The relationship between interest rates and inflation in South Africa: revisiting Fisher's hypothesis

by Mitchell-Innes, H.A.

This thesis investigates the relationship between expected inflation and nominal interest rates in South Africa and the extent to which the Fisher effect hypothesis holds. The hypothesis, proposed by Fisher (1930), that the nominal rate of interest should reflect movements in the expected rate of inflation has been the subject of much empirical research in many industrialised countries. This wealth of literature can be attributed to various factors including the pivotal role that the nominal rate of interest and, perhaps more importantly, the real rate of interest plays in the economy. The validity of the Fisher effect also has important implications for monetary policy and needs to be considered by central banks. Few studies have been conducted in South Africa to validate this important hypothesis.

The analysis uses the 3-month bankers’ acceptance rate and the 10-year government bond rate to proxy both short- and long-term interest rates. The existence of a long-run unit proportional relationship between nominal interest rates and expected inflation is tested using Johansen’s cointegration test. The data is analysed for the period April 2000 to July 2005 as the research aims to establish whether the Fisher relationship holds within an inflation targeting monetary policy framework.

The short-run Fisher effect is not empirically verified. This is due to the effects of the monetary policy transmission mechanism and implies that short-term nominal interest rates are a good indication of the stance of monetary policy. A long-run cointegrating relationship is established between long-term interest rates and expected inflation. The long-run adjustment is less than unity, which can be attributed to the credibility of the inflation-targeting framework.