Since 2002, the international community, especially the United States, has been calling for RMB appreciation in order to reverse its huge trade deficit with China, and strongly demanding RMB appreciation against the US dollar. On July 2, 2005, the People's Bank of China announced that China will implement a managed floating exchange rate system based on market supply and demand and with reference to a basket of currencies. At the same time, it was announced that the transaction price of USD against RMB was adjusted to 1 USD against 8. 165438 0 yuan RMB, with an appreciation rate of about 2 percentage points. In this context, the relationship between the theoretical exchange rate and China's trade balance began to increase the actual remittance of RMB. Many scholars worry that the appreciation of RMB exchange rate will bring a serious impact on China's trade balance, thus affecting the stability of China's national economy. Whether the above judgment is correct is based on our answers to the following questions: the mechanism of RMB real exchange rate change on China's trade balance; Whether the way, direction, degree and payment terms of the impact of RMB real exchange rate changes on the balance of payments are established for a long time; In the short term, whether to use the time lag of playing; Marshall-Miller has a "J" curve effect. The research on these issues is of great guiding significance to the formulation of exchange rate policy in China at present. Based on the data of Sino-US bilateral trade balance from 1994 to 2005, this paper studies the long-term and short-term effects of real exchange rate changes on the trade balance. That is, there is a "J" curve effect in our research; The influence of real exchange rate, real income and other economic variables on China's import and export and trade balance; In order to determine whether Marshall-Lerner condition holds in the long run; Whether bilateral trade is balanced in the short term and the time lag of their respective roles. The methods used in this paper include cointegration analysis, vector m), error pulse model (VEC response analysis and variance decomposition difference correction pulse) and so on. We study the long-term impact of real exchange rate on trade balance through cointegration analysis. In order to study the influence of the real exchange rate and real income on the trade balance, we use the vector error correction model (VECM) to dynamically adjust the machine cycle, and then carry out impulse response analysis and variance decomposition. Through empirical analysis, we draw the following conclusions: 1. The impact of real exchange rate on import, export and trade balance is still relatively small, and even some of them are not statistically significant, indicating that China can not effectively adjust the trade balance through the change of real exchange rate at present. It reflects the current situation that the market-regulated trade balance equation formed by China's exchange rate is low in chemistry, importers and exporters are insensitive to exchange rate changes, and the substitution of import and export commodities is weak. However, the positive parameters corresponding to the positive real exchange rate still show that the Marshall-Lerner condition is established in the long run. Second, the "J" curve effect of pulse analysis shows that the lag period is about one year, which shows that the trade balance between China and the United States is very clear. 3. Variance decomposition shows that the impact of real exchange rate shock on the change of trade balance is far less than that of real income shock. We find that the change of the trade balance between China and the United States is mainly influenced by the change of the real income of the United States, whether in the short term or in the long term; Through the impact on US trade, even considering the difference of economic growth rate between the two countries, the conclusion is still valid. Since 2002, the international community has been calling for the appreciation of RMB. In particular, the United States strongly urges China to raise the exchange rate of RMB against the US dollar in order to reverse its huge trade deficit with China. On July 26th, 2005, the People's Bank of China announced that China would implement a managed floating exchange rate system based on market supply and demand. A basket of currencies will be used as reference information for RMB exchange rate adjustment. At the same time, the exchange rate of RMB against the US dollar was adjusted to 8. 1 1 RMB against 1 US dollar, with an appreciation of about 2 percentage points. In this context, theorists pay more attention to the relationship between the real exchange rate of RMB and China's trade balance. Many scholars worry that the appreciation of RMB exchange rate will seriously affect China's trade balance and affect the stability of the national economy. If we want to know whether the above judgment is correct, we must answer the following questions: What mechanism does the real exchange rate of RMB affect China's trade balance? Does Marshall-Lerner Condition hold in the long run? In the short term, is there a "J" curve effect in China? How does the real exchange rate of RMB affect China's trade balance? In which direction and to what extent does the real exchange rate of RMB affect China's trade balance? How long will China's trade balance be affected after the real exchange rate of RMB starts to change? Studying these issues is of great guiding significance to China's current exchange rate policy. This paper empirically tests the long-term and short-term relationship between the real exchange rate of RMB and the bilateral trade balance between China and the United States. In other words, we try to judge whether the Marshall-Lerner condition holds in the long term and whether there is a "J" curve effect in China in the short term. We try to judge the effect and time lag of economic variables such as RMB real exchange rate and real income on China's imports, exports and bilateral trade balance. We study the long-term effects of real exchange rate on import, export and trade balance through co-integration analysis. In order to study the dynamic adjustment mechanism of real exchange rate and real income affecting the trade balance in the short term, we use vector error correction model, impulse response analysis and variance decomposition. The methods used in this paper include cointegration analysis, vector error correction model, impulse response analysis and variance decomposition. Through empirical analysis, we get the following results: First, the impact of real exchange rate on import and export and trade balance is very small, even statistically insignificant. This reflects the low degree of substitution between imported goods and exported goods. However, the positive parameters of RMB real exchange rate in the standard trade balance equation show that Marshall-Lerner condition is established. This shows that we cannot effectively adjust the trade balance through the real exchange rate now. It reflects the low marketization of foreign exchange pricing, and importers and exporters are not sensitive to exchange rate changes. Secondly, the pulse analysis shows that the "J" curve effect is very obvious, and the lag period is about one year. We find that the trade balance between China and the United States is mainly affected by the real income of the United States, whether in the short term or in the long term, even considering the differences in economic growth rates between the two countries. Thirdly, variance decomposition shows that the impact of the real RMB exchange rate on China's trade balance is far less than the actual income.