Role of trade data discrepancy in the exchange rate-trade dilemma

Previous empirical studies investigating the effects of real exchange rate depreciation on sectoral trade balance have widely ignored the fact that trade data can be substantially inaccurate. By comparing what each side of a bilateral trade claims to have actually traded with the other, the problem of sectoral trade data discrepancy can be assessed. For the very crucial case of the US-China bilateral trade, this research compiles all available sectoral trade data as reported by both countries over 26 years and 1366 sectors of trade. This study first assesses the discrepancy in trade data using the tools of Reports Ratio and Pearson’s Product-Moment Correlation Function. Subsequently, trade balance as reported by both China and the US independently is established for 55 sectors of average discrepancy. After it, the only dynamic theory in concept on this issue, the J-Curve Theory, is tested for the 55 sectors using the Autoregressive Distributed Lag technique of cointegration and Error Correction Mechanism to capture the short and long-run effects. The J-Curve is tested twice for each sector, each time using the data reported by one partner only, ceteris paribus. The results of this research confirm the existence of the research problem and draw many conclusions on the issue. First, there is a profound trade data discrepancy between the US and China, although the data is still heading towards reconciliation over time. Second, the strength of discrepancy depends on many factors, i.e. the direction of trade flow, the level of disaggregation of trade data, and time. Third, the use of different trade data resources for assessing the J-Curve yields relatively different coefficient estimates for real exchange rate, in some cases, the coefficient estimates are contradictory even. Fourth, the effects of Yuan depreciation on the US bilateral trade deficit with China are still apparent and affluent using either data resource, while coefficient estimates claiming the opposite are highly mixed and inconsistent. This study proposes Mutual Confirmation as a technique for improving the reliability of the J-Curve empirical results. That is, an estimation of the effect of real exchange rate depreciation on trade balance should be supported using the data reported by both trade partners separately.

Preview Text FEP 2014 27 - IR.pdf Download (1MB) | Preview

Actions (login required)

View Item