BACKTESTING METHODS FOR VALUE AT RISK MODELS
Value at Risk is a loss estimation that could occur for a given holding period with a given confidence level. Estimated losses are certainly not always the same as the actual losses incurred. An event called an exception if the real nominal of loss is greater than the nominal of estimated loss (VaR)...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/19570 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Value at Risk is a loss estimation that could occur for a given holding period with a given confidence level. Estimated losses are certainly not always the same as the actual losses incurred. An event called an exception if the real nominal of loss is greater than the nominal of estimated loss (VaR). VaR models made from Variance - Covariance, Historical Simulation and Monte Carlo method could produce exceptions, so that the accuracy of the VaR Models must be tested.
Backtesting methods can be used for testing the accuracy of VaR models. These methods test the pattern of exceptions that generated by the VaR models. In this final assignment, we discuss some backtesting methods based on its frequency (Unconditional Coverage) and its independency (Conditional Coverage). For scattering exception will be tested by modified Unconditional Coverage Test
Some numerical experiments are done to show the performance of the backtesting in testing some VaR models. |
|---|