AT&T Domain 2.0 – Who Will Flinch?

On September 23rd, AT&T issued a press release on Domain 2.0, which outlines at a high level its next generation network vision and supplier strategy.  As part of this strategy, AT&T plans to use Domain 2.0 to transform its network through utilization of Software Defined Networking (SDN) and Network Function Virtualization (NFV).   In doing so, AT&T plans to enhance time to market and flexibility of new services as well as beginning a downward trend in capital spending likely beginning in 2016.  The goal is to achieve a downward bias in capital spending through separating hardware from software and control plane from data plane in traditional network infrastructure products while also improving software management and control.  The goals of Domain 2.0 are more focused on a network transition, faster service creation and lower capital spending, as compared to the goals of the 2009 Domain 1.0, which were to reduce cost and time to market for new services by dramatically reducing AT&T’s supplier base to a much smaller set of strategic suppliers/integrators.

While AT&T is certainly sincere in its plans to evolve towards SDN and NFV, I also think the public announcement of the Domain 2.0 strategy serves as a message to its shareholders that through the implementation of Domain 2.0, AT&T expects capital spending to begin to decline starting in 2016.  This is an important message from AT&T as it is likely to have increasing pressure on its cash flow and cash needs beginning in 2014 given higher cash taxes and potentially a significant acquisition overseas.

Wall Street analysts that cover AT&T have often written that the company will likely have to pay higher cash taxes in 2014 unless the current bonus depreciation rules get extended.  The increase in cash taxes beginning in 2014 is likely to raise the dividend payout ratio at AT&T from around 65%-70% to close to 75% next year according to analyst reports.  At the same time, AT&T has a history of raising its dividend annually.  The likelihood of higher cash taxes and dividends next year, thus, will result in lower available cash for stock buybacks and acquisitions.   Clearly AT&T would like a way to reduce its capital spending if possible.

Regarding acquisitions, the press and Wall Street analysts have recently both written that AT&T is likely to make a significant international acquisition in 2014 and some are guessing that Vodafone (post its sale of its ownership of Verizon Wireless to Verizon) is a likely target.  A large acquisition on the scale of Vodafone is likely to require the addition of significant debt on AT&T’s balance sheet (e.g. some analysts suggesting over $70 billion), resulting in additional debt service and a more levered balance sheet.

In my view, the public messaging of Domain 2.0 was targeted to two distinct audiences, namely, AT&T shareholders and equipment suppliers. For its shareholders, AT&T was messaging that it plans to help offset the increasing dividend payout ratio and potential higher debt service due to any future acquisition through a lower capital spending trend after Domain 2.0 begins to be implemented.  For its equipment suppliers, AT&T was messaging that it plans to use SDN and NFV to extract a more agile and lower cost network and to be prepared for this change through lower pricing facilitated through separation of hardware/software and control/data planes (not that the equipment vendors really needed any public messaging from AT&T on lower prices as that is business as usual).   In a sense, AT&T is partly using this press release to remind its equipment suppliers know the “hammer” on pricing is coming and be prepared to help them migrate towards an SDN/NFV architecture or potentially be eliminated as suppliers in Domain 2.0.

Coincidently, Barron’s interviewed the CEO of Cisco Systems, John Chambers, the same week that AT&T issued its Domain 2.0 press release.  According to the Barron’s article, when asked what Cisco’s biggest competitive threat is, Chambers mentioned that at or near the top of the list are service providers who just buy cheap “white box” routers and switches.  While it may be just a coincidence that Chambers made this comment just a few days after AT&T made the public release of Domain 2.0, which implies a migration towards less software intensive routers and switches, its does raise the question on how serious AT&T will be in its migration plans to SDN/NFV and whether Cisco is willing to adjust to meet the needs of AT&T.

Most significant equipment suppliers to AT&T typically have a high dependency to AT&T with sales typically near or above 10% of total company sales (e.g. Ciena, Alcatel-Lucent, Juniper, Adtran etc.).  Given the broad nature of Cisco’s business, however, AT&T is not as material a customer to Cisco as it is to other equipment suppliers.  While not as significant a customer, AT&T is an important customer to Cisco, and Cisco wants to expand its presence at AT&T beyond routing and switching to also include optical equipment given its recent product launches that converge optical and packet technologies into new platforms.  The convergence of packet and optical technologies is also a likely part of Domain 2.0 given this convergence is an enabler of reducing network cost as IP packets and flows can be carried more cost effectively in a converged IP/Optical network.

So the question is, who will flinch in AT&T’s path towards Domain 2.0, Cisco or AT&T.  AT&T has a network vision and higher priorities for its cash in the form of cash taxes, dividends and acquisitions and will seek to reduce capital spending in the future.  While these higher priorities always existed, SDN and NFV now provide an architectural change that could facilitate lower capital spending in the future.  Cisco is the largest and strongest financial company that spans both packet and optical technologies among AT&T’s suppliers; is not heavily dependent on AT&T as a customer as compared to most other suppliers to AT&T; and has a large installed base at AT&T.  Cisco will not easily submit to separately selling its hardware from software and data plane from control plane in Layer 2/3 and converged optical/packet products.  It will be clearly be an interesting process to watch, one that will likely have ramifications for other service provider SDN network migrations in the future.

Disclosure: NT Advisors LLC may in the past, present or future solicited or generated consulting services from any company mentioned in this post.