|
|
Academics and financial theorists have a
concept known as the "cost of equity capital." This is an adjunct
to the cost of debt capital, which is of course the interest rate paid. When
a company is judging whether or not to undertake a project, and how to
finance the project, it must take these factors into account. The project
must be able to pay for the interest on the debt, or the debt won't be paid
and the company will be bankrupt. In a similar fashion, equity holders expect
to get paid too, or else they would just loan the company money instead.
Thus, the "cost of equity capital" is, in a sense, "how much
equity holders expect to be paid."
The cost of equity capital is generally
assumed to be in the range of 13%. Why 13%? After all, the cost of debt (the
interest rate) can be readily ascertained, but there is no such obvious way
to ascertain the cost of equity capital, or what is known as the
"required rate of return." Warren Buffett has said he can't find
any evidence of its existence. This number seems to derive from the average
return on equity for a broad index of publicly traded companies like the
S&P500, which tends to be quite stable in the range of 13% or so.
From a corporate point of view, it doesn't
make much sense to invest in projects that are expected to have an ROE
of less than 13%, since that's below average. Capital will tend to flow to
projects whose returns are expected to be above average, since everyone
is looking for above-average investments. In practice, this flow of
investment makes the sectors in which investment is taking place less
attractive, lowering the return through competition and oversupply. Those
areas in which returns had been poor see less investment, and there is
consolidation and a reduction in competition which can make those areas
better investment prospects again. In this way, the processes of capitalism
direct capital to what is expected to be its most productive use.
This process is rather erratic and slow
moving, however. Witness today the flood of capital pouring into private
equity schemes, with expectations for 20% annual returns reported. To which I
say: yeah, right buddy! On the other hand, if metals prices stay where they
are, the IRR for a new nickel or copper mine can be in the triple digits. Of
course these returns will decline over time, but I think it will not be in
the way most people expect, via falling prices. Rather, copper or nickel
deposits will become so scarce that the cost of obtaining one, via purchase
or discovery, will be exorbitant. This has already begun to happen in the oil
and gas arena. It is virtually impossible to find a new oil or gas deposit of
any significant size.
However, you can't have all projects be above
average, either, by definition. Some projects disappoint, which produces the
below-average returns necessary to produce an average. On the company level,
many investments have the character of venture capital, and any individual
project can be highly risky -- thus demanding a higher rate of return than
debt. Certainly the cost of debt (interest rate) is a lower bound for
returns, as at that point it would make sense for the company to simply loan
money to another company rather than make risky capital investments itself.
Also, it is not at all easy for capital to
flow from one enterprise to another. Let's say one company does not have
prospects where above-average returns on equity can be expected. In that
case, it should return capital to investors, who can then direct it somewhere
else. But how? Dividends are taxed, as are capital gains. And then, the
capital cannot be directly invested into another project, except via direct
investment through venture capital or the stock market.
This is where academics' theories of capital
break down: at the interface between companies and the stock market.
Academics assume that the expectations for company managements (to make a
return on equity at least as good as the average) and for shareholders are
the same. Not so. For if this were true, then the stock market would
consistently trade at 1x book, or at least average some similar figure over
long periods of time. There are arguments that 1x true replacement value is a
proper stock market valuation, and we will leave it for another day to
determine the relationship between book value, ROE, replacement value and
RORV (return on replacement value, an acronym I just made up).
On the contrary, shareholders appear to demand
a return which is the return on purchase price, also known as the earnings
yield or its inverse, the price-earnings ratio. If we take the long-term P/E
of the market of about 14x as a benchmark, then the inverse is about 7%.
However, the underlying companies are still able to compound their capital at
a rate of about 13% -- or at least the portion that they reinvest, which is
about 50%. Thus, of the 13% ROE, about 6.5% is paid out as dividends, and
6.5% is reinvested, producing the roughly 6% observable annual growth in
profits. The 50% dividend payout translates into about a 3.5% average
dividend yield, in the long term. Thus, the equity holder tends to get about
3.5% dividend yield and 6% increase in profits, which translates into 6%
increase in capital value if valuations are stable (ha ha ha),
or about 9.5% annually. These are observable nominal figures, which don't
take into account currency devaluation, which has been considerable over the
past 100 years. Nor do they take into account the fact that the US example
has a very strong survivorship bias. If you were to do the same exercise,
from 1900 to the present, with Russian, Chinese, Japanese or German equities,
the results would be quite a bit different!
If we figure that long-term returns on
equities are about 9.5% (some would say higher, but remember the effects of
inflation!), that is perhaps 200bps above the interest on the debt for the
same basket of corporations, which probably averages about BBB or a little
less. This reflects both the greater certainty of debt holdings (you get paid
every month), combined with debt's higher position in the capital structure
and seniority with respect to the cash flows of the operation.
The main point I want to make here is
regarding the difference between the expectations of equity investors (about
9.5% perhaps) and corporate managements (13%), which I believe reflects the
greater uncertainty of a corporation undertaking what often amounts to a
venture investment compared to a diversified portfolio of publicly traded
equities. Academics' theories tend to conclude that debt financing is always
and everywhere the preferred method of finance, especially when taking into
account corporate taxes. However, this does not take into account the
difference between corporate management's investment parameters, and equity
investors'. We have seen already that the difference between the two can be
expressed in the premium to book value for which equities trade in the
market. Think of it: if a company's stock trades at (perhaps) 4x book value,
doesn't it make sense to sell equity instead of going into debt? When a
company trades above book, then any equity issuance is, by definition,
accretive to per-share book value, while debt issuance is book-value-neutral.
In any case, as the "cost of equity" for shareholders rises (higher
valuation), the "cost of equity" for management falls -- they need
to sell less equity to get their capital. Corporations today almost never
sell equity, except in the case of IPOs. It has been very interesting to
watch the Canadian venture markets, where raising capital via equity issuance
is very common, and a company may do so several times in a year. Managers
have a very clear understanding that the time to raise equity capital is when
the stock price is high, and that it should be avoided (if possible) when the
stock price is low. In other words, their cost of equity changes every day,
which is a rather different view of things than the academics who say that
equity costs 13% forever and ever.
There is an exception to the "no equity
issuance" rule followed by conventional wisdom, and that is in the case
of takeovers. Indeed, in this case, a high stock valuation is considered to
be a "currency" with which companies can buy other companies. But
if a company can take over another via stock, this is functionally the same
as the company issuing equity for cash, and buying the company with cash. And
if that's the case, certainly a company can use cash from an equity issuance
to make investments that would replicate the assets of another company --
indeed, if the target company's market cap is in excess of its true
replacement value, than the acquirer would be better off to issue equity for
cash, and thus create the replacement rather than acquiring the company.
(This raises the issue of greater total capacity in the sector, however.)
One of the benchmarks for acquisition is
whether such a move is "accretive" to per-share earnings, book
value and other metrics. The same benchmarks could be applied to investments
made with cash generated by equity issuance, and indeed this happens very
commonly with companies that raise cash regularly via equity.
On the flip side, since equity capital is
supposedly "expensive" compared to debt capital, always and
everywhere, managements have been incentivized to use cash to buy back stock
at any price, as if its was debt with an interest payment of 13%, rather than
paying down debt. But certainly, it makes more sense for a company to buy
back stock when the stock's valuation is low compared to when it is high. This
is particularly obvious in the extreme case when the dividend yield is above
the cost of borrowing, and it becomes a cashflow-positive
proposition to borrow to buy back stock. The same principle must apply in
intermediate situations as well, and there must be a situation where the
stock's valuation is so high compared to debt that it makes sense to pay off
debt rather than buy back stock, or even issue new equity to pay off debt.
Such a move could be "accretive" to shareholders when relatively little
dilution occurs compared to the per-share improvements in book value and net
income.
Nathan
Lewis
Nathan Lewis was formerly the chief international
economist of a leading economic forecasting firm. He now works in asset
management. Lewis has written for the Financial Times, the Wall Street
Journal Asia, the Japan Times, Pravda, and other publications. He has
appeared on financial television in the United
States, Japan,
and the Middle East. About the Book: Gold:
The Once and Future Money (Wiley, 2007, ISBN: 978-0-470-04766-8, $27.95) is
available at bookstores nationwide, from all major online booksellers, and
direct from the publisher at www.wileyfinance.com or 800-225-5945. In Canada,
call 800-567-4797.
|
|
