xaviq@live.nl
New Member
My question is about how to calculate the VaR using extreme value theory. I would like to thank anybody taking the time to answer these questions.
I Started with the daily data. Fitted the generalized extreme value distribution to the negative daily percentage log-returns, using subperiods of 21 trading days. I used a regression approach for starting values. I have regression and maximum likelihood estimates, with standard errors, and the attained loglikelihood.
How can i now extract both the 1% and 5% VaR, at the last day of the sample?
How would the use of a AR(1)-GARCH model be helpful and applied ?
I Started with the daily data. Fitted the generalized extreme value distribution to the negative daily percentage log-returns, using subperiods of 21 trading days. I used a regression approach for starting values. I have regression and maximum likelihood estimates, with standard errors, and the attained loglikelihood.
How can i now extract both the 1% and 5% VaR, at the last day of the sample?
How would the use of a AR(1)-GARCH model be helpful and applied ?
