Monetary policy and heterogeneous inflation expectations in South Africa

Abstract

This paper examines the relationship between in ‡ation and in ‡ation expectations of analysts, business, and trade unions in South Africa during the in ‡ation targeting (IT) regime. We consider in ‡ation expectations based on the Bureau of Economic Research (BER) quarterly survey observed from 2000Q1 to 2013Q1. We estimate in ‡ation expectations of individual agents as the weighted average of lagged in ‡ation and the in ‡ation target. The results indicate that expectations are heterogeneous across agents. Expectations of price setters (business and unions) are closely related to each other and are higher than the upper bound of the of…cial target band, while expectations of analysts are within the target band. In addition, expectations of price setters are somewhat related to lagged in ‡ation and the opposite is true for analysts. The results reveal that the SARB has succesfully anchored expectations of analysts but that price setters have not su¢ ciently used the focal point implicit in the in ‡ation targeting regime. The implication is that the SARB may be pushed to accommodate private agents'expectations.JEL Classi…cation Numbers: C51, E52, E58.

Forecasting the South African inflation rate: On asymmetric loss and forecast rationality

“…For the sectoral data, we find evidence in favor of asymmetry of the loss function, and against forecast rationality. As expected, given the micro-level results, the asymmetry of the loss function is less pronounced in the case of financial analysts than for the other two groups of forecasters, which in turn, is consistent with heterogeneity of forecast formation across the three groups of forecasters (Ehlers and Steinbach 2007, Reid and Du Plessis 2011, Reid 2012, Kabundi and Schaling 2014. However, all the South African research papers listed here that use the BER survey data study the aggregate data for each of the three groups which ignores any heterogeneity within these groups of forecasters.…”

“…As the analysis of Figure 4 already made clear, the estimates of the asymmetry parameter are larger for the business sector and trade unions than for financial analysts. Also in line with the message conveyed by Figure 4, is the result that the forecasts of the business sector and the trade unions closely track each other, while both series differ to a non-negligible extent from the forecasts of financial analysts (see Ehlers andSteinbach 2007 andSchaling 2014). While the interpretation of the results should not be stretched too far given the limited number of observations, the comparatively large asymmetry parameter (which even increases when we move from Model 1 to Model 4) suggests that forecasts of price setters reflect that underestimation of the inflation rate is more costly than overestimation.…”

“…More specifically, Figure 4 shows that all three groups of BER forecasters tend to underestimate the inflation rate during periods of rising inflation, while during disinflationary episodes, the inflation expectations of price setters tend to return in a much slower fashion to lower levels than the inflation expectations of financial analysts. The sluggish return of inflation expectations of price setters to lower levels after periods of high inflation illustrates why Kabundi and Schaling (2014) might have reached the conclusion that the inflation expectations of price setters are less well anchored than the inflation expectations of financial analysts. Table 3 summarizes the estimation results.…”

“…Specifically, they report that inflation expectations are tied to the lagged inflation rate. In a follow-up study, Kabundi et al (2014) find that forecasts of price setters (business representatives and trade unions) are linked to the lagged inflation rate, but this holds only to a lesser extent for the forecasts of financial analysts.…”

Inflation forecasts and forecaster herding: Evidence from South African survey data

“…They conclude that inflation expectations are not in line with the rational expectation hypothesis. In a recent study, Kabundi et al (2014) also report that forecast of price-setters (business representations and unions) are linked to the lagged inflation rate. This, however, is not true for the forecasts of analysts, consistent with cross-sectional heterogeneity of forecasters.…”