When I first wrote about Infinera (INFN) back in 2007 (see Infinera Corp.'s 'Yeah Baby' Moment), the original title of the article was going to be "Infinera: You Had Me at Photonic Integrated Circuit."
That's because, despite all the claims by other optical transport equipment manufacturers, INFN was really the only one with a solution that cut through the bottlenecks created when traffic needs to jump on and off the optical network and is forced to go through non-optical switches and other points of congestion.
While the technology behind such a solution was, and is, way over my head, the logic behind an integrated, end-to-end transport system was way too simple to ignore. So it is easy to see why INFN IPO went off almost as well as its 10 GB solution was selling at the time. Unfortunately, and not coincidentally, initial public offerings do not tend to happen at the bottom of the cycle, but rather the opposite. Combine that with the advent of the financial crisis, and the rest of my thesis was effectively blown to bits in a very short period of time. INFN business quickly entered what I described on the Buzz & Banter (subscription required) as a "nuclear winter." To prolong the drought, INFN made the strategic decision to bypass the carriers' upgrades from 10 GB to 40 GB channels, and focus all of its efforts on a 100 GB product, which in 2009 was so far forward-looking to amount to little more than a leap of faith. INFN in essence bet the farm that its 10 GB product and installed base were going to be sufficient to carry it through several years of development of a new state-of-the-art technology.
The consequences for the stock were pretty predictable. It has traded between $5 and $10 for the better part of the last five years, as investors where left guessing whether the 100 GB bet would eventually pay off, or whether the company had missed the boat for good. Early last year, management began talking up the prospects for the new 100 GB DTN-X box even though the product was still several quarters away from delivery, and there was still a lot of uncertainty over carriers' acceptance (see European Elections: At Full Speed Toward the Debt Iceberg).
Fortunately for the company, late last year it became apparent that the 100 GB DTN-X box was indeed a winner, and the Q1 business update pretty much confirmed that INFN is now staring at what should be a long and strong product cycle.
Infinera Technology Advantages
To clarify the upside for INFN's business and its stock, I want to touch on its key advantages vis-�-vis the Cienas (CIEN) and Alcatel Lucents (ALU) of the world.
I have already mentioned the first technical advantage Infinera has over its competitors: It provides a solution that can take packets on and off optical transport networks without needing third-party components to convert the signal from optical to electronic and back to optical. In non-integrated solutions, the on/off point is where the network traffic gets jammed.
Second, being an integrated system, carriers can add capacity to their networks by simply adding line cards to their existing boxes. This process does not require any time-consuming configuration and can be accomplished in as little as 24 to 48 hours without any disruption to existing traffic.
Third and most important, being a seamless integrated solution, the DTN-X lends itself perfectly to being managed by Software Defined Network servers. If you follow the networking space somewhat closely, you know that SDN is the new rage, since it makes it much simpler for carriers to adjust their networks to their customers' needs in real time. But effective SDN management is easier said than done, especially if the software must control several working parts of different origin. On the other hand, it becomes relatively simpler and more efficient if the software needs to speak a single language easily understood and coherently applied by the entire data transfer process.
The "Razor / Blade" Model
Infinera's business is pretty much a straightforward "Razor/Blade" model. The company earns its footprint in carriers' networks by selling them DTN-X boxes with a starter amount of Tributary Access Modules (TAM) and Transport Line Cards (TLC). The boxes carry margins only in the mid-30%. The high profit stage of the business kicks in when, over time and as network's demand rises, the carriers fill in the boxes with high-margin (50-60%) TAMs and TLCs. The "fill in" cycle can last several years until a whole new generation of technology is ready for deployment. As a point of reference, INFN still draws a significant portion of its revenues from sales of 10 GB TAMs into the six-year-old DTN boxes.
It's fair to say that, if DTN-X box sales develop as suggested so far by the company, this product line is likely to have much better penetration and success than its 10 GB precursor.
Counterintuitive as it may seem for a high technology company, I tend to approach INFN's stock as if it were a deep-cyclical. That means I want to buy the stock when it is very expensive and I want to sell it when its business reaches peak profitability and the stock appears inexpensive using various earnings measures. At the peak of the last cycle in Q2 of 2008, INFN traded at an EV/EBITDA of 10, on quarterly revenues of approximately $160 million and EBITDA of $45 million. If the business simply gets back to that kind of revenue run rate and profitability by 2015, the stock could reach the mid to high teens before then.
Also consider the following: 1) back in 2008 INFN relied on a single customer for almost 50% of its business. This time around the company has already sold boxes into 27 different customers, and CenturyLink was the largest customer at just above 10% of revenues; 2) widening its revenue base, and/or adding another high-profile tier 1 customer (Verizon (VZ) tech folks have appeared with Infinera executives at several recent conferences) would meaningfully de-risk INFN's business and increase its valuation range; and 3) barring some other macro disaster, this product cycle could last well past 2015.
The number of competitors in this space is limited primarily to CIEN and ALU. Infinera's management has made it very clear that the Chinese market is not open to foreign companies, in the same way that US the market is not totally open to China's Huawei for the same reason. But price competition between these few providers is fierce, especially when it comes to establishing the initial presence in the network.
Second, due to the limited number of buyers of this type of technology, orders can be very lumpy. Having followed INFN for several years my sense is that investors and traders understand this and do not overreact to quarterly misses or beats. That said, it makes it a lot harder to model numbers and valuation expectations.
The first time around INFN's potential was stunted by the financial crisis, and its subsequent decision to bypass an entire product cycle and bet the farm on the transition to the 100 GB technology. By all indications the gamble appears to be paying off. INFN is rolling out a best-in-class solution, and barring another macro disaster the runway for INFN DTX-N boxes and blades should last several years. If INFN merely returns to the 2008 revenue and profit run rates, and its stock is given the same kind of valuation it garnered back then, there is 50 to 70% upside within the next 24 months. More blue chip customer wins and/or a higher multiple could reasonably push the stock back to its all-time high in the mid-$20s. I have a significant long position, which I hedge periodically with long-dated put options as insurance and for cash flow.