Font Size: a A A

The Study Of Financial Risk Measurement Based On High-Frequency Data

Posted on:2014-01-01Degree:MasterType:Thesis
Country:ChinaCandidate:Y DuFull Text:PDF
GTID:2269330425980003Subject:Quantitative Economics
Abstract/Summary:PDF Full Text Request
The study of high-frequency financial data has been a new direction for financial econometrics in recent years. Financial risk measurement is the most important part of risk management with the financial deregulation. This paper studies the models of high-frequency data in financial market, and applies to estimate VaR one common method of measuring financial risk. The main achievements are listed as follows:(1)This paper makes a detailed introduction about high-frequency data, and raise realized volatility and three improved high-frequency volatility, adjust realized volatility, realized bipower variation, and weighted realized bipower variation. Weighted realized bipower variation is better on robustness and valid to estimate the volatility with the consideration of calendar effect,.(2)This paper makes comparison between three ways to choose optimal time interval, and finds that the method of variance is the easiest method to get the best time interval of sampling. We also make an empirical study to find the weighted realized bipower variation is a better estimator compared with other realized volatility.(3)This paper builds an ARFIMA model based on weighted realized bipower variation and comprises with GARCH model to prove the former is a better one.(4)This paper put forwards value at risk model replacing realized volatility with WRBV. We also make an empirical study of the value at risk model with the data of stock index futures, and contrast to traditional risk measurement model. In order to assess the VaR model, the paper presents the Kupiec failure rate test to evaluate the effect and finds the WRBV-VaR model is preferably.On the whole, the VaR risk measurement model integrate different market factor and risk set into a number and is compositive to measure the financial risk. However, this model can predict and control risk under the normal market conditions and is not applicable in irregular market. The model in the paper is limited to measure financial risk, and has to improve itself to get better effects in future.
Keywords/Search Tags:Weighted realized bipower variation, High-Frequency Volatility, VaR
PDF Full Text Request
Related items