G
Gareth
Member
Does anyone else think this question was unusual?
The solutions have terms that are not in core reading, Hull, or any other book I have seen.
6(ii)(a) - the market risk sensitivities idea, seems reasonable, but taking a hedge based on a portfolio consisting of a combination of the daily estimated deltas with respect to each benchmark instrument...you got to be kidding??
(iii)(a) - "There is also the volatility of the Vega which also have a DVAR." - this seems rather a strange statement. I can't quite see what they mean by the volatility of the sensitivity of the portfolio with respect to volatility.... and the DVAR bit i presume is referring to measuring the loss the will not be exceeded by probability x, within a day due to changes in volatility...wtf! So how exactly will we do this? Assume volatility follows a normal distribution etc etc, or do we use a historical method and compute the daily change in implied volatility over the last 500 days, and use this to compute 500 scenarios of how the portfolio will change over the next day etc etc.
So, to me this question seems horribly mixed up and poorly expressed. If someone can actually point me to a book that suggests using such unusual method of measuring market risk, I will be surprised.
The solutions have terms that are not in core reading, Hull, or any other book I have seen.
6(ii)(a) - the market risk sensitivities idea, seems reasonable, but taking a hedge based on a portfolio consisting of a combination of the daily estimated deltas with respect to each benchmark instrument...you got to be kidding??
(iii)(a) - "There is also the volatility of the Vega which also have a DVAR." - this seems rather a strange statement. I can't quite see what they mean by the volatility of the sensitivity of the portfolio with respect to volatility.... and the DVAR bit i presume is referring to measuring the loss the will not be exceeded by probability x, within a day due to changes in volatility...wtf! So how exactly will we do this? Assume volatility follows a normal distribution etc etc, or do we use a historical method and compute the daily change in implied volatility over the last 500 days, and use this to compute 500 scenarios of how the portfolio will change over the next day etc etc.
So, to me this question seems horribly mixed up and poorly expressed. If someone can actually point me to a book that suggests using such unusual method of measuring market risk, I will be surprised.
