Undervalued Finisar Fights to Stay Tops in Optics

It is well positioned to benefit from ongoing robust demand for high-speed fiber optics. However, we do not believe Finisar has an economic moat in this highly fragmented and cyclical industry as a result of product commoditization and strong buying power from customers.

Although we expect Finisar to develop a new stream of revenue from selling 3-D sensing lasers in smartphones, we don’t think it can carve out a moat in this business, given the rampant competition.

Finisar sells into two submarkets: telecommunications and data communications. The advent of the public cloud should inject some secular growth in high-speed fiber optics, which are sold under the datacom segment and used to interconnect hyperscale data centers for delivering and hosting cloud services. The company’s telecom segment is likely to benefit from regional network upgrades in the United States and China, although near-term headwinds in China persist.

Despite these positives, we believe Finisar’s ability to develop a sustainable competitive advantage is limited by an unfavorable industry structure. Optical suppliers face significant pricing pressure from the hefty customer buying power of network equipment providers, which work hard to eradicate lockup risk by demanding components to be interchangeable and modularizing designs. Technological leads among optical vendors are often short-lived and fade by commercial deployment. Cost advantages in production are hardly sustainable in fiber optics, in our opinion.

Finisar is probably supplying laser arrays into Apple’s upcoming iPhone X, which supports 3-D sensing for facial recognition. This could add a profitable revenue stream for Finisar. However, we are not ready to turn more positive on our moat rating, given the strong production capability of key rival Lumentum LITE and the threat of new entrants.

Lack of Bargaining Power Means Lack of Moat We believe Finisar does not have a moat, as we think it lacks the structural bargaining power to extract economic profits from orders placed by network equipment and storage equipment providers and their original-equipment manufacturers. In our view, the fiber optics industry has some traits that are not conducive to a moat, such as low customer switching costs, constant technological change, sporadic purchasing patterns by OEMs, and intense competition for these orders. Additionally, Finisar's high exposure to datacom markets (over 70% of total sales in recent years) and heavy customer concentration (its top 10 clients have contributed over 55% of sales in recent years) further constrain bargaining power. Therefore, despite secular growth in the data communications market, we do not have enough confidence in Finisar's ability to earn excess returns on capital over the next decade to warrant assigning a narrow economic moat rating.

Fiber optics modules or subsystems interconnect the electronic network equipment in a fiber network. Such modules rely on optical components such as lasers, photonic detectors, and integrated circuits. Finisar mainly sells optical modules to two main markets: data communications and telecommunications. The datacom market encompasses data center interconnects and large cloud vendors’ own regional or multiregional networks. The telecom market covers carrier networks for local access, metro/intracity, long-haul/intercity, and submarine fiber cables. Both markets do have a few near- and medium-term growth catalysts, such as secular growth in enterprise cloud and retail colocation services, broadband development initiatives in China, and U.S. metro network upgrades to support converged services between fixed and mobile wireless and 5G deployments. However, such growth catalysts do not necessarily translate into long-lasting economic profits for Finisar because of several unfavorable structural characteristics of the fiber optics industry.

The fiber optics industry is very fragmented, with a dozen small and midsize players by market capitalization ($200 million-$3 billion) that compete fiercely with each other for OEM contracts and have relatively low profitability; EBITDA margins have typically been under 20% for most optical module vendors. Fiber optics suppliers sell fiber optics components or modules such as transceivers to OEMs, which in turn make fiber optics network equipment and sell to telecom carriers. While OEMs have a hard time negotiating with carrier clients and protecting their own profits, they have greater negotiating power with optical vendors, because these modules and components are usually interchangeable or interoperable (usually at the request of OEMs in the design stage) for emerging network equipment, such as 100 Gigabit/second fiber optics. Technological leads among component manufacturers are often short-lived, ranging between 12 and 18 months for ultra-high-speed fiber optics such as 400G/s. However, such deployments are often in pilot programs. By commercial deployment, competitors have often come up with similar offerings and closed the gap.

Fiber optics manufacturers’ market position is further weakened by the fact that the vast majority of their revenue is order-based as opposed to subscription- or service-based. Finisar and its peers rely on winning large contracts of certain modules from OEMs every two or three years, usually when carriers are placing large orders with OEMs for regional or national network upgrades. The modules in demand are usually commonplace, and OEMs work very hard to eradicate any lockup risk by ensuring optical modules are interchangeable or modularizing the network equipment. OEMs usually design their own chips and may manufacture certain parts of the equipment themselves, further reducing any dependence on any optical vendor beyond a given purchasing order.

Finisar and Lumentum are the two largest fiber optics makers, claiming about 60% of the industry sales (Finisar has about 35% while Lumentum is at 25%). More than 50% of Lumentum’s revenue is exposed to telecom, while Finisar derives over 70% of sales from datacom. At the same time, operating scale does not necessarily translate into substantial cost advantage for Finisar to compete with smaller peers. A small component vendor can specialize on one or a few components that match with its manufacturing scale and live off a few clients--Acacia Communications ACIA and Applied Optoelectronics AAOI are good examples. Like Lumentum, Finisar also manufactures laser diodes that can be used for 3-D sensing in consumer electronics. Both companies are largely confirmed to be suppliers into Apple’s iPhone 8. However, we believe it is too early for us to identify a moat-enhancing event, as 3-D sensing among consumer products is in its early stages and profitability for laser providers is not guaranteed.

Furthermore, we believe Finisar has greater customer concentration risk than Lumentum. Finisar’s heavy exposure to datacom and overdependence on orders from its largest customers make its top line highly cyclical, as Cisco and Huawei each accounted for close to 10% of its sales in fiscal 2017, while the top 10 clients represented over 55% of sales. We do not believe Finisar has much bargaining power, given the strong position Cisco has in the communications ecosystem and Huawei’s advantage as one of two leaders in selling to Chinese carriers, whose network upgrades with optical switching have the potential to drive large demand for telecom optics. In total, we do not believe Finisar has the ability to develop a sustainable competitive advantage that would generate economic profits, given the unfavorable structural factors in the fiber optics industry and its customer concentration.

Finisar spends a decent amount on research and development--around a midteens percentage of revenue per year--to maintain its lead in fiber optics technologies and develop better products that enable faster communication speed with superior designs. However, these types of new products are often born out of multiparty agreements that are codified by all major customers and optical suppliers. There is hardly a “secret weapon” type of product launch. Therefore, we think Finisar is keeping pace in the industry, but we still struggle to imagine it or any company sustainably leaping forward on fiber-based communication technology. In addition, while there have been talks of industry consolidation, which we believe would enhance the industry structure, we do not expect M&A among major players to address the fragmentation problem in the near term.

Although 3-D sensing is likely to bring in a profitable revenue stream for Finisar, we do not expect Finisar to develop a sustainable technological lead in laser technology over peers such as Lumentum, which has a longer record of supplying 3-D sensing lasers for consumer applications and has stronger production capability than Finisar for now.

Demand Can Be Sporadic We assign Finisar a very high uncertainty rating because of the risk in the optical industry, both near term and long term. The demand for telecom fiber optics ultimately ties to carriers' plans on network upgrades driven by new generations of communications technology, such as 5G, and demand for enterprise cloud and bandwidth-heavy activities, such as streaming video and Internet of Things. However, carriers could put discretionary network upgrade projects on hold for temporary reasons, usually due to pending mergers and acquisitions or tight budgets resulted from struggling core businesses, as seen in 2017. Such short-term factors may cause OEMs' orders to be more sporadic, which could leave optical vendors vulnerable at the negotiation table and facing capacity constraints as they have to scale back production. In addition, as hyperscale clients like Amazon, Google, and Facebook surpass telecom carriers with stronger and more sustained demand for datacom optics, these Internet giants are taking strategic measures to control their supply. A longer-term trend is that Finisar and its peers are innovating to substantially lower what network operators need to spend on high-speed networking equipment. This provides a windfall for carriers and hyperscale and indirectly end consumers, at the expense of optical competitors who are stuck in this R&D race.

We foresee another set of risks around the company’s 3-D sensing laser diode business. While the opportunity is very promising, we are not confident that Finisar will have technological leadership based on manufacturing the vertical-cavity surface-emitting laser diode. Apple is believed to be a large customer in this emerging opportunity, but it might be pushing multiple vendors to develop competing capacity; Finisar might not be able to secure these potential contracts or earn adequate margins on the business it does win, in the long run.