S&P 500 Returns: According to an
Earnings or the Gordon
Dividend Models
“Valuation is the
central concept of finance. While investors believe that the
process is difficult and fraught
with errors and assumptions…valuation
(ought to be) approached with
rigor and care. Without methods to assess
value, capital budgeting
(investing) becomes impossible.”
A
Quote
It is a
great advantage for investors to be able to directly compare the return of
bonds in actual markets to the return of the S&P500, a portfolio of stocks.
We suggest a simple assumption, applicable to large companies, that converts
reported operating earnings according to GAAP into cash flow. The simple
assumption is this: annual capital expenditures (a cost) approximately equals
annual depreciation (an expense).
This simple assumption converts earnings into cash flow, that can then be used
to value the S&P 500.
S&P 500 Operating
Earnings
Standard & Poors
publishes detailed earnings estimates of the S&P 500, that is used in
finance as a diversified reference portfolio. Crucial to this, that we show in Table I, is the distinction
between “operating” and “as reported” earnings. One would normally think of
“operating earnings” as the direct costs on the factory floor of producing
goods, excluding corporate overheads such as interest paid on debt, marketing
costs, r&d and so on. However, the intent of the S&P definition is to
exclude only extraordinary items such as fixed asset writedowns and gains or
losses on the sale of assets. This enables the summation of diverse company
results into a single figure, S&P 500 operating earnings that includes the
cost of leverage.
For large companies,
yearly capital expenditures approximately equals annual depreciation. Therefore S&P 500 operating earnings approximately equals
S&P 500 cash flow available to shareholders. This emphasis on cash flow is
crucial, because it also enables the assumptions of company level corporate
finance to be applied to the valuation of the S&P 500. We can then simply
capitalize annual S&P 500 operating earnings.
The Operating Earnings
Yield Model
But how? Remember that
the duration (payback) period of a stock is approximately 36+ years. This
implies that in a fluctuating economy, one year earnings results, which we will
also discuss later, are less accurate than some form of long-term average. In
1988, Yale economist Robert Shiller published “Stock
Price, Earnings and Expected Dividends.” Campbell and Shiller found, “Long
historical averages of real earnings help forecast present values of
future real dividends (i.e. stock prices factoring out inflation).”
Thus, as we showed in our detailed 7/1/20 post:
Calculated Level of the
S&P 500 a/o 8/22 for a 7% annual return:
P = (10 year average operating
earnings/annual return) X 1.33
Ps&p
500 =
(135.77/.07) X 1.33 = 2580
As we shall see, a 7%
annual rate of return is what is reasonably required to compensate investors
for the long-term risk of investing in equities.
The Gordon Model
It is a tenet
of corporate finance that the total value of the company to shareholders equals
the present value of the stream of dividends it produces, discounted at the
rate of return required by investors (a cost of equity capital to a company).
For companies with a constant growth in dividends, the Gordon model discounts
those dividends back to the present. Thus:
Ps&p 500 = __kE__ =
.3573 X 220.53/(.07-.04) = 2626
( r – g)
P = price
k = the dividend payout ratio=.3573 per NYU
Stern School S&P statistics 2000-2020.
E = S&P operating earnings=220.53 per S&P report, Table I.
Calculated dividends = $71.64. Actual dividends = $64.80
r
= the rate of return of the
investment=7%
g = the
rate of growth in dividends (earnings)=2% real + 2% inflation
We first encountered this
model in a stock market course; it struck us as very difficult to implement
because what do you assume for (r), what do you assume for (g), and where do
you find (k)? What this model requires is an interpretation in context, and
that context is the realities of corporate finance. As in the above, we assume
for economic growth 2% real and 2% inflation and that investment rates of
return should be a 1% markup over the 10 year bond rate of return.
At a 7% rate of return,
the two models are within 1.8% of each other. We prefer the first model because
it is not dependent upon current dividend statistics. The dividend may be cut, as it was in 2009, in the event
of a severe recession. Both models show that the current 8/1/22 level of the
S&P 500 = 4118 is very unrealistic.
The Context of Markets
Both present value models
need to operate within the context of the economy and markets. On 8/2/22 here
is how equity looks according to the first model:
Logical
Expectation 8/1/22 Actual
Market 8/1/22 *
2% Policy
Rate 2.33%
2% 10 year treasury premium .27%
2% BAA
corporate bond premium 2.38%
1%* Equity
risk premium (S&P 500=4118) (.60%)
7%
4.38%
In response to Fed policy rate increases, the 10 year treasury
is beginning to invert; and market traders (and computers) are buying the dips.
Markets are apparently assuming a return of the low interest rate, low
inflation, low growth environment of the past. The Fed policy rate is currently
way below the current 5.90% rate of inflation in the core CPI, that excludes
food and energy.**
*This is a very nuanced
point. We have a moderate distrust of the economic model propounded by
economists, which essentially assumes some form of economic equilibrium, and therefore
no markets. As a value investor, what is more important is to know what is
happening in actual markets and then to concern oneself with the “right” price
at which to act. 1 A NYT 8/25/22 article states this point
well in relation to climate change. “…the field is learning that simply
tinkering with prices won’t be enough as the climate nears catastrophic tipping
points (our note: i.e. economic regime change) like the evaporation of rivers
(around the world), choking off whole regions and setting off a cascade of
economic effects….Experts working on climate change issues say there are plenty
of ways for economists to help. For example, the damage from climate change is
often specific to geographic characteristics like topography, soil quality,
tree cover, and the built environment. Building on those granular factors
to identify systematic risks may be more useful for policymakers than broad,
top-down economic models.”
1 At
a certain level, business contains a great deal of common sense. First, you
decide if you want to do it; and then you talk about the price. Some people
should learn, its not the reverse.
** This risk premium
refers to a diversified portfolio of stocks.