Exchange rates and prices: Evidence from China

Abstract

This article develops a pricing model that incorporates an industrial organization approach with the traditional quantity theory of money to explain the impact of exchange rates on consumer prices. Using time-series data on prices and exchange rates of China, the model replicates the main features of the observed facts: exchange rates influence consumer prices through changing import prices; money supply and output influence consumer prices following the quantity theory. The estimating results show that exchange-rate pass-through to consumer prices is low and increases from the short run to the long run. The extent of pass-through is likely to depend on markup adjustment and marginal costs.

Keywords:

• China
• exchange rate
• price elasticity

JEL Classifications:

• C32
• E31
• F31
• F41

Acknowledgements

This research is supported by National Natural Science Foundation of China (Grant No. 71003102) and Renmin University of China Science Foundation (Grant No. 10XNK06). The author would like to thank two anonymous referees for their helpful comments and suggestions.

Notes

 1. The exchange rate analysis of its impact on domestic prices varies in two directions within this approach, exogenous exchange rate analysis (Prakken 1978) and endogenous exchange rate analysis (Papell 1994).

 2. If the markup is variable in response to exchange-rate changes, then the law of one price does not hold. It will be discussed later on.

 3. The quantity theory of money is more empirically relevant to the explanation for movements in domestic prices and therefore adopted in this framework of analysis.

 4. Friedman (1956) restates the quantity theory and emphasizes the relationship between the nominal value of expenditures P_dY and the quantity of money M although he assumes the velocity to be a function of rates of interest.

 5. The assumption of stable velocity of money is validated with the empirical exercise as the presence of cointegration indicates that the velocity is stationary. See the discussion on the empirical results.

 6. Goldberg and Knetter (1997) argue that the choice of a measure for price or cost depends on the type of study. Some studies adopt wages as a measure of cost in studying the relationship between prices and exchange rates whilst other studies adopt exporter's price as a measure in the literature (Choudhri and Hakura 2006; Marazzi and Sheets 2007).

 7. Söderlind and Vredin (1996) note that the main disadvantages of cointegration analysis without strong links to economic theory are that it makes interpretation of cointegrating vectors a treacherous exercise, and that it is hard to estimate the correct number of cointegrating vectors. The application of cointegration analysis based on the fully specified theoretical model in this article could avoid the problem and achieve the statistical prediction consistent with theory.

 8. It refers to the long-run weak exogeneity test. If the null hypothesis of weak exogeneity is accepted, it implies that this variable is long-run weakly exogenous in the VAR system. Also, the generalized impulse-response function conditions on long-run weak exogeneity of the variable from which the shock is simulated.

 9. Compared to Choudhri and Hakura (2006), the result in this article indicates a decline pass-through from 1979 to 2005 for China. This finding is consistent with Marazzi and Sheets' (2007) results for the US.

10. Only consumer prices and real output are long-run weakly exogenous in the VAR system.

11. A nominal depreciation is not favorable to China's economic growth and instead leads to economic recession in the very short run. Even in the long run, gains from nominal depreciation are not attractive.

12. The positive standard-error sized shock in nominal exchange rates implies a 2.18% nominal depreciation. Its impact on consumer prices (and real output) has been scaled in terms of a 10% nominal depreciation.

13. The estimate for the long-run pass-through derived from the response function is consistent with the one from the cointegrating regression.

14. Goldberg and Hellerstein (2008) and Nakamura and Zerom (2010) argue that the markup is inversely linked to the price elasticity of demand which depends on the structure of the market and consumer heterogeneity. If demand is elastic, the markup will be low. So if exchange rate movements can change the price elasticity, the markup will change and the pass-through might be low.

15. Nakamura and Zerom (2010) imply that 18% of the incomplete exchange-rate pass-through is attributable to the markup adjustment whilst Hellerstein (2008) finds that the markup accounts for 55%.

16. Hellerstein (2008) finds that the markup components explain roughly half of the incomplete pass-through whilst local-cost components account for the other half.

17. If the curve of marginal costs is upward sloping, then the costs will decrease as the local currency price rises and the quantity for demand falls, given an exchange-rate depreciation. This channel will mitigate the upward adjustment of the price.

18. Nakamura and Zerom (2010) imply that 81% of the incomplete exchange-rate pass-through is due to local non-traded costs whilst Hellerstein (2008) finds that they account for 45%.

19. It is noted that the discussion absents from the role of nominal price rigidities which are not relevant to the framework in the article. See discussions of that point by Goldberg and Hellerstein (2010) and Nakamura and Zerom (2010).