The Impact of Chinese Exchange Rate Policy on the Rest of the World: Evidence from Firm-Level Data

Abstract

This paper aims to gauge the global effect of renminbi revaluation on stock markets. Using data on 12,300 firms in operating in tradable sectors in 44 economies, we find that expectations of renminbi appreciation reduce the relative stock returns of firms providing components or raw materials to China as inputs for that country's exports. We also find some evidence that expectations of renminbi appreciation reduce the stock prices of financially constrained firms.

Keywords:

• China
• exchange rate
• firms
• rest of the world

Acknowledgements

We thank Nigel Chalk, Stijn Claessens, Marcel Fratzscher, Jaewoo Lee, Shang-Jin Wei, and seminar participants at the IMF and the ECB for helpful comments as well as Mohsan Bilal for excellent research assistance. The views in the paper are those of the authors and do not necessarily reflect those of the IMF.

Notes

1. For competing perspectives, see Eichengreen (2007) and Dooley et al. (2009).

2. For competing perspectives, see Scott (2010) and Evenett and Francois (2010).

3. As emphasized by Ali and Dadush (2010), who note that 17% of the total machinery and transport equipment imports of low-income countries are now drawn from China.

4. As documented in the US case by Evenett and Francois (2010) and, in more detail, by Francois (2010).

5. Lumber, coal, and sugar are largely imported for domestic consumption, while iron, non-ferrous metals, and (to a lesser extent) oil and natural gas are heavily used in producing finished metals and other exported items, explaining the difference across commodities.

6. 2009 data would be contaminated by the effects of the financial crisis.

7. The second and third of these variables are constructed following Forbes (2004).

8. In a placebo test, we again looked at the daily stock returns on 19 July 2005 and 17 June 2010. For these two pre-PBOC-announcement dates, (Exports to China/Exports to the world)×Ores has a positive coefficient of 1.08 with a t-statistics of 1.20, in sharp contrast with its negative coefficient in Column 2 of Table 6.