Article: Asset class winners over last 4 decades

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
DaleMaley
Posts: 1592
Joined: Thu Mar 01, 2007 7:04 pm
Location: Fairbury, Illinois
Contact:

Article: Asset class winners over last 4 decades

Post by DaleMaley »

Consistency Matters: Which Asset Classes Have Performed the Best Over the Past Four Decades by Craig Israelsen in February 2011 edition of Financial Planning.

Similar to Callen's Periodic Table, but by decades. Key charts excerpted from article:

Image

Image
Most investors, both institutional and individual, will find that the best way to own common stocks is through an index fund that charges minimal fees. – Warren Buffett
User avatar
fluffyistaken
Posts: 1435
Joined: Fri Apr 04, 2008 1:32 pm

Post by fluffyistaken »

An optimistic typo in the titles of those charts; should be "growth of 10,000" :wink:
User avatar
Topic Author
DaleMaley
Posts: 1592
Joined: Thu Mar 01, 2007 7:04 pm
Location: Fairbury, Illinois
Contact:

Post by DaleMaley »

fluffyistaken wrote:An optimistic typo in the titles of those charts; should be "growth of 10,000" :wink:
Nice catch............I didn't see that...........and apparently neither did editor!!
Most investors, both institutional and individual, will find that the best way to own common stocks is through an index fund that charges minimal fees. – Warren Buffett
User avatar
SSSS
Posts: 1914
Joined: Fri Jun 18, 2010 11:50 am

Post by SSSS »

fluffyistaken wrote:An optimistic typo in the titles of those charts; should be "growth of 10,000" :wink:
Actually it's $1000 contributed per year (presumably at the start of each year) over the course of the decade. $10,000 total, but could be quite a bit different than $10,000 at the start of the decade.
brxn
Posts: 150
Joined: Mon Aug 03, 2009 1:04 pm
Location: USA

Post by brxn »

fluffyistaken wrote:An optimistic typo in the titles of those charts; should be "growth of 10,000" :wink:
$1,000 annual investment, the labeling is correct
dbr
Posts: 46181
Joined: Sun Mar 04, 2007 8:50 am

Post by dbr »

It may be the significant lesson in the data is not which asset classes one invests in but rather in what decade one is investing.
User avatar
bob90245
Posts: 6511
Joined: Mon Feb 19, 2007 7:51 pm

Post by bob90245 »

dbr wrote:It may be the significant lesson in the data is not which asset classes one invests in but rather in what decade one is investing.
True. For a detailed discussion, see my article on Dollar Cost Averaging
Bob's Financial Website wrote:In this article, I will examine how DCA would have performed using historical data. While favorable in most periods, there were certain periods when DCA failed to break even. In other words, the final portfolio value was less than the cumulative amount invested.
Ignore the market noise. Keep to your rebalancing schedule whether that is semi-annual, annual or trigger bands.
Multifactor Advisor
Posts: 266
Joined: Wed Aug 18, 2010 12:42 pm

Post by Multifactor Advisor »

Actually, I think these remote comparisons do diversification some injustice. It makes one assume that you need all of these abstract asset classes (REITs, Commodities, etc.) to achieve any meaningful diversification. When in reality, a simple look at 4 core/high returning equity asset classes shows the same level of decade by decade dispersion:

1970-1979
US Large Value = +11.9%
US Small Value = +14.4%
Int'l Large Value = +11.8%
Int'l Small Value = +23.4%
All 4 = +16.4%
high minus low = 11.6%

1980-1989
US Large Value = +20.8%
US Small Value = +18.3%
Int'l Large Value = +26.1%
Int'l Small Value = +32.1%
All 4 = 25.0%
high minus low = 13.8%

1990-1999
US Large Value = +16.6%
US Small Value = +16.6%
Int'l Large Value = +10.1%
Int'l Small Value = +2.9%
All 4 = +12.2%
high minus low = 13.7%

2000-2010
US Large Value = +3.7%
US Small Value = +13.3%
Int'l Large Value = +6.9%
Int'l Small Value = +12.5%
All 4 = +9.4%
high minus low = 9.6%

In 3 of the 4 decades we se reversals where the leading asset class changes hands, and over the entire period, only Int'l large value stocks did not top the decade long rankings. In each decade, the range between the highest returning asset class and the lowest returning asset class was always at least about 10% annually.

Over the entire stretch we see:

1970-2010
US Large Value = +12.8%
US Small Value = +15.6%
Int'l Large Value = +13.3%
Int'l Small Value = +17.1%
All 4 = +15.5%
high minus low = 4.3%

The "All 4" mix (25x4 rebalanced annually) had a return almost as high as US small value, and much higher than US or Int'l large value, with annual standard deviation lower than 3 of the 4 components (20.8 vs 18.2/26.2/24.3/30.2 respectively). As a matter of fact, the continuously compounded return of the "all 4" was only +14.6% while the simple average of the standard deviations was 25.1. This means that an annually rebalanced portfolio of high returning "core" asset classes had a 0.9% higher return than the average of the components themselves, while turning in an annual standard deviation that was about 20% less. And the range between the best/worst returning asset class over the entire stretch falls to 4.3%, only about 1/3 as high as the average of the 4 seperate periods.

By the way, for those keeping score and wish to use Int'l stocks to lower the risk of a domestic equity portfolio, looking at allocations between 100% US and 50% US/50% Int'l we find that a simple 70% US, 30% Int'l allocation had the lowest volatility. The US only portfolio sat at 21.3, falling to 20.1 for the 70/30 mix, rising back to 20.8 for the 50/50 mix.

The US only mix (50/50 large & small) compounded at +14.4% per year, while the 70/30 global mix did +15.2%. Again, proof that plain vanilla global stock diversification can generate rewarding results.
Roy
Posts: 970
Joined: Wed Sep 10, 2008 9:34 am

Post by Roy »

Hi, Multifactor,

Why are there no returns listed for Growth or Blend funds? Do you recommend Value funds only? (The asset classes used in the article are specified and seem to be "Blend" types, such as S&P 500, or a cobbling of various indexes over time, for certain asset classes, which I'm guessing are different from yours.)

Were the returns you listed attainable by investing in practical and specific vehicles (say, at Vanguard or DFA, for example)? How would an investor have gotten these very returns, especially in decades prior to the creation of appropriate funds? Would he have had to research and invest in individual stocks for each asset class?

What were the Returns for your 50/50 US/INTL?

Thanks.
norm
Posts: 594
Joined: Mon Feb 19, 2007 6:10 pm

Post by norm »

Multifactor Advisor wrote:Actually, I think these remote comparisons do diversification some injustice. It makes one assume that you need all of these abstract asset classes (REITs, Commodities, etc.) to achieve any meaningful diversification. When in reality, a simple look at 4 core/high returning equity asset classes shows the same level of decade by decade dispersion:
Why aren't bonds included in the mix?
Multifactor Advisor
Posts: 266
Joined: Wed Aug 18, 2010 12:42 pm

Post by Multifactor Advisor »

Roy wrote:Hi, Multifactor,

Why are there no returns listed for Growth or Blend funds? Do you recommend Value funds only? (The asset classes used in the article are specified and seem to be "Blend" types, such as S&P 500, or a cobbling of various indexes over time, for certain asset classes, which I'm guessing are different from yours.)

Were the returns you listed attainable by investing in practical and specific vehicles (say, at Vanguard or DFA, for example)? How would an investor have gotten these very returns, especially in decades prior to the creation of appropriate funds? Would he have had to research and invest in individual stocks for each asset class?

What were the Returns for your 50/50 US/INTL?

Thanks.
I would start with value funds and add growth to the extent you are uncomfortable with tracking error. Equities are for long term appreciation, bonds for dampening risk, income, and liquidity. Therefore, in the equity market you should shoot for high returns, in the bond market you should instead aim for low risk.

My point about using value indexes is studies like the one in the OP that place all the emphasis on many different and in some cases exotic asset classes while ignoring the basics around how important the growth/value distinction is.

If I used "blend" or "market" funds instead (4x25) the compound return would have been about 3% per year less.

I used institutional class indexes, but if you lop off about 1%-2% you'd get to the return on retail value indexes (the oldest US value indexes only go back to 1979, so it would limit us to 3 decades instead of the 4 outlined in the OP). Of course, the decade by decade return patters would be similar. Don't get caught up in the specific numbers, it is the patters and conclusion that matters.

As for the availability of value indexes (or any indexes) in the early years -- there were none. But we are talking about risk and return. So unlike hypothetical active strategies, indexes sorted on risk and return are pretty reliable over historical periods.

The 50/50 is the 4x25 I listed for each period. You might wonder about the 70/30 US & Foreign mix I mentioned, which is +15.2%, or 0.3% less than the 50/50. But this is only beccause US stocks underperformed Int'l stocks during this stretch. All developed countries (and asset classes) have the same expected returns over time, so we shouldn't assume a 50/50 has a higher expected return than a 70/30. It just worked out this way for this period.

Norm: I just wanted to focus on stocks. Using 5YR bonds as my bogey, they were the 4th or 5th worst performing asset class in each decade, but did a worthy job of lowering risk as measured by standard deviation (using my 70/30 US & Foreign stock example in a 60/40 stock and 5YR bond mix produced a '70-'10 return of +12.9%. Over this same period, CRSP 1-10 returned +10.0% with about 30% more volatility).
Post Reply