| It is an extensive literature topic examining the systemic risk in the financialmarket which is the risk of the failure of a significant part of the financial system.This is because, in the Wall Street, many old, famous and prestigious financialcompanies have been hit hard, and some even have been destroyed. The Fed and theTreasury and other institutions lost their control over the situation in the hugefinancial tsunami vortex of a loss as to what to do. In order to study its cause andprevent the incident from happening again, scholars from financial system andsupervision have done a lot of researches, one of which is systemic jump risk thatcauses particular concern because it is a core part of systemic risk. Therefore, in orderto absorb the lessons of the American subprime crisis and do some research from thepoint of Chinese financial market to avoid Chinese financial market making the samemistake, the deep research on the systemic jump risk can provide a new perspective tosupervisors.The research on the jump risk of financial system needs to be based on the highfrequency data and stochastic process with the jump. However, ARCH and SV modelsusing the low frequency data can not meet the requirements of high frequency datamodeling, so we have to figure out the new ways. Andersen and Bollerslev (2000)made a pioneering work and proposed a new method to measure the volatility offinancial high frequency data-Realized Volatility (Realized Volatility, RV). On thisbasis, a great variety of non-parametric jump test methods has been researched.However, compared with the low-frequency data, the use of high frequency datacontains more information, and as the sampling frequency increases, the informationcontained will increase, which brought the market microstructure noise, to influencethe computation of RV and complicate the jump test. Therefore, in order to eliminatethe negative influence of market microstructure noise on these researches, the scholarshave to put forward a variety of methods to improve the computation of RV to get the better estimation of the volatility.In this paper, we use the noise adjusted variance swap test as our theoreticalapproach and high frequency data of one minute of Industrial and Commercial Bankof China (ICBC) from2006to2012to detect the jump situation and examine therelationship between the significant price jumps and macroeconomic policies andglobal events. Furthermore, we can further the understanding the characteristics of thesystemic jump risk. Specific content is organized as follows:The first chapter is introduction. It introduces the research background andsignificance of the paper. What’s more, I sort out research ideas, summary themethods and find out the entry point of the study. And then, I give the methods, thetopics and content of the research and the difference from current study.The second chapter talks about the current academic overview on the jump testresearch methods, including the parametric and non-parametric methods in the foreigncountry. And then, I undertake a review the literature of Chinese academy in therecent years.The third chapter puts forwards the theoretical methods of this paper-BipowerJump Test, Variance Swap Jump Test, Market Microstructure Noise and NoiseAdjusted Variance Swap Jump Test.In the forth chapter, I do the simulated experiments and compare the effectamong the methods that I talk above. I find that the variance swap jump test and noiseadjusted variance swap jump test is better than other methods.In the last chapter, I use the high frequency data of one minute of ICBC to do theempirical analysis of volatility of price and jump test. Then, under the identificationof price jump, I put the jump price and government’s macroeconomic policies andglobal economic events together. I find that price jump of ICBC is very commonduring this period and closed to Chinese monetary and macroeconomic policy,dividend policy and global financial events. This research can provide a newperspective to further the research of systemic risk of Chinese banking system. |