Having trouble with understanding the theory behind ICAPM, and would appreciate if anyone can explain it!
I understand we’re deriving it from the original CAPM ----> Ri = Rf + B (Rm - Rf)
I understand that we are solving correlation coeffient formula (Corr = Cov (x,m) / (std dev mkt x std dev investment)) for Cov (x,m) and are subsituting it into the Beta formula (Cov (x,m)/mkt variance), which gets us to:
B = (Corr (x,m))(std dev investment) / (std dev mkt)
Then that gets subsituted back into the CAPM equation.
Ri = Rf + (Corr (x,m))(std dev investment) (Rm - Rf) / (std dev mkt)
After that, they simplify the ERPs:
ERP investment = (Corr (x,m))(std dev investment) (ERP mkt) / (std dev mkt)
Essentially that can be reduced to the following:
ERP investment = (Corr (x,m))(std dev investment)(Sharp ratio of market)
…but after that I get lost.
Why do they assume the correlation coeffient is equal to 1? And can someone explain the concept of market segmentation? …Thank you!
market integration means assets with identical characteristics (risk, payoffs, etc.) should be priced the same regardless of the market they are traded on (price is the same in New York, London, Paris, Tokyo or wherever). As international trade expands and emerging markets open up, more national markets are becoming integrated with one another.
Segmentation is just the opposite. Two identical assets will trade at different prices on different markets. There is a spectrum between perfect segmentation and integration and developed, open markets are further along to perfect integration.
Correlation assumed to be 1 because in a perfectly segmented market, the GIM and the local market are the same, therefore correlation with GIM = 1.0 (in a perfectly integrated market we measure that national market’s correlation with the GIM because that is the relevant benchmark, but if your perfectly segmented, the benchmark is itself.)
You might just wanna check thouch. This is based on my recollection of the material which I haven’t reviewed in 2 months.
Thanks, that is helpful.
So from the example I am looking at, they use the formula to solve for the ERP in multiple scenarios. Firstly, just using the formula I stated:
ERP investment = (Corr (x,m))(std dev investment)(Global sharp ratio)
Secondly, using the hypothetical scenario of fully segmented:
ERP investment = (std dev investment)(Global sharp ratio)
Then they construct a weighted average between the two scenarios (of complete integration and complete segmentation), using the degree of market integration as the weights. Adding back Rf gets them to the expected return, which makes sense. Okay, it’s starting to come together. Thanks again!
I lectured on this in an Economics class at UCI (University of California, Irvine) a few years ago, and I probably have my PowerPoint slides somewhere. I’ll look around.