Description

I believe Calix, Inc (CALX) represents a compelling long opportunity at current levels. The story here is relatively straight-forward. The stock is currently trading near historic low valuation levels in terms of EV/S due to extremely depressed bottom line results YTD in 2017 caused by several factors that should reverse moving forward, and more recently by a negative pre-announcement from their primary competitor. While the stock does not currently appear cheap on bottom line metrics (they’re generating losses), I believe they will return to profitability in 2H18 and their significant earnings power (relative to the current valuation) will become apparent over the next 12-18 months. Trading at just 0.5x EV/Sales, the valuation is compelling for a company that has grown its top line at a respectable 9% CAGR over the past 5 years (since 2012), was consistently profitable on a full year basis from 2010-2015, and has a target operating model of double digit operating margins. As they return to profitability and show progress towards their target operating model, I would expect the valuation to move closer to its nearest comp (ADTN) at about 1.2x Sales, representing a more than doubling in the stock from here.  

What do they do?

CALX is a supplier of hardware/software to Communications Service Providers (CSPs). They are focused on the access (or last mile) piece of the telecom network. Their offerings enable CSPs to provision a wide range of service offerings to end customers through packet-based technologies over diverse types of physical networks (DSL, DOCSIS, GPON, etc). Basically CALX products allow their CSP customers to increase offerings to their end customers while reducing investment in physical networks and with lower operating costs. Their customer base is primarily telcos (as opposed to cable) in North America and they’ve historically served Tier 2/3 operators. While they have about 1300 customers, they do have some customer concentration, with CTL being their largest at 21% of sales in 2016 (and likely somewhat more than that YTD in 2017).

Historical Results

As one can see, the recent poor operating results are not at all consistent with their historical profitability levels.

Verizon opportunity

Before getting into a discussion of CALX’s recent results and my forecasts, I’ll briefly discuss what is likely the company’s biggest top line growth driver moving forward, namely VZ’s NG-PON2 rollout. By way of background, both ADTN and a CALX/Ericsson partnership were selected in mid-2016 for initial lab trials. ADTN specifically press-released their participation in mid-2016, while CALX did not. However, ADTN has referred to being one of two vendors in the trial, and CALX on their Q217 earnings call talked about beginning field trials for NG-PON2 with a ‘large North American Tier 1’. This can only refer to VZ. CALX on the same call stated “this customer can and we believe will become a greater than 10% customer. There's no question in our minds about that.”

It is believed that VZ will end up using both ADTN and CALX/Ericsson in their NG-PON2 rollout. This article discusses VZ’s goal of using multiple vendors and interoperability amongst vendors:

http://www.lightreading.com/gigabit/fttx/verizon-proves-ng-pon2-interoperability/d/d-id/729487

…while this blog indicates a belief that CALX/Ericsson will be the “primary beneficiary” of VZ’s NG-PON2 (albeit they speculate at poor margins):

https://www.fibereality.com/blog/ericssoncalix-de-facto-ng-pon2-model

The timing of the ramp is certainly still an open question. ADTN stated recently that they don’t see a ramp from VZ before 2H18, though VZ (speaking at CALX’s annual user conference in October ‘17) said the following:

"If everything goes correctly... by the end of the year it should be in a state that's a deployable state from Verizon's perspective," said Verizon Director of Technology Vincent O'Byrne, speaking at the 2017 Calix Connexions Innovation and User Conference here. "[We will then] start deploying probably in the first quarter of next year."

http://www.lightreading.com/gigabit/fttx/verizon-preps-ng-pon2-rollout-for-q1/d/d-id/737807

In summary, I assume that CALX is involved here and that VZ likely becomes very material to revenues over time. As discussed more below, I assume some upward inflection in growth in 2H18 due to VZ.

Recent results/model

While CALX’s recent top line growth has been strong at 14% YTD in 2017 (relative to its 5-year CAGR of 9% since 2012), this has been entirely driven by significant growth in its recently negative GM services business, which has been the primary factor driving bottom line losses recently. Focusing specifically on the services business, a couple factors have been responsible for significantly dragging down GMs here, namely their taking on (or perhaps being forced by large customers to take on) significant structurally low margin ‘development services’ business, poorly priced contracts (with poorly defined project scopes) signed in 2016 that largely have been recognized in 2017, and poor internal management/cost controls. 2017 YTD service GMs are -17% and were -27% in Q317. There are a number of reasons to believe this drag reverses moving forward:

• They have guided service GMs to breakeven in Q417 and to reach 30%-40% over the course of 2018.
• Historically, they have done 30%+ service GMs as recently as Q415 and 20%+ in 1H16 before the significant ramp in low-margin, poorly priced/scoped projects signed in 2016.
• They brought on a new Head of Services in December 2016, who previously held a similar role at both Motorola Solutions and Avaya, under whom new projects signed have had positive GMs according to management due to better defined project scopes and internal cost controls. These are still significantly being masked overall by the bad projects.
• They’ve stated that many of these bad projects that have revenue’d this year have year-end deadlines for completion.
• Moving forward they have stated they plan to outsource more of the low-margin development services work to partners, while focusing on higher value-add consulting services engagements.

The other significant factor driving losses recently has been a big ramp in R&D spending, as they’ve invested significantly in 1) their software-defined networking (SDN) platform called AXOS which they believe will be accretive to product GMs over time, and 2) investment in projects for new customers (I believe significantly related to VZ). Non-GAAP R&D expense increased nearly $10M from $21.7M in Q116 to $32.5M in Q117 (a 50% Y/Y increase), but has trended down slightly since then and has been guided to continue to trend down in $ terms moving forward as the incremental spend for investment in the AXOS platform is largely finished at this point.

Finally, I’d note that product revenues will likely grow sub-1% in 2017, substantially lower than in prior years. This also likely has been a drag on sentiment recently. I think this re-accelerates in 2018. While the company did not historically break out product revenue versus service revenue, the disclosures in 2017 included 2016 splits and indicate that product revenues were well over 90% of their business through Q316 (but had fallen to 83% of revenues in Q317 due to the 200%+ growth in services revenues). I think the 5 year revenue CAGR of 9% is generally representative of their product revenue growth rate over that period, given how immaterial services revenue had been before 2017. I believe the biggest reason for this slowdown in 2017 is a tough Y/Y comp (when revenues grew 13% in 2016), while management also attributes it somewhat to a shift in spending by their customers towards services. I find the latter explanation a bit questionable to be honest. Nevertheless, I believe that product revenue growth should move back towards the historical trend for a couple reasons:

• Lapping the tougher comps from 2016
• Much of their growth in services revenue has been in development services (as noted above) for CAF (Connect America Fund) II projects (https://www.fcc.gov/connect-america-fund-phase-ii-auction ). CAF is a program that provides carriers federal funding for providing broadband service to higher cost rural areas. As the name ‘development’ services would suggest, these service engagements are done early in the deployment cycle and are likely a leading indicator for product sales, as CAF II projects move from the development stage to deployment. Carriers first were awarded funding under CAF II in 2015.
• Ramp up of NG-PON2 business with VZ beginning in 2018

In the model above, the key assumptions I’ve made for 2018 are:

• Product revenue returns to growth in 2018, particularly in 2H18 as they begin to see some meaningful contribution from VZ.
• Service revenues decline somewhat as I assume they outsource some development services to partners.
• Product GMs stay at 48%, inline with CY16 and Q317, but below CY15 levels and their target model of 50%+ overall GMs. I assume some margin benefit from their AXOS SDN platform, but also assume it is offset by likely lower than corporate average margins with VZ. Net/net my product GMs could be conservative given I model them at a $600M+ annualized revenue run-rate in Q418, the level at which they target overall GMs of 50%+.
• Service GMs turn positive, but only reach 25% by YE18 (versus company guidance for returning to 30%-40% service GMs over the course of the year). This is purely conservatism as they obviously have shown awful (much worse than expected) GMs this year, so I choose to be significantly below management here.
• Total OpEx approaches their 40% of revenue target model by Q418. Guidance for Q417 actually already implies sub-42% of revenues, so with more revenue scale and the dialing back of R&D, I’m comfortable that they can hit 40% before too long.
• Minimal taxes: these are just foreign taxes. They have a $240M deferred tax asset from NOLs as of YE16, so they will not be a domestic taxpayer for a very long time. I’d note that I expect they will mark down the value of this off-balance sheet deferred tax asset due to the lowering of the corporate tax rate, but that doesn’t change the fact that they will not be a cash taxpayer domestically for a very long time.

Under these assumptions, their operating margins approach 5% in Q418 (and annualized EPS approaches $0.50), putting them well on their way towards their target 10%+ operating model ($1+ in EPS) with continued revenue growth.

Current opportunity

In addition to the factors discussed above that have dinged profitability (and the stock) that I believe will reverse moving forward, most recently competitor ADTN pre-announced a big miss on revenues in Q417. On 12/28/17 after the close, they announced that revenue would come in at $125M versus prior guidance of ~$160M, a miss of over 20%. (on a side note, a nice gift to their institutional shareholders on the last trading day of the year!)

They blamed merger-integration related spending delays at a ‘domestic tier 1 customer’. This customer is CTL, who acquired LVLT during the quarter. In addition the being CALX’s largest customer, I believe CTL is also ADTN’s largest customer at 24% of 2016 revenues. ADTN’s stock has predictably sold off (~11%) since the announcement. CALX’s stock has also been hit – it’s currently down over 6% since ADTN’s pre-announcement (after selling off as much as 12%). This is despite CALX announcing their earnings date this past week without themselves pre-announcing. My discussions with the sell-side in the immediate aftermath of ADTN’s pre-announcement led me to believe that it was one particular program at CTL where ADTN was the sole supplier that was the culprit. That said, given the magnitude of the miss, I believe ADTN must have seen weakness elsewhere beyond CTL given the magnitude of the miss (unless CTL essentially went to zero for them, which seems unlikely). We’ll know more about ADTN when they report this coming Wednesday, Jan 17. But the fact that CALX did not pre-announce only strengthened my view that this was ADTN-specific. While I can’t rule out that some CTL-related weakness could show up for CALX, the stock has already been marked down, which I believe is likely presenting an opportunity here.

Valuation

Given that CALX has recently been unprofitable and I expect it to stay so for a few more quarters, the most tangible valuation metric we currently have to look at is EV/Sales. CALX currently trades at ~0.5x Sales, which is very low relative to:

• It’s historical mean EV/S of 0.9x
• Its nearest comp ADTN at 1.2x (similar revenue scale as CALX)
• Its target operating model of 10%+ EBIT margins
• Its all-time low valuation of ~0.4x

Insider ownership/buying

• CEO Carl Russo has significant equity ownership at 13% of S/O, while officers and directors in aggregate (including Russo) own 16.5% (both as of 2017 proxy statement date in March 2017)
• The CFO bought $75k of stock at the end of November at an average price of $6.84 (15% higher than where it is today). Another data point that suggests they likely didn’t see CTL-related weakness developing in the quarter.

Risks

• CTL merger-related weakness
• VZ opportunity doesn’t develop as expected or is at awful GMs (and/or OpEx needs to ramp substantially to support a revenue ramp there)
• Service margins stay very low/zero/negative
• Currently net cash is healthy at $41M, but continued large operating losses could start to eat away at that
• I model revenues below consensus in 2018. I believe this is almost certainly due to the street assuming higher service revenues than I do. I don’t view this as all that important as I believe investors will focus more on product revenue growth and my EPS numbers are also well ahead of consensus.

Catalyst

• Return to product revenue growth (esp ramp at VZ)
• Return to profitability
• Progress towards target operating model