Buy on Rumor, Sell on News (or Speculate on Rumor, Analyze on News)

French website Capital.fr reported on July 26 that HP had entered negotiations to acquire Groupe Bull for about $990 million. This report was not confirmed
by either company, and no sources were mentioned, thus we have a classic case of an unconfirmed rumor.

The fun part about rumors is that one is not burdened with any nasty little facts as to motivations, expectations, or details of an agreement to get
in the way of speculations, musing and random combining of bits of information. (Somewhat like starting a jigsaw puzzle when there is no picture
of what it is suppose to look like, and you are not sure if all the pieces are there, or even if the ones you do have are from the same puzzle.)

First a few random pieces about Bull …

Bull’s 2006 revenue was about 1.1 billion euros (about $1.5 billion), The company has about 7,000 employees

47% of revenue came from products, 31% form services, and 22% from maintenance

47% of revenue came from France with another 38% form Western and Eastern Europe.

The public sector accounted for 40% of revenue, followed by finance 17%, and Transportation/Utilities 10%

The company sells its NovaScale clusters into the HPC market, these are Intel Xeon or Itanium 2 based systems, running under Redhat Linux or Microsoft
Windows for low-end Xeon systems.

… and about HP …

HP’s revenue for fiscal 2006 (year ends in October) was about $92 billion The company has about 156,000 employees.

Computer systems accounted for 19% of revenue, and services accounted for 17%

HP earned 40% of its revenue from EMEA

HP has a very strong HPC cluster business selling both Xeon and Itanium based systems, running under Redhat Linux, and Windows among other operating
systems.

The Big Question: What does HP have to gain?

This is a big-fish-swallows-little-fish situation. At the end you have a big fish of about the same size as before that is either contented with
the meal or wishing it had just eaten at home. It does not create a new species of fish.

I believe there are three good reasons to for one company to buy another:

  • Buy vs. build – It may be faster and/or easier to buy a company with interesting technology rather than develop it yourself
  • Buy vs. compete – It may be more effective to buy a competitor than try to share a market with them.
  • Holding company strategy – A company buys and sells other companies like investors buy and sell stock. With perhaps some advantages from transferring
    management skills and economies of scale to the acquisitions.

A purchase of Groupe Bull by HP would appear to fall in the second reason, with perhaps some of the first. HP gets additional market space (i.e. accounts)
in the European market. However, there is some risk that the large government sector would lose value depending on whether the new entity was viewed
as a U.S. or European company.

From the HPC perspective the two companies appear to have compatible product lines, at least at the high level. The good news here is that HP would
be well positioned to replace Bull clusters with similar HP clusters over time. The bad news is that at this point a great number of companies
have similar product lines, so customers could just as easily replace the Bull clusters with non-HP clusters.

Bull has stepped up its efforts in the HPC market for the last three years and can be viewed as the top European provider. An acquisition of Bull could
give HP an edge in a relatively open and sizable geography.

Or not. After all, it’s just an unconfirmed rumor.

Posted in