Tuesday, I attended the first day of the Broadband World Forum in Paris. Broadband World Forum and ITU's Telecom World are the two remaining high-impact global telecom vendor trade shows now that Supercomm has faded away.
There was a lot of excitement at the show. An Alcatel-Lucent (ALU) IP Routing salesman said business had never been better. Global themes of cloud computing, mobile broadband, smart phones, rising video traffic, etc. have now filtered down into rising demand on the telecom infrastructure that needs to accommodate this exponential growth in bit traffic. Resulting telecom broadband initiatives at the show included:
100GB telecom switching platforms (and discussions of when it will be adopted)
Mobile backhaul - and whether to use lower-cost microwave or more scalable ethernet platforms
FTTP - fiber-to-the-home will be a much longer process than previously thought
Traffic monitoring - which ultimately leads to heavy users paying more
Digital Home network with all electronic devices wirelessly controlled
I also had an interesting telecom vendor economics discussion with an industry veteran. As opposed to certain software and internet business models with high margins, price setting capabilities, and low fixed costs, the telecom equipment sector is basically "box movers". Price is a given. A limited number of powerful carriers have leverage over multiple suppliers (and system integrators like Alcatel-Lucent squeeze even harder on its suppliers, in order to earn a decent margin when selling to carriers). Technology can be a differentiator, but it really is more of an entry ticket. If the technology is not top-notch, you don't play. Thus, the key to a successful telecom operator is cost. The low-cost producers are more profitable, and the higher-cost producers are less profitable.
For example, Adtran (NASDAQ:ADTN) is a strong telecom switching company that successfully competes against larger competitors like Alcatel-Lucent and Cisco (NASDAQ:CSCO), and has been more profitable than most of its competitors due to its no-frills way of doing business. For example, Adtran cheaply acquired Luminous Networks' (a Quan Ventures II portfolio company) IP about five years ago. It took Luminous' industry-leading resilient-packet-ring gigabit ethernet protocol for metro access, simplified it into a lower-cost solution, and is now a leading gigabit ethernet network supplier. Adtran is now benefiting from the ethernet mobile backhaul boom, again with a low-cost, strong technology offering for major carriers.
In every niche I explored at the show, no one company had a product that another was not at least developing. For example, Juniper Networks (NYSE:JNPR) was the first to announce a 100GB switching platform. But Nokia Siemens Networks (NSN, private) and others will be launching 100GB within twelve months. Moreover, most providers have lambda grooming products and other features that lower the cost of the existing central office switches and add capacity that can elongate this transition period. Juniper had the best differentiation attempt through its own venture-backed "apps farms" where start-ups write customer-demanded apps like video surveillance that are difficult to support internally.
An interesting statistic was that broadband spending is projected to grow by only 3.4% annually through 2015, and total telecom spending would be roughly flat. Thus, vendors need to innovate just to stay even, in spite of the macro opportunity of the broadband boom. Vendors stuck with maturing circuit-switching or ADSL products, for example, will be hurt the most.
Previously, certain groups in the U.S. (namely Verizon (NYSE:VZ), certain cable operators), Korea, Australia, Germany (namely Munich municipal authorities), among others, have had fiber to the premise (FTTP) or home initiatives. A counter to this trend at the Forum was to make the most of the existing copper infrastructure at a much lower cost. NSN demonstrated a 750 Mbps speed over a 500 meter distance using bonded copper cabling. Alcatel-Lucent and Huawei had similar technology demonstrations previously that can be generally labeled as Phantom DSL.
Ikanos Communcations (NASDAQ:IKAN) announced NodeScale Vectoring at the Forum that will enable greater than 100 Mbps speeds over longer distances than previous DSL technologies covering 192 nodes or local connection/distribution points. Bonding VDSL (very-high-speed digitial subscriber lines) today, vectoring VDSL in 2011, or Phantom DSL longer term, are all part of a broader movement to extend the capacity of today's current copper infrastructure to meet future broadband needs. In a world where long-term carrier capex growth is limited by the difficulty of monetizing growing bandwidth demand (as well as a maturing voice revenues), momentum has been growing from a number of large service providers (AT&T (NYSE:T), BT (NYSE:BT), Qwest (NYSE:Q), Windstream (NASDAQ:WIN), Tele2, TDS (NYSE:TDS), etc.) to leverage existing copper to deliver broadband data and video. Over the next few years, the most common strategy will be fiber-to-the-node or local distribution point (FTTN), with copper covering the last 500 meters via VDSL.
Ikanos Communications, a company that we have written about previously, should generally benefit from these high-speed copper and FTTN trends. The VDSL market is growing in excess of 20% annually. On the other hand, Lantiq (private) is a venerable VDSL competitor, that along with Broadcom (BRCM), has recently taken VDSL market share away from Ikanos. Ikanos VDSL market share was 82% on a pro-forma basis in 2008, was 73% in 2009, and will be lower in 2010. In addition, Ikanos has decided to walk away from certain commoditized ADSL business. Thus, we expect Ikanos revenues to decline in 2010, and show single-digit growth in 2011, while profitability should improve substantially next year. The yet-to-be-reported September quarter had substantial restructuring charges and a lower revenue run rate, which led to a plunge in Ikanos stock price. IKAN was trading at above $3 earlier this year, and is now at $1.15, leading to a current market cap that is only 0.4x depressed 2010 revenues.
Disclosure: Long IKAN