INFN - February 2016

What follows is excerpts from my take of the most recent earnings conference call.

Infinera will next report earnings on April 27, after market close. Going forward, I’ll try to present a more complete version of my analysis on this message board, not just excerpts. But you’ll need to be patient with me as I often need a few days to really think through the results and decide what points are important to highlight.


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.

Other Random Musings
Points that CEO Fallon Wanted to Make
CEO Fallon isn’t evasive during the Q&A portion of the earnings conference call. He does, of course, set boundaries regarding what the business will and won’t disclose, but he isn’t evasive. This time, though, I felt as if he steered some of his answers towards messages he wanted to convey, rather than directly answering the question and leaving it at that. I think it behooves us to pay attention to the messages he wants us to hear. The next few sections focus on those messages.

Investing in the Business
I’ve already mentioned that “investing in the business” seemed to be a point he wanted to get across. It is often not one that Wall Street wants to hear, as it speaks to delayed gratification. It should be music to Fool’s ears though. CEO Fallon makes it clear that he sees great opportunities ahead, and he wants to ensure that Infinera is fully-prepared to meet those opportunities. He and CFO Feller specifically identified investing in Transmode to complete its integration, as well as developing sales forces dedicated to governments, Tier 3s, and regional sales forces in Latin America and Asia Pacific. More generally, “We will also prioritize investments in technology and operational infrastructure to enable Infinera to capitalize on future market share opportunities and continue our operational excellence.

Time as a Weapon
Following up on “investing in the business”, one area that was interesting to see in the earnings report was the build-up in inventory, especially finished goods inventory. Asked if the inventory build-up was a sign of diminished visibility into sales, CEO Fallon stressed that the build-up was carefully planned. He then explained that lead times had crept up in 2015, and Infinera sees prompt delivery as a competitive advantage (I’ll return to this point later). So the build-up in inventory is a combination of two factors. First Infinera wants to be sure customers aren’t waiting for products. Second, the number of products available for sale proliferated in 2015 as Infinera worked on tailoring products for certain market segments, which means there are more products to keep in stock.

CapEx versus Bandwidth
One question during the conference call focused on restrained spending by IT companies. CEO Fallon cautioned that some companies may be cutting back on CapEx, but many types of expenditures fall into CapEx. He noted that a common theme in recent conversations with customers is they’re asking how he can help them win deals related to increasing bandwidth demand. That’s a great environment for Infinera. CEO Fallon wants us to focus on bandwidth requirements, not CapEx budgets.

Infinera’s Moat
It would be facile to boil Infinera’s moat down to “We have PICs and you don’t. Nyah! Nyah!” There would be an element of truth to that. But CEO Fallon wants to remind investors that Infinera’s moat includes its vertically-integrated manufacturing process and the intellectual property (both patented and trade secret) associated with that process. He doesn’t currently see competitors making inroads on those fronts. One aspect of vertical integration that CEO Fallon highlights is that Infinera owns its cost structure.

Owning your manufacturing capacity certainly has advantages, as you’re not subject to the profit motives of a third-party contract manufacturer. However, it is generally easier to tamp down and throttle up the capacity you pay for with a contract manufacturer. When you own your manufacturing capabilities, you must pay for them in good times and in bad. That is largely why Infinera was unable to show GAAP profitability at quarterly revenues below $150 million, and has consistently shown GAAP profitability now that quarterly revenues exceed that threshold.

CEO Fallon also believes that the technological lead that Infinera has over its competitors is widening, not shrinking. I’ll return to this point in the “Next Generation Roadmap” section.

Focus on Customers
CEO Fallon wanted to stress that he doesn’t care which Infinera product a customer purchases, as long as they’re getting the right product for their needs. It sounds as if he’d be furious if he found a Cloud Xpress guy competing for the same business against a DTN-X guy, instead of the two cooperating to ascertain the client’s needs. I would suspect, actually, that the Infinera sales force is structured such that no one is focused on a single product.

Differentiated Business Model
Infinera’s business model is differentiated in several ways. We’ve already touched on their vertical integration, an approach most other competitors aren’t using. We’ve also touched on “time as a weapon”, where Infinera sees prompt delivery and speedy installation as a competitive advantage. One of CEO Fallon’s answers to analysts’ questions explored this further. The question sought a direct comparison of revenue recognition timing between Infinera and Ciena. CEO Fallon obviously didn’t want to comment on Ciena’s revenue recognition practices, but ended his answer with this amusing commentary, “We have three weeks or four weeks worth of backlog, they got half a year, it’s a different model. I don’t understand theirs, they probably don’t understand ours.” If this characterization is accurate, one can see how CEO Fallon thinks that speedy delivery and installation can be a competitive advantage against this particular competitor, at least. CFO Feller pointed out another differentiator, noting that Infinera is very selective in the sales it makes to ensure that gross and operating margins aren’t compromised.

Next Generation Roadmap
CEO Fallon offered this commentary in his prepared remarks: “In 2016 we expect to continue to extend our core technology lead and further enhance our product portfolio through the introduction of our next generation Terabit PIC technology. While skeptics have recently suggested that competitors that use photonic technology are beginning to close the gap on Infinera, we disagree and intend to demonstrate that the technology differentiation we offer to our customers is actually expanding.” I would not be surprised if exploration of this next generation technology is the primary focus of the Insight Infinera meeting this fall.

Layer T and Layer C
In my first Infinera earnings analysis – five quarters ago now – I talked about how networking has been traditionally viewed as a seven-layer “stack” where different layers perform different network functions. Infinera’s business has traditionally been in the bottom-most layer (the transport layer), although its hardware and software have begun encroaching on the switching and routing layers immediately above the transport layer.

Industry analysts are increasingly seeing network functions in the top four layers being performed on virtual machines rather than dedicated hardware. A virtual machine differs from a physical machine in that a virtual machine is a software-controlled aggregation of computing resources, which might be spread across several physical machines. In turn, a physical machine may have multiple virtual machines running within it. This has been a very important model in the computer industry for some time now, and it seems that network function virtualization (NFV) is a growing trend. CEO Fallon borrowed an industry analyst’s observation and noted, “You cannot create photons in software.” So the aspects of networking than can be virtualized are called Layer C (the cloud layer) while those that can’t are called Layer T (the transport layer). The dividing line is actually somewhere in the middle of the routing layer – the original Layer 3. CEO Fallon is putting everyone on notice that he intends for Infinera to be a transport provider in the broader “Layer T” sense of transport. “I believe Infinera is ideally situated with our unique intellectual property to deliver the most scalable, flexible and programmable intelligent transport network solution in the industry and to play a significant role in accelerating this new transformation.” This is disruptive to companies like Cisco and even partner Arista, who operate primarily in the old Layers 2 and 3.


Thanks and best wishes,
TMFDatabaseBob (long: INFN)
Peace on Earth


I’m really glad you’re here DatabaseBob. Your posts are always excellent. Like you, when I first discovered this board, I read every post from start to finish. It was definitely a worthwhile exercise.

For those who don’t know, besides being knowledgeable in technical and financial areas, DatabaseBob is also incredibly gracious. He even apologizes when he needs to correct you!

MasterCard (MA) and PayPal (PYPL) Ticker Guide
See all my holdings at


CEO Fallon is putting everyone on notice that he intends for Infinera to be a transport provider in the broader “Layer T” sense of transport. “I believe Infinera is ideally situated with our unique intellectual property to deliver the most scalable, flexible and programmable intelligent transport network solution in the industry and to play a significant role in accelerating this new transformation.” This is disruptive to companies like Cisco and even partner Arista, who operate primarily in the old Layers 2 and 3.

Bob, thank you for that reminder. I think it is also important to note this sort of transformation is going to take a while for the concept to take hold and the industry to catch on.

I think it would be a great exercise for us to find early examples of where layer T is starting to be enacted in a customer’s deployment model.

I am not 100% certain, but I suspect the work they did with Pacnet (now Telstra) got them a little closer to that goal:…

Do we know of other examples since this announcement?


Hi Kevin.

I agree that “… this sort of transformation is going to take a while …”, but I think it will be more from a design and manufacturing perspective. It strikes me that adoption could be fairly swift, once the product’s capabilities are proven. I suspect that those managing networks would rather have fewer physical boxes taking up space, drawing electricity, and throwing off heat. Further, a vendor consolidation is likely advantageous from their perspective. Although I was always on the software side of the house, I’ve certainly heard stories where Vendor A and Vendor B blame each other for problems where a failing process needs to go through two pieces of hardware. When both pieces of hardware come from the same vendor (or it’s just one piece of hardware), those finger-pointing moments are minimized.

Looking at the product descriptions of DTN versus DTN-X, there seems to be some amount of switching capability built into DTN-X that is not present in the DTN. I don’t have networking experience, so I can’t say whether those capabilities are such that a DTN-X purchaser would be able to get away with fewer switches in their network because of the DTN-X features. But, if that were the case, I’d say that it is the beginning of the encroachment. My lack of networking experience also leaves me unsure whether a Layer T box is appropriate at every point in the network where a router, switch, or transport box currently exists.

I could be wrong, but my take on the Pacnet product/service (which was really groundbreaking) was probably more a victory based on Infinera’s Software-Defined Networking (SDN) capabilities, probably with a dash of Instant Bandwidth thrown in for extra spice. I’m not aware that it was a step towards a Layer T architecture.

I sure hope we get some help on these points from our resident network expert, when and if time allows…

Thanks and best wishes,
TMFDatabaseBob (long: INFN)
Peace on Earth