For a while (we mentioned a cloud bubble for components in April, 2015), fibeReality has been discussing the idea of a collapsing optical ecosystem principally as a result of the efforts of the hyperscale data center operators to increasingly push componentry down to unprofitable levels. We currently project that the really positive impact of optics supporting 5G wireless (despite the very aggressive agenda of Verizon Wireless on infrastructure including metal poles, antennas, hardened cabinets, fiber, etc.) as well as the ramp-up of 400G capacity (in spite of the widespread hype of that data rate at OFC 2018) may not occur at least until 2020. So, regardless of the rationale provided by Lumentum’s President and CEO, Alan Lowe, the timing of the vendor’s acquisition of Oclaro this month was not an accident (especially with the industry waiting so long for substantial consolidation). It is similar to the lack of coincidence regarding these other four circumstances: 1) both Lumentum and transceiver contract manufacturer, Fabrinet, diversifying into other sectors like crazy, 2) Oclaro had heavily positioned itself for a buyout (we still believe the marriage would have been much better with Finisar), 3) Finisar selected an extraordinarily unconventional new CEO, and 4) NeoPhotonics has seemed close to being on life support for several months.
At the OIDA Executive Forum during OFC 2018 week, this writer asked Lowe and Finisar’s former CEO, Jerry Rawls, about the negative impact of the hyperscale players, and whether all of this activity and change could just be happenstance. “Lowe admitted that [these data center operators] are extremely aggressive and that it can be tough, and that they have a lot of clout, but they drive a huge volume -- but it is about continuing to make money by coming up with product investments.” While “Rawls responded that it did not drive the selection of the CEO,” he interestingly avoided the topic of these large buyers.
The purchase was facilitated because Lumentum has been one of the darlings of the institutional investors, and they for whatever reason did not fully appreciate the impressive performance of Oclaro (up until recently, four straight quarters of 40 percent or higher Gross Margins (GMs) -- while inexplicably in the past, driving up the market cap of Applied Optoelectronics to a totally inappropriate level. The point is that Lumentum has been able to comfortably purchase Oclaro for some time. Therefore, we just do not buy the response by Lowe as expressed here: “Execution on major M&As had to wait [since becoming a standalone company].”
As partially alluded to earlier, the PSM4s at 100G have been driven down to the point of close to profitless, and our most recent intelligence points to the pricing of CWDM4 at that speed becoming even more alarmingly low in recent weeks. Most critically, there is likely nobody in the industry who would have imagined that it would have been possible to have LR4 under attack.
In addition, it is a huge surprise with Lumentum coming through on shipping six-inch VCSEL wafers with apparently really good yields in order to supply Apple. It was nothing short of a massive accomplishment after worrisome technical matters this past summer (along with producing noticeably good margins). Nevertheless, its total GM for the last quarter was only a little above 40 percent, which may be a cursory sign of a turnaround in its ability to make money in its traditional business.
Lumentum benefited greatly from pretty much being the only source for those 3D-sensors and Apple desperately needed them. Without a doubt, this current situation will not be the case beyond the short term. Of course, Finisar recently announced a terrible quarter, and Oclaro, despite its efforts to focus exclusively on products with the most potential for return, dropped below 40% in its GM in its last report.
On a side note, it was a mystery after four public presentations as to why Lumentum did not promote the unprecedented volume shipment of VCSEL wafers at the large size. We were the first to report it (after our own verification) in a comment on our daily blog as well as in our monthly newsletter. There is obviously a strong indication that the supplier shipped a number of four-inch wafers to Apple as well.
Lumentum also seems to be in a big hurry to get out in front on edge emitters, which also has its own set of technological and cost hurdles. Yet, the key point is that it appears likely that the supplier felt compelled to pull the trigger on the Oclaro acquisition now as its diversification efforts were not expected to be nearly enough to make up for the anticipated hard times in the telecom/data center space over perhaps as long as the next two years.
As always fibeReality does not recommend any securities, and this writer does not invest in any companies being analyzed by us.
To follow us on our totally separate, daily company blog, which is exclusively on fibeReality’s LinkedIn page, please click here.
[written by Mark Lutkowitz]