Black-Litterman model's input

C

crosstheevil

New member
Joined
Jun 18, 2026
Messages
0
Reaction score
0
What is not the input of the BL model?
Investor’s capital market expectations.
Investor’s confidence in these capital market expectations.
Long‐term returns of asset classes.

The answer is Long-term returns. I suppose LT returns here mean the returns derived from the model, not the input. But how can “confidence” be an input? I’m perplexed …
 
There is a confidence parameter you need to input on each set of views, the lower your confidence in those views, the closer the final distribution is to the orignial implied.
All you need to know is that you put a confidence value (from 0% to 100%), and that then blends your views with the ones impled by the market portfolio, to give you the final weights of asset classes.
 
confidence of investor on the views - is by far the MOST important parameter in the BL model.
(you see confidence in the question anywhere - confidently mark BL … or select BL as the model used).
 
Anyway, how do you assign the confidence level (0% - 100%)?, I suppose there is no scientific way, it’s simply intuitive …
 
more than confidence level - it is as the text explains it …
he has view adjusted returns for the various asset classes …
so he increases returns for asset classes where he has higher confidence, reduces returns for the asset classes he has lower confidence in.
 
But if he view adjusts the asset returns, then why isn’t asset returns an input? ( the initial input)
 
long term … vs. these are short term … aren’t they?
and another thing - in most of level 3 - it is ASSET CLASS returns - not particular Asset returns.
 
cpk123 wrote:
more than confidence level - it is as the text explains it …
he has view adjusted returns for the various asset classes …
so he increases returns for asset classes where he has higher confidence, reduces returns for the asset classes he has lower confidence in.
Click to expand...
Confidence is a seperate parameter from adjusted returns.
The way I understand the model, you do not incorporate your statistically expected returns, but the absolute return, then there is a confidence value for that distribution that adjusts it depending on how confident you are in the predictions.
 
Mmmm yes, as red,
You overweight/underweight the Asset Classes with the higher/lower confidence sentiment.
For the Unconstrained BL you base your initial weight on an index or other.
For the BL you take the Expected Return on your base diversified portfolio (or other), you re-attribute each sector (or other) with his share of return and use those sectors return coupled with the confidence sentiment of the investors.
I also find the answer LT returns a bit strange as Expected Returns are involved in the calculation of the weight’s allocation for the BL.
Could you use LT returns as Expected Returns?
Batman might know, Superman? I am not sure,
 
You must log in or register to reply here.
Back
Top