Is exchange rate stability beneficial for stabilizing consumer prices in China?

Abstract

This study examines the relationship between real effective exchange rates (REERs) and the consumer price index (CPI) in China, utilizing a bootstrap Granger full-sample causality test and a sub-sample rolling-window estimation. Considering structural changes, we assess the stability of the parameters and find that both the short-run and long-run relationships between the two estimated variables are unstable. This result suggests that full-sample causality tests cannot be relied upon. We instead employ a time-varying (bootstrap) rolling-window approach to revisit the dynamic causal relationship, and we find that the CPI is affected by the REER for several sub-samples due to the role of exchange rate pass-through (ERPT) under the managed floating exchange rate regime in China. These findings provide further proof of the impact of stable exchange rates on the maintenance of relatively steady price levels especially during the economic crisis and economic reform in China. The policy implication of these findings is that maintaining exchange rate stability is beneficial for controlling inflation during the economic crisis and economic reform.

Keywords:

• Exchange rates
• consumer prices
• pass-through
• inflation
• rolling window
• time-varying causality

JEL Classifications:

• C22
• E31
• F31
• F41

Acknowledgement

This work is supported by the National Social Science Foundation (Grant number: 15BJY155), and Ministry of Education's Humanities and Social Science Research Project (Grant number: 14YJA790049).

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. See Choudhri and Hakura (2006) for a model with staggered pricing that incorporates pricing to market (based on a translog aggregator function) and local currency pricing. They use this model to explain persistence in REER.

2. Choudhri and Hakura (2006) have demonstrated that a stronger feedback reaction would be expected to weaken the association between y_t and y_{t − 1}, and thus decrease λ₁ (the persistence in y_t). However, that the effect on κ₁ is less clear because μ exerts a positive direct effect on κ₁ in addition to a negative indirect effect (through λ'e).

3. Choudhri and Hakura (2006) have demonstrated that κ₁ increases and κ₂ decreases in λ. The effect on κ₁ arises because a larger share of intermediate inputs in costs increases price-level persistence. The effect on κ₂, however, reflects the influence of a decrease in the share of imported varieties in intermediate inputs (i.e. a decrease in 1 − θ) on the pass-through over the short term.

4. Specifically, the critical values and p-values are obtained utilizing the asymptotic distribution constructed by means of Monte Carlo simulations employing 10,000 samples generated from a VAR model with constant parameters.

5. For technical details of the bootstrap test, see Balcilar, Ozdemir, and Arslanturk (2010).

6. We employ data on REER and the CPI as well as the weighting matrix derived from Bayoumi, Lee, and Jayanthi (2006) to calculate CPI-based REER.

7. See Balcilar and Ozdemir (2013).

8. No consistent criterion is available for us to select the window size in the rolling-window estimation (Balcilar, Ozdemir, and Arslanturk 2010). They demonstrate that the optimal window size depends on persistence and the size of the break. Based on their Monte Carlo simulations, they argue that the bias in autoregressive (AR) parameters is minimized with a window size of as low as 20 when there are frequent breaks. We implement different bootstrap rolling-window causality tests employing different month-window sizes. Nevertheless, one may wonder whether the results are sensitive to the choice of window size. Thus, the study undertakes the same exercise using different window sizes to determine whether the findings are affected. We select different window sizes (e.g. 20, 25, 30 and 40 months). The findings are affected, but very little. We want to find the structural changes caused by China's policy regime changes within the shortest time. We therefore select a 20-month window size.