More than 20 years ago, telecom operators started massively deploying circuit switched optical transport networks, benefiting from the increased capacity and cost efficiency of optic fibers. These networks, which are still very wide spread, were originally designed for the transmission and aggregation of a very limited number of TDM services, mainly voice traffic. Most of the optical networks were built and operated using a set of global standards referred to as SONET (Synchronous Optical NETwork), which were characterized by guaranteed delivery and network resiliency. During the 90s, things became more complex as numerous data services started to penetrate the market, leading to one of the biggest revolutions in telecommunications history. Data traffic differs from voice traffic by being less time-sensitive and by having a bursty nature, with dynamically varying levels of utilization of communication channels, as opposed to voice traffic, which requires constant levels of channel utilization. While SONET networks were very reliable and efficient in dealing with traditional voice services, they lacked the flexibility and scalability needed to deal with the flood of new IP- based services and the high levels of data traffic.
Building a new optical transport network was out of the question. First, carriers had to maintain their existing networks in order to support the huge number of voice customers. Second, with the huge investments in circuit switched networks, it was a waste to build new optical networks from scratch. Fortunately for the carriers, a new technology was introduced: Next-gen SONET. Next-gen SONET equipment made it possible for carriers to use circuit switched networks for packet-based traffic as well as traditional services. It certainly earned telcos some time, but they are still facing the inevitable migration towards a network dominated by packets, since packet-based services are not just coming to supplement traditional services but are gradually displacing them. In the face of the strong shift from TDM to IP traffic, next-gen SONET is reaching its limits in regards to the amount of data, variability of services and flexibility to support new applications. In the not so distant future, handling packet services by adapting them to circuit-based networks will become obsolete, as packet based networks enable a much better and more efficient utilization of the optical infrastructure. Transport networks must fundamentally evolve to become packet based, however, carriers will still have to support TDM services for the foreseeable future.
Therefore, the next generation transport networks will be packet switched networks that are able to support legacy services, instead of the current circuit switched networks that are able to support packet services. There are several technologies and architectures which enable the deployment of a packet transport network, however, they all share the same idea: a generalized, converged network that can support and aggregate any type of traffic or application regardless of its origin and nature. Transport networks in metropolitan areas (metro networks) are experiencing the most dramatic shift from carrying primarily voice traffic to carrying a growing mix of data, video and voice traffic. These networks specifically require high-end switching solutions due to the enormous traffic volume.
Orckit Communication (OTCPK:ORCT), which started as an ADSL company in 1990, foresaw the shift and its magnitude already in the beginning of this decade. Orckit’s subsidiary, Corrigent started to develop metro transport solutions that enable carriers to make the inevitable migration from legacy circuit switched transport networks, to next-gen packet based transport networks. By 2004, while all the major vendors were busy promoting next-gen SONET solutions, Orckit commercially launched its first optical transport switch – the CM-100. Optical packet transport didn’t get much attention those days, as carriers and vendors alike were still licking the wounds of the dot.com bubble burst and spent very cheaply on optical networks. However, KDDI, the second largest carrier in Japan decided to be the first carrier in the world to carry out a national upgrade of its metro transport network from circuit switched to packet switched. This upgrade included the deployment of tiny Orckit’s CM-100, raising many eyebrows across the industry.
The exciting project, which has generated more than 200$ mil in cumulative revenues for Orckit, seemed like the first win with many more to follow. After participating in the buildup of one of the most (if not the most) advanced transport networks on earth, everybody expected ORCT to replicate the success with a growing list of new customers. Orckit’s managers were sure they would have a line of carriers willing to upgrade their networks as soon as late 2005. That confidence was based on extensive field trials with more than 6 tier-1 operators, especially in Asia, who seemed pretty happy with ORCT’s product. To date, from various reasons ranging from M&A activity and plain old conservatism, none of these evaluations actually ended up with a large-scale deployment. However, ORCT’s management keeps on claiming they still haven’t lost any of these accounts, and that customers are simply delaying their decision. Without a last minute twist, 2007 will be another year Orckit’s investors would rather forget. In a striking resemblance to 2006, the current year started with big expectation build-up from management, and ended with disappointment. Granted, every small technology company, especially in an emerging field like optical packet transport, is subject to delays, fluctuations and surprises. However, ORCT’s case is especially tragic due to the strong contradiction between the very high hopes and very poor delivery.
In December 2005, more than one year after the official win with KDDI, Orckit’s CM-100 was selected for deployment by WilTel, a U.S. wholesale telecommunications carrier. This deal was supposed to generate substantial sales in 2006, however, Level 3 Communications (NYSE:LVLT), which acquired Wiltel decided not to purchase additional products from Orckit. Then, on the Q405 conference call in early 2006, the company's president said he would be "personally very disappointed" if they didn’t win at least one more customer in Asia in the first half of 2006. The company’s expectations to get at least one win in Asia during 2006 turned out to be premature, as in 2006, the company had meager sales, which sent its shares sliding more than 50%. In the beginning of 2007, the company started mentioning an “imminent” tender in Asia that should be awarded in the second half of 2007. Again, there was no announcement about the imminent customer until the end of Q2, but during an investor conference in the beginning of July, the company’s CFO explained that the selection by the imminent customer was delayed but still claimed they were well positioned with this “lucrative” carrier. In August, on another investor conference, the company’s VP of marketing did not want to comment on specific customers but said the company was still expecting to announce wins (in plural) already in 2007. Only one week after, on August 15th, during Q2 conference call, investors again heard the bad news: The imminent customer decided to go with a legacy product (next-gen SONET) from one of Orckit’s competitors, delaying the decision for its next generation transport network to 2008. During the call, (which ironically was of very low quality and with too many cut offs), company’s management stated that a win this year is still possible. Investors didn’t take that comment seriously, sending the stock to a new low. Frankly, who can blame them?
With all the depressions and skepticism surrounding Orckit, current price levels may represent a good opportunity to invest in one of the most interesting telecom equipment vendors around. Despite the series of failures in expectation management, being a pure-play in such a hot segment has very promising prospects going forward.
Reasons for cautious optimism
The most important factor playing in favor of Orckit, is the huge spending cycle kicking into gear within the telecom industry. For the first time in years, carriers are opening their wallets, upgrading their networks and deploying multiple packet-based services like IPTV and 3G services. Not only is the optical transport segment finally getting the attention it deserves, carriers finally realize it will take a totally new kind of optical transport networks in order to offer next-gen services in a cost effective manner. Such networks must be capable of supporting all the services and applications over one converged network.
The market for metro packet transport is growing from a handful of potential customers to dozens, including the world’s biggest operators. As a result, packet transport, which was once a niche for small companies like Orckit has moved into the mainstream, with many large vendors, including Alcatel-lucent (ALU), Nortel (NT) and Fujitsu, who are planning or already launched packet transport products. Orckit’s most apparent disadvantage is its small size, as it still remains to be seen whether Orckit can compete with the large vendors. Even if we assume the majority of tenders will be won by competitors, each opportunity pursued by Orckit is worth at least several tens of millions of $, with some exceeding 100 mil $. With a current market cap of just under 100$ mil, one win per year is enough to send the company’s shares substantially higher.
In addition, Orckit usually partners with large vendors or system integrators, which integrates Orckit’s products as part of their overall offering. In Japan it has been working with OKI and Netmarks, two of the largest local integrators, who resell the CM-100 to KDDI. Although ORCT’s management may be horrible at expectation management, it should get the credit for foreseeing the paradigm shift ahead of the rest of the market and actually doing something about it. That innovation led it to win the huge multi-year deal with KDDI, whose metro transport network can be crowned as the first and largest packet transport network in the world, with over 2000 network elements and 200 10gig rings – not bad. Only now can we realize how innovative and groundbreaking KDDI was by deploying packet transport networks already in 2004, while the rest of the operators were still trying to squeeze more out of their Next-gen SONET networks. Orckit’s management understood that being first to the party is not enough and maintained its technological lead by rigorous R&D efforts. It has been working on its second-generation product line, the CM-4000, for more than two years and hopes to have it in commercial availability next year. In recent weeks, the company has recruited more than 20 R&D workers, which represents an increase of more than 10% in their R&D workforce. That is not an easy decision for a company with almost no revenue and an alarming cash-burn rate, but technological innovation is the only thing that can differentiate the company from the rest of the market.
With Orckit remaining tight-lipped and vague about their potential customers’ identities, there is no choice but to speculate. From what the company is willing to say, the majority of evaluations are with carriers in Japan, where the need for ORCT’s solutions is much more needed due to the high fiber penetration and intense competition.
KDDI is the usual suspect. To date, KDDI ‘s network is the first and only large scale deployment of packet based metro network. KDDI has concluded its shift to a packet transport network, based on Orckit’s platforms exclusively. All of KDDI’s metro transport traffic, whether IPTV, VOIP, legacy TDM or 3G backhaul runs on that converged network. With 27 mil wireless subs and 9 mil wireline subs, this is an achievement and a powerful demonstration of Orckit’s technology. During the past quarters, KDDI purchased only small quantities of products from Orckit, which led to speculations that KDDI might dump Orckit and move to one of the large vendors. Though it is an option, Orckit’s management is constantly mentioning KDDI’s heavy involvement in the planning and development of Orckit’s second-generation product, the CM-4000. Thus, KDDI seems to be quite happy with Orckit’s solutions and it is likely to view the new platform as attractive. In recent presentations, company officials even hinted that orders from KDDI are likely to be back around mid 2008.
Softbank (OTCPK:SFTBF) is another Japanese carrier, who is rumored to be conducting trials with Orckit’s products for more than 2 years. The carrier was expected to make a decision already in 2006, but the acquisition of Vodafone Japan in March of that year, halted the selection process. More than one year after the acquisition, Softbank will most likely start to converge its wireline and wireless networks, leveraging Vodafone’s large subscriber base (over 15 million).
In addition, Softbank is expected to begin a massive upgrade of their large ADSL subscriber base (more than 5 million subs) to VDSL2 and FTTH. This kind of upgrade necessitates a major infrastructure upgrade, as there is more than a 10-fold increase in the bandwidth needed to support VDSL2 in comparison to ADSL.
Three months ago, Orckit surprised investors by announcing a joint demonstration with T-Systems, at a large cable exposition. Besides the significance of partnering with such a huge integrator, it implies that Orckit is targeting the cable market in addition to the telco market, and certainly makes the addressable market look bigger. Orckit has stated it currently has one large evaluation process in Europe, but remains very confidential about it. The demonstration with T-systems insinuates that the customer is actually a large cable operator, and not a telco.
The “imminent” customer who just recently delayed its decision is an enigma. We know it is a pretty big asian operator most likely outside of Japan. Orckit has offices in both Korea and India, working with local potential customers for over 2 years, but given the monstrous growth rate in the Indian mobile market, an Indian cellular provider from the likes of Bharti and Reliance sounds like a reasonable option. The imminent selection was for a specific layer of the carrier's network, with a larger RFP for the full network (estimated at a few hundred million over several years) expected next year with a selection in '08. Indian operators' announcements about ambitious multi-year network upgrade and expansion plans fits the above description as well. This week, ECI Telecom (ECIL), which is a leading player in the Indian optical market, announced that Bharti had chosen ECI’s next-gen SONET platform for a nationwide optical network expansion. The characteristics of this project, coupled with the timing of the announcement make Bharti the prime suspect.
In summary, Orckit is entering 2008 with low expectations and increasing frustration among investors. Management is still optimistic about being selected for new projects, as customers simply cannot linger for long. Operators are likely to start upgrading their metro transport networks next year, but Orckit’s share in that growing market is still unknown. Personally, I find the nerve-racking anticipation of Orckit’s new win announcements similar to waiting for the last night bus. As time passes by, one part of you keeps thinking that the chances for the bus to arrive are only getting bigger because it has got to arrive some time, while another part is thinking, “you are walking home tonight”. Orckit is a very promising company, indeed, but don’t invest in it unless you are prepared for walking home.
Disclosure: Author has a long position in ORCT