January 1, 2008
2007 Report to Clients*
GROWTH & INCOME FOR RETIREMENT:
Counterintuitive Observations on Portfolio Design
By
Sam Dreher
(January 2008)
Much common wisdom in the investment community
emphasizes growth via equity ownership over income for retirees. Yet an
examination of the dramatic growth in the Dow Jones Industrial Averages
from 1953 to 2006 reveals that even with a tenyear time horizon,
equities can be a dangerous place for retirement capital;
counterintuitively, income(dividends) provides investors with a large,
consistent, portion of investment return. This simple concept,
welllearned, has important applications in portfolio design and related
decisions.
A recently published table of annual Dow Jones Industrial Averages
data going back over fifty years presents investors with the opportunity
to marvel at a dramatic rise in stock prices (1). As shown in the
charts, from the end of 1953 to the end of 2006, the DJIA rose from
280.90 to 12,463.15, an advance surpassing 4,300%. As shown, rising
earnings, rising dividends and rising stock prices in this country have
presented investors with a wonderful wealthbuilding opportunity.
Based on the DJIA’s past appreciation of over 4,300%, before
dividends, the index is a serious contender as an investment vehicle of
choice for future wealth builders. And many investors, now entering
retirement, are encouraged by the investment community to maintain a
large portion of their investment portfolios in growth investments
similar to the DJIA. After all, the logic goes, increased longevity
gives today’s retiree a long time horizon with which to enjoy exposure
to capital growth.
This article is a preliminary examination of the common wisdom that
emphasizes growth over income for retirees, using the results for the
DJIA from 1953 to 2006. If we combine the gross return from growth of
4,340% with dividends received (approximately 1,390%), the total return
is 5,730% for the 53year period. Growth accounts for 75.7% of this
return, with dividends amounting to 24.3% of the total. On the surface,
therefore, growth is more prominent and dominates the final results.
Thus, it is a logical step to embrace growth in investment portfolios,
including those of retirees with “long time horizons.”
Definitions and Assumptions:
To analyze this precept, it is necessary to agree on what is meant by
the term “long time horizon.” On an arbitrary basis, this analysis uses
ten years. For example, buying the DJIA at the end of 1953 resulted in
capital growth of 172% by the end of 1963. Excluding income (dividends
in this case) for the moment, it would seem like a pretty successful
investment program for an individual to grow his capital by 172% after
being retired for ten years. Making a further assumption that this
hypothetical retiree’s spending needs require 5% of his principal every
year, then the results for the 19531963 period are indeed successful:
capital growth is 172%; spending is 50%, and any dividends that come in
are gravy.
A corollary to these parameters of success—again, arbitrarily
assumed—is that an investment failure might be one in which capital
growth was flat or negative ten years after retirement. Given the same
spending rate of 50% over ten years, flat to negative growth for the
period would result in capital depletion of 50% or more before dividends
are taken into account, clearly an unsatisfactory performance for most
retirees ten years into retirement. Refining this assumption further,
investors might also agree that, given a 50%, tenyear withdrawal rate,
capital growth below 25% for the period represents an unsatisfactory
result, or investment failure. Without factoring in dividends or using
any compounding, growth of only 25% over a tenyear period would leave
our hypothetical retiree down 25% after spending needs.
Characteristics of Growth:
Given these admittedly
arbitrary definitions of success and failure, it is possible to examine
the characteristics of growth and income contained in our data source.
As outlined in the table below, average annual growth for the DJIA from
1954 to 2006 was 8.7%, with a rather high standard deviation of 16.3%
and median annual appreciation of 11.8%. These figures point to
potential variability in the results one might expect over shorter time
periods (such as ten years). In fact, the range of growth for the DJIA
confirms this suspicion, with a high of 44.0% appreciation in one year
versus a decline in value of minus 27.6% for one year.
TABLE: DJIA ANNUAL RETURN CHARACTERISTICS

DJIA Price Growth: 19542006

DJIA Dividend Growth: 19542006

Dividend Yields: 19532006

Annual Average

8.7%

5.7%

3.6%

Standard Deviation

16.3%

8.1%

1.3%

Median

11.8%

5.8%

3.6%

Internal Rate of Return

7.4%

5.4%

N/A

Range: High

44.4%

33.4%

6.4%

Range: Low

27.6%

8.2%

1.5%

The discrepancy between average price growth and internal rate of
return (IRR) is further evidence that appreciation might not be as
smooth as the picture of DJIA prices seen in the above chart. Internal
rate of return represents the constant annual growth necessary to
achieve a given value with a given investment over a given period of
time. In this case, $280.90 would grow to $12,463.15 if invested at 7.4%
annually from the end of 1953 to the end of 2006. The IRR is the most
conservative measure of investment results. Due to volatility in
investment returns, IRR can vary substantially from other measures of
investment return, such as average and median.
With these characteristics of the DJIA’s price appreciation in mind,
one might expect some variability in the results attained by investing
retirement capital in the index. This hypothesis can be tested using
tenyear rolling investment periods beginning in 1963. (In other words,
the results from investing for ten years from the end of 1953 to the end
of 1963, the end of 1954 to the end of 1964, etc, without taking
dividends into account.)
On average, the figures look good: from growth alone, the capital
more than doubles every ten years with an average of 120.7% and a median
of 103.0 %.





Histogram: 10Yr Growth DJIA (19632006) 






Average = 120.7% 



Std Dev = 103.0% 



Median = 103.0% 



Range = 30.0% to +323.0% 











1974

1976




1978

1970




1975

1983


2003


1977

1972

1967

2004


1981

1966

2005

1993


1982

1968

1964

1992


1979

1985

1987

1994

1991

1973

1986

2002

2001

1997

1980

2006

1988

1989

2000

1969

1984

1963

1995

1999

1971

1965

1990

1996

1998

50% to +24.9%

+25% to +99.9%

+100% to +174.9%

+175% to +249.9%

+250% to +324.9%

Using the histogram above for a closer look, however, reveals
problems with using growth for reliable investment return. Although, on
average, investors would experience over100% growth every ten years, due
to variability of returns, retired investors would endure sub par
performance of less than 25.0% in eleven of 44 years (1963 to 2006) of
rolling tenyear periods (2). Another way to describe these results is
that retirees investing in the DJIA, without dividends, for the past 44
tenyear periods would have experienced a failure rate of 25% (eleven
divided by fortyfour), based on the parameters set forth earlier. As
predicted, therefore, the variability of annual returns discussed above
can make growth investing a dangerous proposition for retirees. It’s one
thing to have a fiftyplus year time horizon for wealth building; it’s
quite different—and a problem—to have a loss in retirement after ten
years of investing. Although the 4,300plus percent gain shown in the
chart above could be fertile ground for wealth builders, the same
territory could be dangerous for retirees.
Characteristics of Income:
Fortunately for investors, there is another component to investment
return: income (dividends in this case). Return from dividends takes on a
strikingly increased importance when compared to the characteristics of
growth described above. As might be expected, DJIA dividend growth was
more constant, averaging 5.7% annually for the period with a standard
deviation of 8.1%, a median of 5.8% and an IRR of 5.4%. The range of
dividend growth was plus 33.4% to minus 8.2%, exhibiting much less
volatility than growth by itself.
Parenthetically, using the average dividend yield of 3.6% for the
period (also the median yield indicative of low variation over the
years) and combining it with the IRR for price appreciation of 7.4%, one
can derive an annual stock market return of 11.0%, a figure widely
recognized by academics as representing the upper limit of equity
returns over long time periods. Of particular note, under this return
measure dividends account for 32.7% of the total return.






Histogram: Cumulative 10Yr Div Yields on DJIA 




(19632006) 


1971






1969

Average = 53.4% 



1973

Std Dev = 14.4% 



2003

Median = 49.0% 



2001

Range = 30.4% to 86.0% 


1976






1999






2002






2004






1982






1968






1970






1972

1998


1987


2006

1965

1996


1994


1975

2000

1997


1990


2005

1966

1986


1988


1974

1979

1964

1995

1992


1977

1980

1983

1985

1963

1984

1978

1981

1967

1993

1989

1991

30.0% to 39.9%

40.0% to 49.9%

50.0% to 59.9%

60.0% to 69.9%

70.0 to 79.9%

80.0 to 89.9%

As described, dividends represent a larger portion (almost 33%)
of total investment return than they appear to on the surface. Also, as
shown in the histogram, dividends are very consistent compared to
growth. In any of the tenyear investment periods ending from 1963 to
2006, the worst result was plus 30%. In other words, investing in the
DJIA at the end of any year since 1953, investors would have earned at
least a 30% return on investment in ten years. Looking at the parameters
for success and failure set forth above, therefore, it is
interesting that, using dividends without considering growth, the
failure rate was zero: there was never a tenyear period of negative
return and there was never a period of below 25% cumulative return.
Granted, it is not reasonable to look at dividends (or other types of
income) without taking into account the changes that occur to the value
of the investment needed to generate the income in the first place.
This “change in value” is growth, whether negative or positive, and
investment return by definition is a combination of both income and
growth. Hopefully, though, this analysis establishes that income has two
very important roles to play in the total return process:
 Income is a significant portion of the total return, larger than appears on the surface.
 Income’s more consistent nature increases its prominence as a
component of total return, a quality especially important for retirees.
Conclusion
By its nature, growth depends on a buyer other than one’s self
willing to pay more for the same investment. Thus, even if earnings
rise, buyers in the future might not pay enough for them to generate any
growth (increase in price) for the first owners. And vice versa: lower
earnings might receive higher multiples that result in higher prices.
Further, not only is it next to impossible to forecast what multiples
will be paid in the future, it is also indefinite as to when the prices
might be paid. In other words, one cannot predict what another will pay for one’s ownership, nor can she predict when the payment will occur (3).
In contrast to these vagaries of growth, income generated by a
reliable source is steady and predictable. Given this characteristic,
and in contrast to growth that by itself could be said to have a 25%
investment failure rate, it might behoove retired investors to emphasize
income when they are designing their portfolios. For instance, given a
5% withdrawal requirement, a retiree with an investment portfolio
generating 4% current yield could withstand a certain amount of
principal fluctuation, since she would only spend 1% of her principal to
satisfy her spending needs. On the other hand, a portfolio yielding
only 2% would require the investor to withdraw 3% from principal,
perhaps at an inopportune time.
Application:
Charlie Munger (Warren Buffett’s partner) has been quoted as saying
that fully understanding an idea can make a large difference in one’s
life. While these differences between growth and income, and in
particular their relevant characteristics, may be a simple concept,
understanding them well should have profound meaning for retired
investors and their investment advisers. In application, this
understanding may lead to some counterintuitive conclusions and
applications that reach outside direct portfolio management. For
instance:
 We all know that investment return does not equal investor return,
because the available return faces many sources of attrition. Income
taxes are one example. Make no mistake about it, regardless of one’s
political view point, the current capital gains treatment for ordinary
dividends is a wonderful break for retirees. Under current legislation,
this tax break will expire in 2010.
 Properly managed, companies pay dividends from reliable income
sources. Dividends are thereby more important to investors than stock
buybacks, which many times come from extra earnings that are temporary,
or worse, from leverage that could diminish the ability to pay
dividends.
 Inflation is another source of investment attrition, since it
reduces the spending power of a given stream of income. Reliable
dividends that grow with earnings, such as dividends from utility
companies and highgrade industrials, help to diminish the effects of
inflation.
 Investment fees are another source of attrition. Direct investments,
such as bonds versus a bond fund, can eliminate a whole layer of
investment management fees.
 A ladder of individual bonds that enables investors to reinvest
principal maturities at higher rates of return can mitigate the effects
of inflation and also eliminate one layer of investment fees.
 Financial engineering in the form of exchangetraded mutual funds
(ETF’s) and other lowcost investment indices now make possible an
infinite number of portfolio designs. Investors and their advisers
should be able to assemble investment portfolios with the right amounts
of growth and income based on individual circumstances. Using ETF’s,
both income and growth can be diversified, efficiently, by industry,
capitalization, investment style and country and be denominated in any
number of currencies.
 Full understanding of income in the investment process provides
insight into Social Security benefits: when they should be initiated and
how they should be integrated into the investment portfolio and
spending budget.
 Full understanding of income also lends insight into considering immediate annuities and similar income streams.
 Of at least equal importance, a full understanding of income and its
characteristics in the investment process has important implications
for retaining investment counsel. If one’s adviser, broker, consultant
or whatever she is called is not talking about investment selection in
terms that encompass these issues, then perhaps it is time to look for
assistance elsewhere. In an investment environment where pure growth
might have a 25% chance of failure, it’s not enough for the standard
“…your time horizon is long, you need growth, here it is, who’s next…”
solution. Integrating growth and income into the investment/ budget is a
complex process that needs to take into consideration all of the items
listed above and more.
Qualifications:
This article makes broad implications from limited sources. For
example, a study of any number of alternative stock indices could yield
different results. One might examine the S&P 500, S&P MidCap
400, S&P SmallCap 600, Russell 1,000, 2,000 or 3,000 to name a few.
The trajectories of especially the smaller cap indices appear to have
differed from those of their larger relatives. Nevertheless, it seems
stunning, and counterintuitive, that a stock market rising over 4,300%
during the past 50 years can present so much danger to those, such as
retirees, seeking to participate using reasonably long time horizons.
Given today’s technologically endowed investment landscape that allows
investors to inexpensively tailor their investments to meet their
individual circumstances, the idea of arranging, and rearranging, the
balance between income and growth seems worthy of consideration.
(1) Barron’s, editions in 2006 and 2007.
(2) In five of the tenyear periods, returns were actually negative.
(3) In fact, and as a surprise, it was difficult to find any
meaningful correlation between earnings growth, price earnings ratios
and future levels of stock prices, at least using this data source.
*Complete reports available to clients and subscribers in Client Area.

