Ramsey:
<<The biggest unknown, in my mind, is this potential
revenue from I2O.
The co has asked the analysts not to include any I2O revenue in their projections
and it appears that they superficially complied. However, I would have to think that
there has to be some "whisper" off the record number that they must be factoring
into the price. Any guess?>>
I2O is indeed the biggest question mark for WIND's earnings, as evidenced by the healthy discussion on the thread. It continues to be a question in my mind, but like Allen, only with regards to the timing and the ultimate potential of the initiative. There are really two issues wrapped up in your question, and I will try to address both of them: 1)What are the analysts saying/including in their numbers and 2) What do we know about I2O and what is the ultimate potential? Here goes.
1) As far as the analysts, to my knowledge, they are still not including any I2O numbers in their official estimates. The analyst projections I have looked at (Wessels Arnold, H&Q, and Deutsche/Morgan Grenfell) are fairly straightforward. They assume a growth rate for embedded systems in general, say 20% for a few years, with a leveling off after that, and assume that both WIND's and INTS's market share steadily increase, with MWAR's market share staying the same. Their models take market share away from the smaller competitors, as well as from the non-commercial sector of RTOS. They also assume steadily increasing gross margins, as royalty flow constitutes a greater percentage of the companies earnings. I believe their models are wrong on a couple of points. First, WIND is clearly taking market share away from everyone, including INTS and MWAR, and thus a true model should show INTS and MWAR's share declining. Second, the underlying growth rate for embedded systems is too low, and should probably be ramped up to 35% (as a conservative estimate).
The problem with all these models is that they project forward using growth rates and market share, rather than trying to quantify each separate business. This is why I2O has not been included, because WIND is telling the analyst's that even they are not comfortable providing an accurate assessment. We can still try however, and this leads to the second question, what is I2O's ultimate potential.
2. It's still quite unclear, but we got a little more information this time around. There are now 34 announced I2O wins, and Ron promised a more thorough breakdown as to which sectors (RAID, networking, etc.) at the next conference call. I will try to get the information sooner. The growth in I2O design wins follows from everything we have seen, regarding both the Intel relationship and the multiple bus solutions. Remember, Intel sends out a free 30 day trial of Tornado for I2O with every i960 processor, so we can expect even more design wins as a result. Each one of these wins will ultimately produce their own royalty stream, but Intel seems to be the big driver of this flow for WIND river initially, and thus I focus on that.
Intel has said they expect around 2 million units shipped this year, with their estimates ramping to 10-15 million C'98. Intel is usually conservative in their estimates, but this is a good place to start. At around $1.50 royalty per unit, this means we can expect $15-$25 million in royalties (almost pure profit) over the next 18 months. This is conservative, because as more related design wins are announced, it should spur even further demand for i960 chips. Evidence is flowing that things are in place, since Micronics has announced I2O server designs for shippment, and the biggies (Dell, Compaq), should be following suit. The big question is just how big I2O will get. i.e., will the technology migrate to the client side any time soon and show really explosive numbers in the amount of chips required. My instinct is yes, but I leave this question to smarter minds than me. Everything seems to be in place for this to happen.
>> Secondly, per the press release, there are 9,000 developers now. What is that
going to translate into in dollar terms?>>
This is really the essence of the WIND story. Each of these announced design wins ultimately produces a royalty stream, which is almost pure profit. The question is, how big? Some like the NASA pathfinder deal are great for publicity (Perhaps WIND should make a big fanfare this July 4th when the spaceship touches down). But despite the sexiness and publicity, these deals ultimately only carry a small royalty. Some like the INTEL deal have the potential in the billions. The breakdown that Ron gave is as follows: 25% telecom, 25% military/aerospace, 20% computer (I2O etc.), 20% industrial design, with the balance consumer appliance. The average amount from these deals has been around $200,000, but I suspect this will change over time as more multi-unit deals become incorporated. I have my own projections for the deals we know about, but some, like the Qualcomm deal, can not be accurately calculated. The way I value WIND is trying to figure out what each of these businesses is worth, and even from what little we know, the number is truly astronomical. Every quarter WIND has a greater lead, broader penetration of the market, an increasing backlog, further acceptance of I2O, etc. Truly an exceptional business model and an exceptional company.
Of course, I await Prof. Benn's thoughts, since he always sheds light on the situation. Hope this is helpful.
Regards,
Jason Cogan |
