In the field of risk management, certain types of risks, such as market risk, credit risk, liquidity risk, were the main concerns of the asset management of every financial institution. Nowadays, the increase in volatility of financial variables in recent years imposed great risks on the business of financial institution. Especially, volatility and Fat-tail in VAR methodologies are very important.
The main purpose of this study is to estimate the VAR with accuracy and to provide the risk managers with the appropriate Value-At-Risk methods. This thesis examines volatility and Fat-tail in the Korea financial market by using real portfolio data in the market. And this paper propose a new method for VAR estimation which is a mixture of two approaches, where this paper combine Delta-normal (Variance-Covariance) method with Historical Simulation Method. The method is "His + Del method". The method enables us to accurately estimate the tails of a distribution.
This paper also explains the concept of VAR, and then describes in detail the other methods for VAR estimation: Historical Simulation Method, the Variance-Covariance method, Exponential Method and Monte-Carlo simulation. Finally this paper discusses the advantages and disadvantages of the three methods for computing VaR.
