Terms-of-Trade and Trade Balance: Some Empirical Evidence of Asian Economies

Abstract

This study examines the impact of terms-of-trade and oil price on trade balance in Asian economies. Generally, the results of the normalized cointegrating vectors show that the impact of terms-of-trade on trade balance is different across economies. An increase in oil price or permanent oil price will lead to a decrease in terms-of-trade while the impact of an increase in temporary oil price on terms-of-trade is ambiguous. Generally, terms-of-trade, domestic demand, foreign demand, and oil price (permanent oil price, temporary oil price) are important in the determination of trade balance in the short run and long run.

ACKNOWLEDGMENTS

The author would like to thank Tagi Sagafi-nejad and the reviewer of the journal for the comments on an earlier version of this article. Also, the author would like to thank the reviewer of the journal for recommending the use of the BK band pass filter to decompose oil price into permanent oil price and temporary oil price. All remaining errors are the author's.

Notes

¹ Arize (1996) expresses terms-of-trade as import price to export price. The countries examined are Canada, France, Germany, Italy, Japan, the United Kingdom, the United States, Denmark, Finland, the Netherlands, Switzerland, India, Korea, Malaysia, Mexico, and Sri Lanka.

² Barsky and Kilian (2004) argue that oil price is an important factor in the determination of economic growth, inflation, and productivity.

³Volatility is measured by implied volatility. Volatility was in the range of 30 per cent, implying that oil markets were facing big uncertainties regarding future price developments and were sensitive to small shocks and to news (Askari and Krichene 2008).

⁴It has been evidenced that economic activity responses to asymmetrically to the oil price shock. An increase in oil price seems to retard aggregate economic activity by more than a fall in oil price stimulates it. Moreover, an increase in oil price deteriorates terms-of-trade for oil importing economies. Thus, there is a wealth transfer from oil importing economies to oil exporting economies, leading to a fall of the purchasing power of firms and households in oil importing economies. If oil price is longer lasting, it can give rise to a change in the production structure and have an impact on unemployment (Lardic and Mignon 2006: 3911).

⁵See Lardic and Mignon (2006) for various transmission channels of the impact of oil price on economy.

⁶In this study, openness to international trade is measured as trade (exports plus imports) over gross domestic product (GDP).

⁷This model specification is very similar to the model specification by Rose and Yellen (1989). They specify trade balance is a function of exchange rate, domestic demand, and foreign demand. Many empirical studies in the literature of the impact of exchange rate on trade balance use their model.

⁸ Kim (1996: 466) reports that the negative correlation between income and trade balance holds due to transitory income changes. With permanent changes, there exists no such predictable and significant relationship between income and trade balance. Miljkovic, Paul, and Garcia (2000: 327) report that the transitory income shocks are major reasons for the change in trade balance and negative correlation between trade balance and income is primary due to the transitory shocks. They examine five small open economies, namely Austria, France, Italy, Spain, and Switzerland. The aggregate demand shocks are responsible for cyclical variations in trade balance and income and aggregate supply is responsible for the long-run economic growth.

⁹In the year 1999, the Baxter and King (1995) working paper is published in journal.

¹⁰For annual data, it is proposed the following set of parameters, that is, K = 3, ω ₁ = 2π(1/8), and ω ₂ = π (Baxter & King 1995).

¹¹However, in the case of an integrated input series, the Baxter-King filter will produce a distorted output series. Thus, a more appropriate approach is to apply a set of filters in order to judge the robustness of the results (Canova 1998).

¹²In the year 2003, the Christiano and Fitzgerald (1999) working paper is published in journal.

¹³This index is used as its movements with other oil price indexes, namely the United Arab Emirates oil price (Dubai), the British oil price (Brent) and West Texas intermediate oil price (Texas) were about the same. Thus, this index is expected to give a good picture of the movements of oil price in the world (See Section 2 in this study).

¹⁴More specifically, domestic demand of Singapore is expressed by manufacturing production. For Hong Kong, domestic demand is expressed by GDP because the use of manufacturing production does not provide good estimation result. For Japan, domestic demand is expressed by industrial production.

¹⁵See for the lag length used in the estimation of the DF test statistic.

¹⁶See for the lag length used in the estimation of the λ _Max and λ _Trace test statistics.

¹⁷See Table VI for the lag length used in the estimation of the normalized cointegrating vectors.

¹⁸Generally, the plots of cumulative sum of recursive residuals (CUSUM) and cumulative sum of squares of recursive residuals (CUSUMSQ), which are not reported, show no evidence of instability of the error correction model.