Font Size: a A A

Application For Extreme Value Theory In WVaR And Empirical Analysis

Posted on:2018-11-30Degree:MasterType:Thesis
Country:ChinaCandidate:G W ChenFull Text:PDF
GTID:2370330515996132Subject:Statistics
Abstract/Summary:PDF Full Text Request
Financial market risk management has always been one of the core frontiers of financial research.With the persistent innovation in the market,research of financial risk measurement methods is also continuously exploring.Many previous studies have focused on the properties such as mean,variance,correlation,etc.,relatively have less attention on extreme events in financial assets.However,in the financial market,extreme risks in the event,will cause huge losses to investors,especially in financial crisis.Thus,there is an urgent need for an effective risk measurement tool on extreme risk.In the stock market,investors' investment behavior,such as the judgement of financial assets price and their risk attitude will also have a certain impact on the stock market price movements.Therefore,how to build a risk method which can effectively reflect the investors' risk appetite will be a very worth studying topic.Appropriate risk measurement methods should consider the volatility of financial markets as well as focusing on investors' risk attitude.In this paper,we focus on WVaR,which mainly discusses the application of extreme value theory in WVaR and firstly introduces the definition and main properties of traditional VaR and CVaR,followed by comparing the properties between them.Simultaneously,based on the extreme value theory and method,the application method for WVaR of generalized Pareto POT model is established to study the risk measurement of Shanghai Composite Index,Shenzhen Component Index,S&P 500 Index and Nasdaq Index.The results of empirical analysis show that the exponential return series has obvious fluctuation clustering in both the domestic emerging market and mature foreign capital market.At the same time,it has the characteristics of left partial and spike tail,and the generalized Pareto distribution can be better in fitting the extreme daily return data.In contrast to other measures that do not involve investors' risk appetite,WVaR,which includes investors' risk attitude,measures the risk of financial markets more accurately.In the same market,the risk value is not much different and coexistence is found.The market risk of domestic emerging market is larger than the relatively stable and mature foreign market.
Keywords/Search Tags:VaR, CVaR, WVaR, generalized Pareto distribution, peaks over threshold
PDF Full Text Request
Related items