Hi David,
I got the following questions after studying the Foundations of Risk note. I just put them together here so I can easily follow. Hope you do not mind.
1. p12. Could you explain why VAR can be aggregated? I feel assets can have correlations..
2. It seems there are 3 definitions of "tracking error volatility" on p29.
a) In the 1st line, it is defined as the SD of the return on the active portfolio. (BTW, I suspect the "12" in the ensuing formula should be "n" instead)
b) there is another tracking error volatility (TEV) in the box which is defined as SD(TE) = SD(SD(P-B)).
c) However, the example on that page seems to be a tracking error example, so TEV = TE?
I am confused, are these 3 definitions equivalent?
3. on the bottom of p32, is the "factor forecast" expected excess return in the formula above it?
4. on p34, could you elaborate The Dual of Maximum likelihood estimation?
5. Is the calculation in the example on p44 required?
Thank you so much!
I got the following questions after studying the Foundations of Risk note. I just put them together here so I can easily follow. Hope you do not mind.
1. p12. Could you explain why VAR can be aggregated? I feel assets can have correlations..
2. It seems there are 3 definitions of "tracking error volatility" on p29.
a) In the 1st line, it is defined as the SD of the return on the active portfolio. (BTW, I suspect the "12" in the ensuing formula should be "n" instead)
b) there is another tracking error volatility (TEV) in the box which is defined as SD(TE) = SD(SD(P-B)).
c) However, the example on that page seems to be a tracking error example, so TEV = TE?
I am confused, are these 3 definitions equivalent?
3. on the bottom of p32, is the "factor forecast" expected excess return in the formula above it?
4. on p34, could you elaborate The Dual of Maximum likelihood estimation?
5. Is the calculation in the example on p44 required?
Thank you so much!
