Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
Rick Ferri
Posts: 9703
Joined: Mon Feb 26, 2007 10:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Rick Ferri »

Summary of Why Optimal Portfolios Are So Difficult To Create.

Harry Markowitz won a Nobel Prize in Economics in 1990 for his work on a theory of portfolio management for individual wealth holders. Since that time, Modern Portfolio Theory (MPT) has become the bedrock for creating best-practice portfolio selection methods.

Yet, after several decades and extensive literature on MPT, accurately projecting an optimal portfolio forward strikes me as nearly impossible. The problem is the estimates. What goes into forecasting an optimal asset allocation has a big impact on what comes out. This often results in suggested optimal allocations that are quite inaccurate and may lead to portfolio decisions that are counter-productive.

Most MPT analysis relies on historical return, risk, covariance and correlation, at least as a starting point. We know that past data can be period-sensitive, so these estimates are often adjusted based on subjective input factors. Markowitz acknowledges the problem of using inaccurate correlation estimates for portfolio optimization in a recent paper he co-authored.

There are other ways to estimate portfolio risk and return that circumvent correlation estimates. One way is to assume there is no efficient frontier. Each asset class used in a portfolio has its own expected risk and return, and a portfolio of asset classes has an expected risk and return based on the weighted average risk and return of each asset class used in the portfolio. There is no assumed benefit from MPT.

Full article HERE

Rick Ferri
The Education of an Index Investor: born in darkness, finds indexing enlightenment, overcomplicates everything, embraces simplicity.
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Random Walker »

Hi Rick,
I haven't read the article yet, but an off the top of my head question arises from your summary. If correlations are at all less than one, wouldn't it make sense to diversify across risk factors/sources of return anyways? I guess what I'm saying is that isn't it best to bet on MPT and acknowledge we can't know the efficient frontier ex ante? I appreciate that the more we diversify across risk factors/sources of return, the greater the costs. And everyone needs to choose a balance in this arena.

Dave
RadAudit
Posts: 4384
Joined: Mon May 26, 2008 10:20 am
Location: Second star on the right and straight on 'til morning

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by RadAudit »

Thanks for the article.
Rick Ferri wrote:accurately projecting an optimal portfolio forward strikes me as nearly impossible.
Well, that saves me a lot of effort.
Rick Ferri wrote:There are other ways to estimate portfolio risk and return that circumvent correlation estimates. One way is to assume there is no efficient frontier. Each asset class used in a portfolio has its own expected risk and return, and a portfolio of asset classes has an expected risk and return based on the weighted average risk and return of each asset class used in the portfolio.
I just assume I won't ever find the efficient frontier and sort of go with the second approach. Seems to be working so far - but I probably wouldn't know if a 50 / 50 AA won't work until way after the portfolio crashes and burns.
FI is the best revenge. LBYM. Invest the rest. Stay the course. Die anyway. - PS: The cavalry isn't coming, kids. You are on your own.
User avatar
abuss368
Posts: 27850
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by abuss368 »

Hi Rick,

Excellent article. You have saved me the time of finding the optimal portfolio! In all honesty, I learned (and still learning) that the best portfolio is the one that if right for "you" and let's one sleep well at night.

Keep investing simple!
John C. Bogle: “Simplicity is the master key to financial success."
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Random Walker »

"Enemy of a good plan is the perfect plan"
User avatar
abuss368
Posts: 27850
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by abuss368 »

Keep investing simple!
John C. Bogle: “Simplicity is the master key to financial success."
Random Walker
Posts: 5561
Joined: Fri Feb 23, 2007 7:21 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Random Walker »

"Everything should be made as simple as possible. But not simpler"
User avatar
nisiprius
Advisory Board
Posts: 52105
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by nisiprius »

The estimation problem is a bear, particularly with correlations. According to any statistics book, if you have 50 pairs of observations (and you know that they are independent random samples from a bivariate normal distribution), and your sample has a correlation coefficient of 0.5, the true correlation coefficient could really be anywhere from 0.17 to 0.72 (p = 0.01). So in real life, even assuming all the statistics 101 assumptions are met, you'd expect the sampled correlation coefficient to fluctuate all over this range just from sampling error.

In the real world, correlations are wildly unstable... and the instability is probably perfectly consistent with sampling error. The prevalent thinking, however, seems to be that all of the fluctuations are real, and that the changes in correlations are predictable--by a knowledgeable advisor!

Rick, I wish you would discuss the Black-Litterman methodology, which, as nearly as I can understand it, seems... bogus. At least one firm that claims to be using MPT acknowledges that what they are really using is the Black-Litterman methodology.

In this methodology, the numbers, rather than being based on the troublesome data and its troublesome sampling errors, incorporate manager's "views." That is to say, the manager reaches somewhere and pulls out his opinion/prediction/guess about the expected returns of various asset classes. If he thinks emerging markets stocks are going to have an expected return of 10.6% real, that is what goes into the model, regardless of past data.

In that firm's description, "the Black-Litterman model provides a flexible framework to express views about asset class returns, which ultimately will be reflected in the asset allocation." In short, it will assign the highest allocations to the asset classes the manager likes best. While this certainly gets around the pesky sampling errors, and the tendency of MPT to recommend 100% or leveraged allocations to whatever happens to have done best recently, it seems to be an obfuscated form of "garbage in, garbage out," as well as being dishonest if it is presented as being scientific, objective, or "optimized." Actually, I don't understand how exactly Black-Litterman is supposed to be an improvement on simply letting the manager set allocations by intuition.
Last edited by nisiprius on Tue Jul 21, 2015 3:31 pm, edited 1 time in total.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
User avatar
abuss368
Posts: 27850
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by abuss368 »

Hi Rick,

How about a future article on the Yale Portfolio from "Unconventional Success" written by David Swensen for the individual investor?

Would that be possible? I would be interested in your thoughts.

Best.
John C. Bogle: “Simplicity is the master key to financial success."
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Robert T »

.
As Bernstein (The Intelligent Asset Allocator) puts it:

  • "Anybody who tells you that their portfolio recommendations are "on the efficient frontier" also talks to Elvis and frolics with the Easter Bunny"... "Its like trying to generate electrical power by placing a battery and a lightning rod at the last place you saw lightning strike. It isn't likely to strike there again. In other words, next year's efficient frontier will be nowhere near last year's"..."Still, if you're trying to capture lightning in a jar you are better off in Texas than Alaska. There are certain asset combinations and portfolios which are likely (but not certain) to do reasonably well."
I generally agree on the challenge/futility of trying to select the future ‘efficient frontier’ portfolio. However, I do think there is merit in considering correlations (Markowitz framework). Here is an example using Long-term Expected returns from 2006 “All About Asset Allocation” book.
  • 30 yr expected return (%)/risk
    Long-term Treasuries = 5/5.3
    Intermediate corp = 5/5.5
    US Large Cap = 8/15
Based on a simple weighted average of the two portfolios – these have the same expected return/risk
  • 6.8/11.1 = 60:40 US Large cap:Long-term bonds
    6.8/11.1 = 59:41 US Large cap:Intermediate corporate
Then came 2008 – the second portfolio decline twice as much as the first: -24.4 vs -12.2 – but weighted average SD says they have same long-term risk. These portfolios will likely perform in different ways over the sort term during different market environments – deflation, inflation, financial crises etc. (a la Swensen) when correlations change (as in 2008 when correlation between US stocks and treasuries went sharply negative, while correlations between US stocks and corporates were increasingly positive. Need to figure which one you need most protection from, to help stay the course (when it matters the most).

I also think past correlation among factors can (qualitatively) inform allocation decisions eg. Value and momentum. No guarantee – but to reiterate Bernstein..."Still, if you're trying to capture lightning in a jar you are better off in Texas than Alaska. There are certain asset combinations and portfolios which are likely (but not certain) to do reasonably well."

I think portfolio construction is as much art as science. And in practice the ‘efficient/optimal’ portfolio is one you can stick with (which includes considerations on tracking error, simplicity, downside tolerance etc. – which are often specific to individuals).

Robert
.
fortyofforty
Posts: 2083
Joined: Wed Mar 31, 2010 12:33 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by fortyofforty »

It does seem nearly impossible to create portfolios that will lie upon the Efficient Frontier going forward. The only nearly perfect portfolio components are indices that include every investible asset within a class, such as Total Stock Market, Total International Stock, or Total Bond Market. How those are allocated is the trick, but at least we know we're using the best pieces.
Northern Flicker
Posts: 15288
Joined: Fri Apr 10, 2015 12:29 am

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Northern Flicker »

deleted
Last edited by Northern Flicker on Thu Mar 31, 2022 10:39 pm, edited 1 time in total.
User avatar
Epsilon Delta
Posts: 8090
Joined: Thu Apr 28, 2011 7:00 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Epsilon Delta »

Rick Ferri wrote:There are other ways to estimate portfolio risk and return that circumvent correlation estimates. One way is to assume there is no efficient frontier.
This shows the main problem with MPT. People either do not read it or having read it do not understand it.

The efficient frontier is not defined as the Markowitz bullet. The efficient frontier is defined as the set of portfolios that offers the highest expected return for a defined level of risk or the lowest risk for a given level of expected return. MPT shows that under some assumptions the (upper limb of) the Markowitz bullet is the efficient frontier.

It's one thing to question the assumptions of MPT. It's reasonable to question the applicability of MPT. But questioning the existence of the efficient frontier is another thing, you're deep into the foundations of set theory and logic. In particular if the set of possible investments is finite (and it is) the efficient frontier must exist.
User avatar
nisiprius
Advisory Board
Posts: 52105
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by nisiprius »

Epsilon Delta wrote:
Rick Ferri wrote:There are other ways to estimate portfolio risk and return that circumvent correlation estimates. One way is to assume there is no efficient frontier.
This shows the main problem with MPT. People either do not read it or having read it do not understand it. The efficient frontier is not defined as the Markowitz bullet. The efficient frontier is defined as the set of portfolios that offers the highest expected return for a defined level of risk or the lowest risk for a given level of expected return...
If you assume, nihilistically, that in the real world there isn't any way to determine the "expected return," the "level of risk," or the correlations--that they are abstract idealizations in a thought experiment--then I think it is perfectly reasonable to say that in the real world there "is" no efficient frontier.

If the Easter bunny is, by definition, the creature who carries colored eggs in a basket and delivers them to the houses of well-behaved children on Easter Sunday, and if there are 50 million such children and one Easter bunny, and if he begins his journey from his home on Easter Island (returning to Easter Island every time he needs to refill his basket), then what is the optimum routing that will enable the bunny to complete his rounds most efficiently?

Answer: there is no optimal routing, because the Easter bunny does not live on Easter Island.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
archii
Posts: 140
Joined: Wed Aug 07, 2013 8:33 am

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by archii »

abuss368 wrote:Hi Rick,

How about a future article on the Yale Portfolio from "Unconventional Success" written by David Swensen for the individual investor?

Would that be possible? I would be interested in your thoughts.

Best.

+1
Just finished reading his book. Excellent read.
User avatar
abuss368
Posts: 27850
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!
Contact:

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by abuss368 »

archii wrote:
abuss368 wrote:Hi Rick,

How about a future article on the Yale Portfolio from "Unconventional Success" written by David Swensen for the individual investor?

Would that be possible? I would be interested in your thoughts.

Best.

+1
Just finished reading his book. Excellent read.
Hi archii,

I think David Swensen makes a lot of sense. I was fortunate enough to be able to attend one of his lectures and learned so much.

Best.
John C. Bogle: “Simplicity is the master key to financial success."
User avatar
bnwest
Posts: 168
Joined: Thu Mar 01, 2007 6:25 pm
Location: durham, north carolina

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by bnwest »

Rick Ferri wrote:Summary of Why Optimal Portfolios Are So Difficult To Create.

Harry Markowitz won a Nobel Prize in Economics in 1990 for his work on a theory of portfolio management for individual wealth holders. Since that time, Modern Portfolio Theory (MPT) has become the bedrock for creating best-practice portfolio selection methods.
Rick Ferri Signature wrote:Choose a few low-cost index funds in different asset classes, rebalance occasionally and forgetaboutit!
Am I the only one seeing the cognitive dissonance here?
User avatar
Epsilon Delta
Posts: 8090
Joined: Thu Apr 28, 2011 7:00 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Epsilon Delta »

nisiprius wrote:
Epsilon Delta wrote:
Rick Ferri wrote:There are other ways to estimate portfolio risk and return that circumvent correlation estimates. One way is to assume there is no efficient frontier.
This shows the main problem with MPT. People either do not read it or having read it do not understand it. The efficient frontier is not defined as the Markowitz bullet. The efficient frontier is defined as the set of portfolios that offers the highest expected return for a defined level of risk or the lowest risk for a given level of expected return...
If you assume, nihilistically, that in the real world there isn't any way to determine the "expected return," the "level of risk," or the correlations--that they are abstract idealizations in a thought experiment--then I think it is perfectly reasonable to say that in the real world there "is" no efficient frontier.

If the Easter bunny is, by definition, the creature who carries colored eggs in a basket and delivers them to the houses of well-behaved children on Easter Sunday, and if there are 50 million such children and one Easter bunny, and if he begins his journey from his home on Easter Island (returning to Easter Island every time he needs to refill his basket), then what is the optimum routing that will enable the bunny to complete his rounds most efficiently?

Answer: there is no optimal routing, because the Easter bunny does not live on Easter Island.
We do not have an existence proof for the Easter Bunny.
We do have an existence proof for the efficient frontier.

The efficient frontier is tied very tightly to the idea that some portfolios are clearly better than some other portfolios. If you go nihilistic and deny that some portfolios are awful there's not a lot to be said.

More importantly most of MPF is done as mathematics using mathematical reasoning. If you're not going to look at the theory in that context you're going to miss a lot of what Markowitz et al actually said, and you may leave yourself open to con-men who say they said something they didn't.
Browser
Posts: 4857
Joined: Wed Sep 05, 2012 4:54 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Browser »

Contrasting the "MPT way" with the "pre-MPT way":
I feel fortunate that my original investing lessons were taught to me by experienced investors that had not yet embraced MPT. Instead, my original teachers taught that investment selections should be based on a carefully researched assessment of the investment merits’ past, present and future of any asset class under consideration. Simply stated, an asset class should only be included in a portfolio if it makes economic sense at the time.

This is in stark contrast to the tenets of MPT that postulates that portfolios should be comprised of various asset classes that in theory will change in value in opposite ways. Consequently, many practitioners of MPT routinely and automatically recommend standardized asset allocation strategies. These asset allocation mixes are routinely and adamantly recommended...
http://www.advisorperspectives.com/comm ... you-endure

If you are following MPT, then a significant allocation to things like commodity futures, gold, etc. apparently make sense. But if you are looking a little closer and thinking about the features of particular assets, maybe not so much. Perhaps MPT is one of those things where in theory there is no difference between theory and practice, but in practice there is?
We don't know where we are, or where we're going -- but we're making good time.
kolea
Posts: 1322
Joined: Fri Jul 11, 2014 5:30 pm
Location: Maui and Columbia River Gorge

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by kolea »

The market is as much chaotic as stochastic and using statistical methods to model its behavior is iffy to begin with, in my view. But to then believe that such a model can serve to find the optimal portfolio for the future is really a stretch. Finding new and improved methods for determining correlations or whatever are not going to suddenly make the bubble/burst behavior of the markets any more amenable to modeling with statistics. But that is just my view.
Kolea (pron. ko-lay-uh). Golden plover.
User avatar
dodecahedron
Posts: 6563
Joined: Tue Nov 12, 2013 11:28 am

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by dodecahedron »

NYT's Joe Nocera recounts a story from Jason Zweig's book (Your Money or Your Brain) showing that even the legendary Harry Markowitz himself did not use his fancy analytic theory to design his own personal investment portfolio:
It’s a story that says everything about how most of us act when it comes to investing. Mr. Markowitz was then working at the RAND Corporation and trying to figure out how to allocate his retirement account. He knew what he should do: “I should have computed the historical co-variances of the asset classes and drawn an efficient frontier.” (That’s efficient-market talk for draining as much risk as possible out of his portfolio.)

But, he said, “I visualized my grief if the stock market went way up and I wasn’t in it — or if it went way down and I was completely in it. So I split my contributions 50/50 between stocks and bonds.” As Mr. Zweig notes dryly, Mr. Markowitz had proved “incapable of applying” his breakthrough theory to his own money. Economists in his day believed powerfully in the concept of “economic man”— the theory that people always acted in their own best self-interest. Yet Mr. Markowitz, famous economist though he was, was clearly not an example of economic man.
User avatar
Epsilon Delta
Posts: 8090
Joined: Thu Apr 28, 2011 7:00 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Epsilon Delta »

Browser wrote:Contrasting the "MPT way" with the "pre-MPT way":
I feel fortunate that my original investing lessons were taught to me by experienced investors that had not yet embraced MPT. Instead, my original teachers taught that investment selections should be based on a carefully researched assessment of the investment merits’ past, present and future of any asset class under consideration. Simply stated, an asset class should only be included in a portfolio if it makes economic sense at the time.

This is in stark contrast to the tenets of MPT that postulates that portfolios should be comprised of various asset classes that in theory will change in value in opposite ways. Consequently, many practitioners of MPT routinely and automatically recommend standardized asset allocation strategies. These asset allocation mixes are routinely and adamantly recommended...
http://www.advisorperspectives.com/comm ... you-endure

If you are following MPT, then a significant allocation to things like commodity futures, gold, etc. apparently make sense. But if you are looking a little closer and thinking about the features of particular assets, maybe not so much. Perhaps MPT is one of those things where in theory there is no difference between theory and practice, but in practice there is?
Modern Portfolio Practice and Modern Portfolio Theory have almost nothing to do with each other. MPT says that if you know the distributions you can pick an optimum portfolio. It does not say the distributions are stationary, or that they can be deduced by looking at an insignificant amount of history, or that you can know them at all, or any of the other thousands of canards that are added by hucksters using it to sell something or other hucksters denying it to sell something else.

If you look at what MPT says it's just the pre-MPT way expressed in a different framework. The distributions (means, variance and covariance) summarize what is known about the future of all the assets. You use that summary to decide which asset classes make economic sense at the time. The distributions are not supposed to be a substitute for careful research, they're supposed to be a result of careful research.

You can say "stocks have gone up an average of 10% a year for the last 20 years, therefore they will double in the next 7 years" in either MPT or pre-MPT. It's the same nonsense either way.
User avatar
nisiprius
Advisory Board
Posts: 52105
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by nisiprius »

Epsilon Delta, in your own words, then:

In the real world, given the limitations of available data and the limitations of personal behavior,

a) Can an individual retail investor apply MPT in his own portfolio?
b) If so, should an individual retail investor apply MPT in his own portfolio?
c) If so, how should an individual retail investor apply MPT in his own portfolio?
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 10:55 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by lack_ey »

One of the concepts that came out of MPT (much later) is risk parity allocations. In a broad sense, the idea of risk parity is that estimating returns is very difficult, as is estimating correlations (to some degree). Rather than guess at all these difficult-to-determine parameters and find an optimal portfolio on the efficient frontier, risk parity seeks to find a portfolio that takes (more) equal risk across different assets, each with some kind of positive expected return. As long as each asset class has reasonable returns over the long run and they're not all that correlated, whatever the correlations end up being, you earn money across a variety of assets. And splitting risk in this way means not being exposed to big losses from any one source and getting the benefits of diversification. With mean variance optimization, you give up some diversification in order to take a larger bet and concentrate in the assets with the best assumed risk/return and correlation properties.

The general riskiness of assets is relatively consistent across time; standard deviation as a proxy for risk does vary but is more consistent than returns, in any case. And with risk parity nobody really cares if your risk estimates are slightly off and risk isn't actually equal for every asset class. A risk parity allocator already knows that the portfolio is not on the efficient frontier anyway.

All in all, there are limitations and weaknesses of every approach, and I'm not a risk parity fan either. At best with modern portfolio theory, what we get is some intuition about the benefits of diversification.
User avatar
1210sda
Posts: 1865
Joined: Wed Feb 28, 2007 7:31 am

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by 1210sda »

Rick Ferri wrote: From the Article:
"The weighted average method to estimating a portfolio’s overall risk and return may be a throwback in time, but it is conservative and does eliminate errors from inaccurate asset class covariance or correlation forecasts. You can read more about this method in my book, All About Asset Allocation."
Rick, what page is this on?? I'm having a hard time finding it.
1210
User avatar
Clearly_Irrational
Posts: 3087
Joined: Thu Oct 13, 2011 3:43 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Clearly_Irrational »

It's a useful model in that it informs us about the risk/return trade off and how adding various components can affect our portfolio outcome. The numbers can give you an idea of a reasonable range of outcomes but given all the flaws in the data available you'll never be able to actually "optimize" a portfolio. In short, all models are wrong but some are useful. It probably gives a false sense of certainty and precision to many who don't understand its limitations.
User avatar
nisiprius
Advisory Board
Posts: 52105
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by nisiprius »

lack_ey wrote:...The general riskiness of assets is relatively consistent across time; standard deviation as a proxy for risk does vary but is more consistent than returns, in any case...
Is it really? I've never looked at this myself... do you have any data... wait... seems to me I saw... yeah... here it is...

Is this "relatively consistent?" Of course, a standard deviation can't go negative and return can...

Image
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.
Beliavsky
Posts: 1233
Joined: Sun Jun 29, 2014 10:21 am

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Beliavsky »

nisiprius wrote:
lack_ey wrote:...The general riskiness of assets is relatively consistent across time; standard deviation as a proxy for risk does vary but is more consistent than returns, in any case...
Is it really? I've never looked at this myself... do you have any data... wait... seems to me I saw... yeah... here it is...

Is this "relatively consistent?" Of course, a standard deviation can't go negative and return can...

Image
It is consistent in the sense that over a 5 year period, small cap stock volatility is almost always higher than that of large caps, which is in turn higher than bond volatiilty. One cannot make such a statement about the ordering of asset class returns over 5 years.
User avatar
Epsilon Delta
Posts: 8090
Joined: Thu Apr 28, 2011 7:00 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Epsilon Delta »

nisiprius wrote:Epsilon Delta, in your own words, then:

In the real world, given the limitations of available data and the limitations of personal behavior,

a) Can an individual retail investor apply MPT in his own portfolio?
b) If so, should an individual retail investor apply MPT in his own portfolio?
c) If so, how should an individual retail investor apply MPT in his own portfolio?
I don't think you can use MPT to calculate optimal calculations in a useful way. But the theory does support concepts like diversification, various mutual fund separation theorems and a bunch of other stuff that is otherwise unconnected. It also lets you look with suspicion at people who want you to diversify stocks with foreign stocks or Treasuries with corporates, or are otherwise misusing pro or anti MPT statements to sell something or other.

None of this requires MPT, but I find it more coherent than the alternative of disjointed just so stories. This might help improve investing behavior. Or maybe not, many people find expert opinion without justification more convincing than an actual argument.
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 10:55 pm

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by lack_ey »

Beliavsky wrote:
nisiprius wrote:
lack_ey wrote:...The general riskiness of assets is relatively consistent across time; standard deviation as a proxy for risk does vary but is more consistent than returns, in any case...
Is it really? I've never looked at this myself... do you have any data... wait... seems to me I saw... yeah... here it is...

Is this "relatively consistent?" Of course, a standard deviation can't go negative and return can...

[img]
It is consistent in the sense that over a 5 year period, small cap stock volatility is almost always higher than that of large caps, which is in turn higher than bond volatiilty. One cannot make such a statement about the ordering of asset class returns over 5 years.
Yes, that's what I mean by "relatively consistent." In finance, basically nothing is consistent, and this is more so than a lot of other things including returns. Here's another look at 5-year returns vs. volatility (just look at the charts and ignore the text if you want):
https://www.aqr.com/cliffs-perspective/ ... e-long-run

For example, looking at short periods of time, the frequency at which stocks are more volatile than bonds is a lot higher than the frequency at which stocks outperform bonds. (And the capital markets line between the two and the origin isn't drastically bent.)
User avatar
Topic Author
Rick Ferri
Posts: 9703
Joined: Mon Feb 26, 2007 10:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by Rick Ferri »

1210sda wrote:
Rick Ferri wrote: From the Article:
"The weighted average method to estimating a portfolio’s overall risk and return may be a throwback in time, but it is conservative and does eliminate errors from inaccurate asset class covariance or correlation forecasts. You can read more about this method in my book, All About Asset Allocation."
Rick, what page is this on?? I'm having a hard time finding it.
1210
I thought I wrote it in the book. I'll have to go back and look. Sorry!

Rick Ferri
The Education of an Index Investor: born in darkness, finds indexing enlightenment, overcomplicates everything, embraces simplicity.
User avatar
seeshells
Posts: 179
Joined: Sun Feb 23, 2014 11:43 am

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by seeshells »

Unless i'm mistaken, Harry Markowitz defined the only "free lunch" as diversification, that was his widely accepted theory. Suggesting perhaps that any well diversified portfolio is optimal, given ones risk tolerance. As many here believe, and iirc Taylor Larimore always says, there are many roads to Dublin.
CFM300
Posts: 2534
Joined: Sat Oct 27, 2007 5:13 am

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by CFM300 »

dodecahedron wrote:NYT's Joe Nocera recounts a story ... showing that even the legendary Harry Markowitz himself did not use his fancy analytic theory to design his own personal investment portfolio...

"[Markowitz] said, 'I visualized my grief if the stock market went way up and I wasn’t in it — or if it went way down and I was completely in it. So I split my contributions 50/50 between stocks and bonds.'”
However, Markowitz has more recently (2013) said:

"When I was at Rand in 1950, I just did 50/50. That’s all I knew then, it’s not what I would do now and it’s not what I would recommend to a 25 year old. My profession and I have learned a lot."

http://cuffelinks.com.au/the-harry-mark ... selection/
User avatar
dodecahedron
Posts: 6563
Joined: Tue Nov 12, 2013 11:28 am

Re: Is Harry Markowitz wrong? Why Optimal Portfolios Are So Difficult To Create

Post by dodecahedron »

CFM300, thanks for the interesting followup information. This got me curious about exactly what Harry Markowitz is actually doing now and I found a later article (June 2014) on the same site with a fun story (emphasis in red added by me):
His own portfolio is currently equally weighted municipal bonds and equities, the latter with an emphasis on small caps and emerging markets, but with a stable core of blue chips. This is because he feels so many stocks are overvalued at the moment, and his portfolio is also influenced by his age. “I want enough bonds that if I die, and the equity market goes to zero, my wife will have enough capital and income to live well.” His current objective is to reach 100 without appearing on the right-hand column of The Wall Street Journal, with the heading “Harry Markowitz f*cked up”.
So, after all these years, and all the refinement he's presumably done, he is still 50/50 equity/bonds, with the only difference being some sort of fancy tilt to the sub-components of his equities, plus restricting his bonds to municipals (presumably for tax reasons.)

And it certainly is true that folks like him (prominent investment academics whose work has been celebrated and recognized with a Nobel Prize) have an additional special burden associated with their personal investing decisions. If they notably screw up and wind up outliving their money in old age, not only do they suffer financially but they face the prospect of public humiliation and schadenfreude.

http://cuffelinks.com.au/harry-markowitz-investing-100/
Post Reply