汇率波动、生产网络与股市风险--基于中美贸易摩擦背景的分析

本文从生产网络视角出发,研究中美贸易摩擦期间汇率变动对中美两国股票市场的直接影响以及由行业间生产联系带来的网络影响。从静态一般均衡模型可推出具有空间自回归(SAR)模型形式的实证模型,其中以行业间投入产出关系作为空间权重矩阵。实证结果发现,中美双边汇率变动对两国股市的影响在贸易摩擦期间均比之前更为显著,人民币贬值导致中国股市收益率下降,其中约50%是由行业间生产联系带来的网络效应,而美元升值导致了美国股市收益率下降,其中约37%是网络效应。关税制裁波及的行业与未波及行业的股票收益率均受汇率变动影响,但后者受到的网络影响更大,且各行业受到的网络影响主要由其下游行业传递。

Exchange Rate Volatility,Production Networks,and Stock Market Risk during the Sino-US Trade Friction

Sino-US trade friction has created a major challenge to China's efforts to ensure high-level opening-up and maintain production and supply chains. Previous studies have discussed the impact of trade frictions on consumption, employment, economic welfare, and industrial output. This paper examines financial market risks in China and the U.S. against the backdrop of recent trade frictions, and in particular the impact of the expected exchange rate change on stock market returns. An empirical spatial auto-regressive (SAR) model is derived from a static general equilibrium model, in which the production network measured by the input-output table is used as the spatial weight matrix, which is then given economic meaning. Based on the SAR model, the direct impact of exchange rate changes on the stock market and the indirect impact through the production network are estimated. We further calculate the direct and indirect impacts on various industries and assess the average impacts on industries subject to tariff sanctions versus those not subject to sanctions.
The main findings are summarized as follows. First, during the trade friction period, the expected exchange rate change has significant negative impact on stock market returns in China and the United States. Indicating that the correlation between the stock market and the foreign exchange market increased during the friction period. Second, expected bilateral exchange rate changes between China and the United States not only directly affect the stock returns of various industries but also affect a certain proportion of the network because of the production links between industries. The proportion of network effects is 50% in the Chinese sample and 37% in the American sample, which reflects the stronger links between Chinese industries. Third, on average, both industries affected by tariff sanctions and those not affected are influenced by exchange rate changes; the latter is more affected through the network.
The fimdings have important policy implications. First, trade frictions have mutually damaging consequences. Therefore, we should strengthen communication and cooperation between countries. Second, the linkage between the foreign exchange market and the stock market is enhanced during the period of trade friction. Therefore, special attention should be paid to high-level risk and cross-market contagion in periods of economic and financial turmoil. Third, the connection between the entity units (such as a production connection) can transmit the impact of financial shock (such as the impact of exchange rates on the stock market). Therefore, we should take into account the factors at the entity level when attempting to prevent and resolve financial risks. Finally, as an economic power, China should accelerate the improvement of the industrial chain, promote economic internal circulation, and enhance its ability to resist external shocks.
This study contributes to the literature in several ways.First, this paper examines the impact of Sino-US trade frictions on the financial market, studies the financial market risk magnified by inter-industry production linkages and gives some new insight from the perspective of industrial linkages. Second, most studies of trade frictions focus on their impact on entity-level activities. From the perspective of the production network, this paper compares and analyzes the changes in the financial markets of China and the United States in a period of trade frictions, focusing on the impact of exchange rate changes on stock market returns. Finally, this paper uses a new method to estimate and decompose the impact of exchange rate changes, and it obtains several new findings. Using the SAR model, with the production network as the spatial weight matrix, this paper quantitatively evaluates the direct impact of exchange rate changes on the stock market and the network impact caused by inter-industry production links.