Description

Long:  Ceragon Networks (CRNT)

Target Price = $5.00 (35% potential return)

Summary

CRNT is a specialty component vendor for global wireless network operators that is poised to benefit from the build-out of radio access networks for 5G+.  CRNT specializes in wireless backhaul solutions that will be increasingly needed as cellular networks densify and data throughput increases rapidly.  In the 3G and 4G cycle, CRNT reached ~ $450 mm in revenue, but I believe the 5G cycle will be stronger and longer for wireless hauling, particularly fronthaul.  Furthermore, CRNT management has learned valuable profitability lessons from the last cycle, implying much better incrementals for 5G revenue growth.  Conservatively assuming that CRNT only reaches revenue run-rates achieved for 3G and 4G, combined with much better cash flow conversion, free cash flow should be robust at the top of the 5G cycle.  Taking only half of this peak-cycle free cash flow (i.e. an avg of peak and trough of zero) implies fair value of $4.40 per share.  Assuming $500 mm of peak cycle revenue takes fair value up to $5.80 per share.

About

CRNT is a small-cap ($333 mm mkt cap) U.S.-listed company domiciled in Israel.  CRNT specializes in microwave and millimeter wave radio frequency transportation in cellular network and wireless internet applications. The company sells radio and antenna equipment to cellular network operators and wireless internet service providers to manage the transportation of data across a radio access network.  CRNT is a “wireless hauling” specialist.  Wireless hauling consists of the connections between various parts of the radio access network (mid-haul and front-haul) and with the core network (wireless backhauling).  The core network is the central operational node that provides the connection to the broad internet itself.

The company has been around for the 3G and 4G transitions (went public in 2004) in the 2011 to 2013 time frame, and is now prepared to benefit from the 5G transition, as well as 4G densification.

Industry Background

Wireless hauling doesn’t garner the primary attention when building out a radio access network (RAN).  The sexier aspects at the forefront are things such as:  massive MIMO, beamforming, cloud RAN.  Backhaul in particular is generally the last thing operators consider when finalizing RAN design.  That being said, wireless hauling is critical for a network, particularly as data consumption is ever increasing, latency is a key concern and time to market is a competitive differentiator for network operators.  In other words, as uplink and downlink rates increase, the hauling capacity needs to keep pace. As new use cases for 5G evolve, speed to the core network becomes more critical, and wireless hauling in particular is much faster and cheaper to deploy compared to fiber.

Hauling is not always wireless.  In fact, fiber connections are preferred as they generally provide higher bandwidths and zero interference (justifying the higher cost). But fiber is not always feasible given permitting constraints and time-to-market concerns.  As an example, densifying a network into an urban area with millimeter wave spectrum may only be feasible utilizing wireless hauling given the constraints on burying/permitting new fiber.  In fact, over 60% of macro and small cell backhaul links today are wireless[1].

Furthermore, the speed of fiber is less of an advantage today as companies such as Ceragon are now utilizing mmWave frequencies for wireless transport, being able to achieve 20 Gbps and up (up to 100 Gbps using very high frequencies and massive MIMO).  The bottom line, however, is that wireless hauling is directly correlated with overall global RAN spending as it is a critical element of the network.

The Cycle

The prior RAN spending upcycle was with 3G and 4G in the 2011 to 2013 time frame.  During this time, CRNT’s annual revenues were in the $450 million dollar range compared to less than $300 mm today.  While the global RAN market was very strong in 2020 (up 8%, outperforming expectations despite the pandemic) and growth is expected to moderate through 2024, two key points should be noted:

1)      Backhaul spending generally lags base station spending, as backhaul upgrades are done more concurrently with increases in data throughput.

2)      The strength in 2020 was very much driven by China; 2021+ should be driven more heavily by North America and Europe which are higher margin geographies for network and component vendors.

We are now entering the 5G upcycle for backhaul and CRNT should directly benefit.  In addition, where 5G is yet to take hold, 4G networks are densifying in order to keep up with consumption patterns. 

Today, at bottom-cycle, CRNT doesn’t generate any operating profit (at < $300 mm in revenue).  So the stock price should reflect some discount to peak cycle, or in other words, should represent some type of mid-cycle characteristics.  I believe there are three key trends that will drive a longer and larger RAN equipment spending cycle for 5G compared to past cycles that will drive peak cycle higher, and therefore mid-cycle/avg cash flow higher as well:

1)      ORAN (only 1 to 2% of the market per Dell Oro) and Fronthaul

2)      Use cases of 5G outside of just mobile networks:  IoT, Enterprise, Private Wireless, Fixed-Wireless Access (FWA)

3)      Competitive pressures among network operators that will drive speed to market and geographic coverage expansion

Furthermore, CRNT underperformed in the last cycle from a profits perspective (not correctly managing their supply chain and misjudging pricing power on certain high-end products).  This time around, gross margins are higher now at bottom-cycle revenue than they were at top cycle revenue in 2011-2013.  This sets the base for solid incrementals as the cycle matures:

Key Trends that will drive a longer and larger RAN spending cycle for 5G

I will only address the vaguer concept of ORAN and Fronthaul, as the other two drivers are more intuitive.

1)  ORAN

ORAN stands for “Open Radio Access Network”.  The concept behind an “Open” RAN is to disaggregate the base station in a cellular network to allow different parts of the system to be plug and play; i.e. to modularize the system.  The idea is the same as that in cloud computing, where microservices, API’s and virtualization allow for enhanced software innovation while also lowering hardware costs.  The threat of ORAN is pointed directly at the traditional network OEM’s Ericsson and Nokia, as these companies have generally packaged the RAN equipment and software into a unified offering thereby reaping the benefits of the entire profit pool.

ORAN is a very small portion of network spending today because cellular operators are focused more on rolling out 5G services as quickly as possible to compete for subscribers.  But as the networks continue to evolve, ORAN should become more relevant as it allows easier and cheaper future generational upgrades while at same time increasing performance and capability.  As an example, Dish network is building a cellular network from scratch fully utilizing an ORAN blueprint.

The ORAN Alliance was founded in 2018 by AT&T, China Mobile, Deutsche Telekom…among other network operators, with its mission: “to re-shape the RAN industry towards more intelligent, open, virtualized and fully interoperable mobile networks.”  The global standards organizations that develop the protocols for mobile telecommunications (“3GPP”) have adopted ORAN specifications for 5G networks, which have allowed RAN vendors the opportunity to design and commercialize 5G ORAN products and services well ahead of actual deployment.

Wireless backhaul was always sort of “open” in the sense that network operators don’t have to choose an Ericsson or Nokia for their backhaul solution.  They could always choose CRNT if they wanted to, but the network OEM’s generally offered backhaul as an “add-on” to the base station product, making it simpler to stay in-house in backhaul.  According to Dell Oro, Nokia and Ericsson combined had 40% of the wireless backhaul market in 2018. 

But because 5G networks are going to consume much more bandwidth and thus must rely on very high mmWave frequencies, one could beg the question about whether third-party backhaul providers could capture more share going forward.  In other words, will the higher R&D spend by hauling specialists begin to pay increasing dividends? I am not assuming this for the time being; but, there is another interesting outcome of ORAN that I think CRNT stands to benefit from:  Fronthaul.

While backhaul consists of transport from the base station to the core network, fronthaul consists of transport from the remote radio unit to either a centralized baseband unit in the field, or to an edge cloud hosting the baseband processing unit.  The schematic below highlights the structure:

Source:  VIAVI Solutions

Fronthaul is a newer concept compared to backhaul.  As networks densify, it becomes more costly and impractical to deploy multiple fully functional base station units.  Network operators will split up the baseband functionality (essentially the radio CPU) to decrease deployment costs.  Therefore, some processing will take place in the radio unit (e.g. where latency is key) while other data will be front-hauled to the rest of the RAN compute resources. 

Fronthaul obviously makes a ton of sense when one considers the multitude of radio units needing to be deployed for mmWave cells, but fronthaul is still important for network densification in general utilizing midband spectrum.  For example, it will cost less to fronthaul to a cloud RAN (i.e. a baseband/CPU in an edge cloud) where practical.

Fronthaul is where the real ORAN opportunity is for CRNT in my opinion.  Traditionally, fronthaul technology and solutions have been proprietary to the network OEM (i.e. Nokia, Ericsson, etc.).  In other words, this part of the RAN is truly “closed”.  3GPP release TR 38.801 however specifies the fronthaul functional split, allowing vendors to now position themselves for this new market.

While CRNT is not the only 3rd party specialist wireless transport provider, they are the only provider that is offering a fronthaul solution, according to my research.  CRNT will be rolling out its first fronthaul capable products with fully integrated chipsets next month. So, while I believe CRNT will be able to ride the 5G industry wave on backhaul alone, there is extra potential juice to the cycle from its differentiated fronthaul solution.  Again, the preference for Fronthaul will also be fiber, but the advantages of the wireless use case remain the same.

QY Research recently (June ’21)[2] pegged the combined Mobile Backhaul and Fronthaul market at $15.5 billion, growing at a 19% CAGR from 2021 – 2027.  This compares to Dell Oro’s forecast of total RAN spending approaching 0% growth by 2024 due to declining 4G RAN spend.

Dell Oro also forecasts total ORAN revenues to account for more than 10% of the 2025 RAN market[3] compared to 1 to 2% in 2021.

In other words, the market potential for CRNT is there.

Financials and Valuation

Below I present 3 scenarios for peak cycle revenue.  Recall from above that prior peak in 3G/4G was $447 mm.  The other critical assumption is gross margin, where I assume that incrementals are 50% on top of current margins of ~ 30%

I peg total peak-cycle Opex at ~ $100 mm compared to current Opex of $83 million.  I forecast that Opex will grow slower than revenue over the coming years as the company should realize operating leverage from new product R&D being undertaken today.  On Financial Expenses, note that the company doesn’t run with any debt, but they incur bank revolver commitment fees as well as currency hedging costs and translation losses.

Lastly, I take half of peak cycle free cash flow, assuming that trough free cash flow is zero.  In 2020, at $263 million in revenue, the company generated $11 mm in free cash, but burned $6 million in free cash before working capital.  The company currently has over $30 mm of cash on the balance sheet.

Upside on the equity assuming prior peak cycle revenue of $450 mm is ~ 20%, but again, my thesis above is that the company has high potential of exceeding prior peak cycle revenue.

Risks

The two primary risks that I see are 1) customer risk and 2) geographic risk.

1)      Customer risk

Being a smaller company, CRNT has at times had greater than 10% customers in terms of revenue.  Its largest customer in 2020 accounted for 19.7% of total revenues.  CRNT has disclosed that this customer is in India, and I believe it is Bharti Airtel.  India is a major revenue contributor for CRNT as evident in the geographic breakdown below (my estimates for 2021):

In North America, I believe their largest customer is T-Mobile.  Note that CRNT was in a trial with a potential second Tier-1 North America operator.  I believe that this Tier-1 was Dish network, but CRNT lost out to Aviat Networks, the leader in North American third-party backhaul.  Verizon also uses Aviat, and there was some talk about CRNT getting some trials with more NA Tier 1s, but it sounds like most of CRNT’s momentum in NA ex-T-Mobile is coming from WISP’s.  Note that according to conversations with industry experts, Aviat does not have a front-haul solution, but has a very compelling backhaul solution.  There’s no reason to think that Aviat can’t develop a fronthaul solution, but as of now, CRNT is the first mover.

2)      Geographic risk

Over 65% of CRNT’s revenue comes from outside of Europe and North America, leaving less visibility to the short-term trends driving CRNT revenues.  I will summarize near-term trends in each region as highlighted in the latest earnings call.

Europe:

Solid bookings in 1Q ’21 with more than 20% of bookings from 5G related orders.  Participating in numerous 5G proof-of-concepts.  Initial 5G momentum in western Europe with the expectation for this momentum to build to the rest of Europe.

Africa:

Recently completed a large project in August 2020 and won another project in 4Q 2020 which helped 1Q ’21 revenue.  I’m modeling no growth for Africa in 2021.

North America:

Strongest quarter of bookings in 1Q ’21 since 2016.  Seeing increasing demand from their Tier-1 customer (T-Mobile) to expand its 5G network.  Also providing solutions for significant capacity increases for Tier-2 operators and WISPs.  Also increasing their sales efforts to public safety and utility providers for network upgrades (expanding industry vertical uses cases for 4G and 5G).

Latin America:

Showing signs of heating up after a covid-induced spending freeze…signs of latent demand to increase 4G coverage and to start planning for 5G.  Operators are looking to catch-up on the lost capex in 2020…bookings doubled in 1Q ’21 vs. 4Q ’20.

India:

Heavy competition among operators is driving 4G densification and planning for 5G.  Also, great efforts to upgrade the vast population still using 2G and 3G.  Expecting strong results from India in 2021 (with the virus a wild card here).

APAC:

Experienced a slowdown in the last few quarters, but recently signed a significant contract with a Tier-1.  Beginning to sell 5G technologies to China (evidence of the lag between base station/radio capex and backhaul, as RAN spending in China was very very strong in 2020).

Summary

Because of the small size of the company and potential customer concentration risk, CRNT shouldn’t be a large % of a portfolio. But with all the hype surrounding 5G related to the consumer experience and new industry use cases, I think wireless hauling is an under-looked area that will directly benefit from not only 5G momentum, but 4G densification as well.

[1] GSMA ABI Research Report, “Wireless Backhaul Evolution”, February 2021

[2] https://www.mccourier.com/mobile-backhaul-fronthaul-market-size-is-projected-to-reach-us-56340-million-by-2027-from-us-15460-million-in-2020-at-a-cagr-of-18-7-during-2021-2027/

[3] https://www.delloro.com/radio-access-network-market-to-grow-over-the-near-term/

Catalyst

The potential catalysts consist of: a new Tier 1 North American customer win for backhaul and/or fronthaul