Classic Bernstein 3 — Diversifying Portfolio Equities

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
SimpleGift
Posts: 3869
Joined: Tue Feb 08, 2011 3:45 pm
Location: Central Oregon

Classic Bernstein 3 — Diversifying Portfolio Equities

Post by SimpleGift »

PREFACE: This is the third in a series of posts highlighting the classic investing insights of William Bernstein from the 1990s and early 2000s. Many new Bogleheads have never been exposed to his early writings — and while the data sets used may seem antiquated, his portfolio concepts and novel analyses are still helpful to investors today, new and old alike.

Note: Bogleheads with an analytical bent are encouraged to replicate any of Mr. Bernstein’s analyses in this series — using more current data sets or alternative time periods — and share their results in these threads.

Previous topics in the series: 1-Asset Allocation & Time Horizon, 2-Choosing Portfolio Bond Duration.
------------------------------------------------------------------------------------------

In designing an investment portfolio, most investors are faced with the key question: How do I diversify my equity allocation? Whether one chooses a simple mix of two total stock market funds (domestic and international), or seeks to outpace these broad indexes by systematically overweighting various market sectors or return factors (e.g., small, value, momentum, etc.), most equity investors eventually face the perplexities of underperformance, tracking error and “mean aversion.” This post excerpts one of Mr. Bernstein’s early articles about these issues.

In this analysis, Mr. Bernstein first supposes that one has clairvoyant knowledge of the exact mix of asset classes that will produce the highest return over a future period — in this case, for the 27-year period from 1970-1996, and among U.S. large and small cap stocks, the MSCI regional stock indexes (Europe, Japan and Pacific ex Japan) and an index of gold mutual funds. It turns out that the best mix consists of just three assets: 35% Japanese stocks, 28% U.S. small stocks and 37% gold. Assuming annual rebalancing, this “optimal” portfolio returned 16.85% per year over the period:
  • CLAIRVOYANT EQUITY PORTFOLIO = 16.85% annual return
    Image
    Note: Average annual return of the S&P 500 was 12.27% for the period.
    Source: William Bernstein, Efficient Frontier (4/97)
What’s notable in this example is that, though the optimal portfolio returned 4.58% compounded more than the S&P 500, it lagged the index in 12 out of 27 years, or 44% of the time. This annual underperformance was quite significant at times, at more than 20% in several years. His conclusion: Equity diversification is not a free lunch. You must be willing to feel like a fool for extended periods of time in order to reap its benefits.

In his second example below, Mr. Bernstein’s acknowledges that no one has clairvoyance about the best-performing assets of the future as they will almost certainly be different than those of the past. What’s an investor to do? His advice is to just pick an equity allocation, almost any allocation, and stick with it. In this case, he chooses equal amounts of the same six assets (the “coward’s strategy”) and tracks the portfolio's performance over the same 27-year period:
  • COWARD'S EQUITY PORTFOLIO = 15.33% annual return
    Image
    Note: Average annual return of the S&P 500 was 12.27% for the period.
    Source: William Bernstein, Efficient Frontier (4/97)
Remarkably, this naive diversification strategy returned just 1.5% less than the very best possible strategy. Again, even though it beat the S&P 500 by 3% annualized, it lagged the index in 13 out of 27 years — nearly half the time.

To interpret Mr. Bernstein’s main points from this exercise:
  • a) Because we don’t know the best-performing equity assets of the future, it’s wise to spread one’s bets as widely as possible.

    b) Even widely-diversified equity portfolios will diverge from the broad global indexes at times, sometimes significantly and for long periods. To reap the rewards, one needs patience and faith that return divergences usually revert.

    c) Much more important than picking the best performers is choosing a reasonably-diversified mix of equities and having the discipline to stick with it over the long haul.
Thoughts?
Last edited by SimpleGift on Sat May 28, 2016 8:19 pm, edited 5 times in total.
Random Walker
Posts: 4618
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by Random Walker »

A portfolio's annualized return will always fall short of the weighted average annual return of portfolio components due to volatility. Dampening portfolio volatility by mixing weakly or non correlated assets makes a portfolio more efficient by bringing the compounded (annualized) return closer to the average annual return. I can't recommend the following short Roger Gibson read on the rewards of multi asset class investing highly enough. It is derived from his book on asset allocation.

http://www.amcham-shanghai.org/amchampo ... esting.pdf

Dave
User avatar
Topic Author
SimpleGift
Posts: 3869
Joined: Tue Feb 08, 2011 3:45 pm
Location: Central Oregon

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by SimpleGift »

Random Walker wrote:I can't recommend the following short Roger Gibson read on the rewards of multi asset class investing highly enough.
Thanks, Dave. Gibson’s short primer provides another perspective on equity diversification and is a good addition to this thread. I was most struck by his chart below, showing fifteen equity portfolios over the 1972-2005 period. Nearly all of the multi-asset class portfolios (within the red oval) had higher returns with much less volatility than the single assets themselves that make up the portfolios (the four blue squares) — the payoff for long-term diversification.
This small study may also be intriguing for analytical-oriented Bogleheads to replicate with alternative time periods and assets — in addition to Mr. Bernstein’s analysis in the OP.
Random Walker
Posts: 4618
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by Random Walker »

Simple gift,
Ya I'm a huge fan of that chart. It basically summarizes the whole essay, and the essay basically summarizes his whole book :-). Note 2 asset classes generally beat 1, and 3 beat 2. I'm sure that additional asset classes have decreasing marginal benefit as the number of asset classes goes up. At some point each investor has to look at additional cost versus additional benefit to a new asset class.
In deciding on whether to add an asset class to a portfolio, the relevant factors are expected return, volatility, correlations to other portfolio components, and of course costs.

Dave
User avatar
ray.james
Posts: 1389
Joined: Tue Jul 19, 2011 4:08 am

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by ray.james »

Great post, simplegift. For all the issues raised on international allocation/other specific hot/cold asset classes on this board, people tend to ignore with recency bias. Even if an asset class did not beat another in no more than 5 of 25 years years the lower standard deviation is enough to give way close geometric return.

http://www.cdiadvisors.com/papers/CDIAr ... 110814.pdf

Code: Select all

                 Arithmetic    std dev    geometric
Italy             6.49%             29.07%    2.46%
Switzerland      6.28%             19.73%    4.48%

On the other hand of the spectrum, Too many slices by factor loads etc., may not add much to the end outcome.The benefits increase substantial in return/std dev benefit but the additional benefit decreases as number of funds increase. The following is the thread by TREV'H which is one the best data mining. There is second one - a discussion between Small-HI/TrevH/RobertT which I cannot find it by Google. Will find it once I get home.

viewtopic.php?t=38374
When in doubt, http://www.bogleheads.org/forum/viewtopic.php?f=1&t=79939
User avatar
jainn
Posts: 314
Joined: Tue Jun 28, 2011 6:41 pm

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by jainn »

Random Walker wrote:Simple gift,
Ya I'm a huge fan of that chart. It basically summarizes the whole essay, and the essay basically summarizes his whole book :-). Note 2 asset classes generally beat 1, and 3 beat 2. I'm sure that additional asset classes have decreasing marginal benefit as the number of asset classes goes up. At some point each investor has to look at additional cost versus additional benefit to a new asset class.
In deciding on whether to add an asset class to a portfolio, the relevant factors are expected return, volatility, correlations to other portfolio components, and of course costs.

Dave
Is there where having 12 asset classes, ala betterment etc... come into play, automatically keeping you diversified broadly with rebalancing? Doesn't work for taxable accounts though I would think...taxes would kill it.
When you deposit money with Betterment, every dollar is seamlessly invested in up to 12 different asset classes, optimized for your selected asset allocation.
https://www.betterment.com/portfolio/
Random Walker
Posts: 4618
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by Random Walker »

Jainin,
I'm overwhelmingly taxable and have pretty highly torqued portfolio. I'm in accumulation phase so I rebalance with new money. Haven't seen any problems with taxes yet. In decumulation, would rebalance with withdrawals. I have tax managed versions of value funds and use core funds to start off the tilts. If anything, wouldn't more asset classes potentially offer more tax loss harvesting opportunities? Something always seems to be going down :-)

Dave
User avatar
Wildebeest
Posts: 1204
Joined: Fri Dec 27, 2013 2:36 pm

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by Wildebeest »

Simplegift wrote:[In this analysis, Mr. Bernstein first supposes that one has clairvoyant knowledge of the exact mix of asset classes that will produce the highest return over a future period — in this case, for the 27-year period from 1970-1996, and among U.S. large and small cap stocks, the MSCI regional stock indexes (Europe, Japan and Pacific ex Japan) and an index of gold mutual funds. It turns out that the best mix consists of just three assets: 35% Japanese stocks, 28% U.S. small stocks and 37% gold. Assuming annual rebalancing, this “optimal” portfolio returned 16.85% per year over the period:
  • CLAIRVOYANT EQUITY PORTFOLIO = 16.85% annual return
    Image
    Note: Average annual return of the S&P 500 was 12.27% for the period.
    Source: William Bernstein, Efficient Frontier (4/97)
What’s notable in this example is that, though the optimal portfolio returned 4.58% compounded more than the S&P 500, it lagged the index in 12 out of 27 years, or 44% of the time. This annual underperformance was quite significant at times, at more than 20% in several years. His conclusion: Equity diversification is not a free lunch. You must be willing to feel like a fool for extended periods of time in order to reap its benefits.

In his second example below, Mr. Bernstein’s acknowledges that no one has clairvoyance about the best-performing assets of the future as they will almost certainly be different than those of the past. What’s an investor to do? His advice is to just pick an equity allocation, almost any allocation, and stick with it. In this case, he chooses equal amounts of the same six assets (the “coward’s strategy”) and tracks the portfolio's performance over the same 27-year period:
  • COWARD'S EQUITY PORTFOLIO = 15.33% annual return
    Image
    Note: Average annual return of the S&P 500 was 12.27% for the period.
    Source: William Bernstein, Efficient Frontier (4/97)
Remarkably, this naive diversification strategy returned just 1.5% less than the very best possible strategy. Again, even though it beat the S&P 500 by 3% annualized, it lagged the index in 13 out of 27 years — nearly half the time.

To interpret Mr. Bernstein’s main points from this exercise:
  • a) Because we don’t know the best-performing equity assets of the future, it’s wise to spread one’s bets as widely as possible.

    b) Even widely-diversified equity portfolios will diverge from the broad global indexes at times, sometimes significantly and for long periods. To reap the rewards, one needs patience and faith that return divergences usually revert.

    c) Much more important than picking the best performers is choosing a reasonably-diversified mix of equities and having the discipline to stick with it over the long haul.
Thoughts?
I wondered how the his "Clairvoyant Equity port folio" has done from 1996 till 2006 and I would expect that the coward's equity portfolio would have beaten it as well of course as the S&P 500.

The next 20 years from here on out I would love to get 5-7 % per year returns if I am fortunate to clairvoyantly pick the high performing equities and this is not for lack of trying.

My bet is:
• Total Stock market US (10 percent)
• Large Value US (20 percent)
• Small Cap US (15 percent)
• Small cap Value US (15 percent)
• European Equities (10 percent)
• Emerging small value (10 percent)
• Emerging market (10 percent)
• Small value international (10 percent)

And we will only rebalance with dividends and new monies coming in ( due to that our accounts are taxable).

The easy part is that I will not bench mark this against anything so that I'll have no regrets or second guess myself. This is the bet and bed I made in 2008 and I will have to lie in this bed and this is what is going to the next 20, 30 or 40 years.
The Golden Rule: One should treat others as one would like others to treat oneself.
User avatar
Topic Author
SimpleGift
Posts: 3869
Joined: Tue Feb 08, 2011 3:45 pm
Location: Central Oregon

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by SimpleGift »

In a PM this morning, the question was asked whether equity diversification still matters as much in today’s world, when financial markets have become more integrated and asset returns are more highly correlated. I’m not exactly sure how Mr. Bernstein might respond to this question in 2016, but in 2008 he commented:
William Bernstein wrote: Although [greater asset correlation] lessens the value of diversification, it is overwhelmed by two factors that favor it: a steady supply of new non-correlated assets into the investment universe and, even more importantly, the compression of asset-class valuations.

Think about it: In a world with only a few asset classes that vary widely in valuation, it just might make sense to place a bet on those with particularly juicy expected returns and stay away from those with poor ones (as occurred, for example, in the 1930s, with its very low bond yields and very high equity payouts.) But in a world with dozens of very noisy equity asset classes with nearly identical ex-ante expected returns, most of the time you’re crazy not to own all of them.
Today in 2016, with the advent of new ETFs and funds that target all the various factors of return (i.e., momentum, profitability, low volatility, etc.), it does at least appear that a “steady supply of new non-correlated assets” is entering the investment universe.
Random Walker
Posts: 4618
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by Random Walker »

I really like this topic, so I thought I'd find an excuse to bump it up. Another one of my favorite short reads is a 3 page essay by Larry Swedroe: Effective Diversification in a Three Factor World. I simply think it's 3 of the best pages of investment reading anyone can find anywhere. And I think it can open up many Bogleheads minds to looking at diversification in another very important way.

Dave

http://thebamalliance.com/pdfs/Effectiv ... wedroe.pdf
User avatar
Topic Author
SimpleGift
Posts: 3869
Joined: Tue Feb 08, 2011 3:45 pm
Location: Central Oregon

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by SimpleGift »

Random Walker wrote:And I think it can open up many Bogleheads minds to looking at diversification in another very important way.
The equity diversification argument for multi-factor investing is an intriguing one — and Larry’s short article is a good basic introduction. However, I wish I could be more convinced about the investment rationale behind of many of these factor return premia and their future persistence.

The equity market premium (beta) has a strong economic rationale going for it and the (nearly) unequivocal view is that it rewards investors for bearing the uncertainty of the value of future cash flows. In contrast, the debate continues about the risk, behavioral and economic rationales behind many of the other factor return premia (small, value, momentum, profitability, low volatility, etc.). Some factor premia may persist and others may not. Some investors are convinced and others are not.

Personally, I’ve tilted modestly to small and value for many years. But I’m not yet prepared to make the full leap to the multi-factor model of portfolio design and equity diversification. I guess you could say that I’ve sipped the Kool-Aid, but haven’t completely ingested it yet!
Random Walker
Posts: 4618
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by Random Walker »

Simplegift,
I certainly see where you're coming from. Personally, at this point I've diversified across all I know: size, value, momentum, profitability. But I have the real conviction for the ones with the most straightforward risk explanations: size and value.

Dave
User avatar
LadyGeek
Site Admin
Posts: 67144
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by LadyGeek »

A complete list of Simplegift's series is now in the wiki: Classic Bernstein
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.
User avatar
nedsaid
Posts: 13929
Joined: Fri Nov 23, 2012 12:33 pm

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by nedsaid »

It makes sense to me to align an equity portfolio something like this:

U.S. and International.

Within US: Value and Growth, Small and Large.

Within International: Developed and Emerging Markets. Within Developed: Large and Mid/Small-Cap. Within Large Developed, I own both Value and Growth.

The thing is, even if you assume that the returns of all the asset classes listed above are identical, you still get a diversification benefit. There are cycles where International does better than US and visa-versa. Growth sometimes does better than Value and sometimes Value outperforms growth. We see similar cycles with Small and Large. I have also seen such cycles with International Developed Markets and International Emerging Markets. Over time, diversifying over the asset classes described above should smooth the ride of your portfolio, less volatility.

I do believe in the Small and Value premiums and I tilt in an attempt to get excess returns. But if the premiums don't show up, I should still have a less volatile portfolio.
A fool and his money are good for business.
User avatar
Robert T
Posts: 2740
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by Robert T »

.
Another related post that has stuck with me since Oct-2002 from Bill Bernstein (wbern) several posts down in the link http://socialize.morningstar.com/NewSoc ... px#1347979

  • We use the DFA portfolios as benchmarks, and let me tell you, they are not easy to beat on a mean variance basis. Over almost any period you want to look at, including those over 10 years, they are superior to more conventional mixes. Most investors could do far, far, worse than to adopt these allocations as closely as possible outside of the DFA family.
The "DFA portfolios" I understood to be the "DFA Balanced Strategies" reflected in this earlier article by David Booth http://www.seiler-associates.com/downlo ... xFunds.pdf and reflected towards the end of the DFA Matrix Book (page 64) - https://www.ifa.com/pdf/matrix_book_us_2016.pdf

70% US
20% Non-US Developed
10% Emerging Market

From 8/1996-4/2016 the US allocation had size/value loads of 0.25/0.44. The overall portfolio size/value loads are probably closer to 0.30/0.45. I take matching "allocations" to mean matching "factor loads".

I have a more global portfolio (personal preference), and slightly lower size and value load targets, but fairly close. Overall actual return difference with a DFA Balanced Strategy with same stock:bond allocation = 0.2% since start of 2003 to date (similar 2008 downside -28.7 vs. -29.1, and similar SD 16.4 vs 16.5).

Obviously no guarantees, but from The Intelligent Asset Allocator - "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."

Robert
.
User avatar
Taylor Larimore
Advisory Board
Posts: 30019
Joined: Tue Feb 27, 2007 8:09 pm
Location: Miami FL

Dr. Bernstein's Three-Fund Portfolio

Post by Taylor Larimore »

PREFACE: This is the third in a series of posts highlighting the classic investing insights of William Bernstein from the 1990s and early 2000s.
simplegift:

This is Dr. Bernstein's advice in his most recent book: "If You Can" (page 1):
Would you believe me if I told you that there's an investment strategy that a seven-year-old could understand, will take you fifteen minutes of work per year, outperform 90 percent of finance professionals in the long run, and make you a millionaire over time?

Well, it is true and here it is: Start by saving 15 percent of your salary at age 25 into a 401(k) plan, an IRA, or a taxable account (or all three). Put equal amounts of that 15 percent into just three different mutual funds:

* A U.S. total stock market index fund.

* An international total stock market index fund.

* A U.S. total bond market index fund.
Bogleheads can read more about this Three-Fund Portfolio HERE.

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle
mikeguima
Posts: 132
Joined: Mon Aug 22, 2016 3:11 pm

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by mikeguima »

I'm sorry to ask what seems to be a rather silly question, but it suddenly came to my mind and I'm having trouble figuring out:
Low volatility helps CAGR, since the returns are compound. Now, let's suppose I hold a diversified portfolio in order to minimize volatility and achieve slightly higher CAGR in return. If I never sell any of the assets I hold (I rebalance with new funds or dividends/ interest) and hold them for decades, how will they compound? I'll never realise the returns so I can't reinvest them to compound them, right?
Random Walker
Posts: 4618
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 3 — Diversifying Portfolio Equities

Post by Random Walker »

Mikeguima,
Not a stupid question at all. I wrote all of the above posts on bringing compounded return closer to average annual return, and I'm not even sure I can visualize the benefit when the holding period is a long time.

Dave
Post Reply