EXPLORATION THE RELATION BETWEEN VALUE-AT-RISK AND CONDITIONAL VALUE-AT-RISK
In general, Value-at-Risk (VaR) and Conditional VaR (CVaR) are detection tools that used to measure a risk at corporation risk management. Analytically, VaR and <br /> <br /> <br /> <br /> <br /> CVaR predict a loss by calculation quantile of distribution. Therefo...
Saved in:
| Main Author: | |
|---|---|
| Format: | Theses |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/17991 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | In general, Value-at-Risk (VaR) and Conditional VaR (CVaR) are detection tools that used to measure a risk at corporation risk management. Analytically, VaR and <br />
<br />
<br />
<br />
<br />
CVaR predict a loss by calculation quantile of distribution. Therefore, VaR and CVaR can be compared to get optimal risk measures. The optimal risk measures <br />
<br />
<br />
<br />
<br />
can be obtained by the accuracy from VaR-CVaR. The accuracy VaR-CVaR can be achieved by calculation coverage probability (CP) and correct VaR-CVaR. VaR- <br />
<br />
<br />
<br />
<br />
CVaR method use the standard method, by HS-VC-MC method and the advanced method, by VaR-CVaR estimative and improved. This Thesis uses IBM stock price return at 29 Februaru 2008 - 28 February 2013 as return data. VaR estimative is an optimal risk measure for predicting loss, with CP and correct VaR value close to the confidense level which given. The accuracy of VaR-CVaR for time series models is affected by the magnitude of the data association (dependence). The result with <br />
<br />
<br />
<br />
<br />
Autoregressive(1) or AR(1) model shows that VaR and CVaR more accurate at relatively small association, -0.0852. This is due to return have small association value, 0.038. |
|---|