Effects of mineral-commodity price shocks on monetary policy in developed countries

Abstract

This article investigates effects of changes in mineral commodity prices on monetary policy. Using macroeconomic data from three mineral-producing countries (Australia, Canada and New Zealand) and two non-mineral-resource countries (USA and UK), I estimate the impulse response functions of the policy interest rates and the core consumer price index (CPI) inflation rates to mineral-commodity price shocks. I find that the central banks in both groups of the examined countries significantly respond to mineral-commodity price shocks. In responses to an unexpected 10% increase in mineral commodity prices, the central banks are estimated to increase their policy interest rates by approximately 0.8 percentage points. Moreover, the central banks seem to take anticipatory policy reactions to control core CPI variations triggered by these shocks. Thus, mineral commodity prices would act as important determinants of the monetary policies in both groups of the examined countries. These findings would be useful for analysing Taylor rules in their countries. However, effects of the increase in their policy interest rates on core CPI inflation cannot be identified for the examined countries.

Keywords:

• mineral commodity prices
• structural vector autoregressions
• impulse responses
• response decompositions
• counterfactual analysis

JEL Classification:

• E31
• E52
• Q00

Acknowledgements

I am grateful to Takayuki Tsuruga, Hiroshi Teruyama, Ryo Okui, Masahiko Shibamoto and an anonymous referee for their beneficial comments and suggestions. I also thank participants in the macroeconomic and econometric seminars, and the Brown Bag Lunch Seminar at Kyoto University for their helpful comments. Moreover, I am thankful to participants of the Autumn Meeting of the Japanese Economic Association and the Autumn Meeting of the Japan Society of Monetary Economics for their useful comments.

Notes

¹ See Cecchetti and Moessner (2008), International Monetary Fund (2008, 2011), Rigobon (2010) and Gelos and Utsyugova (2012).

² For discussions of the relationship between asset prices and monetary policy, see Bernanke and Gertler (2001) and Mishkin (2011).

³ For weights of each commodity, see the CRB’s website at http://www.crbtrader.com

⁴ For details of the data set, see Appendix.

⁵ Chen et al. (2010) demonstrate that exchange rate growth of Australia, Canada and New Zealand has forecasting power for commodity prices.

⁶ The unit root test indicates that the system is stationary.

⁷ I computed Akaike’s information criterion (AIC) from 0 lag to 24 lags. Results of this study with AIC of 12 lags are almost similar to those with the lowest AIC.

⁸ The ordering of variables including comprehensive commodity prices, not mineral commodity prices, is the same as Kilian and Lewis (2011). As in their study, I assume that mineral commodity prices do not respond to the other shocks contemporaneously.

⁹ Evans and Fisher (2011) hypothesize three relations between IRs of the federal funds rate and the core personal consumption expenditures (PCE) inflation rate in USA to commodity price shocks: a weak central bank credibility hypothesis, a strong central bank credibility hypothesis and a generally uninformative indicator hypothesis.

¹⁰ For simplicity, Kilian and Lewis (2011) decompose an IR of the federal funds rate in the SVAR(p) model without an intercept vector.

¹¹ The IRs of the core CPI inflation rates in the mineral-producing countries in the unrestricted model are included in the 95% confidence intervals of the BGW-type counterfactual model in all horizons.

¹² As in footnote 10, the same can be identified for the non-mineral-resource countries.