Last year at this time I posted a preview of 1997 that contained 10 predictions about what 1997 would mean to WIND. It also contained the controversial reference to a 40% increase in the stock price by the end of 1997. Let's see how these predictions and WIND faired during the year.
My predictions, in Italics, and evaluations:
1. I2O should take off in 1997, adding noticeably to run-time revenues.
Certainly I2O took off in 1997, with virtually every significant vendor of servers and peripheral manufacturers announcing products containing I2O. I even think I am correct in believing I2O will add significantly to the 1997 run-time license fees-if you allow me to invoke justifiable accrual accounting procedures and include the I2O chip production for the months of October through January. WIND's reported numbers for the 4th quarter may not include any of this production due to the generous time Intel is granted to report official production figures. Had I just said "revenues" and not "run-time revenues", then again I would claim I was accurate because of the many Tornado for I2O seats obviously sold during the year.
2. Oracle's 2nd generation NCOS will be important.
I was right in expecting that interesting NC partnerships involving big players would be announced during the year. Indeed, Oracle, Sun and IBM ended the year as the bosom buddies declaring the importance of the NC. I was also correct when I indicated "Early versions of serious NCOS will be targeted to corporate Intranets". However, I believe the date of the actual debut will be in the first part of 1998, rather than late in 1997.
More disappointing than underestimating time-to-production slightly was the lack of specific news regarding expected run-time license fees to WIND. Michael Greene's AEA Conference report indicated that the deal was not a clear-cut as we all presumed, and Dave Lehenky later posted that NCI inherited a VxWorks buyout when it acquired Navio. WIND can't talk about it now, suggesting that negotiations are on going.
3. An endless number of (non Java Virtual Machine) Internet Appliances, partnerships will appear.
I don't have any numbers to prove this, but I believe 1997 was the year that Internet Appliances of all sorts showed up on developers' radar screens, with many such devices making it to market by the end of the year. By the time this prediction was pinned, I had evolved a concept that ubiquitous computing needed ubiquitous communications to ignite the orders of magnitude explosion in types and numbers of useful hidden computers. In trying to articulate the confluence of forces that would bring this revolution about, from time to time I emphasized wireless communications as the best means to connect devices to the internet, and the internet because it provides a zero-cost connection between any two locally-connected devices on earth. In trying to put meat on the bones, I frequently mentioned QualComm because that company displays an understanding of the staggering implications of combining ubiquitous communications and ubiquitous computing, and sits in the epicenter of a significant portion of these momentous forces with its CDMA franchise. My job was made harder because examples and definitions of what I was talking were virtually non-existent.
That all changed on June 23, 1997 when WIND announced a complete Embedded Internet Package. We learned that the Embedded Internet Device (EID) embodied the exact qualities I so lamely struggled to articulate. Tornado for EID, complete with bifurcated applications (both client and server), was fielded to address the huge opportunity for developing Internet Appliances of all shapes and forms.
4. Digital TV/Web access will begin to emerge in large numbers, mainly through cable modems
Thank Bill Gates for making this prediction come true. He revived cable TV companies, causing them to seriously exploit cable modems. He bought WebTV, giving impetus to the digital TV/set-top box sector.
I also slipped in a prediction that DVD, not VOD, would be the multimedia standout in 1997. Actually, I believe I was right. At around 750,000 units, DVD took off faster in its first year than practically any other consumer appliance, like the VCR, microwave, telephone or color TV. Nevertheless, when compared to what the electronics product industry wanted, DVD continues to disappoint. On the other hand, VOD is perpetually on the horizon-just out of reach.
5. Wireless utilization of RTOS' will skyrocket in 1997
This prediction was yet another rehash of the ubiquitous communications/ubiquitous-computing concept. I think it happened, but I'm not sure "skyrocket" is the best choice of words to describe the pace at which these devices are being developed and produced.
As a matter of fact, most of my 1997 predictions can be characterized as being aggressive in timing. (In general, timing is one of the biggest problems for most investors and managers alike, and is certainly one of mine.).
6. Enormous implications, but not yet revenues, of the automobile sector
I got the "not yet revenues" part right, but there is no evidence that WIND was successful in 1997 advancing their cause with the world's automobile companies. Lack of evidence does not mean progress was not made, it means we just don't know. As a result, for now I will take a hit on this prediction, but I may yet be proved right.
7. The military sector will continue its exponential expansion of embedded systems
Right on! The sector was mentioned as being surprisingly strong during conference calls last year.
8. Telecomm/Datacom and Office Automation will continue to provide plenty of revenues of and increasing opportunities for WIND, with an extra spark provided by I2O.
Yes. These sectors provided WIND with much of its growth, despite pricing pressures most network equipment companies experienced during the year. This is confirming of the notion that pricing pressures that lead to increased unit sales are good for WIND under its business model.
9. Medical Systems applications should continue to expand geometrically
Correct. The medical sector was mentioned during conference calls as showing strong growth during 1997.
10. Voice Recognition will continue to be subsumed within telephony in 1997. However, in a few years, VR will become a major driver of embedded systems in most areas of application.
Correct. VR is still a gleam in the eye for embedded systems developers. Talk about timing, DVD has exploded out of the starting gates when compared VR. Look at the balance sheet of any VR company and you will be stunned at the amount of retained losses, no doubt wiping out an army of early investors. Nevertheless, soon VR will be an extremely important complement to hidden computing.
Meanwhile, WIND announced some major developments that were expectedly unexpected. The Mars landing was expected, but not the intensity of world interest in the little WIND-guided rover. This interest provided WIND a platform to begin to establish name recognition as well as to send a message to the entire aerospace/military complex that VxWorks is a faster, cheaper and better solution than any alternative.
The Embedded Internet Package, providing support for the development of EID's, was an extremely pleasant surprise, after having lead ourselves by the nose to the importance of these devices. Previous announcements about partnership developments of a Flash Memory Manager for Embedded Systems with Intel, and a Flash File system for Embedded Systems with M-Systems, blend well with EID's, making them all the more important.
Tornado for DSP was a complete surprise, and has revenue importance downstream. In the meantime, the DSP capability, along with growing pervasiveness of I2O, help establish and maintain WIND's RTOS and tool set in the leadership role.
The development to enable Tornado users to add HP's JetSend to embedded products came out of left field, and should add significantly to WIND's office and consumer products design wins. These are the kinds of wins that keep revenues and earnings growing over 45% per year.
But of all the announcements, all the design wins, the things about 1997 that really stand out are the quality relations WIND formed or strengthened with other companies. The fact that WIND's spring 1998 Developer's Conference is being sponsored by Intel and Motorola speaks for itself. Intel and Motorola wield tremendous weight in the embedded microprocessor market, so their implied endorsement of WIND products and services (by the co-sponsorship as well as I2O and Flash Memory Manager with Intel, and DSP, king of the VME Bus, and dominate microprocessor RTOS vendor for Motorola) should be immensely comforting to investors. WIND's ever-broadening relation with HP, the most innovative developer of commodity embedded products in the world, by far, is heart-warming to the investor. Finally, formal strategic relationships announced in 1997 with companies like Adobe, EFII and Nortel are satisfying and confirming. And speaking of relationships, I have the sense that WIND ended 1997 working with every Networking company in the business, making VxWorks the de facto standard of telephony and data communications.
All in all, 1997 was an excellent year for WIND. Indeed, I believe my optimism at the start of year was exceeded by WIND's exceptional performance during the year. By any measure (EPS, Equity per Share particularly after conversion of the $140 million worth of bonds, pricing power, operating margins, market position, expectations for growth, outlook for I2O, outlook for military uses, EID wonders, notable event being the first OS on another planet) the value of the company is far greater today than it was one year ago.
And that brings us to WIND's 1997 stock price action. I always said the stock price has to be volatile, and it was in 1997, if only to prove my point. Nevertheless, the stock met and even exceeded the 40% objective I wanted to see for the year, before ending the year up 25%. Its tempting to accept the financial market's reduced multiple, applied not only to WIND but most technical stocks traumatized by the Asian turmoil, as natural and perhaps even deserved, but that would be wrong. Irrespective of money flows into or out of U.S. equities, small stocks, bonds, and irrespective of the possible extent of the Asian contagion, if (1) the visibility of WIND's substantial future earnings is brighter and clearer than before, and (2) money is less dear, at every future time frame, then the financial market MUST give WIND a higher multiple than before, limited only by market inefficiencies. The facts are that the market multiple at the end of 1996 was about 94 versus about 68 at the end of 1997. Explain.
It is natural to view the equity market as having recently shifted downward relative to bonds and money flows; less demand for shares means lower stock prices, which translates to lower multiples-no doubt a consequence of the Asian turmoil.
The equity market doesn't work this way, and to think it does will mislead the investor about mid-term prices. Demand and supply for financial instruments do underlie the stock market, but the ebb and flow of money is chaotic and unnecessary to try to figure out. Rather, a sufficient measure of demand for equities is embodied in the ever-changing yields for risk-free bonds. If bonds are down across all maturities, then any financial instrument that promises a given future stream of dollar-based assets will be priced higher by the market than under the previous regime of bond yields-irrespective of past and current money flows, concerns about Asia, recession, deflation, or anything else. Why? Because, the market has only one job, which is does as best it can. The market is trying to level all playing fields, spreading returns as evenly as possible, trying perpetually to preclude arbitrageurs from making excessive profits. Admittedly, the financial market cannot perform its duty perfectly, so it leaves little pockets of extra profits and losses, called inefficiencies. Most market inefficiencies are corrected in very short order, days, weeks or a few months at most. Small, poorly followed companies can experience stock price inefficiencies, with stock prices higher or lower than optimal, for months, perhaps even a year or two.
This means that if the two conditions I delineated about WIND's financial setting are true, then we are witnessing market inefficiency. Certainly money is less dear for any timeframe than it was at the end of 1996, so either the market today sees WIND's future earnings with less clarity, or as being less pronounced, now than it did last year, or one or the other multiples was wrong. If the latter, then given the financial market's recent struggles trying to sort out effects of the Asian turmoil, it is probably safe to conclude that the market was more in error at the end of 1997 than in the beginning of the year. If the former, then if WIND continues to execute along the lines of last year in growing revenues and earnings, the market will correct its mistake and end the year with a higher multiple than it started the year-irrespective of what happens in Asia, the U.S. economy, money flows, rush to Blue Chips or bonds. As long as the bond market keeps money less dear and WIND confirms the previously perceived expectations, then WIND's multiple will rise.
Just to be very clear about this. Personally, I am pleased with a 25% return. Do that consistently for a number of years and WIND will make us all rich. Also, I believe a yearly high over 40%, retrenching to 25%, is unbelievably confirming of what I felt would be appropriate for the year. Nevertheless, the market made a mistake, however small, and investors need to understand why, if only not to be blindsided in the future.
Allen |
