INFN 2Q16 Earnings and Analysis

Excellent earnings. Ghastly guidance. Multiple maladies.

Let me not waste too many words in the introduction. I’m going to need them for the analysis.

Earnings Report Highlights

The earnings press release can be found here:… Seeking Alpha’s transcript of the conference call can be found here:… (Thanks, Seeking Alpha.) Unless stated otherwise, all italicized quotations are from the Seeking Alpha transcript.

2Q16 Revenue: $258.8 million This is 25% higher than 2Q15. Infinera doesn’t categorize growth as organic versus acquired, but, based on Transmode’s pre-acquisition revenues, I’d guess that 15-16% is acquisition-related and 9-10% is organic growth. Management’s guidance had been $250-260 million, so they hit their number cleanly. Wall Street expected $255.8 million.

I’m going to permanently add a table to this section showing several years of quarterly revenue results. I know I find such tables useful in other people’s posts, so I should do it too!

In millions (GAAP).
         1Q       2Q       3Q       4Q     Comments
2012    104.7     93.5    112.2    128.0
2013    124.6    138.4    142.0    139.1
2014    142.8    165.4    173.6    186.3
2015    186.9    207.3    232.5    260.0   3Q15 Transmode acq.
2016    244.8    258.8

2Q16 Gross Margin: 47.8% GAAP and 50.4% non-GAAP Management had guided for non-GAAP gross margins of 47-49%, so they beat their guidance soundly. For the second quarter in a row, this result is above their long-stated intermediate goal of 50% non-GAAP gross margins. As was the case last time, they didn’t adjust their intermediate goal, saying that this level of gross margin currently isn’t sustainable. Some of the gain in margins is due to sales leverage, but it sounds like an important part is due to product mix – more sales of increased bandwidth (higher margin) and fewer sales of new equipment (lower margin, but better for future growth).

2Q16 Operating Margin: 6.2% GAAP and 13.2% non-GAAP Guidance had been for non-GAAP operating margin of 10-12%, so management beat the top end of guidance.

2Q16 Net Income: GAAP $11.5 million ($0.08 per diluted share) and non-GAAP $30.9 million ($0.22 per diluted share) Management’s guidance for non-GAAP earnings per share was $0.15-0.19. Wall Street’s expectations were $0.18 for non-GAAP earnings.

Earnings per Share (GAAP)
         1Q      2Q      3Q      4Q    Comments
2012   -0.19   -0.27   -0.17   -0.14
2013   -0.13   -0.09    0.03   -0.08
2014   -0.04    0.04    0.04    0.06
2015    0.09    0.13    0.06    0.08
2016    0.08    0.08

Earnings per Share (non-GAAP)
         1Q      2Q      3Q      4Q    Comments
2012                   -0.07   -0.05
2013   -0.06   -0.01    0.10   -0.00
2014    0.03    0.11    0.11    0.13
2015    0.16    0.18    0.22    0.21
2016    0.19    0.22

2Q16 Cash Flow From Operations (CFFO): $28.2 million; CapEx: $12.4 million. CFFO numbers above $20 million are pretty good for Infinera.

3Q16 Guidance
Revenue: $180-190 million (vs. $232.5 million in 3Q15)
Non-GAAP Gross Margin: 45-49% (vs. 47.5% in 3Q15; a wider range than normal)
Non-GAAP Operating Expenses: $85-89 million (vs. $77 million in 3Q15)
Non-GAAP operating margin: -2-+2% (vs. 14.4% in 3Q15)
Non-GAAP earnings per share: $-0.02-+$0.02 (vs. $0.18 in 3Q15)
GAAP earnings per share: $0.13 less than non-GAAP earnings per share
Wall Street had previously expected 3Q16 revenues of $273.2 million and $0.22 in non-GAAP earnings. This is an extraordinarily negative surprise.

INFN earnings day share price: $8.31 -33.57% (vs. S&P 500 +0.16%) The awful guidance was the main reason for the sell-off but, in my opinion, not the only one. Prior to the sell-off, INFN was my tenth-largest holding. I’m sure it’s dropped a few notches…

Customer Discussion
The Cloud Xpress customer count increased by 2 to 27.

There were two 10% customers for the quarter: a cable operator and a wholesale and enterprise carrier.

Long Haul
Long haul has long been the bulk of Infinera’s business. Expansion into the data center interconnect (DCI) and metro aggregation markets is more recent (MUCH more recent in the latter case). The DCI market, though fast-growing, is still relatively small. The metro market is a similar size to long haul, and similarly slower-growing. The metro market is a much more competitive market than DCI, with many incumbents desperate to keep market share. This dynamic means that the vast majority of Infinera’s current sales are into the long haul market (which includes subsea).

Although industry analysts peg long haul market growth at 6-8% this year, much of the growth is attributable to China, where Infinera has little presence and local producers like Huawei have prohibitive market share, in part due to government protection. Outside of China, Chief Executive Officer (CEO) Tom Fallon expects the long-haul market to grow 2-3%. In recent years, Infinera has grown its long haul business at three times the rate of the long haul market’s growth. That streak seems poised to end. This seems to be happening for a variety of reasons.
• Some carriers have reached the end of their current deployment cycle
• Some carriers currently have excess bandwidth and are waiting for it to be consumed
• Three of Infinera’s larger customers have recently been acquired (Time Warner, Cablevision, and XO Communications)
• Infinera no longer has the highest capacity equipment, which is an imperative for subsea deployment (it should leapfrog its rival when it releases the Infinite Capacity Engine - ICE).
• At least one customer is waiting for the Infinite Capacity Engine (although CEO Fallon notes that existing chassis are ICE compatible)

This confluence of factors, all hitting at about the same time, has led to diminished bookings and requests for quotes (RFQs). Chief Financial Officer (CFO) Brad Feller offered this description: “… our bookings in Q1 exceeded our internal plan. And bookings for Q2 into early June were in line with plan as well. Unfortunately, however, in the last weeks of Q2, our bookings declined dramatically. And thus far in Q3, the situation has not improved. While visibility is often challenging given our short lead times, we are very surprised by the speed and extent of this recent deceleration of bookings.

When bookings and RFQs in your largest business dry up, that is an unenviable position to be in. Of course, this is exacerbated by Infinera’s vertically-integrated business model. While such a model helps Infinera lower unit costs, it presents a fair amount of overhead that must be covered before profitability can be achieved. As recently as a couple of years ago, it appeared that a minimum quarterly revenue level of $150 million was required to achieve GAAP profitability. Given that Infinera’s guidance is for non-GAAP breakeven on revenues of $185 million, the hurdle has significantly increased. The difference, I’m pretty sure, is the Transmode acquisition. My best guess for the new GAAP breakeven revenue requirement is $220-225 million.

Data Center Interconnect
Data Center Interconnect is still shines brightly for Infinera, but there is a sunspot. For several quarters, management has discussed that competition will eventually come to this market, and this quarter it has arrived. As discussed last quarter, Inphi’s PAM-4 technology is making some inroads in the DCI market, but CEO Fallon noted that there are new entrants with purpose-built solutions akin to the Cloud Xpress. Perhaps most disturbingly, he described Infinera’s advantage over the competitors as being based on Infinera’s “experience set”, meaning “We’ve been … selling to and listening to customers, and putting in features that they value.” I wish that he had been able to say that Cloud Xpress is technologically superior. But he claims that technological superiority will be regained once the ICE technology is incorporated into Cloud Xpress.

Infinera does not disclose the percentages of its revenue that arise from the DCI or metro markets, and CEO Fallon did not disclose anything that could be construed as third quarter guidance specifically for the DCI market. I suspect it will be good, but the volumes here are not enough to “save the day”, at least not yet. If that’s going to happen, it would require better-than-expected business from long haul.

Metro Core, Aggregation, and Access
As he did last quarter, CEO Fallon described Infinera’s results in the metro market as “mixed”. Although highlighting, “… we continue to win deals on our pipeline of opportunities with large customers…”, he noted that spending is “muted” among original Transmode customers. Further, he expressed surprise at the speed – or, more precisely, the lack thereof – of customers moving through the process of testing equipment in their labs, then performing small-scale field testing, and eventually rolling out equipment gradually throughout the network, on the way to full-fledged adoption. Infinera also prides itself bringing its products to market at a sweet spot in adoption curve when the potential customer base isn’t only early-adopters. I get the impression that Infinera may have arrived at the metro party a little early.

Infinera’s Biggest Headwinds
Each of the issues I’m going to present below is likely temporary and/or solvable. The biggest problem is they’re hitting all at once, right now. Is management a victim of forces beyond its control, or are they culpable? Let’s explore these item-by-item.

Long Haul “Saturation”
Generally, I give Infinera’s management good grades for recognizing and anticipating market trends. So I’m a bit surprised they were caught flat-footed here. Just last quarter, CEO Fall on had this to say, “… we have seen some public statements suggesting that the long-haul market is saturated. We believe this is the wrong term, as saturation implies that the market has stagnated, which in our view is absolutely not the case.” (Seeking Alpha’s 1Q16 earnings conference call transcript)

I’m not sure that CEO Fallon would be so quick to issue that statement today. Infinera has a fairly broad customer base, but only a small subset of customers place orders in any given quarter. Historically the absence of one or even two large customers is always offset by the appearance of others. For 3Q16, the absence of some customers appears to be greeted by the proverbial crickets.

Could management have done something? It’s tough to say. I think they may be a little over-reliant on the idea that 30-40% underlying growth in bandwidth demand will always bail them out, and therefore perhaps they didn’t believe the kind of “pause” (hopefully it’s as short as that) we’re seeing today was even possible. Hopefully, too, they are chastened and will put more effort into “managing expenses tightly” on an ongoing basis rather than just in the face of an emergency.

Infinera has, in the past, looked at these as opportunities instead of headwinds, saying that when the combined entity compares vendors, Infinera probably compares favorably. Now, Infinera notes, acquired customers “… have to go check with the head office before they can spend.” Management obviously has no control over acquisitions that occur within its customer base, but it can control the optimism with which it views such events.

Subsea Losses
A couple of weeks ago, TMF poster wnewyorktokyo pointed out a Coriant box with VERY competitive capacity specifications. Infinera’s corresponding ICE-based entry will have greater capacity when released, although not by a huge amount. I don’t know much about other important specifications beyond capacity (e.g., power consumption, data center footprint, cooling requirements), but it is frustrating that Infinera allowed themselves to be leapfrogged. For subsea applications, capacity is more of a driving factor than the others mentioned above. Based on CEO Fallon’s “experience set” comment mentioned earlier, Infinera probably allowed itself to be leapfrogged in DCI too. While the situation is temporary, with ICE on the horizon, I think Infinera’s management team is embarrassed and angry. CEO Fallon said, “We are also going to continue to invest in next-generation technologies, like Gen 5, so that we never end up with a technology gap again because it’s too difficult to gain market share and then lose your position because of a six months’ or a one-year gap. We’re going to bring [up]our cadence of bringing that technology to market. We’re going to continue to improve it to being less than it is today.

Metro Adoption Cycle
I briefly addressed this earlier. This quote from CEO Fallon bothers me. “There’s lots of opportunities in the metro. But we are I guess disappointed and surprised a little bit by how long it’s taking to turn success in labs into success in field trials into success in creating substantive revenue. I feel comfortable that we’re making great progress. But it is taking a lot longer than we had anticipated.” To me, this is very reminiscent of CEO Fallon’s surprise surrounding Cloud Xpress adoption cycles. In the case of Cloud Xpress, I was willing to cut him some slack – after all, DCI is an entirely new market! The same cannot be said of metro. This is an established market and Infinera management should have known what the adoption cycles would look like before they decided to enter the market – either with their own products or via the Transmode acquisition. I find it hard to believe those cycles have extended all of a sudden. I’m hearing “unprepared” when management says “surprised”.

Gen4 Introduction
This is a good thing, of course, and not a headwind. But it is a necessary – even critical – item to which management must attend at a time when there are plenty of other distractions. We’ll learn more about the timetable for rolling out the Infinite Capacity Engine technology at Insight Infinera, which is held annually, typically in September.

Transmode Acquisition
Management has been tight-lipped about Transmode’s results as a standalone entity, arguing that the salesforces have been combined and they’re running as one team. However, at some level, management must be tracking Transmode on its own. According to Infinera’s 2015 annual report on SEC Form 10-K, Infinera does not own 100% of Transmode because not all shareholders tendered their shares. That is a temporary situation and will be corrected over time through a process in Swedish securities law called “squeeze-out”. In the meantime, Infinera must report a “Net gain (loss) attributable to noncontrolling interest”. This line on the income statement shows a loss, and has shown a loss (albeit diminishing) for every quarter since the Transmode acquisition. Assuming that Infinera’s ownership is unchanged at the 95.8% identified in the 10-K, I extrapolate that Transmode lost about $4 million this quarter, or about $0.03 per Infinera share. I wouldn’t be surprised if Brexit (Britain’s exit from the European Union) hurts Transmode’s (mostly European) business in the next few quarters.

Infinera’s Operating Expense Management, Briefly Revisited
Last quarter I reported that, since the Transmode acquisition, sales and marketing have been averaging 12.7% of revenues and general and administrative have been averaging 6.5%. This quarter, sales and marketing was 11.8% of revenues and general and administrative was 6.8%. I hope for some improvement in the absolute numbers next quarter, but I’d be surprised if they decline as a percentage of revenue, given the huge expected revenue decline.

Other Random Musings
Transmode’s Large Customer, Briefly Revisited
CFO Feller indicated that the Transmode customer in question from last quarter is still a customer, but hasn’t increased their sales volume.

Time as a Weapon, But Be Careful Where You Aim
CEO Fallon reiterated that he thinks “Time as a Weapon” is absolutely the right strategy for Infinera and for its customers. It is a strategy that shifts some risk onto Infinera, no doubt, but it is a very customer-centric strategy. Of course, Wall Street hates it because it is difficult for Infinera to give guidance and difficult for analysts to create models based on that guidance. Some of the risk that Infinera bears is inventory. You’ll recall that Infinera has made deliberate investments in inventory to support its “Time as a Weapon” strategy. Especially with the newer generation of Photonic Integrated Chip (PIC) on the horizon, will some of that inventory become obsolete and need to be written down? I’d imagine so, but it’s not too problematic if the amounts are well-managed. I guess we’ll see in the coming quarters.

Based on message board commentary, there seem to be any number of Infinera investors (or former shareholders) who believe that Time is a Weapon aimed at them! Yes, there are challenges today, but look to the next quarter, the next year, the next product release, the next generation. They hear Infinera management telling them to focus not on today, but a future time, when things will be better.

Concluding Remarks
This is a very frustrating time to be an Infinera shareholder. I animatedly likened last quarter’s market reaction – a 22% loss – to a gunfight. This quarter’s 33+% loss leaves me at a loss for analogies, and a bit more subdued.

Where do I see Infinera going from here? My general stance is still bullish, but in a more tempered way, and I definitely see risks to my thesis.

Let’s look at the worst case first. If the slowdown in long haul spending is protracted, that will be very painful for Infinera shareholders. Given the new breakeven threshold due to vertical integration, Infinera will likely post losses as long as long haul revenues remain lackluster. I would expect DCI and metro revenues to be beneficial, but they’re too small today to provide meaningful contributions even with solid growth. Declining sales have the potential to snowball if Infinera’s long-term viability is called into question. I don’t expect that, since Infinera’s balance sheet is solid enough to withstand such a downturn (long-term investments exceed long-term debt), and eventually Infinera WOULD be bailed out by 30-40% growth in bandwidth demand.

Another possible outcome is that Infinera is acquired. That’s not the endpoint I expect or am hoping for, but it is a real possibility the longer the price remains in single digits. I think Infinera’s prowess in large-scale integration and working with indium phosphide is worth something, even if other aspects of the business are discounted. I don’t think those other aspects should be discounted. The drive towards “Layer T” networking and software-defined networking (SDN) are valuable too, although perhaps a little ahead of their time.

I think the most likely outcome is that the long haul downturn lasts a couple of quarters, but resumes with Infinera holding, but not gaining, long haul market share. Growth will need to come from DCI and metro. If the Gen4 (ICE) technology is as good as claimed, I see a solid runway in DCI. Gains in metro are possible, but Infinera will have to compromise its gross margin goals to gain market share. Infinera was able to shake out long haul competitors until fewer were left and prices rationalized. They were able to do that without too much compromise on prices because the DTN-X was nicely differentiated. I’m not convinced at this point that they have quite the same differentiation in the metro market, which leads me to believe they’ll have to compromise on price. I could be wrong; Transmode had excellent gross margins and – legend has it – they were even profitable before Infinera bought them. If, indeed, Infinera needs to compromise on gross margins to win in the metro space, they’re really going to need to keep an eagle eye on operating expenses to be able to show a profit. How attractive will the market find an Infinera that is successfully growing revenues but is struggling with profits? Tough to say, but it probably mostly depends on the revenue growth rate.

Since the turn of the millennium and the associated bubble-burst in tech and telecom, companies like Infinera have been painted with a broad brush depicting a horrible industry where a concentrated group of customers can manipulate suppliers desperate for business. Infinera has broken that mold by focusing on an emergent set of customers – principally Internet Content Providers and Wholesale and Enterprise Carriers. Moving into DCI, Infinera offered those non-traditional customers a differentiated, purpose-built solution for certain applications, and addressed a rapidly growing market. By entering the metro market, Infinera has simultaneously expanded its total addressable market (TAM) but also made itself more reliant on the very customers it deemphasized while gaining market share in long haul. A fascinating proposition! I’m sure I’ll be in the game watching. Whether I’m “in the game” with quite as many shares as I have today is something I’ll have to spend time considering.

I don’t believe that management deliberately uses time as a weapon against investors, but progress happens more slowly than we as investors would like. Management announced that they would enter the DCI and metro markets years ago. We’ve seen their DCI entrance come to fruition successfully. Entry into the metro fray is still in its early days, and I’ve always contended that it will be a difficult battle to displace incumbents. I think that management saw acquiring Transmode as a way to smooth the process, as Transmode already has metro customer relationships. Perhaps that aspect was successful, but I’m not sure that management fully appreciated the effort involved in integrating an acquisition at the same time as they entered a new market. It is frustrating that Infinera allowed competitors to catch up with the capacity of their technology. I think it is evidence that perhaps management is stretched too thinly. I do believe that entering the DCI and metro markets is an appropriate strategy, given the overhead inherent in Infinera’s vertically-integrated approach. But winning the battle in metro won’t be easy, and the current drop off in long haul bookings and the new DCI competition, proves that Infinera needs to maintain vigilance across all its markets. Although the strategy is appropriate, it is proving difficult to execute.

The difficult guidance is part of the cause of the decline in the INFN share price, but I think there is also dawning recognition that Infinera will have to sacrifice margins to win business in the metro market. And that will lead to muted profitability over the next few years, even if the long haul market bounces back.

The release of the Infinite Capacity Engine should prove beneficial to results. Infinera intends to announce a release schedule in September, so we cannot place any hopes on ICE helping third quarter results. Any benefit to fourth quarter results is likely to be muted as customers work through their certification processes. Yes. Investors will need to wait patiently. Or not.

Fool on!
Thanks and best wishes,
TMFDatabaseBob (long: INFN)
See my holdings here:
Peace on Earth


Excellent earnings. Ghastly guidance. Multiple maladies.

Thanks, DataBaseBob, great analysis! Thanks for taking the time.

1 Like

Thank you as always for a great writeup – this is really comprehensive and insightful. For me, the INFN analysis can be summarized thusly:

Is the technology differentiated?
I would argue no, not really. I recognize and appreciate that INFN has been able to hold an attractive gross margin level above peers for some time. Despite that, for a company to consistently spend 20% of revenues (a huge number) on R&D, and then be forced to 1) compete on price in a key market, and 2) watch competitors quickly catch up in key markets such as DCI, says to me that the tech advantage is not persistent. Differentiated tech does not have to enter a price war. With respect to return on R&D investment, the widely-trumpeted CX was purpose-built for the emerging DCI market, and after a couple of quarters is already on roughly equal footing with new entrants? If INFN has to dump all profits into R&D just to stay a couple of steps ahead, it begins to look like a hamster on a wheel, spinning fast but not really going anywhere…

Does management have a clear plan?
Again, I would argue no. If we assume that INFN management was being forthright (I am willing to give them the benefit of the doubt on this as I have talked with the team several times and in prior years they have not sugarcoated bad news), then this swiftness of the market change caught INFN completely off-guard. To echo TMFDatabaseBob, it is discouraging that INFN management would so badly misunderstand the dynamics of the metro marketplace given their identification of it for some time now as a key component to future growth. The metro initiative by any objective measure has been very poorly executed. If longer than expected lead times for sales cycles and more intense pricing pressure are the norm for metro, management should not be surprised as they had a long time to evaluate the competitive landscape since metro is an established market.

CEO Fallon’s newfound desire to manage expenses more tightly is a welcome sight, but cost-cutting is limited, low-hanging fruit and a poor substitute for a well thought out and executed long-term strategy. For a long time now, customers have complained about the slow ‘cadence’ of upgrades from INFN. They are taking steps to address that, but again this sounds to me like R&D is a gas guzzling engine that sucks up every drop of revenue it can. In previous conversations with management, I’ve been told that R&D spending would come down somewhat to a more normalized state (I was previously projecting R&D falling to the low teens after 2017) after the metro market products were running, but this seems unlikely now. So cost cutting in response to declining sales does not change the outlook much for me.

Are slowing spending patterns an industry dynamic or specific to INFN?
My view is yes and no. It is clear that INFN is losing market share due to the aforementioned ‘catch up’ of competitor tech. I’m not discounting much of the outlook shortfall as customers waiting on ICE, mostly because of the speed with which the outlook deteriorated. It is also clear that INFN cannot differentiate itself at present in the metro market. Industry-wide longhaul slowdown (ex-China) and lumpy metro spending are not to be blamed on INFN, but part of my original thesis on INFN was the 1) tech advantage, and 2) focus on non-telco customers would both allow INFN to generate returns on capital in excess of what is the norm in a tough industry. That appears to not be the case, especially given that the big telco players that make up the metro customer base seem by all appearances to be playing hardball on price.

So all told, perhaps this splash of cold water on the faces on INFN management will result in a better strategy, and perhaps the upcoming ICE will be a more persistent tech advantage. But IMO that would be a substantial departure from what I see now.


Saul, thank you for your kind words. They mean a lot coming from you.

bobloblaw, thank you too, both for your kind words and extra insight. While I think we largely agree, I think your view is a bit harsher than mine, and I want to clarify the differences I see.

I DO think that the highly integrated PIC provides a degree of differentiation, and that is why Infinera has been as successful in long haul as it has over the years.

I’m not sure that I’d characterize competitors as having quickly caught up. It largely took them until the end of Infinera’s upgrade cycle to reach a competitive state. Soon Infinera will up the ante. I suspect that it will take the lead in DCI. In long-haul… I have to say, I gave the specs on that Coriant box a good looking over, and the ICE equivalent is only slightly better capacity-wise. That thing is a beast!

CEO Fallon did indicate that they are gaining market share in metro. Of course, it’s slowly, and off a small base. I think it may be too early to say they’re executing poorly. But the “surprised” remark clearly didn’t sit well with me.

I actually like CEO Fallon’s approach with throttling R&D in line with revenue growth. I think it enhances predictability at a company whose “Time as a Weapon” approach reduces predictability. However, the combination of lower revenues, new market initiatives, next generation roll-out, and desire to increase next-gen cadence combine to make it unlikely that he will be afforded the option to throttle R&D - and I’m not sure he should stick to 20% for the next few quarters if he could. I remember when I looked at PMC-Sierra after Skyworks announced their intent to acquire them. PMCS’ R&D was a gas-guzzling engine. Infinera is well-tuned by comparison (PMCS was a low hurdle, admittedly, but I think Infinera’s R&D is probably better than average, not worse). Yeah, I remember commentary a couple of years back about the 20% cap being lowered at some point in the future. I can see that happening if Infinera has success in metro. But I don’t see it on the near-term horizon. If anything, I think it may need to go the other way in the near term.

Again, thanks for your thoughts and kind words, and best wishes,
Fool on!
TMFDatabaseBob (long: INFN)
See my holdings here:
Peace on Earth