I am looking at WSJ numbers and they are (for FY ending May)
1996 $4.2B
1995 $2.97B
1994 $2.0B
1993 $1.5B
1992 $1.2B
MSFT for the same period shows (FY ending June)
1996 $8.7
1995 $5.9
1994 $4.65
1993 $3.75
1992 $2.76
You can see that for the last 5 years, Oracle has had faster growth than MSFT, except the period where you got the high revenue from Windows 95 (and all the other software upgrades to 32-bit applications), which is a one-time burst. The problem for people who are expecting the sales growth of late last year and early this year to continue are too optomistic (and therefore are over-valuing this stock).
Part of this is that Microsoft operates in so many markets. The bulk of their sales are in OS and applications. These are markets that can grow only two ways for Microsoft--growth in computer sales and raising prices. The first is growing at only about 20% world wide and unlikely to do better in the future. The second runs into some monopolistic questions (and potential market loss to companies like Corel). Since these probably account for half or more of MS sales, this serves to lower the overall MS growth rate. Except for driving everyone into NT, Microsoft does not have a good prospect for outgrowing the industry by very much in these areas.
On the other hand, the overall market that Oracle serves is growing at an overall growth rate of 40-50%, and is separating into a number of categories (OLTP, OLAP, Internet support and other object related) that should continue significant growth for the forseeable future. Except for a few defensive measures (like the NC), Oracle is fundamentally in the middle of a market that shows better long term growth than Microsoft should see.
I know that Microsoft is moving into the DB field, but they are still a small factor. Microsoft is significantly behind in a number of categories (partly because they bought an older code base from Sybase). Although they have some good performance numbers on small to medium size machines, they have no presence on large machines and should not have any before 1998 (when we may have cluster support in NT); Oracle and Informix are already shipping engines (with significant installations) that support clusters and MPP (Massive Parallel Processing) computers like the IBM SP/2. The capacity limit for SQL Server is 200G, while competitors like Oracle and Informix have been shipping multi-terabyte engines for some time (and customers are using them). I could go on, but this should give some indication of how much Microsoft has to do to catch up.
I have been speculating that Oracle might catch Microsoft in a 5 year time frame (since I started voicing that a year ago, I have 4 more years). However, over a 10 year time frame, I expect Microsoft to regain the crown because of its OS monopoly. |
