Research On Measurement And Risk Identification Of China's Financial Stress Index

Financial stress index is a risk index with wide coverage and strong representativeness.It has stronger continuity and timeliness,and plays a very important role in the process of measuring financial risk.It can not only describe the operation of the financial market in a multi-dimensional and systematic way,but effectively describe the extreme values of the financial market.Firstly,this paper divides China's financial market into four sub markets,and constructs a financial stress index measurement system including four sub markets and11 risk indicators.China's financial stress index is calculated by using methods of the dynamic critic and the equal proportion to represent the risk situation of China's financial market,and uses the nuclear density analysis method to fit the distribution curve of the financial stress index.The distribution characteristics are analyzed in detail.Secondly,using the B-N data decomposition method,the financial stress index is divided into three components: random impact,determined trend item and periodic item.Taking the determined trend item as the standard,the high-risk and low-risk periods of China's financial market are divided.Finally,this paper estimates the high and low risk conversion probability of the financial stress index by using the Markov regime conversion model,and compares it with the results identified by the deterministic term,and summarizes the differences between the two.The innovations of this paper are as follows: Firstly,the trend term obtained by B-N data decomposition method is used as the threshold of high and low risks to identify China's financial stress index.This method makes up for the deficiency that Markov regime transformation model can not give the threshold of high and low risks,and enriches the relevant theories of financial risk identification;Secondly,two different recognition methods are compared,and the advantages,disadvantages and applicability of the two methods are summarized.Research conclusions: Firstly,compared with the equal variance method,the dynamic critical method can not only accurately describe the main risk events in the sample period,but also has more obvious advantages in describing the extreme values of the financial market.Therefore,the financial stress index measured in this paper has strong rationality and is in line with the operation trend of China's financial market.Secondly,according to the results of B-N data decomposition,random shock and deterministic term play a very important role in the change of financial stress index.Random shock term affects the change trend of financial stress index,and deterministic term reflects the average risk level faced by China's financial market in the long term.Thirdly,the high-risk and low-risk conversion probability obtained according to the district system conversion model shows that in the sample period,the probability of maintaining China's financial stress index in the low-risk state is 0.9644,while the probability of switching from high-risk to low-risk state is 0.0986,which shows that China's financial market is mainly in the low-risk period.According to the identification results based on the deterministic items,the high-risk periods of China's financial market include 2004,2008-2009 and 2014-2015,and the low-risk periods are 2012-2013 and 2017.This conclusion is close to the recognition result of the mainstream Markov regime transformation model.Finally,this paper compares the advantages and disadvantages of the two identification methods.The zone conversion model can accurately estimate the low-risk probability and high-risk probability of each financial stress index,but it lacks the explanation of the risk conversion theory,and the identification method based on the deterministic term can make up for the deficiency of its risk conversion theory,The combination of the two methods can improve the effectiveness and accuracy of recognition.