Investment mistakes usually fall into one of three
categories: analysis, behavior or bad luck. In October 2011,
after the shares of Hewlett-Packard Co. had been halved from
about $48 earlier that year, I made a case for the stock. That
was a mistake. There was no bad luck. I made several errors in
my analysis. In this column I want to drill down into my
mistakes and provide a new analysis of what is still an
In 2011 I got three things wrong about HP: printers,
services and culture. To better understand the company,
its helpful to use an analytical framework based on two
companies in different time periods: computer maker IBM Corp.
circa 1993 and film giant Eastman Kodak Co. since 2006.
Kodak was responsible for pioneering work in digital
photography as early as the 1970s. In the 90s, when
digital photography was introduced commercially, Kodaks
35mm film sales at first continued to grow, as digital cameras
were an expensive novelty. But as digital cameras got better
and cheaper, and thus more popular, sales of 35mm film started
to decline. If you were a value investor analyzing Kodak, the
stock would have appeared cheap on past earnings. And if you
assumed that Kodaks cash flows would gradually decline
years into the future, youd have been dead wrong. Kodak
turned into the value trap of all value traps. Once digital
cameras went mainstream, Kodaks sales went off a steep
cliff, falling from $13 billion in 2006 to $5 billion
today. Cameras are replaced every few years, and the cost
savings from not buying expensive film any longer were
substantial. Also, the new digital cameras were simpler, and
the learning curve was not steep.
I challenge you to find a single roll of 35mm film at your
local grocery or drugstore. You might look on the same shelf
that has the eight-track tapes and vinyl records. There is
absolutely no reason for 35mm film to exist in the digital
If the HP of today is Kodak in 2006, then value investors
like me who dare to discount its future cash flows (even if
they are declining) are fools. But Ill argue that for the
most part, with the exception of a few of its businesses, HP is
IBM in 1993, not Kodak in 2006.
In 1993, IBM was in trouble. Its stock was trading in the
low teens, down from the high of nearly $44 hit in 1987. Its
bread-and-butter high-margin mainframes were being attacked by
cheaper Japanese competitors Hitachi and Fujitsu. Mainframe
sales were down almost 30 percent in 1993. Though mainframe
sales were only about 20 percent of total sales, they were the
core of IBMs business; they drove revenues for
maintenance, software and services. If mainframe sales
disappeared, so would the bulk of IBMs other revenue.
Cheaper and better products from direct competitors were
only part of the problem. The computing landscape was
drastically changing, as PCs powered by Wintel (Microsoft
Corp.s Windows software and Intel Corp.s chips)
threatened mainframe dominance. Suddenly, a much smaller and
cheaper box that did not require a forklift to install and a
clean room to live in was replacing a lot of mainframe
functionality. IBM came out with what many believed to be a
superior product to Wintel PCs the PowerPC, with
its own processor and operating system, OS/2 but the
company was still losing the PC war.
The media were very negative on IBM, and rightly so. A
Fortune article from the period had this to say:
IBMs difficulties include overdependence on
high-margin mainframes, when computing power has become a
desktop commodity; . . . emphasis on hardware, when software
and services are ever more important; and an inability to get
quickly to market with the new machines that periodically
remake the industry.
Ironically, a Fortune piece in May 1994 described how HP was
positively thriving at the expense of IBM:
Over the past decade, HP has quadrupled in size, created
$10 billion in shareholder value, and transformed itself
from an also-ran No. 7 among U.S. computer makers to a strong
and profitable No. 2. It eclipsed DEC [Digital Equipment Corp.]
This is my favorite part: No one thinks HP will
succumb to the malaise that got its rivals. While IBM has long
stood for mainframe computers and DEC for minis, the products
with which HP is synonymous pocket calculators and
computer printers are in little danger of becoming
As the media were writing IBMs obituary, IBM stock
must have been a dream for short-sellers, especially if they
shorted it in the $20s, $30s or $40s. Considering the articles
mentioned above, I can imagine how easy it was to make the case
that IBM, which at the time had more than $20 billion of
net debt and 300,000 people working for it, would go to
But in 1993, IBM did the unthinkable: It hired a former
McKinsey & Co. consultant whose last gig had been running a
food and tobacco company, RJR Nabisco, who knew nothing about
technology. One of the first and most important decisions Louis
Gerstner had to make was whether to follow through with a
breakup of the company or keep it together. Previous management
had already set a plan in motion to break IBM into several
pieces, and consultants were already working on naming the
divisions to be spun off.
If this story sounds eerily familiar, it should: HP
yes, the company that was in little danger of becoming
obsolete is going through similar pains today.
Not unlike Gerstner, HP CEO Meg Whitman inherited a
significant decision about whether to break up her
companys largest division, PCs, or keep it. Wall Street
loves to hate ex-consultants (they probably hate them in
advance, as consultants are bound at some point to turn into
politicians). Just as Gerstner was ex-McKinsey, Whitman was
ex-Bain & Co.. Predictably, the pundits criticized her for
being an ex-consultant and lacking relevant experience.
This brings me to the first mistake I made: culture.
Similarly to HP today, in the early 90s, IBM was riding
the coattails of its past success and saddled with a
dysfunctional corporate culture in which the individual
divisions were engaged in constant turf wars with one another.
IBM was slow and bureaucratic. Its human resources systems were
antiquated and inflexible if employees wanted to
transfer from one division to another, they had to be fired and
In the case of HP, I underestimated how much its culture had
been gutted by its past few CEOs. I apologize in advance for
using this analogy, but HP reminds me of a talented kid whose
parents passed away, who then went through several abusive
foster parents. The good news is, the kid has finally found a
great and caring guardian, but the question is, how irreparable
is the damage done by the previous foster parents?
HP employees have been demoralized by cost cuts and
acquisition integration. Endless layoffs havent helped
morale. But HP was inefficient; it had too many people working
for it. The layoffs will eventually end. IBM went through
layoffs too: By 1994, Gerstner had reduced IBMs workforce
to 219,000. As I have written before, HP needs a good parent,
not a visionary. Whitman is just that. Like Gerstner, she
is trying to gradually change the corporate culture. She moved
all executives from corner offices into cubicles; she sits in a
Whitman took a page out of Steve Jobss playbook and is
trying to refocus the company on fewer products. For instance,
HP made a mind-boggling 2,100 types of laser printer in 2012.
Whitman will cut that number in half by the end of this year
and probably even more in the future. HP needs a lot more
(A side note: When I wrote my first article on HP in October
2011, I tried to buy a new PC from the companys website.
The website was a mess, offering too many choices, and after 20
minutes I gave up and bought a PC from Dell. Recently, I was
looking for a computer for my father, and HPs website had
been simplified and become easy to use.)
Historically, HPs PC and printer businesses had
different sales forces despite sharing the same customer base;
Whitman is combining them. The company also favored investing
in external research and development through expensive
acquisitions that destroyed an enormous amount of value.
Whitman is reversing this. She is putting more money into
R&D and has largely ruled out acquisitions for the
It is clear now that the company that was the founding
father of Silicon Valley was turned by its three previous CEOs
into a sorry empire of dozens of misguided and misintegrated
acquisitions, which could barely function as one unit. Being
large has its advantages, and one of them is buying power, but
HP did not capitalize on this. For example, there are 1,500 HP
decision makers around the world buying media. Whitman is
fixing that too.
My second analytical mistake was to underestimate the
problems in HPs printer business, which will
significantly decline over the next five years. The bulk of the
decline will not occur in business printers and supplies (laser
printer toner) but in consumer printers and, more important,
consumer ink. Business demand for printing has not dramatically
changed over the past five years. You see it in Xerox
Corp.s numbers: Black-and-white is declining a few
percentage points a year, while color is offsetting it with
mid-single-digit growth. Computer and tablet screens strain our
eyes, so printing and reading on paper, though archaic, is a
habit that will take a long time to die. Also, when we print at
work, our employer is paying for the ink and paper, and because
we are spending someone elses money, there is little
motivation to change our antiquated and costly habit.
The consumer printer business will die at a much faster
rate. Consumers have used ink-jet printers for kids
homework and photos. Though kids still print homework (at least
for now), with tablets, cell phones and Facebook, photo
printing is on a dramatic decline.
HPs consumer printer business model was simple: The
company sold ink-jet printers at a very small profit (or maybe
a loss) but made a killing by charging high prices for ink. It
was a great business, but the previous management milked the
ink cow too hard they raised prices for ink to the point
of destroying already-fickle demand. The current management has
been slowly rectifying this in emerging markets, where HP has
raised prices for ink-jet printers but lowered prices for ink
and toner, but it cannot reverse the trend. HP doesnt
disclose it, but by my estimate, ink for consumer printers is
two thirds of HPs printer-supply business, and it is on
the way to meeting its maker, fast.
HPs consumer printer business looks just like
Kodaks 35mm business, and its rate of decline is likely
to accelerate significantly in coming years. In my latest
analysis I shrink consumer printing profitability by two thirds
in three years but more on that later.
My third mistake had to do with services. The bulk of the
service business is Electronic Data Systems, which HP acquired
under Mark Hurds leadership for $12 billion in 2008.
I did not realize how grossly mismanaged HPs service
business was. How grossly? In 2012, HP wrote down
$8 billion of the EDS purchase price as its margins
collapsed to almost nothing.
Though on the surface it seemed like EDS was a good fit, and
it still may be, HP, as a company that invents and makes stuff,
simply did not know how to run a more mundane service business.
The service business in general doesnt face any
significant secular headwinds Accenture and Affiliated
Computer Services (owned by Xerox) are prospering but to
make the numbers and please Wall Street, HP starved this
business for investment and seriously mismanaged it. The
margins in the service business are unlikely to recover to the
artificially high levels under Hurd, but considering that they
are at 1 percent today, the bar for success has been set very
I do not put the acquisition of Autonomy or the
deterioration of the PC business on my mistakes list.
Considering that HP stock dropped more than 20 percent on the
announcement of the disastrous $11 billion acquisition, I
assumed that the market had already priced in a complete
write-down of Autonomy.
When you read the headlines, it sounds as if HP is dangling
from a thread, just a few months away from bankruptcy. It
isnt. During its horrible 2012, when sales
were down 6 percent, HP generated more than $5 billion of
free cash flow and its net debt fell to $5 billion
($10 billion if you count its underfunded pension
In my original article about HP, I thought it had earnings
power of more than $5 a share. I was wrong. Though the services
business will probably recover in a year or two, albeit to
lower margins than I expected, the decline in the printer
business will only accelerate. My base-case HP earnings power
is somewhere between $3.50 and $4 a share.
However, more important, even after I tried to smother each
of HPs businesses with a pillow I assumed that
profitability and revenue growth of each business would
decline, profitability of the printer business would drop by 80
percent, and the PC business would earn only half of what it
earned in 2012 I still got earnings power of $2.50 or
Wall Street has a very short memory, and at some point HP
will lose its stigma (just like IBM did). If the market grants
it a price-earnings ratio of 12 not a shabby number
then, based on my worst-case earnings, HP will trade at
$30, and by my base-case earnings, the stock will double from
its current level to $40 to $50 a share.
I can only imagine how toxic IBMs stock was in 1993.
You had to be able to admit to your clients that you owned a
has-been that had lost its way, that had just taken
$3.4 billion in charges (that was real money then).
IBMs stock fluctuated, stagnated and then skyrocketed.
IBMs mainframe business, though challenged by PCs, is
still around today. HP will likely be no different it
will be around for a long time. Will its sales decline further
next year? Possibly. Will they fall off a cliff, like 35mm
Vitaliy Katsenelson (firstname.lastname@example.org) is CIO at Investment
Management Associates in Denver and author ofThe Little
Book of Sideways Markets.
See related story:
How HP Can Navigate the Information Superhighway