Using the lilies of the pond metaphor for the network effect, where in the summer do you think we are? When will the "doubling" become noticeable.
The lily pond metaphor explains geometric growth often found in networks. Network effects is a different concept (see below). In terms of the lily pond metaphor, WIND has a number of market segments showing signs of sustainable geometric growth. I don't know all the ones beginning to be significant, but clearly I2O, DSL, cable modems, set top boxes, digital cameras, network processors with/without TMS (not IXP), postscript printers/multifunction devices and auto navigation systems are illustrative of ones that probably are. Next in line would be Internet Appliances, Internet/Java devices, wireless phones/devices, intelligent network processors and server appliances.
You can assess where we are in the summer with any one of these "lily ponds" as follows:
1. Determine the size of the current market in units, and the manufacturer's BOM for the processor chips -- approximate the latter by taking one-half the retail price.
2. Obtain a market research estimate of units in three to five years; these are often publicized.
3. Calculate the CAGR. If the number is large, say, 75% or greater, then we have a lily pond developing rapidly. Two questions remain, who owns the pond and where what part of summer is it?
4. Estimate WIND's market share from hints by the company and press releases by WIND and their customers. If the share is, say, 50% or greater, then WIND can be expected to be a major beneficiary of the pond, i.e., the "owner".
5. Calculate WIND's market share times the size of the current market in units times BOM times the value-add in percentage terms. The value-add varies from, say, .5% in a highly competitive market to about 5% in a high-value segment like TMS. This number tells you what part of summer it is.
I2O clearly meets all the criteria of a late-summer lily pond. I suspect that cable modems do also, as do set top boxes, digital cameras and printers. In particular, the DSL pond looks very late summer to me. I would have thought that the IXP pond is early summer, but that may have to be rethought in light of Intel's recent announcement that they have started full production of the IXP 1200.
While the company will never forecast an inflection point caused by one or more lily ponds entering late-summer growth, there is much the company can do to help investors anticipate future opportunities. This subject is broad enough to warrant many future posts, but suffice it to say now that it is helpful when the company defines a pond, its market share, and a rough indication of ASP. The company's decision to start reporting exact I2O royalties in Q2 FY00, and by ending the one-quarter reporting lag that confused prior I2O announcements, was exemplary. That same kind of precise detail should be considered for other lily ponds.
In past years, we've expected earnings growth to outstrip revenues because I2O and other royalties would kick in. Now, it appears from guidance that management is geared more toward revenue growth. Is this a design to dominate?
Yes. More than anything else, it reflects the need to travel at Internet speed when all about you are also. In my view, when first announced, it reflected a concern about the slowdown in revenues in early CY 1999. In an earlier post, I explained the slowdown as a lagged effect of the Asian crisis. Nevertheless, I was and still am entirely supportive of the shift toward more investment. Nothing is more important than doing everything possible to guarantee that WIND is king of the mountain at the point when size and importance of the mountain becomes clear.
Earnings reports also used to be more predictable, a comfort to my investment style. Next week and going forward this year and next, do you think WIND can manage all of these: revenue and earnings expectations; exceed both; accelerate growth rates of both?
Yes, for lots of reasons. Post-Asian crisis re-acceleration is a mighty force, especially when coupled with a sector beginning to sprout its wings. That alone should be sufficient to make meeting "properly managed" estimates a shoe-in. First Call's consensus estimates for WIND are better positioned now than at any time over the last six years. I'll try to explain why in a separate post.
Digesting the merger and adjusting to new management at the top represents the only risk I can think of for not making the numbers. However, if the unthinkable happens and WIND misses the 8 cents this week, then I would expect the company to announce something in the CC that would knock your socks off, causing the stock to go up anyway. I can think of about 10 announcements that could do that, but many more certainly are imaginable.
Allen
RE Network Effects.
The lily pond metaphor exemplifies the miracle of rapid, geometric growth often characteristic of networked systems/economies. Geometric growth is not to be confused with three other important factors, all of which key off of value-add and can be the causal agents fueling geometric growth.
1. Increased value-add by moving up the value-chain. This is what is implied when the company refers to "verticals" or "total solutions."
2. Changes in value-add brought about by changes in the competitive landscape. The ISI merger made a dramatic change in value-add because often ISI was the alternative to WIND. Value-add was immediately increased by the economic difference between ISI and the next best alternative, which often was much greater than the economic difference between WIND and ISI.
3. Changes in value-add from Network Effects. Network effects come into play only when the value of a solution to each node increases whenever another node implements a compatible solution. Metcalf argued that when network effects happen, they increase with the square of the number of nodes. He was thinking of person-to-person communication when he stated his law. Each person can communicate with each of n-1 other connected people, resulting in n*(n-1) possible communications, which is approximately n^2. In computer-based networks, where each node might well communicate simultaneously with many other nodes, network effects may well grow even faster than the square of the nodes.
There are many more success factors that can prove critical, including economies of scale, the size and scope of the sales channel, support, etc. However, the lily pond and changes in value-add deserve special attention. |
