i’m quite frustrated w the charts on page 277 of volume 3. if anyone could try and explain would be nice…

With the increase in risk, how much of the various assets in the portfolio would you hold.

The key take-away from these graphs is that Plain old Mean-Variance optimization based on historic expected return and covariance matrix will generate highly undiversified asset allocation (extremely high weights in one asset) Whereas Black Litterman approach will yiled better diversified asset allocations.

got it, thanks guys: i think i’ll look at it like … the combination of asset class weights that result in a given level of standard deviation