Skip to content

The Effects of Monetary Policy on Real Farm Prices in South Africa

Literature › Works

Goodness C. Aye; Rangan Gupta (2011). The Effects of Monetary Policy on Real Farm Prices in South Africa. University of Pretoria, Department of Economics Working Paper Series, 2011-19. University of Pretoria, Department of Economics.
Type working paper
Scope Post-1994 South African economy
Method econometric
Period studied 1970–2010
Regions South Africa
Themes agriculture land, macro fiscal finance debt, methods data sources
Citation key aye2011

Argument

The paper asks whether monetary policy (actual, anticipated and unanticipated, positive and negative) affects real agricultural producer prices in South Africa, using monthly data for 1970:01-2010:12. It combines a monetary misperception model with a VAR and models unanticipated shocks as both ordinary and recursive residuals. Actual, anticipated and recursive-unanticipated money growth have statistically significant effects on real farm prices. Asymmetry between expansionary and contractionary policy is supported only for recursive anticipated policy. The effects are quantitatively small, explaining roughly 1.5-6.5 percent of real farm price variance, so other (real-sector) factors presumably dominate.

Evidence and method

Monthly South African Reserve Bank and Statistics South Africa data, 1970-2010: M1, M2 and M3 (M3 reported) and the agricultural producer price index deflated by the CPI. Unit-root (ADF, Phillips-Perron) and Johansen cointegration tests were run, with results available on request. Evidence is regression coefficient tables, Wald/F tests, impulse responses and variance decompositions.

Findings

  • Headline result: monetary effects are significant but explain little of real farm price variation. (p. 2)

    The effects observed in this study are quantitatively small and accounts for only a very small percentage (1.5 percent - 6.5 percent) of the variation in real farm prices.

  • Prior South African evidence (Naude 1995) found no significant effect of unanticipated money shocks on relative agricultural prices. (p. 3)

    He found that the unanticipated money supply shocks from the three monetary aggregates (M1, M2 and M3) did not have significant effect on any of the two relative agricultural producer prices

  • Real farm prices and money supply show no cointegration, so a short-run VAR was used. (p. 5)

    Using the Johansen (1991) approach for cointegration test, no evidence of cointegration between real farm prices and any of the money supply measures could be

  • Actual money growth significantly affects real farm prices, but asymmetry is not established. (p. 8)

    The results show that actual money supply growth has significant effect on real farm prices. This is evidenced by the significance of the 7th and 10th lags as well as the joint significance

  • The impulse response of real farm prices to actual money is short-lived, with a modest variance share. (p. 10)

    The variance decomposition shows that actual money supply growth accounted about 4.4 percent of the variation in real farm prices.

  • Ordinary unanticipated shocks are jointly insignificant (neutral), whereas recursive unanticipated shocks are jointly significant. (p. 16)

    unlike the ordinary unanticipated monetary shock, the recursive unanticipated monetary shock had a significant impact on RFP.

  • Recursive anticipated policy has asymmetric effects, with contractionary policy contributing more variance than expansionary. (p. 21)

    Whereas the value is 4.5 percent for the positive shock, it is 6.5 percent for the negative shock.

  • Conclusion: the outcome for unanticipated shocks depends on the residual specification, and effects are small. (p. 25)

    Thus, the outcome of the impact of unanticipated monetary policy shock depends on whether the shock is modelled as ordinary or recursive residuals.