Comparing Macroeconomic Forecasts For South Africa From 2001 to 2017: Do We Need Official Forecasts?

Abstract

This paper compares the accuracy of the GDP growth and inflation forecasts made by private sector forecasters who participate in the annual “Media24 Economist of the Year” forecasting competition to the official forecasts of the public sector, namely, the National Treasury (NT) and the South African Reserve Bank (SARB). We include the inflation expectations as gathered through the Bureau for Economic Research (BER) quarterly survey, since these estimates constitute an integral part in informing the course of the SARB's monetary policy decisions. Furthermore, we compare the accuracy of the aforementioned entities to the forecasts of an adaptive-naive model that generates forecasts through the extrapolation of past actual inflation observations. This is undertaken using a combination of descriptive statistics and quantitative measures which allow for the analysis of both absolute and relative accuracy. We also undertake a non-parametric test for forecast dominance. Furthermore a deeper investigation of the relevant accuracy of public sector official forecasts compared to those made by private sectors participants, reveal that on average, over the full sample period, and with respect to both current and one year-ahead forecast, the BER have been more accurate than both the NT and the SARB.