While we would find it impossible to ever recommend securities of any kind because of the irrational ups and downs of the prices, we have a great deal of respect for many of the financial analysts to whom we provide consulting services on a regular basis, and they tend to be conscientious, informed, and trustworthy. Nevertheless, those people in the industry who have direct access to our second blog by following us there, realized that we felt compelled to publish the following article the other day: “Mind-Boggling Amount of Ignorance About Optics on the Street?” It was a directly a reflection of a “very close to an obscene” valuation of Applied Optolectronics, which has been hovering around the $2 billion mark for a while, and in our opinion, has been spurred on by too many individuals influencing optics stock prices, who are disingenuous, unknowledgeable, or unwilling to make basic distinctions between companies. Even with the article, which was connected with that post, which made the case for AOI’s market cap being too high, we thought was based on a superficial analysis. The worst part of AOI’s present market cap is that it is more outrageous than even that of Acacia Communications about a year ago.
While Acacia later had some credibility issues, such as its dealings with a particular customer, and on its marketing script, at least it was at the time of the IPO, the only merchant DSP vendor for system vendors providing pizza boxes for data center interconnect. Conversely, AOI is more of a me-too, low-end transceiver module provider, with some cost advantages, including product development and vertical integration, which will not be long-lasting, because of price pressures at 100G, including from both small and large competitors.
We have also written in the past about the PSM4 at 100G space being very close to an unprofitable market situation. In addition, there has been an insufficient awareness (perhaps partly based on AOI's narrative alteration) of how vital a role 40G will continue to play in AOI’s total margins in the long term. While the supplier is apparently trying to get a jump on the 4WDM opportunity, such as in apparently having a much greater cost advantage over Inphi’s ColorZ solution, again, it will ultimately, still face price pressures from a host of vendors, which are part of the MSA.
In some cases, the biggest factor which gets at most insufficient attention is that AOI’s largest customer, representing about half of its total revenues, Amazon, is producing its own 100G CWDM4 devices. Undoubtedly, AOI chose to pre-report its very impressive performance during the quarter because of concern that the hyperscale operator working with Macom and Fabrinet for that product might get additional attention. On the other hand, it was kind of a similar situation of Lumentum raving about high margins for 3D-sensors with just one customer as AOI’s exceptionally high margins were also right in the faces of Amazon’s network planners.
Therefore, Amazon may be even more determined to produce its own CWDM4 gear even more quickly to get its costs down as low as possible. In fact, in being a different sort of animal than Google, Microsoft, and Facebook, Amazon may decide that it could get an ever greater advantage by increasing the manufacturing volume, by having the Macom/Fabrinet combo selling to other large customers and/or selling the CWDM4 gear itself. With the varying amounts of direct competition between Amazon and these other hyperscale players (although there are instances of cooperation, such as on submarine cables), it remains to be seen how it would all shake out.
[written by Mark Lutkowitz]