Intro

Casa systems is a company that was created in 2003 and launched its first commercial product in 2005.  It was founded by Jerry Guo, the current CEO, who was a research scientist at Bell Labs and holds a Phd in electrical engineering from the University of Wisconsin and an MSc from Tsinghua University (a leading Chinese university). Jerry and his family own ~19% of the company, with John Malone’s Liberty Ventures owning 5.3% and Summit Partners 41%.

Casa Systems’ primary business is designing network equipment such as CCAPs and network nodes that are purchased by cable companies. This equipment supports the transmission of data, video and voice from cable company “head-ends” to business and residential customers.  This equipment is effectively the plumbing of the cable industry. Notably these systems are becoming increasingly important to deploy for MSOs, as they compete with telephone carriers looking to supplant MSOs with 5G wireless technologies.

Not Pretty (in fact quite ugly)

Now with that said, we recognize this is a somewhat controversial long recommendation as the company has done a poor job of managing the Street.  Since their IPO in December 2017, the company has missed guidance two-times and are in the penalty box as it relates to managing investor expectations.

They completed an IPO in December 2017 and put out fairly bullish guidance for 2018.  They then hit the road, reaffirmed guidance several times and made a lot of noise about how solid their end-markets look at investor conferences. Furthermore they often discussed a large and imminent opportunity associated with re-architecting their products for wireless networks that would see their TAM expand by ten-fold to +$20 billion annually by 2021.

Sounds good, right?   The market agreed and ran the stock up to $33 per share by April 2018.  The company then chose to take advantage of this run-up and sell 7.35mm shares at $25.00, with executives and sponsors selling some shares.  Despite being handled poorly by the underwriters, the secondary itself wasn’t problematic. It’s just that the company’s visibility into its end-markets began to decline soon after, and the company failed to communicate this to investors.

With hindsight, the optics on the secondary transaction looked a lot worse than the reality.  In fact the company’s visibility on sales is truly minimal and may be as long as 10-12 days or as short as a day in the case of software capacity expansions.  As a result they generally have little visibility into any particular quarter or year. It appears their conviction on annual guidance was solely driven by historical buying patterns in prior years and expected increases in the rate of broadband capacity demand growth with MSOs. Unfortunately for Casa, MSOs gave only verbal commitments on how much capacity they may need and the reality turned out to be less.   The company rarely discussed this dynamic in investor meetings and only spoke about their confidence with their customers’ historical buying patterns. Their omission of this dynamic is certainly a mark against them, however they seemed to genuinely believe the demand patterns of prior years would hold.

By late June 2018 it was clear to the company they were going to miss their original guidance of $380mm-$395mm of revenues and $1.08 to $1.19 of EPS for CY18.  As a result the CFO was fired in early July and they then reported an inscrutably bad earnings report in early August which came in far below the street’s already tempered expectations.  Their only explanation was that MSOs are delaying spend and running their networks much hotter than they have in the past.  This resulted in less need for new capacity purchases. The new guidance called for $330-350mm in revenue and $0.80-$0.88 in EPS.

Notably, at this time in August they characterized the revised guidance as so ridiculously low that it would easily be met and potentially exceeded if certain things lined up. They then re-affirmed this lowered guidance on the Q3 call in November, posting solid numbers that suggested they would likely come in at the very high-end of the reduced guidance range. The company then went on a road-show of conferences that extended into December 2018, expressing confidence in the guidance, the business and the long-term opportunity for the company.

But on January 17th of this year the company announced preliminary results that missed their new lowered guidance from August. This was disappointing to say the least. For FY18 they now expect $292-298mm in revenue, a higher gross margin of about 70%-73%  (was high 60s to low 70s), adjusted EBITDA of $94mm-98mm (Street was around $100mm) and adj EPS of about $.082-$0.87. While there wasn’t a material downgrade in EBITDA or EPS, the revenue outlook was materially below what they had guided to -- with the only silver lining being higher margins.  The new guide suggested an over 40% decline in Q4 yoy revenues. Furthermore, the better margin guidance implies a decline of 65% in hardware revenues YOY and a 35% decline in software. This was nothing less than an epic blunder, especially considering that Casa sells to a somewhat concentrated base of corporate customers who generally have decent visibility on their purchasing.   So at this point we are working on the assumption that Casa minimal visibility in into their revenues and we would recommend others do the same.

When we spoke to the company, they suggested it was broader weakness in the MSO market that led to last minute orders not being finalized. According to the company, some of these were literally expected during the 2H of December 2018 and seemed highly likely to come through. They also repeated their view that a delay in a new type of MSO capacity called DAA (Distributed Access Architecture) was a partially to blame.  Their reasoning is that conventional capacity (CCAP + standard node) would be somewhat redundant in the coming months/years as MSOs switch over to a DAA network architecture. This seems accurate based on our checks, however why it caught Casa so off guard is odd, as it became clear from our own work during Q3 2018 that DAA would not really pick up until 2h19. In many cases we gleaned these insights from the same trade-shows and conferences where Casa was an active participant.

Fundamentals Going forward

Our discussions with MSOs was largely in-line with what Casa management communicated. This gave us some comfort that the company wasn’t being dishonest viz-a-viz guidance, but simply lost whatever visibility they thought they had with their customers.  While MSOs had expected (and budgeted) for a base amount of capacity spend in Q4 2018, there was a notable deceleration in broadband data throughput in late Q3 and more so in Q4 that obviated the need to buy more hardware such as CCAP units. In some cases these MSOs expected a sharp uptick in capacity demand during December that never materialized. Many found this to be odd, as Q4 tends to be the biggest broadband data growth period.

Another dynamic is that CCAP hardware purchased in the 1h of 2018 is shipped with latent capacity in the form of inactive line-cards and line-card slots.  So as demand in a particular area picks up, MSOs will first use inactive line-cards (by buying a software activation key). They will then subsequently purchase new line cards to put into an existing CCAP, which generates small amounts of hardware revenue but again more software revenue.  This was in lieu of the expected hardware CCAP purchases that would be required if broadband capacity demand met expectations. As demand growth softened in Q4, there was clearly some mix shift towards software as there wasn’t a need to add as many CCAP units across the MSO footprint. This appears to be the reason Casa’s Q4 margin guidance is coming in well ahead of expectations, but revenue was much softer than guidance.

It is worth pointing out that MSOs don’t have good explanations for this broadband data deceleration, but they were expecting yoy growth of 40-45% like it had been in prior years. Notably, according to CableLabs (an industry trade group), they estimate 2018 downstream broadband throughput growth was about 33-35%, with the first half of the year being much stronger than second half of 2018.  This was a fairly large deceleration in demand and appears pretty far off the mark from what Casa and it’s MSO customers had expected. MSOs cite the following potential reasons for this slowdown in growth: 1) stagnating engagement on social media platforms (# of users and hours spent per user flat), 2) Compression Technology, 3) decelerating user growth for OTT video platforms and more importantly hours spent watching OTT videos, 4) mix shift of OTT video consumption to SD and other small screen formats (phones, tablets and computers) instead of televisions and other HD display formats.

Through conversations with executives at Netflix, Youtube, Google and Apple we believe the reasons cited above may have likely played a role in this deceleration. However, most believed this is simply a temporary slowdown and they expect broadband data demand growth to accelerate again eventually.   They noted over the last decade there were several periods where broadband data demand growth slowed but eventually reaccelerated, as demand has many variables that can change quickly causing this ebb and flow.  Notably, executives at Netflix and Youtube confirmed the deployments of new iterations for VP9 and H265 (compression technologies) in the 2h18.  The expected impact was an increase in efficiency gains of at least 20% over large swaths of what’s consumed on their platforms in the US. They also confirmed that there was significant increases of OTT video mix shift to smaller screens and other SD oriented platforms which could also have an impact as some MSO executives had suspected.

At a high level Google and Netflix executives held the view that broadband data demand would grow by up to 10x over next 3-5 years, driven by things we have heard a lot about lately.  These include growing AR/VR platforms, growth cloud-based services, booming OTT media consumption and just the general inexorable demand for more data intensive services for businesses and homes. Also, on a recent earnings call the CEO of Charter spent a lot of time talking about where they’re growing their capex budget dollars; broadband data capacity. Charter is Casa’s largest customer and we believe and acceleration of spend at Charter on this type of equipment will be good for Casa.  Furthermore we believe the market fundamentals over the medium term are solid for Casa, despite the near terms headwinds and diminished visibility.

So why own the stock?

We think that despite the poor visibility picture in 2019, the MSO capacity market will stabilize and eventually grow modestly beginning in 2020. More importantly, the shift to DAA architectures will likely be somewhat additive to Casa’s current cable equipment TAM and become a modest tailwind towards the end of 2019.  Our conviction level on a longer-term stabilization (and eventually moderate growth) for this part of Casa’s business is fairly high for a couple of reasons.

First, the market share dynamics are fairly sticky, with Casa actually being the largest market share gainer over the last 10 years. This is because they’ve consistently unveiled leading-edge technology every few years while remaining price-competitive, allowing them to poach clients from Cisco and Arris.  We think Casa’s technological lead will continue based on our feedback from large MSOs. They believe Casa has always been 1.5-2.5 years ahead of other players and this will remain the case, as it’s a very R&D focused company that takes a software-centric approach to design, unlike Cisco and Arris that are hardware centric. Furthermore, the demand for broadband data capacity will continue to grow long term and this creates a basic level of long-term visibility around the overall market for cable and broadband equipment.  It’s also worth pointing out that MSOs like Comcast and Charter know that 5G fixed wireless into homes is an existential threat. As a result they view broadband connectivity as their primary future business and their ability to increase data speeds as the main way to compete. This bodes well for Casa longer term.

Finally, we believe Casa’s ability to benefit from selling wireless equipment, especially as 5G build-outs occur, is not reflected in the stock.  The company has developed and deployed EPC edge-solutions as well as small-cell access points with leading global wireless carriers, who are in the process of certifying the products. Once certification is complete Casa can begin to book revenue from wireless products; some of this will be recurring as is typical of network core solutions while small-cell deployments will be one-time sales.  Today these customers include Sprint, Telefonica and China Mobile, but we also expect to hear from AT&T and Verizon sometime during 2019 based on our conversations with executives at these carriers.

While we are not fully confident on management’s estimates of a $20B annual TAM for 5G wireless equipment, we believe that if it turns out to be half that and Casa is able to garner a 1% market share over the next 20-30 months, this would result in $100-120mm of incremental annual revenue.  We think the market would place a higher value on Casa’s wireless business than its core MSO segment due to the larger TAM, recurring nature of the software sales, fragmented customer base and ability to serve more geographies (Asia, Africa and Latin America). Furthermore, the percentage of capex spend that Casa’s wireless product represent for carriers relative to MSO spend on CCAPs is much smaller and so we think this will likely support a better margin and growth profile for the wireless business.

Valuation:

The company is clearly in the penalty box from a valuation stand-point and therefore at current levels we believe the stock is attractive.  The company is trading at less than 8.5x 2019e EBITDA of about $100mm, net of the cash on the balance sheet. This a discount to the approximate price CommScope paid for Arris (8.7x EBITDA), despite Arris deriving nearly 60% of their revenue from a secularly declining CPE business (set-top boxes) that “directly contributes” only 8% of profits as defined by the company.   In contrast, Casa is likely sitting on a meaningfully large and nascent opportunity in wireless that investors really aren’t paying much for now.

We believe Casa will likely book $300mm-$320mm in revenue and $100mm in EBITDA in 2019, with $20-30mm of wireless revenue.  In 2020, we think wireless revenues will be $50mm-$60mm and total revenue will be $395mm-$405mm, with EBITDA around $155mm-$165mm. The growth in 2020 will be driven by wireless gains as well as a more DAA architecture investments by MSOs, as they begin to compete more aggressively with wireless carriers deploying 5G standards.  The company generated approximately $153mm in EBITDA in 2017, so we don’t view our 2020 target as particularly ambitious. Moreover, with Summit Partners and Liberty Ventures as large holders we think there is a good chance the company will be taken out by a larger equipment maker or new sponsor as they begin to show progress on wireless revenues and the core MSO market stabilizes.

The company is FCF positive and we believe it could generate about $60-70mm in 2019 and about $100-120mm in 2020, so the balance sheet is stable and we think the FCF will make it attractive to buyers. We also believe the market will ascribe an 8.5x-10.5x EBITDA multiple on $155mm-$165mm in 2020 EBITDA. This would put the stock around $16-$22, before adding about $2 of cash per share that should build through 2020 (assuming no buy-backs). This gets us to about $18-$24 per share, which would generate a 60%-110% return over the next 18-24 months.