The Performance of VaR Measurements－The Empirical Studies of Currency Exchange Rates
碩士 === 輔仁大學 === 金融研究所 === 88 === Since financial asset returns don’t often follow the hypothesis of normal distribution（such as high kurtosis and fat tails）, this study considers not only normal distribution but also GARCH model. In addition, we try to evaluate VaR by mixture of normal distribution...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2000
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/19850987812501983696 |
| Summary: | 碩士 === 輔仁大學 === 金融研究所 === 88 === Since financial asset returns don’t often follow the hypothesis of normal distribution（such as high kurtosis and fat tails）, this study considers not only normal distribution but also GARCH model. In addition, we try to evaluate VaR by mixture of normal distribution since there are some abnormal events that induce large volatility of asset price. Our empirical studies also compare their performances with the performance of Monte-Carlo simulation, historical simulation. And we try to find an objective risk-disclosing way by comparing the performances of these methods.
The empirical results are as follows: 1.In general, GMB（mixture of normal with GARCH and simulation） has a good ability to forecast risk because it combines the properties of GARCH、mixture of normal and simulation. GMIX（mixture of normal with EWMA and simulation） also has a good results but worse than GMB. That is because GMIX evaluates variance by EWMA（exponential weighted moving average,λ=0.94）. 2.HIS（historical simulation） and EQW（equal weighted moving average） are easy to calculate and are good risk measurements if the structure of data didn’t change. 3.Most methods over evaluate the risk when there are thin tails at out-of-samples period.
|
|---|