Research On The Mechanism Of External Shocks On RMB Exchange Rate Fluctuation Risk And The Construction Of Early Warning System

Since the opening to the outside world,China’s foreign exchange market and domestic economy have been continuously affected by external shocks,and the external shocks show a trend of normalization.In order to adapt to the environment,stabilize the RMB exchange rate and maintain domestic economic development,China successively adopted two important RMB exchange rate system reforms in2005 and 2015,and completed the transformation from unilateral appreciation to two-way fluctuation.From normal intervention to gradual fading out and relying on market regulation,the RMB exchange rate is more flexible and more adaptable to environmental changes than ever before.On May 14,2020,General Secretary Xi Jinping stressed the construction of a new development pattern of "double cycle",which is an important strategic measure to deal with complex international environmental changes,and also puts forward new requirements and expectations for us.Therefore,this paper takes the Risk Spillover of external shocks to RMB exchange rate fluctuations under the strategic competition pattern between China and the United States as the background,takes the exchange rate pricing theory as the guidance,and adopts the research three links of "external shock source-risk action mechanism-risk transmission mechanism" as the research framework of the full text to explore the spillover effect and transmission mechanism of RMB exchange rate fluctuations due to external shocks in an all-round and multi-level way.And use the advantage that deep learning can approach any complex function with arbitrary precision to realize the construction of early warning system,and finally determine the research focus and policy suggestions in the following three aspects.Firstly,this paper studies the Risk Spillover Effect of external shocks on RMB exchange rate fluctuations.By reviewing the evolution process and current situation of RMB exchange rate,the connotation and characteristics of external shock and the spillover effect of external shock on RMB exchange rate fluctuation,this paper deeply analyzes the action principle of external shock on RMB exchange rate fluctuation risk and the role of various factors in each link.Considering the common characteristics of heteroscedasticity and volatility clustering in the edge distribution of financial data,six common Copula,two rotating Copula and two time-varying Copula models are used to construct the measurement model of the tail dependence between external shocks and RMB exchange rate fluctuations,which realizes the detection of the dependence between RMB exchange rate and external shocks and event risk,and solves the problem of capturing the asymmetric dynamic dependence between external shocks and RMB exchange rate.In addition,before the occurrence of fluctuation aggregation,using VAR model,arma-garch,density rolling prediction and backtracking test and other technologies,it is found that some fluctuation signals will be released before the arrival of the crisis,which also provides theoretical and practical support for the subsequent construction of early warning system.Secondly,this paper studies the transmission mechanism and transmission path of external shocks on the risk of RMB exchange rate fluctuations.In view of the lack of systematic research on the risk transmission mechanism and transmission path of RMB exchange rate fluctuation in the traditional model,based on the traditional exchange rate pricing theory,this paper summarizes the typical cases of RMB exchange rate fluctuation and the constructed risk index system,and introduces a series of variables such as consumption preference,capital stock and price stickiness mechanism,Build an open economy DSGE model in line with the actual situation of China at the present stage,realize the fluctuation and change of RMB exchange rate and domestic economic indicators under the action of external shocks,so as to understand the action mechanism and transmission path of external shocks in RMB exchange rate fluctuations,and further deepen the understanding of external shocks on the transmission mechanism and transmission path of RMB exchange rate fluctuations,Provide data and theoretical support for China’s stable transformation of monetary policy from quantitative to price.The empirical results show that the US interest rate shock,US GDP shock and US inflation shock have a significant impact on the RMB exchange rate and China’s macroeconomic market.Among them,the impact of US interest rate shock on RMB exchange rate and China’s macroeconomic market is the most significant and lasting,while the impact time of US GDP is shorter than that of US interest rate shock,but longer than that of US inflation.In terms of impact degree,the impact degree of US GDP is lower than that of US interest rate,but similar to that of US inflation.Thirdly,this paper studies the external impact early warning index system and the RMB exchange rate fluctuation risk early warning model based on LSTM network model.In order to achieve the goal of early warning of exchange rate fluctuation risk and preparing for shocks in advance,based on the construction of foreign exchange market pressure index system and the quantification of external shocks,using the advantage that Markov mechanism transformation model can describe dynamic nonlinear time series data in the form of probability distribution,the crisis state decomposition and inflection point identification are realized,Compared with the traditional method of risk dichotomy according to the fluctuation amplitude,the model can identify the risk inflection point more accurately,and can further identify the occurrence probability of each fluctuation risk on the basis of discovering the risk.At the same time,in view of the insufficient ability and unsatisfactory effect of the traditional fr,STV and other models on the early warning modeling of RMB exchange rate fluctuation risk caused by external shocks in the new era and new environment,this paper introduces the LSTM network model in the field of in-depth learning.Because the LSTM network has the advantages of multi-layer structure characteristics,feature filtering and automatic extraction of screening characteristics,It can realize the advantage of fitting data relationship with arbitrary accuracy.In addition,when fitting the data relationship,LSTM network does not require too strict the distribution of the original data as the traditional regression model,which makes the risk early warning model based on LSTM network have better effect and adaptability to the data.When using LSTM network model for RMB exchange rate fluctuation risk early warning modeling,through the practical test of external impact data and RMB exchange rate data in recent 24 months,22 early warnings and 2 false alarms are successfully realized,which shows that the system can better realize the early warning of external impact on RMB exchange rate fluctuation risk,and has certain practical value.Finally,through the analysis of the spillover effect of RMB exchange rate and the research on the transmission mechanism of external shocks,combined with the fitting results of the early warning system and the aspects that need to be strengthened in the actual management state of RMB exchange rate,suggestions are made from two aspects: the policy suggestions to prevent the spillover of RMB exchange rate fluctuation risk and the technical suggestions to build the early warning system of RMB exchange rate fluctuation risk,Provide technical guarantee and support for the practice and implementation of the new development pattern of domestic and international double cycles and mutual promotion proposed by the general secretary.