Here is a quick explanation of parametric value at risk (VaR) as a means to illustrating its strengths/weaknesses. Please note:

• The essence of parametric VaR is "no data:" while historical data is surely used to select a distribution and calibrate its parameters, a parametric VaR leans on a statistical distribution to infer losses
• In this illustration, I use the normal distribution which is typical. The normal is nice because we only need two parameters (expected return and volatility), but we don't need to use the normal! We can use other distributions; e.g., lognormal.

Screencast:

Comments

1. Rob 05 Feb 2010

   Very nice video, but i have one doubt say if I am working on finding the VaR of a portfolio and have monthly data, And I want to scale the VaR over one year time horizon. what should be the ‘dt’ term in the equation.
   Thanks,
   Rob.

2. Rob 05 Feb 2010

   Nice Video,
   I have a doubt on how to scale the VaR over one year time horizon. Say I have monthly returns of a portfolio. In this case what will be the term ‘dt’ in the above mention equation.
   Thanks, Rob.

Leave a Comment

Your Name*

Your Email*

Your Comment*

Notify me of follow-up comments?

Submit Comment