Thursday, May 24, 2007

2 Fund Seperations

In a two fund monetary seperation, is the risky fund always Mean Variance Efficient?
If so, is it necessarily the tangency portfolio (defined by the line between the Risk free and the Frontier) or can it be any mean variance efficient fund?

3 comments:

Barney Hartman-Glaser said...

Yes the risky fund must be a frontier portolio, see H and L section 4.2 and 4.12. And in response to the second question, the answer is yes again, try drawing a picture and you see that the tangency portfolio always gives the best trade off between mean and variance.

Nish Rajan said...

Thanks Barn. I think my confusion came in the fact that I thought the frontier was a hyperbola (in ER and sigma space) when infact with the risk free asset its actually the cone?

If that were true, then the second portfolio has to be the tangency one because its the only combination of risky assets thats MV optimized. Market clearing on risky assets (given rf in zero net supply) would then make the tangency pfolio the market pfolio. Does that explanation make sense to you?

Barney Hartman-Glaser said...

Yes I think that is correct.