARCH (1, 1) model of the exchange rate, we simultaneously estimate the effect of cumulative intervention on the mean and variance of the exchange rate.We find that central bank intervention in the foreign exchange market increases the mean but reduces the variance of the exchange rate. The explanation leans towards speculative bandwagons and a 'leaning against the wind' strategy. Although there is no attempt to distinguish through which channel intervention operates, we argue that this is more likely to be a signalling effect rather than a portfolio balance effect. This effect operates mainly through the supply and demand of foreign exchange in the market.