Fisher’s hypothesis, survey-based expectations and asymmetric adjustments: Empirical evidence from South Africa

Our study re-examines Fisher’s hypothesis for the South African economy in the post-inflation targeting era and presents two empirical novelties over preceding works for the same country. Firstly, we examine Fisher effect by making use of survey-based inflation expectations data for financial analysts, business sector, trade unions and households, hence making our study more disaggregate in nature. Secondly, we examine both short-run and long-run asymmetric cointegration effects in Fisher’s relation using the nonlinear autoregressive distributive lag (NARDL) model as an econometric framework. For the full quarterly sample of 2002:01 – 2019:04, our study finds interest rates respond more aggressively to falling expectations than rising ones, with a full Fisher effect found for financial analysts, partial effects for households and business, and no effect for trade unions. However, when the data is split into two sub-samples corresponding to pre- and post-financial crisis periods, we observe changing dynamics in which interest rates respond more aggressively to rising inflation, with partial effects being also found for trade unions. Policy recommendations based on our study are offered.