I saw this on 2011 Mock PM #29, having to calculate covariance of 2 markets based on market factor sensitivities and residual risk. I don’t remember seeing this in the 2013 curricculum - does anyone else?
Thanks!
I saw this on 2011 Mock PM #29, having to calculate covariance of 2 markets based on market factor sensitivities and residual risk. I don’t remember seeing this in the 2013 curricculum - does anyone else?
Thanks!
under cap mkt epecataions. it’s there
annoying but there. See 2013 mock #27
That’s a sneaky formula. You never know but I’d guess that it’s just there to make the mocks more challenging.
The LOS it’s under, 18C, is one of the most ambiguous descriptions of all time. “Demonstrate the application of formal tools for setting CME…etc…”
Anyone have a good way to remember that formula? It’s a fucking bear…
isn’t it just the formula to calculate the standard deviations of 2 correlated assets, just saying Var instead of sd^2?
Right. Use variance times beta coefficient for each.
you’re mixing up covariance and variance / std dev.
you also need to use beta instead of weight. just calculate the error term first from the given numbers then calculate it with the new correlation
Portfolio Variance formula. Check. Beta instead of weights. Check. Where does the error term fit in?
you will be given the varriance (A) of the index and 2 variances (B & C) and their betas that are correlated to the index the question will ask you to calculate a new variance given a new correlation.
Step 1) find the error term ( just a plug in)
Variance A = Variance B x Beta B squared + variance C x beta C squared +2 x SD B x SD c X beta b x beta c x corelation + ERROR TERM ( this is the plug in that you will have to find first)
step 2) calculate the ne variance given the new estimates.
Variance A = Variance B x Beta B squared + variance C x beta C squared +2 x SD B x SD c X beta b x beta c x corelation + ERROR TERM ( you add the error term found in step 1)
Gotcha. I’ll try that one out. Thanks!
The CFAI text made it look awful.
If you looked at the problem in the mock in a different way … you actually do not need to do anything at all.
2 ways that I know of:
Way 1:
They gave the current variance = 0.2704 with 0 correlation -> 0.52 (or 52% std dev).
they told you that correlation was now 0.25 … so std dev must go up. Only one answer choice 56.4 met that criterion.
So you have your answer right away.
Way 2:
How to calculate to make certain?
New term that gets added = 2 x SD B x SD c X beta b x beta c x corelation
= 2 * 1.02 * 1.045 * 0.0784^0.5 * 0.1024^0.5 * 0.25 = 0.04775
Add to 0.2704 = 0.3181
sqrt (0.3181) = .564 = 56.4%