Black Litterman model steps in to fill the gap between market views as implied by current and historical prices and an advisor's forward looking views. However, advisors have to assess future market moves and explain them to the clients. Often, advisors and their employers (fund managers) are reluctant to take a position that is radically or even moderately different from the consensus view that is already reflected in current market prices. This causes their recommended portfolios to be close to efficient portfolio under MPT.

An active manager that is truly active about managing evolving risks by incorporating his views into portfolio construction can be an investor's best friend.

Portfolio models already work with asset classes to minimise inapplicability of models and various indices fill a vital gap. this also solves the problem for investors who do not have the bandwidth and know better than trust half cooked research. That means of course that there is no substitute for comprehensive research and individual asset /stock selection for each investor and abnormal returns for those whoi invest in knowledge.

Samit Ahlawat, Black Litterman approach does not suggest a new solution to the efficient portfolio problem. My understanding of current situation in portfolio management is that the very term "efficient portfolio" must be reviewed and revised in order to reflect recent development, namely technological and AI.

In my opinion the main shortcoming of the MPT is its extreme sensitivity to the errors of the market parameters estimations. With simulated scenarios (as long as we exactly know the "genuine" expected returns and covariance matrix) it works pretty well. Interestingly, Markowitz and Kelly often deliver similar optimal portfolios (though conceptually they are totally different).

But hardly can we (straightforwardly) apply Markowitz to the real market data since the "optimal" Portfolio might be far away from real optimum due to parameter estimation errors, let alone that they ever change.

Ilia Drozdov, Black Litterman approach augments the MPT not by suggesting a new solution but by providing a method to incorporate individual views within the MPT framework. AI and other methods for prediction can be integrated into Black's approach. But as you pointed, most of the supporting research work is in its infancy stage.

A remarkably out-of-date submission about modern portfolio theory in practice. It seems a stylized perception of Markowitz practice, like zombies, will never be able to die an honorable death. The key problem with Markowitz per se is instability and ambiguity. Black-Litterman is not a solution as we have proved in Michaud, Esch, Michaud (JOIM 2013 1st quarter). BL is nothing more than disguised and dangerous Markowitz.

@Richard, why not add a reference to your paper? I would be interested in reading that.

Michaud, Richard O. and Esch, David N and Michaud, Robert, Deconstructing Black-Litterman: How to Get the Portfolio You Already Knew You Wanted (August 1, 2012). Available at

SSRN: http://ssrn.com/abstract=2641893 or

http://dx.doi.org/10.2139/ssrn.2641893

this is pre publication. Published paper at the Journal Of Investment Management, 2013, 1st quarter.

@Richard, thanks. I will make sure I read it. Much appreciated.

Indexing will prove ephemeral, over-rated, & sub-par in years ahead.

@Graeme, IMHO, ephemeral: no. They still have a lot of years to go. Over-rated: yes. Sub-par: definitely.

@Richard, I think your paper did resume the situation quite well: “Essentially, BL optimization solves estimation error in MV optimization by ignoring it.” And I have to agree.

Interesting work, thanks for sharing. I appreciated your point of view.

Good Read. MPT was formulated in Markets long ago. Today real diversification in assets classes is hard to find.

We rebalance every month with recent correlations and a momentum strategy in our fully algorythm Based index portfolio

In the 70's it took work to draw out the charts and provide the numbers for valuations relative to market price action. Now every software program offers this completely free on every trading platform. Asymmetries of information gap has become much narrower. I think making money in the markets is now about understanding systems working with a software developer perspective.

The guys from the Mythbusters should dissecate those theories in one of their next TV shows....

It's great to see a Black-Litterman author in here, in this thread (Richard M., I remember your name from my last school project, although I can now see I didn't exactly quote these articles in my work, which I provide a link to below). I really enjoyed learning and working with the Black-Litterman-model - https://www.dropbox.com/s/1n4srhkvr9l6bqr/report3.pdf?dl=0

In my opinion answer is simple, cause MPT was created with an assumption that risk and expected return are stable over time - market is efficient or tends to. We know that this condition is not passed. We may not avoid tendencies of risk and return, but we can reduce risk related with it by rebalancing as a tool for our portfolio (to try fit portfolio under our investment aims). If it will be no opportunity to use tendency in the our performance - rebalancing has no longer impact on the investment result. Against to above MPT will be still in memory of everyone, cause everyone expect stable market behaviours. The same, everyone try to diversify our investments which is in real half part of reducing risk. Based on my research is equal-weighed diversification enough for every amateur, where MPT is a simple tool if you want to get more than benchmark. I am happy that I had a chance to create simple tool and implement MPT and I called it 'mappstock'. Nice article, thank you for sharing.

I am sure i will shock many in this thread and for those who think my views may have interest in understanding quantitative methods when i make the following statement that i very much believe: markowitz invented modern finance with the notion of the markowitz efficient frontier in march 1952. more than sixty years later (almost) no one understands markowitz. if you think it requires the notion of a normal return distribution (or log normal, t-distribution, anything you can think of) you are wrong. if you think markowitz is dependent on a quadratic utility (or any other utility function you can think of) you are wrong. It is simply a description of informed investor behavior in large capital markets. consequently it is as relevant as a framework for understanding the behavior of many capital markets circa september 23 2016 as it was in march 1952.

I agree completely with Mr. Michaud.

I completely agree with you Mr. Richard Michaud . I can only add, that Markowitz's article from 1952 pushed me to create video, when we can observe Efficient Frontier, and feel advantage how information about risk is important by comparison between two portfolios: one equal-weighted (no need any information) and one from efficient frontier line ( link: https://www.youtube.com/watch?v=fCncxWzuq1k ). It is not about what portfolio is the best, but how to manage information that we have to reduce risk of performance. Best Regards.

Thomas Brody When I saw the article I thought to myself that either article is wrong or I'm doing wrong things. Then I saw my results in investing ... and I said to myself: maybe it is better how it is. Maybe it is better when investment managers, investing authorities and other people say that Markowitz can't be true (becouse they don't understand him or can't use MPT in a real life). I like this kind of attitude. And I'm really happy that not everybody understands MPT and can use it. Therefore I can earn on average more than 3% a month on SNP500... every month since many years. GWF Shape is not purely Markowitz's MPT - but it proves that you can earn and earn and earn ... based on math, calculations and 0% of luck or guessing.

@Wojciech - I saw your video. Nice piece of work. It's really close to something really big. I bet that in one or two months you will develop a working solution. Or maybe you have done it already? I would love to see how it works. Cheers!

When I saw the article I thought to myself that either article is wrong or I'm doing wrong things. Then I saw my results in investing ... and I said to myself: maybe it is better how it is. Maybe it is better when investment managers, investing authorities and other people say that Markowitz can't be true (becouse they don't understand him or can't use MPT in a real life). I like this kind of attitude. And I'm really happy that not everybody understands MPT and can use it. Therefore I can earn on average more than 3% a month on SNP500... every month since many years. GWF Shape is not purely Markowitz's MPT - but it proves that you can earn and earn and earn ... based on math, calculations and 0% of luck or guessing.

@Wojciech - I saw your video. Nice piece of work. It's really close to something really big. I bet that in one or two months you will develop a working solution. Or maybe you have done it already? I would love to see how it works. Cheers!