Red Hat’s financial future: No guesses about what’s next

You probably noticed that my favorite charity
case actually made a profit this quarter. As
Tina Gasperson said on these pages, Red Hat proudly announced an “operating profit” of $600,000.

At the same time Red Hat also got around to writing down some of its bloated “blue sky” valuations giving the company an effective net loss of $27 million. Still a cash flow profit is a real profit. Let’s give the guys credit. So why did many leading analysts
spend most of Wednesday slashing the stock and the company? The reason is
simple. Even Red Hat’s management isn’t sure where things are going next. I
think they deserve credit for admitting the fact.

You’d think that after announcing a profit, the team at Red Hat would
get a little credit. The $600,000 they made is several million more
than VA Linux made this quarter. Its nearly a billion dollars better than
Lucent’s bottom line. Any tech company that could make money in the last quarter
is manned by geniuses, in my view.

But my view isn’t widely shared. The usual suspects spent the day
dissing the company and its prospects. Hambrecht & Quist and Goldman
Sachs analysts were particularly negative. They downgraded the
stock and their comments were unusually harsh. For one thing, the company’s
sales had dropped. What money the firm made came out of investments and
through slashing costs. That’s not how “growth” companies are supposed to
behave.

More troubling was the withdrawal of what the analysts call
“guidance.” This is the whispery stuff that companies feed analysts, to tell them
not only what they “think” the company will make in revenue, but
where the company will make it and why. In effect, it amounts to
tipping the business plan a little to give the analyst community some basis to
determine the value of the firm going forward.

This time, Red Hat’s Kevin Thompson wouldn’t play. This “zero
visibility” upset Goldman Sachs and Hambrecht’s Prakesh
Patel. Patel said, ” We believe [this] signals significant
uncertainty in Red Hat’s ability to capture revenue growth or deliver continued
profitability. We have significantly lowered our estimates.” Patel downgraded the
stock to “neutral,” which in analyst speak is pretty darn negative, and
suggested that the company’s current stock price was selling at a “premium” (i.e.
overvalued by) 48%. Ouch.

Other First Call analysts also began to backtrack on
their previously favorable predictions about the company’s long -term prospects.

Although these kinds of comments make it very hard for institutional
investors to hold stock or for Red Hat to go out into the capital
markets any time soon, I don’t blame Thompson one bit for holding back. To be
absolutely fair, Red Hat faces a set of variables that would tax the
prognostication skills of Nostradamus.

Here are the issues:

• The big account mountain: If Red Hat’s experience over the last
  year or so has taught the company anything, it is that whatever money there
  is to be made in Open Source involves constructing and supporting
  enterprise-level solutions for big companies. These companies not only require code,
  they expect extensive customization, and lots of assistance with
  planning and deployment. And they pay for it.
  I’ve said this before, IT services are a real business — in many
  ways they are the only viable Linux business.

But an approach tied to big businesses comes with a big cost. Large
companies move slowly, and by committee. Budget committees are often
bound by annual budget cycles. Even after an item has been budgeted, actual
deployments can take up a year to plan and another year to roll-out across the
enterprise. I’m currently involved with a case study concerning the adoption and
deployment of a new OS by a major accounting firm. The process has taken three
years! In the meantime vendor sales personnel and developers still
need to get paid. This adds up to a major cash flow challenge that has
wrecked more than a few promising companies.

Timelines aren’t the only challenge; as IBM has demonstrated, and as
Microsoft has learned, there’s a lot more to the enterprise market than
building and deploying software. The market requires vendor to provide
expertise that extends far beyond just developing good code. Global
enterprises expect independent software vendors to provide assistance with R.O.I. calculations, with the selection of local development partners, with training, and with deployment planning, all frequently as part of the sales process (i.e. vendor underwrites
the cost). Red Hat has
a decent basic enterprise package but the bulk of its products
are really structured to smaller businesses rather than large enterprise
accounts, so is their sales approach.

So — who knows how much Red Hat will have to wait, how much it will
have to spend and what enterprise deals it can land in the next 12
months? I don’t blame Red Hat management for declining to give an estimate.

• The inscrutable telecoms market: I’ve
  said it. Intel’s Craig Barrett said it. So have lots of others — there’s hardly a more
  natural market for Linux and Open Source than the telecoms marketplace. To a
  very large extent, telecom hardware resembles nothing more than old-style
  mainframe systems. It’s a dead approach that’s ripe for change. Everything about
  Linux– its cost, its Unix ancestry and above all else, its open nature
  makes it an ideal platform for changing the telecommunications
  marketplace.

Red Hat has done a better job than anyone else of partnering with
telecom hardware developers. The company has developed important relationships
with Lucent, Cisco and Nortel. But much of the value of Linux is that
enables communication service providers to dump the old-style telecoms hardware
built by the guys that Red Hat has partnered with.

More importantly, this market is in a complete turmoil, brought about in some small part by the Linux revolution. Linux hasn’t fueled orders for new systems as much as it has given corporate hardware purchasers a yet another reason to sit back and wait a year. The result has been a disaster. Nortel
is cutting thousands of jobs. Lucent employees are abandoning ship as the company’s bonds are downgraded to junk. It’s quite possible that one or more telecom giants could disappear in the next 18 months. That alone makes it hard to know the future of this segment for Red Hat.

• The database challenge: That Red Hat plans to get serious about
  the enterprise database market should come as no surprise. Enterprise
  accounts, in particular, are wedded to their database management systems. Red Hat
  has to provide a credible solution to win credibility as an enterprise
  platform. The investment community loves the idea of a Red Hat based
  database solution, but the relationships with IBM and Oracle could be
  treacherous. A wrong step and the company could wind up as little more than a
  temporary junior partner. How this plays out over the next year is anyone’s
  guess.
• Redmond: It matters. Does Microsoft emerge from antitrust
  litigation with little more than a wrist slap? Do the Europeans take it on
  themselves to punish the company? Will Win2000 become a hit in the enterprise
  marketplace? Does Microsoft’s massive investments in consulting expertise pay off in
  new deals? Whatever one thinks of Redmond, it still defines the
  marketplace, at least for the next year. What is not known is whether its
  initiatives in embedded systems, home computing or distributed computing will
  succeed or fail. That makes the next 12 months hard to predict for anyone.
• Oh, and the economy: The Y2K crisis was real. It just took 18
  months longer to wreck the IT community than anyone expected. But
  where are we in the business cycle? Under normal circumstances demand
  should be starting to build up, but it’s hard to see what the “next big thing”
  is going to be. Memory chips are a glut on the market. Server sales are
  way down. Corporate pocketbooks have been slammed shut for a year. So
  who knows what the demand for new equipment from IBM or Intel is going to
  look like? Who can tell what demand for Open Source IT products and services
  will be by this time next year?

Red Hat doesn’t know, and it’s not going to feed the analyst community
a lot of speculative numbers until it does.

Good for them for being honest about it.