This is first post in series of implementation of VaR in banks. Almost three fourth banks uses the historical simulation rather parametric or Monte Carlo methodologies for calculating bank's VaR numbers. It is not that this is best method to calculate the VaR or it captures risk correctly and completely than Why should this be so – what are the advantages of historical simulation over the other two approaches?
1) The main advantage is that historical VaR does not have to make an assumption about the distribution of the risk factor returns. Other methods can incorporate the multivariate distribution and skewed tail risk factors. But in historical simulation method this is assumed that historical returns factored in all these skewed returns and multivariate distribution.
2) It is easy to calculate and less computation intensive compare to Monte Carlo method.
3) Regulators ,Basel committee, have accepted this method for the calculation of regulatory capital requirement of the bank. Though Basel committee guidelines have other provisions like back testing add-on of VaR model to capture the risk that can be missing in VaR.
4) This method captures fat tails as fat tails occurs more frequently than it has been assumed.
However there is lot of controversy on using the historical simulation method. Some of the above benefits have been challenged and discussed in the How historical simulation made me lazy paper.
Success of historical method depends upon the underlying historical return that is getting used to calculate the return. Data can be stale or this historical time series may not be making much sense in today's time due to change in economic scenario (expansion or decline) , volatility , regulation change.
Various flavor of historical simulation has been developed. we will discuss them in subsequent posts.
1) The main advantage is that historical VaR does not have to make an assumption about the distribution of the risk factor returns. Other methods can incorporate the multivariate distribution and skewed tail risk factors. But in historical simulation method this is assumed that historical returns factored in all these skewed returns and multivariate distribution.
2) It is easy to calculate and less computation intensive compare to Monte Carlo method.
3) Regulators ,Basel committee, have accepted this method for the calculation of regulatory capital requirement of the bank. Though Basel committee guidelines have other provisions like back testing add-on of VaR model to capture the risk that can be missing in VaR.
4) This method captures fat tails as fat tails occurs more frequently than it has been assumed.
However there is lot of controversy on using the historical simulation method. Some of the above benefits have been challenged and discussed in the How historical simulation made me lazy paper.
Success of historical method depends upon the underlying historical return that is getting used to calculate the return. Data can be stale or this historical time series may not be making much sense in today's time due to change in economic scenario (expansion or decline) , volatility , regulation change.
Various flavor of historical simulation has been developed. we will discuss them in subsequent posts.
No comments:
Post a Comment