There have been a series of eerie misstatements and misrepresentations lately that I believe account somewhat for WIND's recently leaking stock price, adding to the momentum pushing up INTS stock price. First, there was the misstatement by the Investors Business Daily (previously reported on the thead). It is understandable why the misstatement occurred, but I don't understand why it wasn't corrected in Tuesday's edition?
Second, the H&Q spot report following the earnings report was extremely upbeat with a, read between the lines, suggestion that INTS may be better positioned in the long term with its pRISM+ product than WIND is with Tornado. Consider the following lifted verbatim from the report:
----------------------------------------------------------------------------------------------------
"Will the real CORBA please step forward? Earlier this week Wind River announced that a third-party tool vendor, Highlander Communications, would be porting Visigenic's CORBA-based Visibroker technology to the Wind Tornado development environment. While this move in part resembles the approach ISI has taken with pRISM+, there are substantial differences and we believe this collaboration actual serves as a further endorsement of ISI's decision to build pRISM+ around a CORBA backplane. Essentially CORBA provides customers with a scaleable, platform independent solution that permits the exchange of objects across applications and networks. Since software developers are rarely centralized in any one location nor standardized on any one platform, the ability to communicate in a distributed environment is increasingly critical as complex projects become more difficult to monitor and control. pRISM+ currently provides this capability to developers as will Tornado when the new CORBA tool is released sometime in the December quarter."
"An important distinction however is the fact that the pRISM+ architecture is actually based on a CORBA backplane or foundation, enabling tools from a variety of vendors to openly exchange information between one another. This fully modular approach allows customers to pick and choose which component tools they require either from ISI or other third-party vendors. The interooperability of the tools is an important advance as it speeds the iterative development process, which in turn, reduces the length of time typically required for application development. Wind River's approach, while widely supported by over 100 different tool vendors, is based on a proprietary API or Application Programmers [sic] Interface, which simply means that each vendor must customize their tools for use with the Tornado environment. We do not believe the near term implications of Wind's open, but proprietary, approach are negative, but believe that over time the third-party developers will prefer to write once to the CORBA-standard rather than to several proprietary APIs."
--------------------------------------------------------------------------------------------------------
This statement implies that WIND has a proprietary, but open, tool interface (which is true), but INTS has a CORBA, open and standard, tool interface (which is not true). pRISM+ still needs a proprietary API, just likeTornado. The only difference is that INTS is attempting to use CORBA as the middleware for interconnecting tools through its proprietary API, while WIND uses a more-efficient and always available means called RPC (which is a open standard that sits on top of TCP/IP).
To be perfectly clear about this, there is no such thing as a CORBA-standard for third-party developers to write to. They have to write to some proprietary standard for communicating messages among various tools, whatever transport system is used to ferry the messages. CORBA enables communication between distributed objects, but the objects still have to agree about what needs to be communicated, and that's essentially an API. Assuming that all API's get the job done reasonably and efficiently, third party developers would hardly favor INTS API over WIND's, or anyone elses. Instead third party developers are driven entirely by greed. They only want maximum market access, and will do anything to make their tools work in the dominant market.
CORBA is one of many standards competing for distributing objects, and it cannot be assumed to be available on even the majority of networks. For example, MSFT has a competing approach DCOM, and we all know it is foolish to ignore MSFT, especially when the host development platform showing the greatest growth in embedded systems is Windows 95/NT. Because CORBA is complicated and extremely heavy, it has never received much attention from embedded systems developers, which tend toward small and efficient everything. While some development shops may prefer CORBA, possibly because their networks are maintained by corporate IT shops running database applications on Unix, I believe it would be unnecessarily limiting to restrict the market for development environments to those using, or willing to use, CORBA. For this reason, I suspect that CORBA may prove to be a liability inhibiting INTS sales rather than an impetus expanding them - even if the company can overcome inherent technical problems their customers may face trying to work with so much weaponry for so little gain. For example, what is an INTS salesman supposed to say to a Windows NT DCOM shop? Providing CORBA support when wanted is a plus; requiring CORBA it is a serious negative.
The H&Q analyst noted that revenues were healthy across all product groups, including 17% year-on-year growth for MATRIXx and 24% growth for embedded software revenues. But the latter must include services, because we know product license revenues, including royalties, were flat year-on-year. Since MATRIXx increased 17%, this means that embedded systems products declined when compared with last year. The fact is that INTS is still losing product revenue in its vital embedded systems area, which is ignored by the analyst. pRISM+ must lead quickly to increased sales if INTS is going to remain competitive with WIND, and if it is not to follow in the footsteps of Ready Systems, Microtek Research and possibly MWAR. And there is still the question of why royalties haven't picked up enough over the last year to compensate somewhat for slowing product license sales, since mature design wins should have pumped up run-time license fees.
The bottom line is this. How much more is it worth to the investor to know with near certainty that a company will produce EPS at least $1 to $1.10 in a year and a half versus knowing that the company will produce an estimated $.88, if things go well. The answer is: if you had the same confidence that INTS will produce $.88 as WIND producing $1.1, then roughly speaking INTS per share market value should be something less than 80% of WINDs, or $42 x .8 = $33.6. It deserves to be less because WIND has higher expected out-year earnings growth. However, if INTS fails to get embedded systems product revenues growing again, it will fall back to a much lower value and stagnate. Presumably, this depends on whether or not the CORBA play will pay off in the hard-nosed embedded systems market place?
As I said in a previous post, let the games begin.
Allen |
