An In-Depth Look At BEA Systems

  • February 28, 2005
  • Melanie Hollands


BEA Systems
needs to avoid becoming another Sybase


I don’t see much
growth at BEA Systems these days, although BEA has the rosiest picture of most
companies in the J2EE space. BEA is the market leader in terms of deployments,
and the market leader in terms of technical innovation. It has great name
recognition as the industry leader. Its engineers are deeply embedded in the
technical community, and it is common knowledge that BEA’s WebLogic platform is
the easiest product in the space to use and maintain.

 

So long as J2EE
remains a viable technology, then BEA will be right there. J2EE is pretty well
entrenched.

 


Strong in
its J2EE niche, but it’s not enough

 

However, BEA has
gotten squeezed from both sides — from IBM on top and JBoss (an open source
competitor) below. BEA should have joined forces with the BI (business
intelligence) or content management guys; that would have been interesting. It’s
getting a bit late for that now, and I don’t think BEA will do it. JBoss has
become J2EE-certified and should give BEA more competition along with IBM,
Oracle, and Sun Microsystems. In addition, Oracle has very aggressive with its
10g app server, which was rolled out earlier in the year. Furthermore,
management regards BEA’s integration business to be a success, but I don’t see
the metrics to support the verbiage.

 

At a conference
last Summer, BEA noted that the financial industry has adopted J2EE, and the
financial industry doesn’t adopt anything on a whim. The speaker made the
comment that “J2EE is poised to be the COBOL of the future,” which elicited lots
of groans and uncomfortable looks from the audience. But what he meant was that
J2EE isn’t going anywhere and it will continue to grow and be used for new
projects for years to come.

 

Hypothetically,
if Sun (which was discussed in this column

in January) went belly-up and sold off the technology to IBM, with
the community development behind it, it would take a while to crash. Take a look
at what’s killed other technologies: PowerBuilder, Delphi, Pascal — and what
was it? Well, VB (visual basic) pretty much jumped up and crushed these guys. So
why doesn’t Microsoft do it again? J2EE owns the server side of the equation:
all the parts we don’t see — the part that talks with the database; the part
that integrates with the external systems. It’s like Linux: Most folks are using
it and they don’t even know it. Dotnet (.NET) doesn’t own the server side yet.
It’s got a strong Web services story now and it’s only going to get better with
the next release of Dotnet, so we’ll have to see. Java’s got a strong lead, so
it’s Java’s race to lose.

 

The reality is
that most shops with end up with both Dotnet and Java. Dotnet doesn’t run on
UNIX, but a lot of large corporations use UNIX systems for their heavy duty
mission-critical applications. Heck, there’s still C/C++ apps out there, so
there’s room for another development environment.

 


open sourcing more of its products. As
the company finds parts of its application that don’t necessarily give it a
competitive advantage, it gives them to the community. Which is pure genius.
We’ve seen how fast open source software can move to marginalize mature markets:

·        
MYSQL in the DBMS
space;

·        
JBoss / Tomcat in
the J2EE space;

·        
Linux in the OS
space, and

·        
Eclipse in the
IDE space.

 


Competition
from Three Areas

BEA Systems has three main competitors – two in the J2EE space, and one in the
Java environment:

  • WebSphere (which is part of
    IBM);

  • A group of Open Source / low
    lost vendors (Orion, Pramati, JBoss, etc), and

  • Microsoft Dotnet (.NET).

 


color:#222222″> IBM is No. 2 in market share in J2EE app servers, mainly by
existing customers picking up J2EE. In-the-trenches reports are that WebSphere
is difficult to develop with, difficult to maintain, and expensive (it entails
paying big, big money for licensing). But again, these are problems for
ground-level developers who are not the same folks that make the enterprise
purchasing decisions.

 

But this doesn’t
stop CxOs (which is really where the decisions are made) from using WebSphere.
(A CxO is Corporate (anything) Officer: CEO, CIO, CTO, COO, etc. These are the
folks with the purse-string power that make the decisions between buying BEA and
IBM, and have, for the most part, never heard of the low-cost alternatives.) The
adage “Nobody ever got fired for buying IBM” may still be true. IBM is deeply
entrenched in Java and invests a ton of money in basic research for Java
technology.

 

If Sun ever “gave
up” Java in some sense or another, I’d expect IBM to either grab it completely
or at least increase its stake and influence in the technology. Java is a huge
win for IBM, more so than Sun: IBM has many different hardware platforms (PC,
mainframe, minicomputer etc.) and each has its own programming platform. Now,
Java presents a unified development model, and gives IBM huge leverage to
connect all of its hardware platforms to all of its existing products (DB2,
MQSeries, WebSphere) without having to write platform specific code all over the
place.

 

But I don’t think
Sun would ever entirely give up on Java; it is likely to always retain ownership
of its core. Sun has seen what happens to languages and technologies that are
turned over wholesale to standards bodies (C++, CORBA, and HTML to name a few).
The end-game is innovation at a glacial pace (everything has to start in a
committee, a committee has to talk about it for a few years, then argue about it
for a few years, and then talk about how to implement it, and by then we’ve all
since moved on to something else), and esoteric features only useful to
academics (because guess who sits on the standards bodies committees: really
smart folks, aka academics) that vendors don’t want to waste cycle implementing
because their user base isn’t asking for it. This is what happened to CORBA. The
standards body threw in the kitchen sink for CORBA 3.0, and only one vendor
implemented the entire specification. Once spec fragmentation sets in, and the
user community doesn’t complain, the spec is useless.

 


10.0pt;font-family:Arial;color:#222222″> The open source / low cost vendors fill
in a huge J2EE middle market. BEA Systems and IBM are regarded as being
primarily for large mission-critical, big-ticket, enterprise applications.
There’s a huge market for departmental / corporate J2EE app servers that don’t
need the “-ilities” (i.e., scalability, availability) of BEA and IBM. In
addition, BEA and IBM have not figured how to market to, and communicate with,
this bunch of folks. There’s a huge opportunity for BEA and IBM to move in if
they can change their image and price appropriately. It’s relatively easy for an
upscale brand to extend downward, but it’s much harder for a lower-scale brand
(such as JBoss, Pramati, and Orion) to go upscale. Slight caveat: this is my own
personal view and is based on exactly zero professional technology product
marketing experience.

 

In order to enter
the middle market in J2EE, BEA and IBM would have to drastically pare down their
existing offering in terms of licensing and pricing. To play in the middle
market, these companies would need to have:


  • A flat pricing scheme
    of some kind, instead of a “per CPU” license.


  • Flexible licensing    .
    An end-user has to be able to pickup and move the software to another machine
    at will. A key shouldn’t be locked to a physical machine and require the
    end-user to return the existing key and generate a new one.


  • Documentation    .
    Middle market users typically don’t have high expectations of live phone
    support because they aren’t paying for it. They need all the information so
    they can install, configure, diagnose, and fix their own problems. A company
    like BEA can’t just tell someone “Well, if you need help with “X,” you should
    engage our professional services staff.” If these folks had money for their
    professional services staff, they would have bought a full-blown app server.


  • User community    .
    Along with the whole “fix your own problems,” BEA needs to cultivate user
    communities of champions through online collaboration mechanisms (email,
    newsgroups, blogs, etc) to help spread the info about its product in a
    grassroots fashion. Middle market folks tend to be distrustful of vendors but
    fiercely loyal to their own fellow developers.

 


Microsoft Dotnet.
Another threat for BEA Systems, aside from competition from IBM and open source
vendors, is Microsoft’s Dotnet (.NET). In three to four years, it looks like
most IT shops will be about 50-50 Dotnet versus Java. Some folks might say this
is bad for Java, and maybe it is. But it’s good for developers. Since Dotnet has
come on the scene and started to gain traction, it has lit a fire under Sun.
There have been more innovations in the past two versions of Java than in all
previous versions of Java since.

 

But, BEA Systems
(and IBM too, for that matter) is having problems moving out from the core J2EE
app server business into other areas: application integration, portals, etc. And
just like Java itself, they’re having problems attracting new, non-J2EE
developers.

 

Everyone in the
J2EE space is gunning really hard to be the one to which hordes of unwashed VB6
programmers move. But, the VB audience isn’t used to dealing with all of the
low-level technical details that are exposed (and must be dealt with) with J2EE,
so there’s no clear winner yet. There’s a big push on the tools side to make
J2EE point and click / drag and drop easy, but it hasn’t caught on yet.

 

The reality is
that both Java and Dotnet will survive. But for either one of them to have a
future, they will need to interoperate. Whoever causes the problems will be
viewed as the bad guy and will likely be out the door. Heterogeneity is a fact
of life for IT shops everywhere, and developers don’t have the time and
resources to deal with unruly children who don’t play well with others.


 


font-family:Arial”>In his 2004 deposition to the Department of Justice in
connection with Oracle’s offer for PeopleSoft, Ellison mentioned BEA Systems as
a possible takeout by Oracle. Buying PeopleSoft has made some productive use of
Oracle’s huge cash flow; but the company still has considerable cash flow that
could be invested productively. (Although, as
mentioned in this column about Oracle

last week, a better use for Oracle’s cash flow would be to pay a
dividend.)

Oracle is a
company that can do one thing very well and has pretty much sucked at everything
else. Acquiring BEA Systems could happen, although BEA probably isn’t cheap
enough for Oracle. Furthermore, I’m unfamiliar with what sort of relationship
Oracle President Chuck Phillips had with BEA when he was an analyst at Morgan
Stanley (or what he thought of the company). From a technical perspective, BEA
would be a better stack fit for Oracle than PeopleSoft because it’s a platform
and not apps. Buying BEA could integrate everything for Oracle: PeopleSoft was a
horizontal acquisition, whereas buying BEA would be much more about filling
holes. That said, acquiring the expertise on the apps side to write the
next-generation software (via PeopleSoft) is a plus for Oracle. Does a company
like BEA Systems fit that description? I think much less so. So my guess is that
acquiring BEA is less likely.


10.0pt;font-family:Arial”>

Would the “costs”
of no longer being neutral outweigh the benefits of being “more complete”?

Not really. BEA
doesn’t really matter because Oracle has more than enough customers, and the
only other real complete option is IBM, which isn’t neutral, either. But IBM
buying BEA would seem like a waste of money.

 

Fair value for
BEAs looks to be around $8 to $9; so I’d put takeout value in the $9 to $10
range. The stock is currently trading at around $8.25, so maybe a buyer pays a
15 percent premium (and who has to pay a premium these days?). But at current
levels, that’s more potential for upside than when the stock was trading at $13
(when there was more potential for downside). The $9 to $10 range provides some
support for the stock price. And who knows — maybe business rebounds for BEA,
which could then command a higher takeout price. However, opportunities for
growth at the company seem limited these days. So there might be questions about
how much business BEA could bring with it to an acquirer due to concerns over
growth and possible concerns over neutrality.

 


Valuation
and risks

 

According to some
analysts, BEAS stock is cheap. However, I disagree. The growth outlook is
limited and quarterly results last year didn’t propel the stock that much
higher, although in a tough operating climate coupled with a tough market
environment (where expectations in the tech sector continue to be very
negative), at least they didn’t cause a big sell off.

 

I don’t see any
positive fundamental catalysts to propel the stock meaningfully higher although
takeout speculation will be there to an extent. BEA’s only hope is to be bought
out at some point. There are too many big players in the space, and I don’t
think BEA can put together a complete enough offering, so survival as a
standalone looks tough. I think someone will buy BEA, though; it has a great
platform.

 

I’m not crazy
about BEAS technically, either. I’d like to see it pop its head over the 50 DMA
with volume of at least 15 to 20 million. That would cancel out the recent sell
off on February 16. Follow that with a close over $8.55 and that might do the
trick. If that happens, then I could give BEAS a shot at $10; but the reward
doesn’t look like it’s anything beyond that.

 

In terms of
takeout value, perhaps I’m being overly conservative with the $9 to $10 takeout
range. But I don’t think any potential acquirer is going to pay big premiums
right now — there’s no reason. Take Oracle, which

took down its bid for PeopleSoft last
year. That’s a barometer of where things have been, and (despite Oracle’s
subsequently increased offer for PeopleSoft) are likely to remain for a while,
in software (and tech) M&A.

 

Melanie
Hollands


melaniehollands@yahoo.com