It has appeared couple of time in exams, just like to understand it better
Unconstrained Black - Litterman (UBL)
Uses market cap weight of asset classes in a global benchmarks - Neutral starting point
Adjust up/downward to reflect investors opinion (according to Bayesian procedure)
Unconstrained (I guess no negative weights here)
Advantages - Well diversified (since global bencmark weight are starting point), overcomes estimation error problem in conventional MVO
Black-Litterman
Equilibrium market returns as Neutral starting point
To have equlibrium mkt return, first need to define Equilibrium mkt weights & covariance - these weight (considered optimal) are taken same as in UBL - from a diversified global B/m - back solve return - I guess this is why it is also called as REVERSE OPTIMISATION )
Weights are given no special insight. Use same weight as in global bencmark.
On these market equilibrium returns have views & confidence levels of investors & ADJUST
They call it then _ View - adjusted market equilibrium returns _
Run mean variance optimization get Efficient Frontiers & portfolio
Advantages :
Considers current market condition & investors expectations (improvement over historical return based MVO) - collaborating market equilibrium returns (i guess present mkt scenario) with investors views helps to dampen the extreme views investor may hold which could have dominated optimation
Well diversified (same as UBL - anchored to global B/m weights)
Overcomes estimation error problem in conventional MVO
REF
Run optimation many times - hovering around the point estimates return & covariance. (I think they use Monte carlo simulation here?)
Each simulation trial generates set of return & variances - simulated efficient portfolios
Intergrate many simulated efficient portfolios into one become REF (here integrating into one leads to averaging of each input)
Advantages: Stable, diversified than traditional MVO, Disadvantage - no therotical support
Please share your inputs if you have understood these processes differently…
We call this unconstrained Black–Litterman model, or UBL model, because the procedure does not allow non-negativity constraints on the asset-class weights.
Words here are confusing…But yes i realized unconstrained every where meant negative weights are allowed but should equal to 1
So you are saying they are the same but investor views are reflected by adjusting weights in UBL (e.g. overweighting areas you expect to outperfom) but returns are adjusted in BL (simply increase the return if you tink it will outperform rather than overweighting)