Interesting detailed MKM analyst's comments about Acacia, posted over on that Yahoo chat board. Same analyst has a neutral rating on Infinera and an $8 price target citing "Infinera: Long-Haul Still Weak, and Metro Far Away"... unfortunately accurate, so far. He used to be bullish on Infinera a few years ago, but their delay with ICE looks like killed his previously enthusiastic price target for INFN. Too bad.
"MKM Partners networking analyst Michael Genovese this afternoon initiated coverage of fiber-optic component supplier Acacia Communications (ACIA) with a Buy rating, and a $64 price target, writing that it is the "only optical components name we would recommend to long-term investors with a 3+ yr time horizon."
Acacia is poised, he believes, to replace the “digital signal processor,” or DSP, that various networking vendors develop “in house” for their equipment.
"In our view, Ciena is the only in-house DSP maker that can challenge Acacia for technological leadership over the long term,” writes Genovese, referring to fiber-optic networking giant Ciena (CIEN).
Acacia’s pace of innovation will outstrip most of these other vendors, he writes:
Optical transceivers operating at 40+G speeds and 50+km distances require electronic DSP chips to overcome dispersion. Several of the large Optical systems vendors have in-house DSPs they refresh every 2+ years, with each new DSP costing $50mn-$75mn to develop. Acacia is the leading merchant DSP vendor. It sells six DSPs, and it introduces a new one every 12-18 months. Merchant solutions have ~1/3 of the DSP market share. We expect this share to grow over time driven by purchases from vendors that have in-house DSPs. Some of the larger systems players are looking to diversify DSP sourcing for application flexibility and because of the high cost and long development cycles for in-house DSPs.
In case you were wondering who those system vendors are, they include Ciena and all its competitors, as Genovese outlines in a chart in the report:
Some of the things that have recently hampered the company are looking like they’ll improve, such as the slump in demand from China that has hit all the vendors, writes Genovese:
Acacia missed 1H17 expectations because of weak Chinese demand, lumpiness in its DCI opportunity and a contract manufacturer product quality issue that caused supply constraints. All of these factors are now improving […] Acacia is seeing improved order rates, compared to 1H17, from its large Chinese customer (ZTE) driven by inventory normalization and formerly delayed backbone expansion projects that are now moving ahead. It is also seeing improved order rates in the DCI market, particularly from a Tier 1 Hyperscale that it sells 400G modules to directly.
He likes the company’s lead in moving up to higher and higher speeds with its DSPs:
The company has a new 1.2Tbps DSP chip scheduled to be generally available in mid-2018. It is also the strong market leader in the emerging CFP2-DCO module category with over 15 customers. ACIA has roughly 1,000bps higher GMs than its peers due to selling DSP-ASICs and from Silicon Photonics. We think it merits a valuation premium to the group.
(For more on Ciena’s efforts with its WaveLogic DSP chipset, see my recent interview with Ciena CEO Gary Smith.)
Genovese sees the silicon approach of Acacia winning out over the exotic materials such as indium phosphide used in most of fiber-optics component manufacture:
We believe the demand trends in Optical transceivers, including faster and faster speeds and smaller and smaller form factors, bode well for SiPh over time. Acacia views CMOS-based optics as key to lowering overall system design costs. Photonic integration reduces the number of discrete components and saves on packaging costs. Acacia’s transceivers do not need lenses, do not need to be hermetically sealed, and have especially compact waveguide designs. Wafer-level testing (testing chips while still on the silicon wafer) also helps save on costs. Over the long term, we believe SiPh is the best positioned technology to deliver compact, low- power, low-cost modules. As the technology continues to develop, SiPh should enable greater component integration densities over time. It also offers co-packaging opportunities with electronics such as DSP-ASICs, amplifiers, and drivers. All of this promises to reduce the power consumption and costs of coherent modules going forward. Some of the near-term impediments that need to be overcome are the lack of uniformity in design, processes, and techniques being used by today’s SiPh industry players. However, as volumes increase we expect to see more uniform tools and processes take center stage.
As far as estimates, Genovese sees Acacia making $417 million in revenue and $1.90 per share in the fiscal year ending this December, just slightly above consensus for $416 million and $1.87.
For 2018, he models $513 million and $2.59 per share, which is also higher than consensus for $416 million and $1.87 per share."
FierceTelecom article cites Cisco and Ciena taking leadership in 100G and 400G WDM as well as DCI. A passing mention of Infinera as participating in the DCI space.
Sounds like Infinera is well positioned in DCI, thanks to Dr Elby. Now it's up to their recently hired couple of Enterprise-related Sales stars to spread the word about the wide variety of situations where Infinera DCI appliances, Flexponders, and Meshponders can make their local and regional networks fly at warp speed.
Pretty much the same story at the Deutsche Bank Technology Conference.
CFO comments that the Street is treating this stock like a "prove yourself" situation but that they've already been proving themselves for years. I think he's confusing proving yourself with respect to technology path versus maintaining sufficient development cadence so as not to have product gaps. Especially when this is not the first time they've tripped over their own feet late in reaching a sufficiently competitive cadence. imvho, sometimes statements like that can be worded in more of an admittance tone, rather than "I think" news. That and giggling about ICE back in early 2016 when they're still a year away from their target release in DCI and another 6 months to sequentially roll it out in their other systems.
They belatedly up'ed their ICE development cadence in order to keep up in DCI. Since pretty much their entire product line now hinges on ICE, LH and other Metro systems get the benefit of a faster upgrade cycle where the CFO states doesn't have nearly as fast a refresh rate. Maybe before, but I think he's relaxing too much with respect to Ciena. They're refreshing their entire lineup via Wavelogic, as well.
Thanks again Dr Elby... you da man in DCI and speeding up Infinera's PIC and ICE cadence. Hopefully their development and production teams can get in shape and not end up making a habit of a few months' delay on top of their target, like with ICE Gen1.
Moore's Law done profitably is supposed to increase integration (that's in demand) much faster than development difficulty, So far, the former hasn't been that strong and the latter has been much more than planned, for Infinera.
Well this is good news. A big system deployment. Need a few more of these.
Seaborn Lights Seabras-1 Subsea Network to Offer Cloud Scale Services On Demand with Infinera XTS-3300 Meshponders
SUNNYVALE, Calif. and BOSTON, Sept. 12, 2017-- Infinera, a provider of Intelligent Transport Networks, and Seaborn Networks, a leading developer-owner-operator of submarine cable systems, announced the ...
Update on CenturyLink closing of merger with Level3.
Mid to late October 2017 (2-4 weeks later than originally targeted). They were shooting for end of September but legalese had one month of padding.
Interesting that California PUC announcing additional negotiation $$$ demands, at this time. I guess that's how hardball is played. Ask for plenty in concessions and make the party sweat as the deadline clicks closer.
Number of open job reqs at Infinera has gone down a bit in the last few months. 98 total openings.
Not sure if it's worrisome or not, but a majority of the reqs were posted over a month ago. Nobody filling the positions. Majority of openings are for various scientist or engineering positions.
Business Development Group(1) Development Group(80) General(2) Information Technology(2) Manufacturing Group(5) Marketing Group(3) Procurement Group(1) Sales Group(2) Technical Support Group(4)
Some of the CFO responses at the recent Citi 2017 Global Technology Conference questions from the analyst covering Infinera seemed to indicate it's going to be a slow road with no upward surprises in store, at this point: Sounds like at best a return to $10M/quarter ramp in sales. Unless Metro starts succeeding beyond Transmode's previous performance, we won't see Infinera back at it's early 2015 revenue highs, until late 2018 at the earliest. For sure 2019, or else these guys have finally shown to the investment community that really don't know how to (really) keep up with the cadence of their Competition and submit a high percentage of winning bids. DTN-X (Windstream CTO wishes it had come out earlier). CloudXpress2 (same thing, with Customers going elsewhere). Continuing to come out late and insisting they don't need "help". They do need help via an outside cash investment to pay for increased cadence.
- Not sure if and when legacy major Customers going through M&A will return to their previous product buying levels.
- Revenue recovery currently expected to be slow. This is a 2019 story, now.
- XTM2 sales starting in 2H2017, ramping in 2018
- DTN-X line module with ICE sales starting in late 2017, not expecting significant sales until late 2018
- Gen 2 ICE expected to be ready in 18 months (late 2018) in order to speed up new product cadence to catch up with the Competition
- no indication yet whether Verizon purchase of XO will be positive or negative effect on Infinera's legacy sales to XO, either with respect to filling deployed DTN-X footprint or cross-selling into the Verizon network (finally). Their last opportunity was 3 years ago, but they got heavily underbid by Alcatel (now Nokia). In the worst case, Verizon simply used Infinera's bid as a way to squeeze more out of Alcatel. There never was or will be a chance to penetrate the 2 biggest US TIer 1's, for some strange reason. Dr Elby and the Board probably knows why.
- still waiting for CenturyLink/Level3 merger to pass. Having two of their best Customers pausing sales has been killing them for going on one year plus. 10% Customer for Infinera amounts to $18M+/quarter in LH optical equipment sales from each of those companies. $36M+/quarter hopefully will come back outright, some time in 2018.
2018 is The Year for all the planets to align for Infinera. Hopefully shareholders have another chance to bail out and opt to run far away from this unsettling rollercoaster ride. It'll be a tough decision to stay hoping for a buyout or multi-product cycle continuous ramp in revenues. Both seem like long shots, based on previous performance and conference call comments unless they start beating Ciena (Waveserver) and ADVA (open, white box) to market with next gen systems.
------------------------------------------------------------------------------------------ Per Historical INFN stock reaction to new product announcement and actual rollout... it looks like it took about 2+ years for the DTN-X sales traction to wake up the stock price... 3 years for it to peak and sell off.
------------------------------------------------------------------------------------------ It seems DCI sales aren't as strong as DTN-X... CloudXpress initially shipped with 10G client ports back in Dec 2014. 100G version came out mid 2015. CloudXpress2 selling mid 2017. Analysts were poking at when first CloudXpress will hit $100M/year in sales. No answer from Infinera. Now with two models, maybe after 2 years they've surpassed that century mark.
So expect about $12M QoQ ramp in DCI revenue
------------------------------------------------------------------------------------------ Transmode sales were about $140M/year run rate, back when merger closed August 2015. Transmode plus Infinera Metro sales not broken out separately as far as I can find, but CEO comments indicate it's not taking off, so far.
So, $100M + $140M = $240M/year in sales coming from new or acquired Metro products
=================================================== Current revenue run rate total is about $700M/year
$700M - $240M = $460M/year coming from LH sales before ICE. That jibes with my last attempt to calculate what's going on that they've fallen back to 2012 LH-only sales levels
------------------------------------------------------------------------------------------ 09/2011: DTN-X announced 1H2012: DTN-X actual sales rollout
Stock hits an all-time low of $4.57/share on 10/01/2012, probably due to fears of DTN-X cannibalizing DTN sales
2013: slow rise to $11/share range 1H2014: wild swings in the $8-9/share range 2H2014: wild swings in the mid teens range 07/2015: peaks at $24+ 08/2016: pretty much downhill all the way, back to $8
INFN shareholders’ interests seem hardly aligned with management, given that the CEO has done quite well, while the company’s EPS dropped through the floor. If the next earnings call is more of the same, everyone should start to scream about this
Some Infinera M&A
Completed, some awhile ago: XO, Yahoo + Verizon (not a Customer)
Cablevision + Altice, Europe (not a Customer)
Charter + Time Warner
Due end of September 2017, might drag into Q4 due to pending California PUC approval CenturyLink + Level3
I'm sure there are others... I think Windstream acquired somebody. Point being most of this M&A effect should be a non-issue by now. Once CenturyLink pushes through its merger, hopefully no more talk from Infinera pointing to slowing M&A LH sales
Infinera ICE seems to currently have a pretty slim chance at AT&T and their Open ROADM initiative. Integration goes totally against that philosophy:
AT&T Ciena Fujitsu Alcatel-Lucent USA (now a subsidiary of Nokia) SK Telecom Orange S.A. Rostelecom Cisco Saudi Telecom Company Coriant Telecom Italia (TIM S.p.A) Juniper Networks
"The Open ROADM Multi-Source Agreement (MSA) defines interoperability specifications for Reconfigurable Optical Add/Drop Multiplexers (ROADM). Included are the ROADM switch as well as transponders and pluggable optics. Specifications consist of both Optical interoperability as well as YANG data models."
Citi Investor Presentation host asked a lot of what I thought were really good questions poking at Infinera's development cadence, opportunities coming back with their Customers going through M&A, how CloudXpress/CloudXpress2 is going relative to other solutions.
Dr Elby quipped that he wished CloudXpress2 rolled out earlier, because pent up demand is healthy and DCI/Cloud new product cadence is around 3 years, so they lost opportunity and really have to accelerate next gen ICE to catch up with the Industry cadence and get in-step with the Competition. ICE might become a pluggable product separate from CloudExpress, if that part of disaggregation becomes a significant sweetspot that Infinera shouldn't disregard/discount.
It's interesting that Infinera appears to hesitate at rolling out anything other than complete system solutions.
CFO seemed to say the usualabout going after only deals that make sense and pass on opportunities that will "never" be profitable, for them.
One thing left unsaid: If there was no technology barriers, they should have accelerated spending and gotten ICE product out in the 2016 timeframe to dampen the effect from all the M&A slowing down their Customer buys. They could have accelerated development cadence when revenues were strong. Now they are belatedly accelerating cadence and suffering revenue collapse. Mr Fallon said they won't make that mistake, again. It's interesting that with Dr Elby on board, he was probably a dissenting opinion pushing to accelerate engineering development earlier but had to defer to the majority focusing on fiscal stabiily. Looks like his road would have been less painful but too late now.
Infinera learning the hard way to accelerate development when revenue is good and forget about the CFO pushing to post earnings in the short term. They're not big enough yet to focus just on earnings when they don't have enough cashflow horsepower to keep all of their products first to market . This mandatory refresh of their entire product line right when their revenue collapsed due to M&A is bad news they should have seen coming as lining up badly if they delayed ICE rollout until 2017. "Never Again" with them learning at the expense of the stock price performance.
California PUC planning to approve CenturyLink/Level3 with conditions to spend at least $242M in California over the next 3 years.. Last hurdle to pass in order to close the merger.
Next PUC hearing is on 09/14/17 but apparently the earliest the proposed settlement can be heard is 10/12/17.
Other dark cloud over the deal is how deep things get with allegations against CenturyLink with respect to stuffing product sales.
If not for these ethics issues at CenturyLink, closing at the end of September would have been a sure thing, especially the way Level3 as smoothly closed out previous deals with their top notch CFO. This time it's mostly up to CenturyLink, unfortunately.
Hopefully they clean that up quickly get their optical equipment upgrades figured out and ramped up , again.
Institutional investors and hedge funds own 86.70% of the company’s stock.
Number of analysts covering traction-challenged Infinera seems to be increasing... might give them some pressure not to continue to stall out and/or kill their revenue with all of this attention. Some analysts are already audibly losing patience with their skidding and bouncing between guard rails trying to regain (revenue) control.
A lot of it is merger related out of their hands, but now that all those events have long passed (except for Level3+CenturyLink closing this month), that set of obstacles should be in their rear view mirror.
The only worry I can see is if the Verizon's acquisition of XO somehow ends up meaning Infinera loses filling in the rest of that installed base, in the worst case. I can't see that happening, unless somebody at Verizon calling the shots really hates Infinera. At least Infinera will get the message soon that that door is prety much closed for them, just like China.
This could very well be the best trading service around…Want in for FREE? Check out the*stock-dork .
One of the things Infinera appears not to have counted on was intense pricing. You can't win them all but Infinera's getting a wake up call on what it takes to win contracts. Their equipment may be too good to win anything except the research-related, Internet2, and high performance installations unless ICE allows them to do the limbo on pricing down to the bone to win footprint. WIth the likes of Huawei, ZTE, Nokia (Alcatel) going kamikaze on pricing. If CIena and Cisco chime are chiming in as well, it's pretty bleak for this guy going forward.
Some news today about ADVA: "Germany's ADVA Optical Networking (Frankfurt: ADV) has landed a fiber rollout deal with FibreCo, a network operator based in South Africa. FibreCo will deploy ADVA's FSP 3000 metro and long-haul technology to offer connectivity services to its business, government and carrier customers. Initial work will center on a 100Gbit/s ROADM network carrying data more than 780km between Johannesburg and the undersea cable landing station at Mtunzini."
All we have to do is wait until the end of October to hear Fallon tell us about unforeseen circumstances, and postponed capital decisions caused by vendor acquisitions and mergers ad nausem (basically the dog ate his homework again).
Boy, this stock just cannot get any respect from the Street, for some reason: - Company is sweeping through and upgrading its product line, albeit it 1 year or more behind the Competition - Q3 guidance indicated their revenue has bottomed out and they are confident in forecasting a slight recovery - 3 weeks left in Q3 and then another 3 weeks till the next conference call - Their legacy revenue stream is stable enough that they always hit within their guidance. They've never pre-announced a shortfall.
Admittedly, there are some warning signs about their PIC and ICE strategies - CEO commenting that they're "always for sale" (but nobody is interested) - delay in implementing a PIC-based module for their XTM2-Series. Response is that they first wanted to be backward compatible with their installed base. Indirectly, one can conclude that their R&D is really constrained and still has to prioritize based on available $... which systems do they think they can afford to be late-to-market with more than others.
My ad nauseam mantra... they continue to act like they need external capitalization help. Ciena has been first-to-market in everything that Infinera does and it's a wonder why Infinera still doesn't feel like it needs to lean forward more and at least align their cadence more closely, if not beat them. They've never just said they can't afford to be first to market... it's always been the story that they'd rather target mass adoption. Somehow that rings hollow... their confidence in PIC and ICE is enough to drive their market early on. Their product cadence defines their revenue growth which times their R&D budget to support their widening product line.
It looks like the Company finally got religion that their ICE cadence must accelerate, in order to ramp revenues faster to support their widening product line. Otherwise, they'll always be just a one or two system juggling act that drives their revenues while stalling on the rest of their products.
Interview with CEO Gary Smith, who commented about the optical equipment industry and why he thinks Ciena is considerably undervalued:
Ciena's newer initiatives, sales of its "Wavelogic AI" chipset, and its software, Blue Planet, are two areas that show there are opportunities for the fiber-optic company beyond just how many units of equipment it can sell, according to chief executi