To all - yet another bunch of stuff regarding valuation level of stock market currently.
Here are two Barrons articles -- one from last week, one from this week.
(The Federal Reserve economist (Leonard Nakamura) who came up with these ideas was mentioned at least once before in another Barrons thing. I cannot remember what the subject was that time, but my response to it was similar to my response to this new stuff -- it DOES makes sense).
*******************************
August 16, 1999
Economist's Toil and Trouble Tell of
Mini-Bubble, But It's Hardly as Bad as Bears
Would Have You Think
By Gene Epstein
The stock market is way overvalued. We all know that. The price/earnings ratio
is far above its historic range, and a major correction is bound to happen sooner
or later. I myself pray daily for a long, drawn-out trading-range market, one that
oscillates nicely between 9,000 and 11,000, so that earnings will have some time
to catch up to prices.
Now comes Philadelphia Federal Reserve economist Leonard Nakamura to say
that the market's P/E, properly adjusted, is not nearly as high as the standard
measures indicate. Maybe you've heard this all before. But Nakamura doesn't base
his case on one more version of bubble accounting; he draws instead on solid
economics that is surprising for its simplicity.
To begin with, the denominator in his adjusted P/E conforms to the industry rule
of tracking earnings for the previous four quarters. So it doesn't presume to make
any New Age prediction about future earnings growth. But as the chart on this
page shows, its 30-year history creates a very different pattern from the notorious
course of the standard P/E. While the conventional price/earnings ratio is currently
in the stratosphere compared with its own past, Nakamura's measure is barely
above its historic range.
The main difference is that the standard P/E puts "book profit" in the denominator,
while Nakamura employs a variant of "economic profit." And it turns out that
economic profit comes much closer to being the genuine article, since it eliminates
the distortions that corporate accountants introduce when they book a profit.
For starters, earnings got bloated in the 1970s compared with the current decade
because the double-digit inflation of that period caused phantom inventory profits.
Through the method called "inventory valuation adjustment," economists remove
this bias by washing out the part of inventory profit that is attributable to inflation.
Another distortion arises from the Reagan tax cut of 1982, which permitted
greatly accelerated depreciation charges that could be deducted from earnings.
While the change in the law was good news for corporations, it's caused post-'82
profits to look artificially lower than those of pre-'82. Economists avoid this
problem by using a method called "capital consumption adjustment," which is a
synonym for depreciation. C.C.A. proceeds according to the rate of true
depreciation instead of the rate permitted by tax law.
These two adjustment methods -- inventory valuation and capital consumption --
have the effect of boosting the current decade's earnings compared with earnings
in the 1970s. And if today's profits are higher, then today's P/Es are lower. But
then Nakamura puts an unabashedly New Age spin on the concept of economic
profit by introducing an adjustment of his own: He adjusts earnings for the recent
surge in R&D.
Generally accepted accounting principles require that investment in research and
development be treated as a current expense. This is probably a wise policy,
because if companies were allowed to depreciate R&D costs over an extended
period, then they might be tempted to lump ordinary expenses with R&D in order
to disguise a loss. And of course, for tax purposes, firms prefer to expense R&D,
since doing so amounts to the most accelerated depreciation loophole of all -- a
100% write-off in the first year.
But from an economic point of view, says Nakamura, there's really no good
reason why such investment shouldn't be capitalized. True, R&D differs from
spending on plant and equipment because fixed assets can usually be sold for
other uses if their current use turns out to be unprofitable. But the cost of
researching and developing such items as Viagra, the Gillette Mach3 razor blade or
Windows 98 is the same animal in every other way. In Nakamura's words, what's
involved is the purchase of "intangible capital," and just as with capital of the
tangible kind, money is being put at risk in order to reap a profit over the long run.
One might also object that physical capital eventually wears out, which gets back
to the very word "depreciation." But trade secrets and brand names have a service
life that is usually limited by time, not to mention patents and copyrights, whose
durations are limited by law. In all these cases, a firm is buying assets that are not
used up in the first year of production, which means these are capital assets that
should be depreciated over time.
This point matters for the trend in earnings because private sector spending on
R&D has been growing at a faster rate in the 1990s than in any other decade. But
because firms are expensing rather than capitalizing these costs, this decade's
book profits are lower than they would otherwise be.
Nakamura adjusts for this bias by performing a simple "straight-line" depreciation
of R&D over a six-year period. The six-year assumption is fairly conservative; he
could plausibly assume seven or eight years, with result that current profits would
look even higher and current P/Es even lower.
As the chart indicates, the economic profit price/earnings ratio stood at 21.2 at the
end of the first quarter (the most recent period for which calculations were
possible), up from 20.7 at the end of last year. The previous end-of-year high was
19.8, which was hit in 1972. By contrast, the book-profit P/E was a lofty 27.8 at
the end of the first quarter, compared with a mere 18.4 on the last trading day of
'72.
So instead of the big bubble telegraphed by the conventional measure, at most
we're dealing with a mini-bubble. And at worst we may be due for a
mini-correction, which means stocks might do no harm to this lovely economic
expansion.
Israeli Economist Sees Big Rise for Dow
Speaking of New Age Economics, superbull Mike Astrachan (pronounced
"Astrakan") thinks stocks may be dirt cheap. Over the next two to four years, he
believes that Dow 20,000, and maybe even 30,000, is a distinct possibility. (Given
the wonders of compounding, a 30% annual appreciation from the present level
would bring us to 30,000 in four years.) But he still sees near-term risks and
would prefer to be a buyer when he sees the Dow pull back to 9,000.
Of course, anyone can go out on a limb like that, but the Israeli economist and
investor has a few plausible arguments and can boast an awesome track record.
More than three and a half years ago, when I first wrote up his forecasts ("An
Economist Sees the Dow Hitting 6000 and Japan in Twilight as a Superpower,"
December 11, 1995), he's been uncannily accurate in calling successive legs of
this bull market, including the bearish hiccup last fall.
The standard P/E for the broad market (red line) is based on "bank
profits," a measure that suffers from various distortions. A more
accurate P/E (gray line) puts "economic profits" in the denominator.
Notice that the book-profit P/E is currently in the stratosphere, while the
economic-profit P/E is barely outside its historic range. (Book-profit P/E,
based on operating earnings, puts the S&P 500 Index in the numerator;
the numerator in economic-profit P/E is the market value of all
nonfinancial corporations.)
His basic message is still the same: What's new about this new economy, and the
key factor that has been driving double-digit earnings growth, is the zeal to
achieve greater productivity through constant restructuring. Since the figures on
corporate layoffs are still quite strong, this factor should continue to lift stock
prices to shockingly high levels.
Astrachan's main concern: that accelerating wage growth will give labor all the
benefits of rising productivity, leaving nothing for profits. But he expects that
higher interest rates, including a Fed tightening, will soon cool off this overheated
economy.
E-mail:gene.epstein@barrons.com
Copyright ¸ 1999 Dow Jones & Company, Inc. All Rights Reserved.
********************************************
August 23, 1999
A New Stock Valuation Model Suggests the
Market May Not Be So Pricey After All
By Gene Epstein
Some might call this a Pollyanna model of stock-market valuation. But I prefer to
dub it a model of modesty, since it's based on known information and strictly
eschews any Pollyanna-ish assumptions about the unknown.
What it says is that from the end of 1993 through 1998, the market may actually
have been undervalued, and that as of the first quarter '99, the degree of
overvaluation was relatively small, at least by historic standards. So from this
perspective, we've recently been living through the time of a mini-bubble and not,
as conventional models would have it, the Big One that could bring ruin on us all.
Last week I
unfurled a
price/earnings ratio
for the broad
market that was
based on the
concept of
economic profits
rather than the
book profits that
go into the
denominator of the
standard P/Es. The
ratio, which was
developed by
Philadelphia Fed
economist Leonard
Nakamura,
showed that the
P/E for the first
quarter was barely
outside its 30-year range, in sharp contrast to the same reading on the
conventional P/E, which was far above it.
My discussion provoked a firestorm of protest from readers who accused me and
Nakamura of being accounting whores for the speculative rabble. But we weren't
denying the truth of Internet mania; we were saying only that the run-up in the
broad market might be more justified than conventional indicators say. Herewith
an additional massage to this measure, inspired by the Federal Reserve's
stock-valuation model.
As readers familiar with the Fed approach know, what it basically does is compare
the earnings yield with the yield on the 10-year Treasury note. If the former is
lower than the latter, that indicates overvaluation, and vice-versa. The earnings
yield is just the price/earnings ratio flipped over: E/P, taken as a percentage. So,
for example, four-quarter trailing economic profits ran 4.71% of total market
value as of the first quarter, versus a Treasury yield at that point of 5.25%. Result:
5.25 divided by 4.71 equals 1.11.
Now, that's above one, which is not so good, but as the chart shows, it's still well
within its historic range. Of course, one reason we get this result is that 5.25% on
the 10-year note was not far above its three-decade low.
One way to think about this model is to introduce what's missing, which is the
risk-adjusted value of the future earnings stream. The reason you buy the market
even though its recent earnings yield is lower than the risk-free rate offered by
Treasuries is that you expect the future earnings stream to do a lot better than
Treasuries.
Accordingly, in order to interpret this chart, let's ask two questions. First, should
today's risk adjustment be on the high side of this three-decade span or on the low
side? Answer: More likely on the low side, given the tamer nature of the business
cycle. Second, should today's assumption for earnings growth be higher or lower
than past assumptions? Answer: Probably higher, based on the record of the past
few years.
These two assumptions taken together mean that the risk-adjusted value of the
future earnings stream looks a bit better than it used to. So if anything, the mild
overvaluation signal featured on this chart is somewhat less troubling than those of
the past.
But of course, much of this turns on whether you buy the concept of economic
profit. Like any measure, it's an approximation, but it's much more accurate than
the book profits used in the conventional P/E. Its principal weakness may be that
it's difficult to make it very current, since there's a big delay in the availability of
the figures that go into it.
Critics Respond
One critic noted that the numerator of this P/E is the market value of non-financial
corporations rather than the more commonly used S&P Index and wanted to
know whether this made any difference in the result. In fact, it makes virtually no
difference, since these two measures have tracked each other closely. Another
asserted that such indicators as price-to-book-value still say the market is
overvalued. But actually, if R&D is capitalized, as economic theory says it should
be, then book value is boosted in the same way as earnings are.
Still another critic argued that the liberal use of stock options has been causing
recent earnings to be overstated. While this argument has been exaggerated, there
is undoubtedly some truth in it. But on the other side of the ledger, there is still a
great deal of understatement that Nakamura still hasn't taken into account. For
example, there's a lot of R&D going on that isn't counted as such. Think of
financial corporations: Virtually all they have is intangible capital, but the huge
sums they spend on developing new products are expensed rather than capitalized.
E-mail: gene.epstein@barrons.com
Copyright ¸ 1999 Dow Jones & Company, Inc. All Rights Reserved. |
