RiskMetrics (EWMA)

Discussion in 'P1.T2. Quantitative Methods (20%)' started by idalton, May 1, 2008.

  1. Hi Sam,

    I don't have experience in backtesting EWMA, sorry. Did you look at
    http://www.riskmetrics.com/publications/working_papers/backtesting.html
    (free with registration)

    Re: tolerance, I just assumed is was ~ significance (e.g., only want to fall outside tolerance interval 1%) but maybe I shouldn't make that assumption about their specific use of tolerance...thanks, David
  2. nmmsantos@gmail.com

    nmmsantos@gmail.com New Member

    Hi Sam,

    I'm doing a Master in Economics, and have a little paper to do about EWMA. Right now I'm "freezed" because, even after reading Chapter 5 of RiskMetrics to find de optimal lambda, I don't know how to compute it in Excel. I don't know if it is too much to ask if you can share your Excel file with us?
    Thanks in advance.


    Best Regards,
    Nuno
  3. nmmsantos@gmail.com

    nmmsantos@gmail.com New Member

    Hi,

    I think I already got to find the optimal lambda, but there is one question that remains. If I change the number of days from 734 to 1509, the optimal lambda changes from 82,47% to 99,86%. Am I doing something wrong??
    I send my excel file with both situations.

    Thanks,
    Nuno

Share This Page

loading...