Calix Inc (CALX) 2015 Q3 法說會逐字稿

Greetings, and welcome to the Calix third-quarter 2015 earnings conference call. At this time, all participants are in a listen-only mode. A question-and-answer session will follow the formal presentation. (operator instructions)

I would now like to turn the conference over to your host, Mr. Tom Dinges. Thank you, you may begin.

Thank you, Karen. Good afternoon, everyone.

Before we begin the call, I want to remind you that in this call, we will refer to forward-looking statements regarding future events and expectations, including but not limited to statements about our anticipated growth and growth drivers; our development of new products and product features; the success and/or market adoption of our products; our future performance; and our expectations related to our financial condition and results of operations, including revenue, gross margins, and earnings per share.

We have based these forward-looking statements on our current expectations, assumptions, and projections. These forward-looking statements are subject to risks and uncertainties that may cause our actual results or actions to differ materially, as described in our filings with the Securities and Exchange Commission, including our most recent annual report on Form 10-K and our quarterly report on Form 10-Q to be filed shortly after this conference call. All forward-looking statements are made as of the date of this call, and we do not intend to update this information.

Also on this conference call, we will be discussing both GAAP and non-GAAP financial measures. We provide non-GAAP financial measures to enhance an understanding of our operating performance excluding certain items, such as stock-based compensation expense and acquisition-related expenses, which we believe are not indicative of our core operating results.

We use these non-GAAP financial measures to evaluate our ongoing financial results and to establish our operational goals. These non-GAAP financial measures are used in addition to, and in conjunction with, our results presented in accordance with GAAP.

A reconciliation of our GAAP and non-GAAP results are included in our earnings press release, available on our website and at www.sec.gov. All numbers that are discussed in today's conference call are non-GAAP unless otherwise noted.

This conference call will be available for audio replay in the investor-relations section of the Calix website at www.calix.com. In addition, our earnings press release, supplemental financial data, our earnings presentation, and a [briefing] book on AXOS, our new architecture, has been posted in the investor-relations section of the Calix website, which you may want to review in conjunction with our press release and conference-call remarks.

I would now like to turn the call over the Calix President and CEO, Carl Russo. Carl?

Thank you, Tom, and good afternoon, everyone. Joining me on today's call is William Atkins, our Executive Vice President and Chief Financial Officer.

We just returned from our 2015 user's group, and there is a lot to discuss. After William's remarks, I will cover all the exciting news, highlighted by the announcement of our breakthrough operating system, AXOS, which marks the beginning of the era of software-designed access.

Before I turn it over to William, there are a few items that I would like to highlight up front. In Q3, we were encouraged by the progress across our markets, and the Calix team executed well again this quarter, delivering revenues of $112.3 million and earnings of $8.3 million.

While we saw good performance from our larger customers in Q3, some of them have been cautious in their communication of their remaining CapEx plans. Therefore, I would like to caution you that we see Q4 revenue being down from our strong Q3 results.

We also note that one of our existing customers has engaged Calix to execute a number of turnkey network-improvement projects that we expect will result in significant revenues over the course of 2016.

This engagement continues the expansion of our professional-services team, which focuses exclusively on the project management and implementation of networks based on Calix products.

With that, I will now turn the call over to William to cover further details, after which I will discuss the announcements from our user group, as well as provide more detail about software-defined access and the revolution that AXOS brings. William?

Thank you, Carl. We last provided you with guidance regarding Q3 on the 28th of July and in that guidance, we estimated revenues of between $107 million and $111 million, a gross margin of between 49% and 50%, and operating expenses in a range of $50 million to $51 million, thus resulting in EPS of between $0.05 and $0.09 per share.

Actual revenue for the quarter was $112.3 million, and EPS was $0.16 per share, with revenues above the top end of our guidance and earnings per share also above guidance. We incurred some expenses in relation to our Occam litigation, and EPS would therefore have been slightly better, at $0.17 per share, if we exclude these litigation expenses.

Driven by products and customer mix, gross margins came in at 49.3% and compared favorably to Q3 2014's level of 44.8%. We continue to be pleased with the continuing growth in our gross margins, with 2015 year-to-date gross margins of 49.8% being above last year's year-to-date gross margins of 46.1%.

We remain focused on growing gross margins over time, but we also reiterate our view that there are likely to be fluctuations in those margins from quarter to quarter.

Operating expenses came in at $47.2 million, below the bottom end of our $50-million to $51-million guidance range, and up $4.5 million from the same quarter a year ago, with this year-over-year increase primarily due to additions in headcount, particularly in research and development.

We recorded $0.6 million of Occam litigation-related expenses in the quarter, with operating expenses therefore been $46.6 million without taking into account these costs. A slower-than-anticipated pace of hiring contributed to our under-shooting OpEx guidance.

We do not break out costs that relate to turnkey network-improvement projects, but I also note that we do defer costs directly attributable to these projects, and that these will be reflected in our gross margins, rather than in our operating expenses, as we close out those projects and recognize the related revenues and the associated costs of revenues.

We ended the quarter with a total of $93.9 million in cash and marketable securities, down from Q2's $99.5-million figure. Q3 operating cash flow was in line with guidance, at a positive $5.5 million.

The primary driver of our overall use of cash in the quarter was our share-repurchase program, which amounted to $7.7 million of investments in our stock. We expect to continue to repurchase Calix shares under that share-repurchase program, and we also expect neutral operating cash flow in Q4.

Revenues for the quarter were $112.3 million, an increase of $6.5 million, or just over 6% higher than last year's third-quarter level of $105.8 million.

International revenues were $14.4 million in Q3, up from $13.8 million in Q3 2014. International revenues were 13% of Q3 total revenues, and 11% of year-to-date total revenues.

We had one 10% customer in the quarter. We have also seen increased demand from one other existing customer for the turnkey network-improvement projects that Carl and I noted earlier, with the revenue impact from those projects likely to be more heavily felt in 2016. We believe these projects will generate meaningful business and that this will also result in deferred revenues, which will be recognized as they close out.

Turning now to our balance sheet. Receivables DSOs remained at the same healthy 37-day level, as in each of the first two quarters, and were slightly better than the 39 days recorded for Q3, 2014. Inventory levels increased to $43.8 million in Q3 from Q2's $40.7-million level and were in line with last year's Q3 level of $43.8 million.

Inventory turns increased to 4.6 times in Q3 from 4.0 times in Q2 and were equal to Q3 2014's 4.6 times. During the remainder of the year, you are likely to see an increase in inventory from its current levels, as we continue to ramp up to meet the needs of turnkey network-improvement projects. As expected, and as per federal government guidelines, all broadband stimulus projects were completed in Q3.

I will now review how Calix performed over the first three quarters of 2015 relative to same period last year.

Year-to-date 2015 revenues increased by $12.9 million to $302.5 million, or up 4.4% from last year's $289.6-million level. And gross margins came in at 49.8% versus 2014's number of 46.1%.

Operating expenses increased by $17.6 million to $142.6 million for the first three quarters of 2015, versus 2014's year-to-date level of $125 million.

After deducting Occam litigation expenses of $2.4 million, the 2015 year-to-date operating-expense figure would have increased by $15.2 million to a $140.2-million figure, up from an equivalent figure for last year of $125 million.

Year-to-date 2015 earnings per share were $0.15 per share, versus last year's first nine-month equivalent EPS of $0.16. Without the Occam litigation expense, 2015 year-to-date EPS would have been $0.20.

I'm now going to turn to our Q4 guidance and to its implications for the full 2015 year.

Revenues for the fourth quarter are expected to be in a range of between $102 million and $106 million, with a resulting midpoint of $104 million, being down 6% from the $111.6-million level achieved in Q4 of 2014.

We are guiding to a 47% to 48% range for Q4 gross margins, broadly equivalent to last year's Q4 level of 48.1%. 2015 Q4 gross margins will be impacted, amongst other factors, by customer and product mix and by our ability to close out some of the projects that we described earlier in this call.

Operating expenses are expected to be in the range of $51.5 million to $52.5 million, up from last year's Q4 level of $47.1 million. Occam litigation-related expenses that will not be reimbursed by its insurers are expected to be approximately $1.5 million.

Please note that a significant portion of the sequential increase versus Q3 2015's $47.2 million is comprised of the usual one-off year-end expenses that we incur in Q4, including year-end sales-commission accelerators, the cost of our annual user group conference, and an adjustment to reflect the alignment of our financial calendar with calendar year end.

Over the past three years, Q4 expenses have been 2% to 10% higher than in Q3, with these types of specific year-end factors driving a lot of this increase. As evidenced by our announcements earlier this year regarding G.fast and NG-PON2, and by our recent announcement of the new AXOS system architecture, we continue to invest in incremental R&D resources to position ourselves to take advantage of the transformation that is underway in our sector. And Carl will take you through our recent product launches later in this call.

The expectations that I've just finished taking you through result in a guidance range for Q4 earnings per share of negative $0.07 to negative $0.03. If you were to exclude anticipated Occam litigation expenses, EPS for Q4 would be negative $0.04 to $0.00 per share.

Having issued guidance for Q4, I can therefore now share with you the implied [out turn] for the full 2015 year. For 2015, revenues will be in the $404.4-million to $408.4-million range, up from 2014's $401.2-million level.

Gross margins should be in a 49.1% to 49.4% range, up from 2014's 46.7%. And annual operating expenses are expected to be $194.1 million to $195.1 million, up from 2014's $172.1-million level.

At this point, I'm handing the call back over to Carl.

Thank you, William. Before I discuss AXOS, I would like to update you on the bevy of new products coming from Calix.

We added our new [GigaHub and GigaPoint] products to complement our very successful GigaCenter solution. Available in November, our Giga family of premises solutions can now tightly address every indoor deployment scenario globally.

To this, we added new powerful [comp] and software applications that, as part of the GigaCenter solution, allow our customers to deploy winning services to their subscribers at the converged, unlicensed wireless edge.

Fresh from its launch at the Society of Cable Television Engineers Expo in New Orleans just two weeks ago, the revolutionary E3-8G remote optical line terminal was a big hit. The E3-8G allows our customers to deploy fiber much deeper in their networks, thereby dramatically increasing their serving radius on the head end, central office, or data center.

Fully hardened and completely sealed, the E3-8G can be strand-mounted, pole-mounted, or placed under [grade]. Now, there really is no limit to what a PON can do.

We launched a series of new VDSL2 vectoring enhancements too numerous to mention here. However, as our customers have grown to expect, Calix delivered innovation to this product space, as well, showing our breakthrough in system-level vectoring technology, allowing 96 ports of VDSL2 to be vectored in a single rack unit E7-2, all without the cost or encumbrance of a separate vectoring engine.

As we have seen all summer, leading up to the user group, G.fast is hot. Our customers had the opportunity to lay hands on our G.fast systems for both indoor and outdoor deployment, as well as a complementary set of G.fast-enabled Giga family of solutions to complete the subscriber experience. We even demonstrated bonded G.fast, which allows our customers to deliver 1.5 gigabits per second over two copper twisted pairs.

As exciting as these product innovations are, there was a much larger revolution demonstrated at the user group, and that was the beginning of an entirely new era -- the era of software-defined access.

Last quarter, I shared with you how the architectural concepts of SDN and NFV that are currently driving significant change in the data center would ultimately affect every networking sector.

At the intersection of unified access and SDN is something entirely new, and that is software-defined access. And AXOS was built from the ground up to meet this demand.

To be clear, this is not about taking our existing systems, adding a shim, and making them SDN or NFV compatible. This is about creating an access infrastructure that is -- fast, allowing our customers to invent and product-ize new services in weeks; always on, eliminating down-time due to system failures or scheduled maintenance; and simple, creating an access infrastructure that can run itself.

This is the culmination of our decade-long pursuit of the unified access infrastructure. And as such, it is the most significant platform announcement in our Company's history.

An AXOS-enabled unified-access infrastructure will seamlessly connect the device-enabled subscriber to applications and content in the cloud and will place the communication-service provider at the fulcrum of subscriber demand and content supply, thereby enabling our customers to build high-value businesses. AXOS meets this challenge.

AXOS, or Access eXtensible Operating System, is the first open, Linux-based network-operating system and software-application platform built for access. AXOS is completely hardware-independent, thereby allowing all software to be developed and run independently.

AXOS fully abstracts all network services into separate processes and a consistent set of APIs, enabling full containment in isolation without any impact to other processes.

AXOS is stateful, as it maintains system state separate from the processes themselves, which allows for powerful self-auditing and self-healing capabilities. Furthermore, this enables individual processes to be upgraded on the fly, without impacting the system, the network, or subscriber services. In short, a fast, always-on, and simple network for our customers to provide services to their subscribers.

AXOS's open, component-based architecture, combined with its power portability and programmability across all layers, allows for optimizations in multiple dimensions. We can choose where a given service runs, thereby allowing the entire access infrastructure to be optimized on a real-time basis, from the subscriber edge all the way back to the data center.

Finally, the open component-based architecture all manifests in a single software image, allowing for maximum reuse. This translates into maximum agility and minimum operational expense for our customers, while at the same time yielding maximum R&D efficiency for Calix.

To reiterate, as all-access infrastructures converge into a single architectural approach, we believe a new access operating system is required to enable the winning communication service provider of the future. From the subscriber premises, all the way to the access edge of the data center, the operating system for access is AXOS.

And saving the best for last, AXOS is available now in several of our products, including our new G.fast systems. Let the future begin. Karen?

(operator instructions) Simon Leopold, Raymond James.

So guidance is somewhat surprising to me, and I guess the simple question is, where's the CAF II money? I thought that was going to start flowing in. So your tone sounded like you're trying to be conservative, given a lack of visibility, but this is very unseasonable for you, so I need to better understand what's going on in the fourth quarter.

So let me address, and I'll let William add some comments, Simon, to CAF II. Actually, this is sort of the front end of exactly that question. So here's what's happening with our customers.

We now know how much all of our customers are taking from CAF II. And in fact, what we're seeing is they're now down into such detailed planning that it's actually had a negative effect on their year-end spending.

So, yes, we're very encouraged by CAF II and what we're seeing into 2016, but it's actually had a slowing effect in Q4, and that's been called out.

In addition, we've had some of our customers, as you're probably aware, just on their last quarterly conference calls change their CapEx for Q4. So it's a combination of those two things.

William, is there anything you want to add?

Not really. I think the only thing I would elaborate on, in terms of Carl's response in CAF II is, as you know, we see CAF II as being substitutive rather than incremental. It does take time for the operators to formulate and then essentially procure for their roll-out plans, because it's such a granular process. It literally is down to a zip-plus-four-type postal code designation for the work. And so that does take time.

So there may be a CAF II planning element to the conservatism that we're demonstrating here, together with some of the noise that you and others have been hearing coming out of the sector.

And we may very well see orders, but that's not revenue.

And then the other point I would make is that, while it's more of a 2016 event than a 2015, these turnkey network-improvement projects that we've been asked to undertake, because they inherently involve a deferred-revenue and deferred-cost element, are going to be recognized as and when those projects close out.

So we're working on stuff now that is going to involve subsequent revenue recognition. That's the best way I can characterize it.

That planning aspect makes sense to me, but it then begs the question of timing. Do we see a March quarter that is exhibiting a lack of seasonality and therefore should be, let's say, flattish or possibly even up from your December quarter, given the shift-out in the timing?

It's, frankly, too soon to tell. If we could guide out that far, we would. I understand what you're driving at. Remember, CAF II is a multiyear process, right? We're talking about a six-year program. So we can't really call it one way or the other. We see it as a sustained spend over that period of time. Carl?

Just briefly, I think the challenge is, the timing in the quarters in 2016, we can't call yet. But I would reiterate that, as we've said before, we certainly see CAF II as putting a floor underneath CapEx and reducing the variability going forward. We're encouraged by what we're seeing, but we just don't have enough yet to be able to call that, Simon.

Okay. And then, AXOS, it's really intriguing, so I'm kind of interested in trying to get a better understanding and probably could ask a ton of questions. But let me just go with a simple one. It seems to invite the opportunity for some displacement, sort of the white-box alternative. And so I'm wondering how you think about the potential that you get into a software business and customers procure elements from white box, unbranded ONTs, ONUs. Just help us understand what the broader implications are to the business model.

Okay. Let me give it to you at the strategic level. I can't tell you -- I can't break it down for you in a margin or a revenue piece, but let me help you understand what we're doing here and what it's about.

So to your point, AXOS is, in fact, completely hardware independent. There is an enforced, flat hardware abstraction layer that allows us to move the operating over top of any hardware. Therefore, it dramatically speeds our ability to come to market. So that's the first piece.

The second piece is, as you asked, yes, we can put it on hardware that we would not have put it on before, including white boxes. Now, let me give you the contrary.

The access space is not as big, nor is it as congruent as the data center, where it's easy to build white boxes. In the access space, we still have stuff that -- well, recently was underwater in South Carolina.

So, yes, there's going to be white boxes, we think, over time. But let me give you a slightly different example of what we have done.

For those of you who are more astute observers of our leadership page, you may have noticed two years ago that we actually separated hardware from our systems organization. So Michel Langlois, who actually heads up our systems products, does not have hardware reporting to him, nor has he for two years. It's, in fact, handled by Mehdi Bradaran, who has both hardware and product operations.

The reason I say this, is we have actually been internally white-boxing ourself for some time.

Thank you. That's very helpful. I will yield the floor.

George Notter, Jefferies.

I wanted to go back to the guidance for Q4. Can you help me understand how many customers we're really talking about here? Certainly I can put a circle around the CAF II customer and kind of figure out which ones are yours versus customers that you haven't traditionally sold into. But is it just one or two customers that are causing the adjustment in guidance? And then also, I seem to remember that this happened last year to some degree, as well. I'm wondering what kind of similarities we can draw from this experience to last year. I'm trying to better understand why last year's numbers turned out to be quite a bit bigger, I think, in Q4 than the numbers you're contemplating here. Thanks.

Let me just give you a high level from my perspective, and then I'll let William make comments.

To go back over, it is not a broad set of customers, but it's more than a handful. But the impact is really driven by a smaller set in the big numbers.

So as you know, there have been some service providers that have actually gone out and made comment of tightening CapEx in Q4. That effect we have seen.

But the secondary effect that I want to reiterate is also what William spoke to. The detailed planning around CAF II is now in full bloom. And it just ties people up.

And so what we're seeing is sort of this slowing before things start to move, in our opinion. Because as we look through the dollars that were accepted, the time lines that they have to be deployed in and turned up, it's very clear that there's a floor under 2016.

So it sort of feels like that the car at the front of the line tapping on the brakes, and then they're going to accelerate away while we're sitting here with a little bent metal. William?

I think the only thing that I would add to what Carl has said, George, is a bit of an obvious statement, but worth repeating, which is that Q4 and Q1 are the two hardest quarters for us to forecast.

With Q4 in some years, we've seen a budget flush, and others, we haven't. This year, we're seeing people talking about essentially CAF II being substitutional, but they're also taking time to roll out their plans. And then in Q1, as we saw earlier this year, it takes time for people to pull together their CapEx plans. So the next couple of quarters are the hardest ones for us to forecast.

Got it, great. And then I also wanted to ask about international. I guess it was a good chunk better this quarter, but can you talk about where we're at, in terms of international, the opportunity to get growth out of that side of the business. What sorts of things are you changing there from a selling and marketing perspective, and anything changing there in terms of your level of investment?

Actually, George, in this, an up quarter, and last quarter a down quarter, you're actually going to hear me say the same thing, which is, we get the level of investment that we're making, and we are going to focus on productivity of that organization. And look, we were [incurred] last quarter, and some of it came through this quarter, and we'll see how it goes.

We had a lot of international customers this year, a record at the user group. So there's a lot of good forward signs, but we're not changing -- we set our plans, and we're executing our plans, and that's what we're doing. So William, I don't know if you care to add.

I think the only thing I would add there, George, as I think I've said to you and others, we see international as being a game of inches.

Carl was careful to note when he described some of our product roll-outs -- for example the GigaHub and GigaPoint products, as well as the GigaCenter -- now being globally available. That is a good, very positive new development for us as a company, and specifically for our international operations.

You look at G.fast -- that's done with an eye to more densely populated regions, multiple-unit buildings, etc. And these tend to be, in this country, more with the larger telcos. And then internationally, obviously, countries have got a more densely populated demographic.

So when we are doing our R&D, it's obviously just not with a view to expanding our product portfolio, but also in terms of penetrating new markets and expanding our share of those markets. Carl?

Yes, and one other piece -- I'm pointing at William, so he's handing it back to me -- he's bringing up a really good point. And we've talked about this for years.

As we converge on what we call our strategic spine of products, but specifically the E series, and now with AXOS and the GigaCenter family of solutions at the edge, it's putting us in a situation where our products are taken up globally.

So if you were at the user group, you would see customers globally really talking about the same product families.

And that wasn't always the case in the past.

Got it, thank you.

Paul Silverstein, Cowen and Company.

Carl, as I've listened to these calls over the years, I often think of the famous Roseannadanna line, it's always something. That said, let me ask you on the AXOS software. Can you give us any additional color on -- I know it's early, I know you just started shipping -- but I trust you had customers involved to some degree or another in the planning or the product and the conception. Can you give us any color on customer engagements, interests, involvement of the development? Can you give us any color on visibility, in terms of timing of impact, prospective revenue? Again, I know it's early.

And also, does this change the equation with respect to the other tier-ones that you've been appropriately cautious in engaging historically, in terms of not trading off revenue at the cost of profitability. But I assume this should have a very high margin, the software we're talking about. But any color would be appreciated. And then I've got a question on the turnkey that you referenced in your prepared remarks.

Okay. So, look, this has been a little over four years in the making. This is not something that we started recently.

We've been in pursuit of, in essence, a unified-access infrastructure that we felt would be worthy of what the next generation of communication service providers would want to deploy so that they could shift their business models. This does that. It cuts right at the business-model shift that the service providers are pursuing.

And so having said that, the best way I can characterize an answer to your question regarding customers is, you sort of have to look at the more forward-looking ones that are trying to change their business models, versus others. So that's the first piece.

The second piece is, I would make no comment about customers at this point. I would only encourage you to stay tuned would be my best answer on that.

And then as far as the business model for Calix, let me -- you have a broad background in this industry. Let me tie out a couple of pieces, here.

There are many different directions we could go with the business model at Calix because of AXOS. But that will be determined by customer readiness and market readiness to go forward.

So we will see which way folks want to purchase and how they want to deploy the systems. But to your point, this is pure software. This is pure, open, programmable software.

And on a personal note, it's a beautiful thing, as how it's structured and how it's built. I've pursued something like this for my entire career. So I'm certainly excited to be at this point. But 99.97% of the news about AXOS is in front of us, not behind.

So, Carl, I appreciate the excitement. Can you -- has this been through beta trials? Has this been through any customer trials (multiple speakers)?

So it's been running quietly. We've actually had it in networks running for more than 18 months now. So (multiple speakers) we would not make this announcement unless we knew what we had, and -- look, we think we know what we have, but there's nothing like putting it in environments and seeing what we've got.

All right, but I guess that speaks to the issue of whenever you roll out a new product, regardless of the product market, regardless of the vendor, there's always issues. But if you've had this in beta for 18 months, I trust you have a high degree of confidence that it works as advertised and you're not going to hit significant bumps on the road.

We have a high degree of confidence that not only it works, but it works, in fact, as advertised, that it is fast, that it is always on, that it is simple.

In addition to that, the new G.fast systems that are coming to market are built with this operating system. And it has been a remarkable undertaking to watch the speed of development.

So it not only meets the requirements of what we believe the next-generation winning service provider will require, but it actually has dramatically changed the way we do our own business.

All right. If I could ask you all, on the turnkey that you referenced, can you give us more color on what you're referring to, how big this project -- or I think you referenced it as projects. I assume these are derivative of CAF II, or related to CAF II?

Actually, it's an unusual circumstance, which is why we chose to highlight it. It is, in fact -- as William said, it's a turnkey project.

So as you know, we've been slowly but surely expanding our professional-services and our consulting-services organization to take on projects that can help our customers transform their networks.

As such, we just recently reached an agreement with this particular customer. It is not CAF II-related, actually. It's an entirely separate item.

Okay. And any sense for how large, or that's not something you care to share?

Gee.

(laughter) I'll leave that to William.

Gee, that's the question we weren't prepared for.

No, I think -- you know, Paul, I appreciate the question and totally get it. But just given the nature of the projects and the fact that they're sort of back-end-loaded within each project, in terms of delivery of professional services and products, etc, we're just not at a stage where we can call out the size.

Meaningful for us to call it out as being something that you'll see move through our financials, not just our P&L, but also our balance sheet over time. And you're going to see us essentially indulging in this activity. But really it's going to be more of a 2016 event, but it's going to start to affect our numbers in 2015, which is why we wanted to call it out.

All right. And I apologize to my peers on the call, but one last question on this. You're clearly indicating -- and here's the question -- that this, all other things being equal, for calendar 2016, that this particular customer engagement is incremental to your normal-course business?

That would be correct. We're saying that it's another existing customer, is the way I would put it. That's how I would phrase it. And we're calling it out because of the nature of the project and its potential size. That's what I would say.

All right. I'll pass it on, thanks, guys.

James Faucette, Morgan Stanley.

Hi, it's Meta Marshall for James. A quick question on -- a little bit more strategic of just what as -- the carriers are doing, or CAF II planning. If you could just speak to what you're seeing, as far as their decisions for [vectoring] versus VDSL versus G.fast -- just what speed upgrades you're seeing them make, would be helpful.

Look, CAF II is mostly a rural deployment, and as such, it's going to be almost exclusively the province of VDSL2 and loop shortening. There may be some other technologies deployed, but that's what you're going to see -- I would say over 90% of it. So the other things are really not worth noting.

Okay, but then where, like, are you seeing more G.fast kind of take place, and then what environments are you seeing those?

So G.fast would be away from CAF II, and G.fast -- so G.fast has an amendment 1, its first-generation version, which is -- we're demonstrating now, has a very clear deployment scenario, which is in MDUs and MTUs, where you are trying to get the last hundred feet, two hundred feet, three hundred feet, or where you can't easily get a fiber lateral -- so in other words, from the street to the building.

You can then deploy G.fast over hundreds of meters, and it yields very, very high speeds, hundreds of megabits or more, and in the bonded case that we spoke to in our prepared remarks, 1.5 gigabits.

Great. And then (multiple speakers).

Do you want to add, William?

I just wanted to add one thing, if I could -- sorry, if I can just leap in. Just one thing I would say about G.fast is that, in addition to being the kind of technology that you can roll out on greenfield or even brownfield-type developments, it's a particularly relevant technology for operators who've already essentially laid fiber past buildings, where they have not got that fiber drop into the building and into each individual unit, as Carl indicated.

So it's a great way for an operator who's essentially already covered a neighborhood with fiber to go back and take care of the narrow-band islands that would otherwise exist within that neighborhood.

Great, that's helpful. And just one last question on the turnkey. Understanding that you guys aren't giving size -- and just given that you said it was back-end-loaded, does that mean that it won't necessarily hit early in 2016, we would just see the deferred revenue in early 2016 and we would see more revenue in late 2016? Or did I mishear and it's already in 2015 and will hit early in 2016? Thanks.

What I would say -- the way -- we were careful in our prepared remarks to refer to it as an engagement for a series of projects. And so what that basically means is, as we ramp up that engagement and that series of projects, there inevitably is going to be activity that's not going to be recognized as revenues as we ramp up.

Once we ramp up, each of those individual projects will take maybe a few weeks, maybe a few months. And at the end of that period of execution, you will then see revenue recognition.

So it's not like it's one big lump sum that is going to be recognized at some point in 2016. It's an incremental sort of mosaic of projects which is being ramped up now.

So initially, yes, there's going to be an aggregate back-end loading as we go into 2016 from 2015, if you will. But once we're in 2016, you're going to see revenue recognition during the course of the year. But we can't call out the pattern of that revenue recognition in terms of quarter-to-quarter-type allocation of revenues. Does that help?

Yes, that's very helpful, and I'll hand it off. Thanks so much.

Tim Savageaux, Northland Capital Markets.

A couple questions that are sort of interrelated, which is -- I wonder if you could describe any trends that are notable amongst your smaller-carrier customers, the tier-three customer base, in terms of their overall activity levels and technology choices, and what have you. And stepping back more broadly, you've made reference to the CAF II putting a floor under your customers' spending plans. I wonder if we might infer a floor under your current revenue run rate as a baseline from that, and whether, as we look at the network project that you mentioned or overall growth potential in 2016, whether you might be able to make some comments on overall market growth rates in the space, and maybe your performance relative to that. And I'll maybe follow up with one. Thanks.

Thanks, Tim. First of all, let me address your tier-three question.

The tier-threes have been relatively consistent, and there are things that see showing up in the regional spaces, including munis and electric co-ops. They've been consistent for quite a while now. There are a set of them that are moving forward and investing, and they're consistent -- we just literally left the user group, and obviously there were many there -- and they're consistent in their investments, and they're consistently investing in fiber. And they are pretty rapidly ramping up, from a GigaCenter standpoint, their desire to own the antenna that they provide their broadband service over inside the premises for their subscribers.

So that's really consistent and hasn't changed. On CAF II, without me trying to reduce it to numbers -- you know, I just look at variability of CapEx from our customers. And to me, CAF II just puts a floor underneath the troughs. So there's certain places where it can't go below.

That doesn't put me in a position to say how high it might go. And I'll let William make comment, if he would choose. And the only other thing I would say is, I do think we are well positioned with our technologies and our products, especially with AXOS, to do well in these markets and expand globally. So that would be my answer. Anything you'd care to add, William?

No, I think you summed it up very well, Carl.

Okay. Tim, other questions?

Yes, just to follow up, if my might. I guess I'm trying to get toward some sense of overall growth expectation for 2016. These last -- the Q3 report and the guide kind of represent two poles, if you will, in terms of growth and profitability. I don't know whether it's fair to expect something between those poles in terms of a baseline. And somewhat related to that, I wonder if you can make any comments on market share, either at your largest customer or throughout that tier-two customer base, and whether you see if there's been any changes there. Thanks.

So let me answer it this way. The growth is not what we would want, but we're growing. So I think that's the easiest way to answer both of those questions. I don't know what else more I would add.

I'm certainly very encouraged by our position in our customers and our prospects for expansion. But it feels like 2016 looks pretty strong, but we'll see. William?

What I would say is -- this is, as you know, a tough business to forecast, Tim. So there's a reason, again, that we don't guide that far out.

In terms of your question about market share, we'll know when the dust settles and everybody publishes their 10-Ks in Q1, in terms of how we've done at our largest customer relative to our closest peer. We have won market share in every year, I think, at least for the past four to five years, as far as I can recall.

I think what's going to happen over time is that you may see the underlying business being done, whether it's with that particular customer or elsewhere, being quite different over time potentially. And professional services that we do is quite different from, say, some peers who, for example, do [turf] business and the like. So it's going to be harder and harder potentially to compare the underlying businesses.

So we'll know in Q1 how we're doing on market share. We really don't have any indications currently.

Okay, thank you.

George Notter, Jefferies.

I appreciate the followup. So I just want to be clear on the professional-services business that you're doing -- is this turf business, where you're doing installation-type services, or is this some other kind of consulting work? I want to make sure exactly what you're doing here. And then I inferred from your commentary that it's going to be margin-dilutive. Is that the right way to look at this?

Thanks for reiterating so I can reiterate my reiteration.

Our professional-services offerings and our professional-services team is focused on providing value-added services to deploying Calix equipment only. We do not do turf business. We will not do turf business. We have no value-add in that business. And as such, that business, we view, as being very low margin.

When it comes to our equipment, where we can provide value-add because we understand the equipment and we can help our customers and we understand their processes, it's much higher margin than turf business would ever be.

And as such, when you look at the whole project and you add in the professional services and the equipment, it might be slightly below our corporate-average margins. But if it is, it's not by any anything that moves the model.

It doesn't move the needle, George. That's the key point. Obviously, there's a professional-services element, but remember the project is substantially about delivering systems.

Got it, okay. Fair enough. And then, as I step back and look at the level of investment you guys have in the business, this year as a percentage of sales, I think your R&D is going to be up a good chunk versus recent years. Just thinking about all the new product flow you've got here, I guess it certainly makes sense. But do you envision a point where you start to get more leverage on the investment and on R&D going forward? Or do we continue to invest aggressively from an R&D perspective? I guess I'm wondering if there's something about the industry or your business that's requiring this level of investment, or is it something that can relax going forward? Thanks.

So, great question, George, and so let me put it into two timed surges.

The first surge was obviously to get ourselves onto a platform, and that platform is AXOS. As I alluded to in my comments about what we've been able to do on subsequent systems product developments that are built on AXOS, they're at a very different R&D efficiency. But as you might imagine, that's only one part of our R&D stream.

Now, having said that, we obviously intend it to become the largest part of our R&D stream, and ultimately all of our R&D stream. So as you flow that through, you're going to see increasing efficiencies.

What does that mean from an R&D standpoint? That R&D -- the growth rate in R&D is going to slow. That's for sure.

The other thing I would merely caution you on is, there could be things in our future that might be market-segment or specific-customer driven that might cause momentary R&D. But otherwise, my expectation is that the growth in R&D will slow over time. That's exactly right.

Thanks.

Mr. Dinges, there are no further questions at this time. Would you like to make any closing remarks?

Thank you, Karen. Calix will be participating in the UBS Global Technology Conference on the 18th of November in San Francisco and the Cowen Networking and Cybersecurity Summit on the 16th of December in New York.

In addition, Calix will be participating in a number of investor meetings during the fourth quarter. Information about future events will be posted on the events page in the investor-relations section of calix.com.

We remain focused on executing against opportunities ahead of us, and we look forward to meeting with you at one of these upcoming events. Once again, thank you for joining us today. Good-bye for now.

This concludes today's conference. Thank you for your participation. You may disconnect your lines at this time.