网络舆情对人民币汇率的冲击效应--基于中美贸易摩擦事件

Effect of Network Public Opinion Shocks on Exchange Rate:Evidence from the China-US Trade Friction

2018年中央经济工作会议上,政府强调把稳定外汇市场作为基本政策之一,但自美国总统特朗普上台以来,其展现出诸多个性化施政方式,并在2018年不顾中方劝阻,掀起中美贸易争端,这势必对外汇市场产生较大影响。因此,需进一步深化研究汇率变动的影响因素,从而有利于外汇市场的平稳健康运行。行为金融学理论表明市场并非完全理性,考虑到网络舆情蕴含市场的非基本面信息,其变化在一定程度上影响投资者行为,因此以中美贸易摩擦事件的网络舆情为研究对象,爬取新浪微博中关于中美贸易摩擦事件的文本,通过信息词典构建基于利好信息和利空信息的网络舆情信息指数。在分位数Granger因果检验的基础上,构建分位数向量自回归模型,研究网络舆情对人民币升值、平稳和贬值等不同阶段外汇市场的冲击效应。研究结果表明,网络舆情对不同阶段汇市的影响存在差异,对不同分位水平汇率的影响也不尽相同。①整体上,网络舆情对汇率的冲击呈现正负交替现象,个别分位点持续正向冲击,且尾部分位点的冲击明显强于中位点;②在人民币贬值阶段,舆情信息对汇率的冲击均较小且衰减较快,但尾部冲击异于其他分位点;③在人民币平稳阶段,利空信息对不同分位水平汇率的影响具有非对称性特点;④在人民币升值阶段,舆情信息对汇率的冲击具有强度大且衰减慢的特点。通过网络舆情与外汇市场关系的研究,发现处在极端情况下的外汇市场更易被网络舆情左右。研究结果为研究汇率变动提供了新的证据,肯定了将网络舆情信息纳入汇率变动影响因素的重要价值;通过使用分位数回归技术,能够有效捕捉偏态数据的尾部信息,为监管部门进行极端风险管控提供决策参考。

At the 2018 Central Economic Work Conference,the government stressed that stabilizing the foreign exchange market is one of the basic policies.However,since Donald Trump was elected president of the United States,he has shown many personalized ways of governing.In 2018,Donald Trump set off China-US trade friction despite Chinese dissuasion,which is bound to have an impact on the foreign exchange market.Therefore,it is necessary to further study the influencing factors of exchange rate change,which benefits the foreign exchange market running stably.Behavioral finance theory reveals that the market is not perfectly rational.Taking into account that the network public opinion contains the non-fundamental information of the market,its change will influence the behavior of investors to a certain extent.Therefore,this study takes the online public opinion of the China-US trade friction event as the research subject.First,we extract the text of the China-US trade friction event in Sina Weibo.Second,we build the network public opinion information index based on positive information and bad information through the information dictionary.Third,we use quantile vector autoregression model on the basis of the quantile Granger causality tests,and then study the shock of the network public opinion on the foreign exchange market in different stages(appreciation,stabilization and devaluation stages).The results show that in the shock of the network public opinion,there are differences on the foreign exchange market in different stages and the exchange rate in different quantile levels.①As a whole,the shock of network public opinion on exchange rate shows positive and negative alternation.And the shock of individual quantile is always positive.Moreover,the shock of the tail quantile is significantly stronger than that of the median.②In the stage of RMB devaluation,the shock of public opinion information on exchange rate is small and attenuated rapidly.But the tail shock is different from the other quantiles.③And in the stage of RMB stabilization,the shock of bad information on the exchange rate of different quantile levels is asymmetric.④In the stage of RMB appreciation,the shock of public opinion information on exchange rate has the characteristics of large intensity and slow attenuation.By studying the relationship between network public opinion and the foreign exchange market,this study has found that the foreign exchange market in the exceptional cases is more likely to be effected by online public opinion.The results provide new evidence for the study of the exchange rate,and confirm the important value of introducing the network public opinion into the fluctuation of the exchange rate.By using the quantile regression model,the tail information of the skewed data can be effectively obtained,which can be conducive to provide practical guidance for regulators in their risk management in exceptional cases.