Someone asked me whether I thought the stock market was
overvalued, undervalued, or fairly valued. They were of the
opinion that the Shiller P/E - which is now 23 - overstates how
expensive the stock market is. They think today's plain vanilla
P/E ratio is more accurate. They also think today's return on
equity may be sustainable. I don't share either opinion. I think
the Shiller P/E is a better measure of the market's price level
than the P/E ratios you see in the press (which are either
trailing P/E ratios or - worse yet - forward P/E ratios based on
analyst estimates). And I don't think today's returns on equity
I think it is more likely than not that return on equity will
trend down rather than up over the next five to 15 years. At
least in America. I can't speak for other countries.
These are important points for investors. They mean I think
earnings are - if anything - above normal rather than below
normal. And, if earnings are above normal right now and P/E
ratios are already normal or higher - that means, prices are
already normal or above normal.
I think that's true. But there's a catch.
I don't know anything about making macroeconomic predictions.
I read the Buffett article on return on equity that was recently
reprinted in "Tap Dancing to Work." I thought the Buffett article
made sense when he wrote it. I still think it makes sense.
However, almost everything Buffett said in that article turned
out to be wrong. For that reason, there's a good chance even if
everything I say here is correctly reasoned it could turn out to
be wrong. The future could look nothing like I expect it to look.
Buffett's article made a lot of sense to me. And yet returns on
equity did go up because of changes in tax rates, changes in the
amount of GDP going to corporate profits versus labor, and
changes in the use of leverage.
These are complicated issues. They are especially complicated
issues when we are talking about the entire economy and how it
works. For example, Japan has a high debt to GDP ratio. Everyone
knows that. What a lot of people don't talk about is their low,
low, low corporate debt and high cash balances. In that case,
yes, the government is indebted - but the people and companies
are saving. How can we know what returns on equity will look like
in the future?
You see the same thing in everyone from Italy to the United
States. Italy has government debt problems. But Italian
households do not have debt problems. American households may
have debt and their government has some debt - and a very large
deficit - but their corporations (and now their banks) have
strong balance sheets.
So most discussions you see of any country's economic future are
superficial. They make a quick comment about how much debt that
country's government has, its corporate sector has, its banking
system has, its households have, etc., but they do not go through
the entire economy explaining how some balance sheets are
stronger or weaker than their counterparts in other countries.
For example, American investors looking at European banks for the
first time immediately notice weaknesses in those banks. Until
looking at European banks, I doubt most American investors
noticed how good the funding of their own banks is. After looking
at European banks, they realize this difference.
But do they incorporate this into all their macroeconomic
I doubt it. I doubt the average investor can think very sensible
thoughts about the future of return on equity in the U.S. versus
Australia versus Japan versus Italy versus Brazil, very easily if
at all. I know my thoughts on the subject tend to be superficial.
They reflect nothing more than the conventional wisdom about
those places and their futures. This wisdom is often wrong. And
where it is not wrong - it is usually already priced into stocks.
The problem with predictions about return on equity is that they
often focus on just one aspect of the overall economy. The
reality is that future nominal returns on equity for American
companies depend on:
1. The amount of capital they have invested in the business
2. The amount of money in the economy
3. The amount of actual demand for their products (in real terms)
4. The amount of leverage they use
5. The amount they pay in taxes
6. The amount they pay to labor
Think of Japan. Japanese companies have very low returns on
equity. How predictable was this? To some extent, it was
predictable in that they often had low returns on equity even
during strong economic growth, and in that, they clearly
over-invested in terms of the amount of tangible capital they
added to their businesses versus reasonable expectations of
We can see a danger of China doing the same thing now. They have
very high capital spending relative to their population growth.
That makes the risk of having a glut of invested capital (and
eventually tangible equity) high.
It's worth noting that in both cases I have said that an excess
of tangible investment may eventually lead to low returns on
equity even if much of the expansion is debt fueled at the
moment. In other words, I'm saying that too many added assets are
a problem for equity investors because even if those assets are
debt supported now - they may become more and more equity
supported over time if they turn out to have been a mistake. We
can see this with many Japanese companies. To the extent they
have remained in place for two decades, they have tended to
replace debt with equity and thus either lower dividend payments
and stock buybacks to investors below what would be expected
and/or lower returns on equity even while returns on assets have
not actually worsened. In plain English, they've de-leveraged.
The result can be - as it has been in Japan - both contracting
multiples and low shareholder returns even compared to corporate
I bring up Japan not because I expect the United States to be at
all similar to Japan, but because I would otherwise have to bring
up ancient American financial history. I would have to talk to
you about what returns on equity were like in the 1940s and
1970s. To modern investors, these periods sound as foreign as a
place like Japan. But they are certainly possible futures because
they were actual pasts.
I can't predict future returns on equity for American companies
any better than Buffett did. And he did not do a good job as it
However, I believe they will be lower in the future than they are
now. I think returns on equity at the moment are quite high.
There is very little evidence that incremental returns on equity
could be remotely as high as average returns on equity. If you
have a company with $100 million of tangible equity earning $20
million on that equity - I promise you that in most cases that
company does not believe it could earn another $250,000 if it
added $1 million more of equity. I think many companies in such a
position actually doubt they could earn $100,000 on that next $1
million of equity.
That doesn't mean earnings have to be above normal right now. If
the future does not involve much added equity - it can include
the same sorts of earnings relative to equity that it does now.
In fact, that is why I think return on equity has been
surprisingly sustainable in the recent past. It has not been
necessary to add much equity to American businesses.
Corporate profits have done a lot better than wages for a while
now. People give a lot of explanations for why that is. In most
cases, I'm not terribly convinced by any of those explanations.
All I can see to explain what has happened to corporate profits
relative to wages is that over time there has been more trade and
more technology and both of those things make labor more
efficient. In the U.S. - which has been a slowing growing economy
during that time - it makes sense to limit the amount of (very
expensive) labor you use and earn more money using similar
numbers of employees rather than trying to add a lot of employees
to grow. Growth rates have not been strong on the top line for
many American companies relative to what they've accomplished on
the bottom line.
I certainly think that real return on tangible equity - reported
returns less inflation - are incredibly high right now. I think
it would be crazy to assume such returns will increase in the
future. I think it is reasonable to assume they will decline. If
history is any guide, it is possible that returns on equity for
American companies could decline to half of what they are now -
at some point. I am not predicting that will happen. But it did
happen a couple times in the 20th century. It has also happened
in other countries for extended periods.
For those reasons, I think people who look at the market P/E
ratio - the median P/E on the 1,700 stocks Value Line covers is
16.5 right now - and think that is low - are wrong. There is no
reasonable argument to be made that stocks represent a good value
in absolute terms or in historical terms. Stocks are overpriced
both in terms of the return you would expect going forward and in
terms of where they are priced relative to the past. The only
price measure that makes stocks look reasonable is a simple P/E
approach. If we used price-to-average-earnings, price-to-book, or
price-to-sales, they would appear overvalued. So the only
argument for stocks earning decent returns in the future is a
narrow argument. You have to make a very specific argument - that
the price-to-sales ratio, the price-to-book ratio, the Shiller
P/E, the dividend yield, etc. - is wrong and the current P/E is
right. It might be. But it is one measure. And unless the current
level of earnings is perfectly normal, there is no argument in
favor of buying stocks right now.
Stocks represent a relative value.
Stocks in other countries - outside the U.S. - are mostly not
that cheap. Commodities are - in almost all cases - clearly
expensive. In fact, most are further from their historical values
than U.S. stocks. And, of course, bonds are the most overvalued
asset of all. So, the argument is that American stocks are a
better choice than foreign stocks, commodities, and bonds.
I don't disagree with that. I think if you are a buy and hold
investor you are probably safer in American stocks than in
foreign stocks, commodities, or bonds.
However, are you safer than if you were in cash?
I don't think so. I am not advocating going into cash to time a
market. But I am advocating - and have always advocated - going
into cash when you can't find investments that you feel
comfortable holding and which you believe offer a good chance of
10% annual returns. Right now, I think it unlikely that investors
can find many such investments. I personally am 50% in cash. I
would love to be 100% in investments I thought would return 10%
or better. I have had trouble finding such stocks. I keep
looking. And I am sure I will eventually be 100% in stocks again.
It could happen very quickly.
Here is my problem with the argument against the Shiller P/E. I
don't believe the Shiller P/E is necessarily a better price
measure than any other price measure. However, I know the P/E
ratio is the worst measure of all. If there is one number that is
the least likely to help you pick a stock or time a market, it is
the P/E ratio. Every investor looks at that ratio. It is the most
likely number to get you into trouble. It is the most likely
number to sell you on a myth.
I don't know if current returns on equity being sustainable is a
myth. I would say the same thing about the price of a barrel of
oil right now. It might be perfectly correct. But it has now been
- in real terms - much higher for much longer than it had been in
the past. That starts to encourage companies to change behavior.
I have analyzed
in the past. They and their largest competitor, RCL, have cut
fuel consumption per passenger each year for several years now.
They are - when they set their minds to it, especially if they
design ships with fuel economy in mind - quite capable of
reducing fuel consumption. There was no reason for them to make
any effort to cut fuel consumption in the entire decade of the
1990s or the early part of the 2000s.
The same is true on the supply side. People always talk about how
much oil there is, where it is, etc. That's always a much
squishier subject than the press makes it sound. It's ultimately
an economic question. At $20 a barrel, there is a lot of oil
nobody would be concerned with thinking about. It may exist, but
if it costs $60 to get out of the ground - it makes little
difference whether it exists or not, it's not something anybody
is thinking about. That changes if oil tends to trade at $90 a
barrel for a long time.
It may sound like this has nothing to do with return on equity.
But it does. The problem with return on equity at big companies
generally is similar to the problem
has right now. If Apple's return on equity is normal right now -
and if Apple won't shrink in the future - it's clearly a cheap
and wonderful stock. I'm not sure it is a cheap and wonderful
stock. I'm not saying it isn't. I'm saying we have very little
basis to say it is. Why not?
Because Apple's history of high and stable margins, returns on
equity, etc. is very short - almost non-existent. It is has a
strong, short history of positive business momentum. It's an
amazing upward climb. But it has no long-term average basis - no
sustained period of plateau - on which to make any projection
into the future. All it has is a period of constant improvement.
The trajectory is unsustainable. The question is whether the
point where that trajectory ended - we'll call that the present
day - is sustainable. I won't argue it isn't. But I will argue we
have no basis on which to say it is.
Let's consider this problem one of repeatability across time? Can
Apple repeat its performance year after year?
Apple's cash pile is very large. Meanwhile, it has not bought
back lots of stock, etc., every year. It has no "safety valve" in
its past that would stop return on equity from declining
dramatically in the future. All signs point to poorer returns on
capital in Apple's future unless it constantly innovates. This is
not an argument against Apple. It is an argument that constant
acceleration is required to maintain a constant rate of return on
its equity. It is an argument that in such a treadmill-like
position - one in which the rate of asset growth is so high, and
the rate of return on those assets is so high to always encourage
more asset growth and more competition - you really need constant
acceleration to maintain a constant speed.
I am speaking metaphorically. But I think it's an illustrative
analog. The temptation to invest more when the return on
investment is so high creates difficulty in maintaining a
constant rate of return in the same way that experiencing drag at
all times would make maintaining a certain speed forever a more
remarkable feat than it might first appear.
That is why I am more likely to look at stocks like John Wiley,
Weight Watchers, Dun & Bradstreet, etc. They have very
limited (in D&B's case absolutely no) hopes of future growth.
They can't put a lot of new capital to work in the business. But
they don't have to put a lot of new capital to work to earn the
same returns. If we applied the metaphor to these companies - I
would say that it is possible to maintain a constant speed
without any increase in acceleration. I can certainly misread the
situation and miss strong forces that will slow the business
down. But I can see reasons why the addition of lots of assets -
either by these companies or their competitors - is low relative
to the very high returns on investment these businesses provide.
What I'm saying is that I can reason why the status quo may be
Will Apple maintain its status quo?
Will the average American public company?
I think the chances are much, much higher they will not. I think
it's reasonable to fear that over the next five to 15 years a lot
of companies will add capital that relative to their increases in
earnings will tend to drive down average returns on equity. I
think this is a fat period for returns on equity, mostly because
it has been possible to do the same with less. There is a lot of
evidence of this in every company's annual reports. They have
seen weak demand and yet they have increased earnings. How did
they do it?
They did it by making themselves more efficient, more profitable,
That's a one-time gap closing. It's nice. But it's not
permanently nice. It's the equivalent of the run stocks had from
the early 1980s to the mid-1990s. That was a wonderful run - and
it made perfect sense, just like the run from the 1940s through
the 1960s made sense - because prices were too low at the start
and they were just right at the end.
I am not saying that we are definitely at a peak in terms of
return on equity. I am, however, saying that you should exercise
as much caution with assuming the current level of return on
equity is sustainable as you would in assuming the current oil
price per barrel is sustainable. It may be. But it is hardly an
unremarkable level in history. It is not historically "normal".
Today's returns on equity could be a good predictor of future
returns on equity the same way today's oil prices could be a good
predictor of tomorrow's oil prices. In both cases, the present is
unusual compared to the past even if it is predictive of the
future. So, we need to be extra cautious when taking the present
as a guide to the future.
Finally, the Shiller P/E doesn't even use a super-long average
the way using the average real price of oil for every year since
the 1860s would. The Shiller data goes back to around 1870. But
the calculation only uses a 10-year trailing average. That means
a Shiller P/E for today is covering the period 2002 through 2012.
That's hardly a low return on equity time relative to the full
history of U.S. business. A simple price-to-tangible-book metric
would be far more damning than the Shiller P/E ratio in
suggesting stocks are overvalued. In fact, since probably about
2010, the stock market has had an above average price-to-book
ratio. For example, even in early 2009 (at the market bottom)
price to book was much higher than it had been for the market in
the 1974 to 1982 period (basically the most recent modern era in
which American stocks generally were consistently and undeniably
In a sense, the Shiller P/E ratio is agnostic about return on
equity. It actually treats the last decade as normal in all
circumstances rather than price-to-book which obviously takes
care of the issue of high and low returns on equity. This is
important in places like Japan. The price-to-book ratio is low
there. However, the Shiller P/E is often not. In fact, it has
often been higher than other major markets. That's the argument
people have in Japan - it has low asset prices but high earnings
prices. I'm generalizing. Japanese small caps are pretty cheap on
a lot of measures. It tends to be the big cap parts of the index
that exaggerate how expensive Japanese stocks are generally. But
you get my point. In Japan, price-to-book is low but the Shiller
P/E is high. In America, price-to-book is very high, the Shiller
P/E is high, and the regular old P/E is normal. Personally, I
don't see how that combination of price ratios can possibly argue
for cheapness and I doubt it even argues for normal prices.
The only solid argument for American stocks is that they are
cheap relative to other assets. That's a perfectly defensible
argument. Arguments for any absolute value in American stocks
depend on assuming today's high returns on equity are perfectly
sustainable - not even that they will just slightly trend down
over time - and even then, that only gets you to the idea
American stocks (according to their P/E ratios) are normally
priced. I don't see any measure that says they are cheap.
In the end, I have no idea what returns on equity will do.
However, the fact that they are so high suggests you should
exercise special caution when looking at P/E ratios, etc. In
other words, I would put lower than average faith in today's P/E
ratios because they are based on higher than average returns on
equity. Again, this is really no different than my attitude about
oil prices. If I was analyzing an oil company that benefits from
higher oil prices I would give them the detriment of the doubt
and insist that they trade at even lower P/E ratios than I would
normally like to see. Likewise, I could be persuaded to give
consumers of fuel like cruise lines and airlines the benefit of
the doubt and assume that if they were trading at a normal P/E
ratio today it might actually be a below average P/E if we
understand that oil prices are particularly tough right now.
In exactly the same way, I would look at any stock's P/E in the
U.S. with the understanding that return on equity is especially
easy to come by these days. In the future, it may not be. So I
would exercise extra caution. I would probably be inclined to
look at both the P/E ratio and the price-to-book ratio. For the
market as a whole, I think the Shiller P/E ratio is at least as
good as the regular old P/E ratio (which has almost no value in
predicting future returns). And, most importantly, I think the
Shiller P/E ratio is better at helping investors avoid the
absolute biggest disasters by keeping them out of the market when
the market is obscenely overpriced.
I can offer my own personal opinion. It is only an opinion. It is
based on facts but also personal prejudice, my own anecdotal
experience of having trouble finding stocks to buy right now,
I think U.S. stocks are overpriced. I don't think they will
perform well in the future. I think American stocks are more
expensive now than they've usually been in the past. And so I
think they will perform worse in the future than they did in the
That does not mean there will be a crash. That does not mean
American stocks will not outperform American bonds. And it does
not mean you are better off in cash than you are in stocks. It
might mean that. But I certainly can't argue it does mean that.
What I can argue is that it is wrong to assume you will make 8% a
year or more in stocks going forward. I don't see any reason to
believe that is true. That doesn't mean you shouldn't invest in
stocks. It does mean you should go into any such investment
knowing you are paying a higher price for stocks than most
investors have at most times in the past.
Historically, this is a pricey time in American stocks. You have
to use arguments based on forward earnings estimates, low
interest rates, etc. to argue it is only a normally priced time.
The problem with those arguments is that forward earnings
estimates are always too high and interest rates are almost
always higher than they are now.
Betting either - and I think in this case, you would need to bet
both - on forward earnings estimates being accurate and interest
rates staying at their current levels seems like a big gamble.
I don't have any better knowledge than anyone else about how
actual earnings will differ from estimated earnings. Nor do I
have better knowledge than anyone else about how tomorrow's
interest rates will differ from today's interest rates.
But if I was forced to guess, I would put the odds at better than
50-50 that forward earnings estimates are too high. And I would
also put the odds at better than 50-50 that future interest rates
will be higher than present interest rates.
The big problem is - like I said - I think a split decision goes
to the bears in this case. Even if interest rates stay low,
forward P/E ratios are the only "normal" P/E ratios at the
moment. Trailing and Shiller P/Es are already elevated. So are
price-to-sales, price-to-book, and other such measures.
If we did a thought experiment where we said to get a
historically average return in stocks from this day forward you
would need both forward earnings estimates to be accurate or
conservative and you would need future interest rates to be the
same or lower than today's interest rates - we can see there is
less than a 25% chance of stocks giving you returns equal to what
they did in the past.
I base this on the (incorrect, but used for the sake of
simplicity) assumption that the accuracy of forward earnings
estimates and the direction of future interest rates are
independent events. In each case, it is more likely than not that
the current state (the current forward earnings, the current
level of interest rates) is above the actual future level. If you
needed two coin flips to be "heads" and you knew both coins were
weighted in favor of "tails" you have less than a 25% chance of
winning both flips.
That is a simple, almost childish way of looking at market
returns. But I think it's a more practical description of the
situation we face than the more elegant explanations you see.
I think you are betting on both outcomes. I think we know enough
to know both outcomes are weighted against you.
That means the odds are at least 3-to-1 that stocks will provide
annual returns in the future that are lower than what they
returned over their entire recorded history. If you are using a
"Stocks For the Long Run" type approach to what you think you
will earn in American stocks - if you are hoping for something
like 9% a year - I think the odds are at least 3-to-1 against
That's not the same as saying this is 2008 all over again. But it
is part of the reason why I am still 50% in cash right now. I
don't like cash. I love stocks. I want to own stocks. If they
were - on average - promising returns of almost 10% a year, you
can bet I'd be all in.
Even now, I think I can find better than average stocks. So if
the average stock promised 8% or 9% or 10% returns - why would
any of my money be on the sidelines?
It probably wouldn't. It probably is because even some above
average stocks now only promise 8% or 9% or 10% returns. Average
stocks promise even less.
Again, that's my opinion. I think it is an opinion based on some
evidence. But it is far from fact. And you have to keep my
personal prejudices in mind.
I consider myself a value investor. That means I like lower than
normal prices. Also, I've been investing since the late 1990s.
With the exception of really just a couple years, I've never been
"a bull" on the overall market.
Compared to most folks, I've tended to be bearish. That is a
personal prejudice that must be discounted whenever you hear me
speak. Remember, you need to discount when I make bearish sounds
just as you discount when Professor Siegel makes bullish sounds.
When you hear me make really bullish sounds, that's the time to
listen. Right now, I sound like I always do. I sound a little
annoyed I can't find enough good stocks. I sound a little
resigned to mediocre future returns for most individual
But I felt that way for a few years before 2008. It's not a
So, in addition to hearing my opinion on stocks - that they're
overvalued - you must remember two things:
1. I'm biased towards bearishness
2. Overpriced and about to plummet are not the same thing
I think my opinion is valid. But it's a biased opinion. And even
if it was a biased but correct opinion it would be an opinion
about price levels, not about whether stocks are about to go
I try to remember when I'm biased. And if I don't entirely trust
my bearishness - you shouldn't either. That said, I am bearish.
Ask Geoff a Question
About GuruFocus: GuruFocus.com tracks the stocks picks and
portfolio holdings of the world's best investors. This value
investing site offers stock screeners and valuation tools. And
publishes daily articles tracking the latest moves of the world's
best investors. GuruFocus also provides promising stock ideas in
3 monthly newsletters sent to