Everything in the Q2 report dealing with design wins, product license fees and royalties surprised me. How can the number of design wins decrease significantly on a sequential basis yet product license sales jump an unbelievable 16% over Q1? Without any doubt this gives credence to a remark Khan made a few months past about the meaninglessness of a total count of design wins. The answer given by management is that a slowdown in pSOS design wins more than compensated for increases in VxWorks design wins. Similarly, VxWorks royalties continued to increase, yet a slowdown in pSOS royalties lead to a sequential reduction in overall royalties. Finally, it was pointed out by Jerry that product license fees are a precursor to follow-on royalties, suggesting that royalties are soon to pick up again. What’s going on?
First, it is absolutely clear that pSOS rapidly is becoming a niche product, which is great. ISI was not acquired because WIND needed another RTOS. I suspect the 1000 design wins mentioned in Q4 (pre-merger) reflected a huge wave of customers migrating from pSOS to VxWorks. While the migration is essentially zero-sum for the combined company, it probably created significant extra expenses for customers, a portion of which must have accrued to WIND.
But math is math, and there can be no evading the obvious conclusion that the only possible way for product license revenues to jump hugely with significantly fewer design wins is for the average selling price (ASP) to have increased sharply. A shift in mix from pSOS to VxWorks will contribute to such an increase, but I suspect not nearly enough. Something else caused the ASP to increase, and there is only one thing that could be. WIND’s year-old thrust in vertical solutions is paying off, and paying off big. The 20 optical switch design wins for the quarter probably entailed ASPs five to 10 times ordinary ASPs. The 74 Java Internet Appliance design wins (even without the full TIA rollout) must have involved considerably higher ASPs than ordinary. This is very good news. It also supports Jerry’s claim that higher royalties should follow, because vertical royalty ASPs are commensurately higher than non-vertical royalty ASPs.
But there is more to the story. A critical aspect of the health of WIND’s business model is its ability to maintain gross margins on product sales. Gross margin for the quarter increased slightly from 75.8% to 76.2%, which appears to be on the positive side of insignificant. That appearance is wrong. Q2 displayed a recovery in pricing power that approximates what it regularly accomplished before the merger. (Recall that, as a standalone company, ISI did not have product margins comparable to WIND.)
The proper way to look at product gross margins is first to separate out services, and then to factor out royalties. While WIND only started reporting on the royalty component in Q1 of this fiscal year, I am able to extend the breakout back to FY 1995, with the breakout for the combined fiscal years 1999 and 2000 being the least accurate by far. WIND historically enjoyed gross profits on product license sales of about 85%. The introduction of Tornado increased WIND’s pricing power enabling the gross margin in increase to a peak of 87% in FY 1998.
The Asian crisis hit product license sales in mid FY 1999 and, worst, WIND’s financial results were combined with ISI’s after the merger. These resulted in the reduction of gross margin on product licenses to somewhere in the high 70%’s or very low 80%’s through FY 2000. We know concretely that this gross margin was 78.8% in Q1. But what was this gross margin in Q2? The answer should surprise you. It was an outstanding 84.1%. In words, this means that in the first complete quarter after the merger WIND has almost fully restored its pricing power of old. For the naysayers who see gremlins in the form of peer competition, Windows CE, Java from Sun, and now Linux, there is no indication that any of these apparitions are pressuring WIND’s business model, nor have they ever.
The recovery in WIND’s pricing power was expected after completion of the merger, because of the huge shift in WIND’s value-add sans ISI. Remember pricing power is a direct reflection of value-add and nothing else. However, the speed of the recovery is astonishing given that pSOS designs are still being sold.
The sense I got from the numbers and the CC is that WIND is at the beginning of some truly remarkable vertical markets, almost all connected to the Internet. WIND is the only soup-to-nuts supplier of the software foundation to Internet appliances, network devices and server appliances – all three of which are beginning to explode. As the sole software company capable of providing, servicing and supporting these solutions on a global scale, WIND benefits from rapidly increasing network effects.
The news about WIND’s rejuvenated pricing power and what it presages is so good that it overshadows the unfortunate difficulty in Japan, and what was clearly a questionable policy of appeasement in integrating ISI’s sales force. Despite major design wins with world-class companies like Sony, Japan slipped noticeably due to a stated need to shift to a direct sales and support business model. Frankly, I have never understood WIND’s or any other U.S. company’s Japanese business models, because of the apparent need or convenience of engaging distributors in that country. More times than once, the sell-through to WIND’s Japanese distributor(s) caused difficulty with the numbers, starting the first quarter after going public in 1993. Consequently, I have no suggestions or opinions about what needs to be done in Japan, other than fix the problem and end up with the ability to do business optimally with the likes of Sony. I hope it all gets fixed by Q4, but I will be infinitely patient as long as positive relationships with companies like Sony continue unabated.
The uncontrolled mushrooming of the cost of sales was blamed on a shift to a quarterly commission plan. Growth in the Marketing and Sales expenses yr-on-yr of 49% was worse than Q1’s 46% increase, but at least there was a slight reduction in Marketing and Sales as a percent of revenues, consistent with the argument that WIND is growing out of the problem. I am certain this item will receive close scrutiny by all analysts following the company, as it represents the biggest fly in the expense ointment.
R&D expenses increased yr-on-yr even faster than Marketing and Sales, but that is consistent with WIND’s stated intention to “invest” and the obvious extra expenses of rationalizing product lines. Personally, I will remain extremely patient concerning R&D expenses, possibly forever. The only constraint is for management to stay in touch with the requirement of the financial market in order to maintain an acceptable stock price – assuming the stock price becomes acceptable one day.
G&A expenses held flat sequentially and, at 30.8%, grew less than revenues yr-on-yr. This suggests we are over the hump digesting the ISI acquisition, at least from an administrative point of view.
Finally, at 74 days, DSO’s were at the lower of end of guidance, and, coupled with the Book-to-Bill being greater than one, is indicative of a healthy sales environment. Be a little cautious about the Book-to-Bill indicator, however, because it may be dominated by professional services contract, since these contracts typically span far more than the 90 days encompassing the sequential quarter. During a period of services buildup (which I hope we are experiencing), Book-to-Bill can be much larger than unity at the start of a quarter that may still be more back-ended than comfort dictates. Another way of saying this is that the ratio of 90-days-backlog to bill would be much more indicative of the ease of making the next quarter’s numbers.
Overall, the market should be very pleased with WIND’s market position and opportunities, especially in light of the insider selling during a risky, post-merger quarter. If WIND is able to reduce the Marketing and Sales as a percent of revenues LIKE THE COMPANY DID consistently for years after going public in 1993, guidance of 38% revenue growth this year and 30% revenue growth next year should result in EPS growth exceeding 60% next year. Thereafter, 30% revenue growth should produce EPS gains in the mid-40% for years.
Someone on the Belkin thread wondered why guidance was limited to 30% revenue growth for next year, in light of all the lily ponds showing early signs of significance. Management is as interested in the potential for lily ponds to impact revenues as anyone, and they can and will help us identify possible ponds forming. But precise timing of revemue inflection points makes reading tealeaves seem like hard science. Consequently, do not expect WIND management to risk the wrath of the stock market by issuing aggressive guidance under such uncertainty. Without such guidance, sell-side analysts would be foolish to forecast an inflection in revenues.
About the only public forum for anticipating explosive results stemming from exciting niche markets, or lily ponds, is this thread. As I said, the company can be expected to help us identify high-growth, high-value markets they dominate, and they can be counted on to provide the metrics needed for making forecasts, but they can not be expected to include even their internal extrapolations in guidance to analysts. (The DSL story presented at the Analyst Day is an example of the company expicitly providing values for all necessary parameters governing one lily pond, the DSL lily pond. At analyst meetings, expect management to update these metrics as well as provide similar metrics for other budding lily ponds.)
Allen |
