Model Risks
About the Lecture
The lecture Model Risks by Edu Pristine is from the course Archiv - Operational Risk. It contains the following chapters:
- Model Risk
- High Level Categories of Model Risk
- Model Specification Risk
- Other Risks
- Quantifying Model Risk
- Managing Model Risk
- Procedure for vetting and reviewing a model
- Concept Checkers
Author of lecture Model Risks
Edu Pristine
Customer reviews
5,0 of 5 stars
| 5 Stars |
|
5 |
| 4 Stars |
|
0 |
| 3 Stars |
|
0 |
| 2 Stars |
|
0 |
| 1 Star |
|
0 |
Excerpts from the accompanying material
... with trying to capture an observed phenomenon using a financial model. If the output of a model is a ...
... heavy tailed, we might mistake a lognormal P/L process for a normal one, and so on, and it is rarely easy to identify the 'correct' process. Missing risk factors: We might ignore factors such as stochastic volatility or fail to consider enough points ...
... not replaced it when a superior model became available. We might run Monte Carlo simulations with a poor random number generator or an insufficient number of trials. Model Implementation Risk: ...
... period sample for modeling. Risk models are prone to calibration problems for volatilities and correlations. Programming problems: Errors arising from poor programming, ...
... 'Normal' so that expression will be distributed as with n - 1 degrees of freedom. This eventually results in a closed loop confidence interval for VaR that takes care of uncertainty about the volatility parameter. Two parameter models: An example of such a ...
... of Monte Carlo process as follows (it can be extended to n assets): Means would be drawn from normal distributions. Variances and covariance from Chi-square distributions ...
... Check results and test the proposed model. Independent Risk Oversight: This unit should encompass risk measurement as well as risk management, should be independent of line execution Areas, and ...
... Identify, evaluate and key assumptions: Users should explicitly set out the key assumptions on which a model is based, evaluate the extent to which the model's results depend on these assumptions, and check them as ...
... extremely revealing, and simple histograms or plots often show up errors that might otherwise be very hard to detect. Re-evaluate models periodically: Models should be recalibrated ...
... the temptation to overlook a model just because it's use has started spinning profits. Hear out concerns of mid-office and back-office in this context and do not over depend on front office traders. Avoid situations of model ...
... Independent Risk Oversight (IRO) unit. Ideally a middle office unit responsible for overall risk management and measurement. The unit head should report directly report to CEO and may even sit on the ...
... numbers generated in the model are biased. ii.Implementation risk can occur when different users input different parameters into the model. iii. An incorrect model application occurs when a well-specified model ...
... other units for it to function properly. Which of the following statements is most likely incorrect? A. The compensation of IRO should be based on the compensation of other units. B. A model should be properly vetted ...
... random number generator is a form of incorrect model application. A model specification error occurs when the ...
