Description

Ubiquiti Networks (UI-US)

Summary / Thesis Overview

UI is interesting. I have a lot of respect for what Robert Pera has built and accomplished. It’s truly admirable. Also, I’m a happy customer myself and they really do make good products at cheap prices. But their explosive growth over the last few years looks like its about to hit some pretty bumpy road and the premium valuation they trade at doesn’t seem prepared for it either. Further, while UI has been written up before (good write-up on VIC, check it out), with a comparable thesis, the problem back then (in 2014) was that there was truly a lot of juice left to squeeze out of its unknown market potential. It’s a lot harder to say that today at the size they are currently at. Plus, new competition is coming to the fore. Finally, I like that I think buyout-at-a-premium risk is actually fairly low, given Pera owns 90% of this already, doesn’t deal with investors in any capacity already, and is share-buyback-ing his way to 100% ownership without having to pay a premium to anyone. 

Company Background - Understanding the Business

UI was founded by Robert Pera in 2005. Pera today might be more well-known for his ownership of the Memphis Grizzlies in the NBA (or his bizarre antics in this role, including challenging Michael Jordan and Tony Allen to 1-on-1s - see here: https://sports.yahoo.com/grizzlies-owner-robert-pera-challenges-michael-jordan-to--1-million-game-of-1-on-1-for-charity-214032512.html?y20=1). Regardless, Pera was only able to buy an NBA team because of the success of UI (formerly under the ticker UBNT). The short backstory is that Pera got a job at AAPL out of college as an engineer working on AAPL’s Wi-Fi routers (yes, AAPL makes Wi-Fi routers, called “AirPort”; obviously they aren’t a focal point of the business). While there, he realized that the Wi-Fi access points were underpowering the antenna strength relative to maximum allowed levels under the FCC’s laws limiting electromagnetic emissions (leaving a chunk of signal strength on the table). AAPL didn’t feel that putting more development into making better Wi-Fi routers was worth their time, so Pera left and founded Ubiquiti - at age 27 - to pursue this route.

UI’s MO has always been making better equipment for cheap. It’s a very engineering-driven culture, if not effectively Pera’s personal science-project sandbox. With this approach, and with the help of a very unique business model in this area of the world, Pera has built a billion-dollar-plus rev business, valued at >$10B, and has become exceedingly rich, especially quickly and especially young. Genuine kudos to him, its truly remarkable. 

In terms of the actual business, UI develops networking and other related equipment. Originally, the core business was wireless networking equipment specifically aimed at wireless internet service providers (WISPs). 

[Sidebar: WISPs provide internet service via a fully-wireless infrastructure, something far more common to rural areas underserved by wireline operators. WISPs exist all over, but are far more common in emerging markets vs. the US. Another way of understanding it, in light of recent hysteria around 5G and “fixed wireless”, is that WISPs provided the “original” fixed wireless. The idea is that the WISP procures broadband access from a fiber provider (probably via a datacenter or tower connection) and then builds out infrastructure to relay that wired fiber connection into a residential area (often rural and thus underserved by traditional wireline operators). The infrastructure is similar to what mobile operators use – base stations (to broadcast the signal), microwave backhaul (to connect the base stations to the original source point wireless), and customer premise equipment (to receive the signal). Everything is done wirelessly. Downside is typically a weaker or slower connection vs. what’s possible with a wired connection directly to the home, so WISPs don’t really compare to cable/fiber-based wired broadband.]

Beyond the SP-focused original business, UI has built out a broader range of products as well, including cloud-controlled / “controller-less” enterprise Wi-Fi (WLAN) equipment (access points), consumer-focused WLAN equipment, networking switches and edge routers, and IP video / surveillance equipment. A kind way of viewing UI would be to say NTGR’s product list, but at CSCO’s quality, aimed at something above consumers. At this point, the “enterprise” business (which is everything that isn’t service provider-focused) comprises 2/3 of rev.

As noted, UI has a unique business model and culture (which has pros and cons). UI spends minimally on sales & marketing, with no formal channel program or direct salesforce. Go-to-mkt is largely made up of distributors and web sales, and web-based marketing. Customer support is purely message board / community-based. So what UI gives up on GM% by making an enterprise-grade product at a fraction of competitor prices, they make up in having bare-bones operating infrastructure relative to other enterprise IT outfits. 

UI’s “special sauce” is its relationship with its highly-engaged community - UI has a large community of fanboys, with whom employees stay highly engaged via UI’s message boards, which makes it all work. As noted, support is handled at a community level, where questions / problems are asked and solved/answered by “the community”. Further, product/feature requests via the community are often quickly and efficiently incorporated, which the community loves and is key in keeping the ecosystem happy and engaged. A company that actually very literally listens to, and enjoys engaging with, its customers, constantly iterating products based on that engagement. Unsurprisingly, given the culture, R&D is fairly decentralized, with significant autonomy among employees/teams to drive product invention or evolution. Add it all up, the culture, product quality, and aggressive price point allow the company to drive sales without a traditional enterprise sales motion.

To recap, the unique model positives are largely: 1) highly-efficient sales motion - almost by-definition UI runs a self-serve model, which is effectively unheard of / unique in such a sophisticated area; 2) highly-sticky customer base – customers typically are fanboys, vs. being on the receiving end of whatever vendor is being pushed by a given VAR; 3) minimal sales expenses and support infrastructure allow for significant underpricing vs. competitors for similar-quality equipment. The GM% hit is more than made up for on opex savings and UI enjoys EBIT% well ahead of peers.

Of course, as is fairly integral to the short thesis here, the “collateral damage” negatives are primarily: 1) inability to crack real enterprise accounts - to get in the door in vast majority of accounts, need to go via VARs, who often are the gatekeepers, regardless of product quality; 2) lack of a formal support function renders them unusable in most of-size accounts, regardless of product quality; 3) high inventory / inefficient distribution - UI has carried high inventory levels for a long time. Without a sophisticated supply chain (which includes GTM), UI is in a constant mismatch on product availability (too much inventory of certain products, not enough of others)

Recent History / Setting the Context

A big part of UI’s DNA is moving into adjacent areas that are interesting to the company and the community. Technically, the entire enterprise business was an adjacent move, given the business really started out focused on WISP equipment. As enterprise grew, the enterprise business itself expanded beyond Wi-Fi to switches and edge routers. More recently, UI has moved into more sophisticated video equipment (networked IP cameras, etc.). 

At the same time, UI has had a convenient uptick in their enterprise business relative to the timing of new Wi-Fi standard introductions (which bring new product cycles with it) and the melt-down at long-time number-3 player Ruckus Wireless, who went from being public to being owned by Brocade, to being owned by Arris (who bought BRCD), to being owned by Commscope (who bought ARRS), all in the span of 30 months. Multiple acquisitions can do that. Additionally, a new entrant has come directly into UI’s territory (with both WISP and enterprise Wi-Fi equipment) - CMBM - all while 5G is on the cusp of enabling cellular-based fixed wireless. 

In other words, the UI short case has always been some version of there is only so much juice they can squeeze out of the way they do business. Time and again, Pera has been able to find some way to squeeze a bit more juice than others expected, and I give him all the credit in the world for that. But there are now more credible threats coming into UI’s turf, and recent growth has been meaningfully aided by temporary factors, which sets up the business to slow down from recent trends, with a valuation that won’t be forgiving in such a scenario. 

Pera takes no compensation, period (no salary, no stock grants, zippo) and stopped doing earnings call altogether nearly 2 years ago and there is essentially no real IR function. So there is no real investor communication of any real kind beyond SEC-mandated filings/requirements. Also, given UI’s strong profitability, but no real M&A angle (given how unique the company is), UI has instituted a dividend and has been buying back shares (good for them, seriously), so Pera now owns ~90% of the company (which is why he doesn’t get stock grants - they’re meaningless). 

Primary Thesis Points

1) Recent gains largely based on temporary tailwinds; hard to see real structural angle moving forward

It’s hard to tell exactly who they are gaining share from, which makes me wonder about the quality of their share gains in general. UI’s enterprise growth rates have been far above their peers, but their peers are still growing at solid rates (i.e., unclear that major peers are “losing” customers” vs. UI finding “non-traditionally-defined enterprise customers”). In other words, if they aren’t taking customers from CSCO/HPE or other “Enterprise” vendors, then they aren’t really gaining share in the traditional TAM, which makes further potential unclear (they could already be majority of their “real” TAM). To the extent they’ve gain significant share from a competitor, its mostly been at the expense of RKUS, who was bought and sold 3x in the last few years, which has caused a lot of disruption to their business (i.e., UI’s share gain coming from a specific situation, vs. as part of a larger trend).

Further, the enterprise WLAN market had a great run from the end of 2018 through most of 2019, as the new Wi-Fi standard (AX, aka Wi-Fi 6; previous standards being A, B, G, N, AC, and now AX, i.e., the 6th standard) has brought a product refresh across the ecosystem. During UI’s FY19 (ending June), the enterprise WLAN market accelerated to ~8% growth from LSD/MSD previously. The market slowed back down towards the end of 2019, and so did UI’s enterprise revenue. Plus, COMM announced a stronger-than-expected close to the year for RKUS. A stabilization from that vendor could make the recent trends of outsized growth a lot harder to come by for UI. 

As is typically noted with UI, but I will note as well regardless, some (large) % of TAM is totally unavailable to them, regardless of product quality. As noted above, gatekeepers are usually VARs/SIs/MSPs, with businesses outsourcing a large % of IT decision-making to those parties; while there are MSPs/SIs that love UI, anyone set up as a traditional channel partner is simply not getting paid to sell UI. Worse still, with or without the VAR hurdle, the lack of formalized support is a dealbreaker; some MSPs will take that risk since they are the choke point anyway (i.e., the customers call the MSP if there is a problem), but it’s far from the norm. This is a fact of life in enterprise IT - support is a non-negotiable point. In fact, the entire RHT business model was effectively adding formal customer support to cheap, but effective unsupported software; point being, being cheap and effective is worth less than being formally supported. 

Further, even assuming “better quality”, there are many angles to selling enterprise IT that are hard to get around. Most orgs buy WLAN with campus switches, so every campus switch provider also provides WLAN equipment. Some campus switch providers also provide datacenter switches. Some datacenter switch providers also provide edge or core routers. Some firewall providers also provide basic campus switches, some with WLAN equipment as well (or partner for it). Bottom line being, where is the decision point being made? At the DC switch level? At the campus switch level? At the firewall level? Router? What % of WLAN sales are ex-bundle? No one knows, but the answers probably don’t go in the right direction for UI penetrating “real” enterprise.

And of course, there are extreme levels of competition in this market with minimal differentiation. While this benefits UI from a low-cost-provider perspective, it makes it difficult to win real enterprise accounts with all the aforementioned necessities. Some vendors get a better rep than others on the hardware side, but its largely hard to differentiate. RKUS had real, differentiated antenna tech; look how that turned out. Plus, the tech these days is largely standardized anyway (wireless standards bodies, etc.). Sure, UI has built incrementally better versions of things, but its all based on non-owned tech (e.g., using QCOM’s Wi-Fi chips but layering on their own firmware). This isn’t to take away from their IP, which they obviously have, but they are also dependent on others IP; its not ground-up proprietary, so the improvements are often not game-changing enough.

Going further down this road, UI’s controller-less approach was once novel, but no longer. While UI speaks often about its cloud-based, uniform platform for managing equipment (this is referred to as “controller-less” or “controller in the cloud”), HIVE (since acq by EXTR after several uneventful years public) was a major innovator here and went nowhere because everyone copied the tech and provided choice of deployment (usually a physical controller is need to control for quality at a certain level of scale regardless). The industry has largely gotten on board with this deployment methodology, so its lost its unique flair. 

Point is, its hard to really see an angle to UI beyond “its cheap”, and enterprise IT decisions, while certainly sensitive to cost in many ways, are more sensitive to many other things where UI comes up very short, and more so than in previous years.

The service provider side of the business has real risks

2) New competition being brought to UI

In addition to the competitive elements noted above - which aren’t new - there is the entry of CMBM into the market. Their IPO sets them up better from a capitalization standpoint and presents a real threat to the SP side of the business that they haven’t really had before. Further, increased broadband buildouts globally, as well as 5G fixed wireless no longer being the sole domain of WISPs, present further problems for the SP side of the business. Effectively, any broadband provider expanding into an area currently covered by UI’s customers’ customers becomes a problem (every new passed home by a telco/cable company = bad for UI), while 5G fixed wireless could potentially kill a decent portion of the WISP market. Further, as is often noted with WISPs, no one knows the size of the WISP markets – it’s a huge unknown. While this could’ve gone either way 5 years ago (i.e. maybe its huge, maybe its tiny, we don’t know), UI’s SP rev has peaked at ~$460M annually and has had a hard time eclipsing that level. It seems that this is the peak for them and it doesn’t look like this area of the market is improving for them. Bottom line – WISP equipment market is more likely to shrink, not grow

3) Minimal risk of large-premium buyout to crush the short

Pera owns 90% of the company, holds no conference calls and doesn’t talk or deal with investors. Buybacks continue to increase his ownership. While the bar is very low to remain public, there’s a non-zero chance UI gets delisted based on minimum ownership requirements. 

More importantly, I’m not sure that a take-private by Pera at a premium makes sense. What does Pera not have now that he will gain by going private. Like I said - he already does nothing with investors and has full control. One could even argue there is a personal value to him of having a public currency (he borrows against his stock, so that’s at least one). Based on recent history alone, he seems more likely to squeeze out the remaining shareholder base with regular buybacks, as opposed to tendering at a premium. I can’t imagine why he’d feel the need to pay the other 10% of shareholders 25% above current prices to get rid of them. What’s the point?

It’s totally unclear what Pera’s motivations are at this point. Sure, its all his wealth, but if he just collects his dividend for the rest of time that would be fine too (close to ~$100M/annually just from the div). It’s hard to tell if he is happy with his $10B sandbox or if he wants to build something even greater/bigger – my gut tells me the former, not the latter. Obviously, he’s very into his Grizzlies ownership; while it may not eat into his time significantly or meaningfully, it doesn’t make me feel better about his long-term ambitions for UI from here

CEO/Founder Pera eventually takes the company private

Pera owns ~87% of the company

No earnings calls, presentations, or formal IR function

FCF has aggressively gone toward buybacks (which has supercharged Pera’s ownership despite not receiving any compensation – equity or otherwise – in ~7 years)

Key Risks / Negatives /  Bear Case

1) UI keeps doing what they’ve done and defies all the odds

More juice is found to squeeze and real DD growth continues with strong incremental margins. 

-- Even if that happens - which the trends seem to point that things get harder, not easier - it doesn’t mean that the valuation is all that justified up here in the nosebleeds. I don’t short on valuation, but I’m less worried about the downside of the short (i.e. upside of the stock) given I don’t see much of a multiple expansion angle, and much more likely multiple compression even with a level of continued success.. 

2) Pera takes UI private

Obviously, at a premium

-- As discussed above, not sure that he’d even care to do this, and I can’t imagine he’d care enough to pay a real premium.

Primary Catalysts

1) Slowdown in enterprise segment growth to Low-DD or single-digit-or-worse growth

2) Contraction of service provider segment

Secondary Catalysts / Accelerators

1) De-listing

Valuation Thoughts

UI trades on next FY’s rev/EBITDA/FCF at 7x/19x/27x. For a company with tremendous market growth, market penetration and market share potential, this makes sense, along with high-teens-plus topline growth and rapidly expanding margins. Conversely, basically everyone that is comparable to UI trades at more-or-less half of the above multiples (3-5x on rev, 8-12x on EBITDA, 10-15x on FCF). As I often say, everyone can choose their own multiple, but directionally, if things slow down, the valuation should compound things in that direction, whereas if UI continues to do decently well, it may contract anyway given how rich the valuation is today for what they do and the realities of the markets they play in. 

Catalyst

Primary Catalysts

1) Slowdown in enterprise segment growth to Low-DD or single-digit-or-worse growth

2) Contraction of service provider segment

Secondary Catalysts / Accelerators

1) De-listing