Author Topic: Meb Faber: Which Asset Allocation Model is Best?  (Read 7223 times)

hodedofome

  • Handlebar Stache
  • *****
  • Posts: 1319
  • Age: 40
  • Location: Texas
Meb Faber: Which Asset Allocation Model is Best?
« on: May 18, 2016, 03:37:47 PM »
http://mebfaber.com/2016/05/18/institutional-asset-allocation-models/

If you’re like most investors, you’re asking the wrong questions.

I was chatting with a group of advisors this week down in La Jolla and a question arose. I’ll paraphrase:

“Meb, thanks for the talk.  We get a steady stream of salespeople and consultants in here hawking their various asset allocation models.  Frankly, it can be overwhelming. Some will send us a 50-page report, all to explain a strategic shift from 50% equities to 40%.  I want to do right by my clients, but I have a hard time reading all the various research pieces and models, let alone reconciling their differences.  Any thoughts?”

The advisor followed up by emailing me this summary of all of the institutional asset allocation models by the Goldmans, Morgan Stanleys, and Deutsche Banks of the world.  And as you’ll see, they are HIGHLY different.  Morgan Stanley says only 25% in US stocks, while Silvercrest says 54%!  Brown Advisory says 10% in emerging markets and JPMorgan 0%.

Returning to the advisor’s question, if you know me, you’re aware that of course I have some thoughts on the topic!  In fact, I wrote a book on asset allocation, and would happy to send you a free copy (go to freebook.mebfaber.com to download).

So what is an advisor to do? What’s the most effective asset allocation model?

Turns out, that’s actually, that’s the wrong question.  The correct starting question is, “Do asset allocation differences even matter?”

In the summary article which the advisor sent me, there’s a link to a data table showing the asset allocations of 40 of the nation’s leading wealth management groups. Below, my analyst Jonathan and I teased out all the data from the table to examine three allocations:

The allocation with the most amount in stocks (Deutsche Bank).
The average of all 40.
The allocation with the least amount in stocks (Atlantic Trust).
Below is the equity curve for each.  Unless you have hawk-like vision, you’ll likely have a hard time distinguishing between the curves, and this is for the most different.  The other 40+ firms live somewhere in the middle!!

Below are the returns for each allocation over the entire 1973-2015 period.

Most aggressive (DB):  9.72%

Average:  9.60%

Least aggressive (AT):  9.19%
 


There you have it – the difference between the most and least aggressive portfolios is a whopping 0.53% a year.  Now, how much do you think all of these institutions charge for their services?  How many millions and billions in consulting fees are wasted fretting over asset allocation models? 

Let’s try one more experiment…

Overlay a simple 1% management fee on the most aggressive portfolio and look again at the returns.  Simply by paying this mild fee (that is lower than the average mutual fund, by the way) you have turned the highest returning allocation into the lowest returning allocation – rendering the entire asset allocation decision totally irrelevant. 



And if you allocate to the average money manager with an average fee (1%) that invests in the average mutual fund, well, you know the conclusion.



This is one reason my company launched the first, and still only, ETF with a permanent 0% management fee. It owns a handful of other ETFs and all-in fees are 0.29%.  For your core buy-and-hold allocation, you (or your clients) should pay as little as possible.

So all those questions that stress you out…

“Is it a good time for gold?” 

“What about the next Fed move – should I lighten my equity positions beforehand?” 

“Is the UK going to leave the Euro, and what should that mean for my allocation to foreign investments?”

Let them go.

If you’re a professional money manager, go spend your time on value added activities like estate planning, insurance, tax harvesting, prospecting, general time with your clients or family, or even golf.

If you’re a retail investor, go do anything that makes you happy.

Either way, stop reading my blog and go live your life. 


steveo

  • Handlebar Stache
  • *****
  • Posts: 1943
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #1 on: May 18, 2016, 05:01:27 PM »
Basically pick a reasonable asset allocation, get the lowest fees possible, use indexes and forget about it.

forummm

  • Walrus Stache
  • *******
  • Posts: 7389
  • Senior Mustachian
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #2 on: May 18, 2016, 06:36:46 PM »
The popup ad on his blog was for his book: "Global Asset Allocation"

Ah, irony.

MustacheAndaHalf

  • Magnum Stache
  • ******
  • Posts: 2754
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #3 on: May 18, 2016, 10:39:09 PM »
Meb Faber strips away differences to make all portfolios fit a very limited model of stock performance.  For example, Larry Swedroe advocates small/value funds as a tilt to improve performance or lower stock allocation.  Meb Faber mentions Swedroe in an index, and removes small/value funds.  Calling something a Swedroe portfolio after you remove small/value isn't accurate.  It's distorting data to fit the conclusion.

Warren Buffet beat the S&P 500 for decades.  It's documented, it's a fact.  Yet if you look at Meb Faber's "Buffet" portfolio, it performs like the S&P 500.  That's because Meb Faber replaced the man most famous for beating the S&P 500 for decades... with the S&P 500.  He calls the S&P 500 the "Buffet" portfolio.

In my view, Meb Faber makes the data fit his conclusion.  He strips away small/value from a Larry Swedroe portfolio, and turns Buffet into the S&P 500 to demonstrate how similar everything is... but it's not, it's just data twisted until it no longer resembles reality.  I'm skeptical that Meb Faber is pulling the same trick in another context - making data fit his conclusion.
« Last Edit: May 18, 2016, 10:47:18 PM by MustacheAndaHalf »

dachs

  • Bristles
  • ***
  • Posts: 257
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #4 on: June 03, 2016, 03:26:42 PM »
The "Buffet Portfolio" is Buffet's recommendation for his trust once he passes away. 90% stocks, 10% bonds is not what Berkshire Hathaway does, obviously.

I remember something like having just enough money in cash/bonds so you can survive 5 years of a bear market without having to sell stocks. In case you apply the 4% "rule" that would mean 20% cash/bonds.

k9

  • Bristles
  • ***
  • Posts: 250
  • Age: 40
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #5 on: June 04, 2016, 07:08:39 AM »
And it doesn't work when the bear market lasts more than 5 years, obviously.

But, yeah, Faber makes it very clear that what he calls a Buffet portfolio is what Buffet advised for small investors : 90% in an index fund, 10% in cash.

MustacheAnsAHalf, I don't know what makes you think Faber ignored the SCV/emerging tilt in Swedroe's portfolio.  I just checked his book, and here's what I read at appendix F "The Larry Swedroe Portfolio" :

Quote
Larry Swedroe is one of my favorite writers and researchers. With 15 books to his name, his focus on evidence based investing fits in well with how we view the world. We debated about including this portfolio in the book since it requires using a value tilt, but think it is an important example of how a simple smart beta strategy may be beneficial to the overall portfolio.
The biggest difference between the allocation and others in this book is that he allocates to small cap value. Small cap value stocks have outperformed broad small caps by about four percentage points a year, which is a lot. The value stocks have slightly more volatility and a higher drawdown as well.
[...]
Swedroe’s unusual allocations are another example of a consistent performer due to including a mix of stocks, bonds, and real assets [Faber considers TIPS are a real asset]. We include the performance of including small cap as well as small
cap value to illustrate the improvement in performance. The portfolio with the value tilt results in the highest Sharpe ratio of any portfolio in the book.

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28255
  • Age: -999
  • Location: Seattle, WA
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #6 on: July 01, 2016, 03:13:57 AM »
The correct starting question is, “Do asset allocation differences even matter?”

In the summary article which the advisor sent me, there’s a link to a data table showing the asset allocations of 40 of the nation’s leading wealth management groups. Below, my analyst Jonathan and I teased out all the data from the table to examine three allocations:

The allocation with the most amount in stocks (Deutsche Bank).
The average of all 40.
The allocation with the least amount in stocks (Atlantic Trust).
Below is the equity curve for each.  Unless you have hawk-like vision, you’ll likely have a hard time distinguishing between the curves, and this is for the most different.  The other 40+ firms live somewhere in the middle!!

Below are the returns for each allocation over the entire 1973-2015 period.

Most aggressive (DB):  9.72%

Average:  9.60%

Least aggressive (AT):  9.19%
 


There you have it – the difference between the most and least aggressive portfolios is a whopping 0.53% a year.  Now, how much do you think all of these institutions charge for their services?  How many millions and billions in consulting fees are wasted fretting over asset allocation models? 

Let’s try one more experiment…

Overlay a simple 1% management fee on the most aggressive portfolio and look again at the returns.  Simply by paying this mild fee (that is lower than the average mutual fund, by the way) you have turned the highest returning allocation into the lowest returning allocation – rendering the entire asset allocation decision totally irrelevant. 



And if you allocate to the average money manager with an average fee (1%) that invests in the average mutual fund, well, you know the conclusion.



That's terrific info for people fretting over their AA.

60/40, 40/60... Pick something that lets you sleep at night and forget it.  Tweaking your IPS every three months gets you nothing but extra stress.

Thanks for sharing that article!
We are two former teachers who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and are now settled with three kids.
If you want to know more about us, or how we did that, or see lots of pictures, this Business Insider profile tells our story pretty well.
We (rarely) blog at AdventuringAlong.com. Check out our Now page to see what we're up to currently.

MustacheAndaHalf

  • Magnum Stache
  • ******
  • Posts: 2754
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #7 on: July 01, 2016, 02:58:30 PM »
Had Faber listed both the S&P 500 and Buffet's performance separately, it would have disproved his thesis.  Faber wants to conclude all portfolios have similar results, so he forces Berkshire Hathaway to be represented by the S&P 500 - even though it beat the S&P 500 for several decades.

Warren Buffet is the most successful value investor of all time, becoming the richest person in the world through his investment choices.  His wealth and fame are not from buying and holding the S&P 500.  And if you follow what Buffet actually said, it was that 99% of investors should use the S&P 500.  But that doesn't mean Warren Buffet - he's the 1% that has pulled off beating the S&P 500 for several past decades.  Faber misleads the reader into thinking two things are the same, when they should be understood separately: Warren Buffet's performance, and the recommendation (by many people) to follow the S&P 500.  But note Faber doesn't separate them - he hides Buffet's inconvenient performance and swaps in the S&P 500 instead.

arebelspy

  • Administrator
  • Senior Mustachian
  • *****
  • Posts: 28255
  • Age: -999
  • Location: Seattle, WA
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #8 on: July 01, 2016, 03:42:40 PM »
He uses Warren Buffet's recommendation for his wife's portfolio after he dies. 

He doesn't claim to show Warren's results, but the portfolio Warren recommends.

Warren is an extreme outlier.  If you think you can match Warren, well, good luck to you.

But if you know you can't, when looking at various AAs (specifically to see how much AA matters, as is the premise of this article), I don't think it's disingenuous at all to not use an individual result from one person that's the biggest outlier ever in the field, and instead use the portfolio they recommend for everyone else, including their own family, to see how that portfolio did.
We are two former teachers who accumulated a bunch of real estate, retired at 29, spent some time traveling the world full time and are now settled with three kids.
If you want to know more about us, or how we did that, or see lots of pictures, this Business Insider profile tells our story pretty well.
We (rarely) blog at AdventuringAlong.com. Check out our Now page to see what we're up to currently.

MustacheAndaHalf

  • Magnum Stache
  • ******
  • Posts: 2754
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #9 on: July 02, 2016, 09:51:01 AM »
He uses Warren Buffet's recommendation for his wife's portfolio after he dies.
He doesn't claim to show Warren's results, but the portfolio Warren recommends.
And Warren Buffet is still alive, with his assets in Berkshire Hathaway.  Why doesn't he show the S&P 500 separately from Warren Buffet?

... If you think you can match Warren, well, good luck to you.
Where did I say that?  Putting quotes in my mouth to mock me... where's a moderator when you need one?

Warren is an extreme outlier.  ...

But if you know you can't, when looking at various AAs (specifically to see how much AA matters, as is the premise of this article), I don't think it's disingenuous at all to not use an individual result from one person that's the biggest outlier ever in the field, and instead use the portfolio they recommend for everyone else, including their own family, to see how that portfolio did.
This isn't about me, it's about Faber's book.  Faber repeatedly cites outliers.  For example Swensen of the Yale Endowment, who made significant profits off private equity.  Faber allocates no private equity to a portfolio he calls the Swensen portfolio.

AdrianC

  • Handlebar Stache
  • *****
  • Posts: 1095
  • Location: Cincinnati
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #10 on: July 02, 2016, 10:09:15 AM »
Faber allocates no private equity to a portfolio he calls the Swensen portfolio.

Faber is probably showing a version of the portfolio Swenson gives in his (excellent, IMO) book "Unconventional Success: A Fundamental Approach to Personal Investment". That's the one Tyler and Bogleheads have as a "Lazy Portfolio".

Kaspian

  • Handlebar Stache
  • *****
  • Posts: 1535
  • Location: Canada
    • My Necronomicon of Badassity
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #11 on: July 02, 2016, 01:04:04 PM »
The correct starting question is, “Do asset allocation differences even matter?”


That's terrific info for people fretting over their AA.

60/40, 40/60... Pick something that lets you sleep at night and forget it.  Tweaking your IPS every three months gets you nothing but extra stress.

Thanks for sharing that article!

Basically pick a reasonable asset allocation, get the lowest fees possible, use indexes and forget about it.

Exactly this!  I see so many (especially new investors) fretting and wringing their hands about this and tax minutiae and just not getting started because of information paralysis--trying to get everything into a 100% perfect model.  My advice of "Just get fucking started--save as much as you possibly can, put it in low cost funds, forget about it, and reconsider the finer details once you have an egg worth worrying about--there is no exactly perfect portfolio!  If it reaches your goals, it's good.," falls on deaf ears because it sounds too simple.  Maybe people think it's like studying for a physics or calculus exam or something--it has to have a crystal clear system and answer.  When the real failure involves either not getting started or fucking around with your portfolio once it's going.

Can we please send all the asset allocation trolls to this thread from now on?  "I might want bonds.  Forget it, I don't want bonds.  Is 40/60 good?  No, I'm going with 100% equities."  Geesh, they usually know what they're going to go with before posting and know they're going to start little battles (which are completely needless.)  ...Also, there's about 200 threads about the matter in this forum alone for anyone who cares to do his/her homework.
« Last Edit: July 02, 2016, 01:08:07 PM by Kaspian »

AdrianC

  • Handlebar Stache
  • *****
  • Posts: 1095
  • Location: Cincinnati
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #12 on: July 02, 2016, 01:49:08 PM »
The popup ad on his blog was for his book: "Global Asset Allocation"

Ah, irony.

Funny! And when Faber writes: "This is one reason my company launched the first, and still only, ETF with a permanent 0% management fee. It owns a handful of other ETFs and all-in fees are 0.29%.  For your core buy-and-hold allocation, you (or your clients) should pay as little as possible", he's plugging his GAA Cambria Global Asset Allocation ETF:

http://www.cambriafunds.com/gaa.aspx

It holds 31 different ETFs. That's some complicated asset allocation right there.



k9

  • Bristles
  • ***
  • Posts: 250
  • Age: 40
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #13 on: July 03, 2016, 05:05:27 AM »
He uses Warren Buffet's recommendation for his wife's portfolio after he dies.
He doesn't claim to show Warren's results, but the portfolio Warren recommends.
And Warren Buffet is still alive, with his assets in Berkshire Hathaway.  Why doesn't he show the S&P 500 separately from Warren Buffet?

... If you think you can match Warren, well, good luck to you.
Where did I say that?  Putting quotes in my mouth to mock me... where's a moderator when you need one?

Warren is an extreme outlier.  ...

But if you know you can't, when looking at various AAs (specifically to see how much AA matters, as is the premise of this article), I don't think it's disingenuous at all to not use an individual result from one person that's the biggest outlier ever in the field, and instead use the portfolio they recommend for everyone else, including their own family, to see how that portfolio did.
This isn't about me, it's about Faber's book.  Faber repeatedly cites outliers.  For example Swensen of the Yale Endowment, who made significant profits off private equity.  Faber allocates no private equity to a portfolio he calls the Swensen portfolio.

The book is about asset *class* allocations, not stock picking and other deals only available to a few institutional investors. So he talks about the advice great investors (Buffet and Swensen) gave to individual investors.

If you're going there, even his Talmud portfolio implementation is wrong. Do you really think guys who wrote the Talmud advised 33% of REITs and 33% of US treasury bonds ? Thoses things didn't exist back then.

k9

  • Bristles
  • ***
  • Posts: 250
  • Age: 40
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #14 on: July 03, 2016, 05:10:59 AM »
Exactly this!  I see so many (especially new investors) fretting and wringing their hands about this and tax minutiae and just not getting started because of information paralysis--trying to get everything into a 100% perfect model.  My advice of "Just get fucking started--save as much as you possibly can, put it in low cost funds, forget about it, and reconsider the finer details once you have an egg worth worrying about--there is no exactly perfect portfolio!  If it reaches your goals, it's good.," falls on deaf ears because it sounds too simple.  Maybe people think it's like studying for a physics or calculus exam or something--it has to have a crystal clear system and answer.  When the real failure involves either not getting started or fucking around with your portfolio once it's going.

Can we please send all the asset allocation trolls to this thread from now on?  "I might want bonds.  Forget it, I don't want bonds.  Is 40/60 good?  No, I'm going with 100% equities."  Geesh, they usually know what they're going to go with before posting and know they're going to start little battles (which are completely needless.)  ...Also, there's about 200 threads about the matter in this forum alone for anyone who cares to do his/her homework.
Well, Faber's book demonstrates asset class allocation doesn't make a big deal when looking at CAGR, but it also shows that all those AA have a different behavior throughout the years. Losing money 10 years in a row is something most investors have a hard time dealing with, even if, on the long run, they'll be okay. For instance, if I recall correctly, of all the presented AA in the book, only 2 would have had a positive return in the 1973-1981 period (namely, the PP and the Marc Faber portfolio).

But, yeah, chose an AA, stick with it and you'll be okay.

MustacheAndaHalf

  • Magnum Stache
  • ******
  • Posts: 2754
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #15 on: July 03, 2016, 11:46:36 AM »
Maybe a counter example would demonstrate the problem better.  Let's say you agree with Swedroe and decide to tilt towards value, and tilt towards small.  Does that asset allocation differ from the S&P 500?

How do equal value and small tilts (large cap, large value, small cap, small value) perform vs the S&P 500?
1996-2015, $100 grows to $474 in S&P 500 and to $547 with value/small tilts
1976-2015, $100 grows to $6717 in S&P 500 and to $12,645 with value/small tilts

Yes, those are different risk levels.  But that's why Faber is wrong - different allocations can yield different results.

k9

  • Bristles
  • ***
  • Posts: 250
  • Age: 40
Re: Meb Faber: Which Asset Allocation Model is Best?
« Reply #16 on: July 04, 2016, 04:08:28 AM »
Faber agrees with you :

Quote
While covered more extensively in our other three books and white papers, consider tilting
the equity exposure to factors such as value and momentum. Trendfollowing approaches work great too.

That's in the conclusion of the GAA book.