 |
In
This Edition... |
 |
-
A Look at the Financial
Sector's Most Meaningful P/E Ratios - Undervalued, or Just Cheap?
-
From the Blog -
A Funny Thing Happened on the Way to Lower Lows, and an Economic Report
Card
 |
 |
What's
a 'Normal Valuation' For a Financial Stock? |
|
 |
If
there's one thing 2009 forced all investors to reconsider, it's
the definition of 'cheap'. What one man has considered 'cheap' this year,
another
man has just considered 'undervalued'. Worse, the fine line between
cheap and undervalued - particularly for financial stocks - has
been sliding back and forth along its scale.... not just this year,
but for several years.
Those dynamic
changes underscore a frustrating problem for investors trying to get a
bead on whether or not a particular stock is worth its price or not. Well,
we've got the beginning of a solution to the problem today.
What's a
fair value for the average financial stock? Great question. Even if
you boiled it down to a simple P/E ratio, the 'ideal' P/E ratio is still
a moving target.
For
instance, investors thought nothing of buying a stock priced at
an earnings multiple in the 20's before late 2007. The economy was
booming, and financial companies (lenders and banks in particular) were
making money hand over fist. Now - in the shadow of the worst market
debacle in decades - the same P/E in the 20's is deemed way too expensive
to own. What changed? Nothing but perception.... but perception
is all that needs to change.
Therefore, our
goal is to find out the most reasonable average P/E, just to have a
benchmark for judging a particular financial stock's price (and any
stock,
for that matter). Maybe a P/E of 20 is perfectly normal. Or, maybe it's
ridiculously high.
So how does
a disciplined and patient investor determine if a stock is overvalued or
undervalued? By taking a step back and looking at even longer-term
averages of key ratios like price/earnings data.... and we do mean
long-term.
Most
fundamental analysis of a stock's valuations compare this year's
results to the prior year's, or compare next year's projected
results to the current year's. The problem is, any one of those
comparison years could be so skewed, that to hold it up as a standard for
other years is practically meaningless. Case in point - most companies
have put up significantly better numbers in 2009 than they did in 2008.
But, a 10% improvement (as an example) in revenue hardly offsets the 50%
decline in revenue suffered last year. It's all relative.
Instead, a better
P/E benchmark can be established over a five year period.
Think about
it. If you look at a company's data going back five years, you've looked
at the company in five different scenarios.....the middle of a bull market,
a top, a bear market, a recession, and a recovery. The 'middle ground'
P/E ratios over this time paint a much more meaningful picture of realistic
stock valuations.
And that's just
what we did.... looked at the average P/E ratios for the major financial
stocks - by industry - for the past five years. Though it's still
not a prefect methodology, these figures give a much more realistic idea
of valuations that are sustainable, and which ones are not.
In the interest
of time and space, we're not going to show you all the math for
this study. We'll just walk you through one set of numbers, and then tell
you the rest.
Let's
start with accident and health insurers. There are only eleven listed stocks
in this category, so it's a fairly easy study to do. All we want to do
is compare operating P/E ratios for this year, last year, and next
year, to the five-year average operating P/E... the 'benchmark'
we're seeking. (FYI, the data came from Reuters.)
Right away we
can see the biggest challenge with this research... skewed and irrelevant
data created by persistent losses taken at some point over the last five
years. Eastern Insurance Holdings' average five-year P/E of 88.9 clearly
isn't normal. The same goes for Triple-S Management. Even Conseco's average
P/E over the last five years doesn't mean much.
Though not scientific,
for the purpose of developing a meaningful benchmark P/E level, those three
numbers have been removed from the calculation of the average long-term
P/E. The result? The accident and health insurers' typical P/E over
the long-term has been 19.07. Some of these stocks are cheaper than that,
and some are more expensive. Some will be cheaper than that in the future,
and some won't. No matter what though, there's now a reasonable 'too high'
level to work with.
Like we said,
it's not a perfect tool, but it's better than the skewed before/after
technique most investors are using now.
Now
that you understand the methodology, nearby you'll find the five-year average
P/E ratio for all the major industries in the financial sector. In some
cases we again had to make manual adjustments to offset skewed or misleading
data. The integrity of the research, however, was not considerably
affected. In fact, the benefit of the research was enhanced by cleaning
up the data and producing meaningful benchmarks.
A couple of
things stick out right away. For starters, savings and loans stocks have
been surprisingly expensive. At the same time, most insurance stocks have
been pretty cheap in recent years. It's not exactly clear what affect the
insurer implosion had in that regard, though it may not matter either -
they were still net profitable during the past five years. Those
aren't the only factoids you can glean from the data though.
The next step
in using this information is to compare your stock picks to the long-term
P/E benchmarks we calculated. Granted it's not day-trading data,
but that's ok - it's great data for long-term investors, as it
lets them really know where a stock stands in terms of long-term norms.
A more complete
study would look at long-term norms for other key ratios as well, in
addition to making technical and cyclical considerations (stocks tend
to exceed normal valuations in the latter portion of bull markets). Even
the brief research example above, however, offers a tremendous edge.
In case you
missed them, here are the latest blog entries....
|