Hi there,
I would appreciate if someone could help me clarify some of the questions I had on Portfolio Value at Risk using the Variance-Covariance approach. Details as follows.
Number of Stocks: 5 (all in US Market so no FX impact)
Data: 5 years of price history
To Calculate: 95% Daily VaR
STEPS:
-
Calculate daily return on stocks for the five year history
-
Get a 5 * 5 correlation matrix based on the calculated returns (Using Correl() function in excel)
-
Get the standard deviation using stdev() on daily returns
Question: Since I am using the five year history to calculate daily returns, do I need to scale the standard deviation?
- Calculate individual VaR using the current market values for each stock position.
1.65 * std dev * current mkt value of position
-
Transpose indivdual VaRs to get a column vector and matrix multiply with correlation matrix using MMULT()
-
Matrix multiply the result from previous step (a row vector) with indivdual VaRs
-
Take square root of the result
Makes sense?