Both Verizon and AT&T see LTE as a fixed, as well as a mobile, broadband option. Both have rolled out fixed wireless offerings, and Verizon even added voice to its HomeFusion system last month. Such products are mainly geared to replacing DSL over time, reducing maintenance costs and fees to other carriers outside the main carriers’ regions of coverage. But Verizon’s recent upgrade of HomeFusion, and its deal to take full control of Verizon Wireless from joint venture partner Vodafone, has stoked talk that it may use LTE to reduce its investment in its FiOS fiber network, or even sell it off to become a fully wireless-driven company (plus its important cross-marketing agreements with four cablecos).
But surely this looks like investors trying to talk up yet another major transaction, with which to move share prices? The logic is so unsound as to be dumb.
The thinking revolves around valuing the wireless business very highly. As Vodafone’s 45% share of it was worth $130bn, the whole unit must be worth at least twice that, if it were standalone. Since Verizon is only worth $132bn, this suggests that the cellular network on its own, less perhaps that $62bn of fresh debt arising from the Vodafone buy-out, would be worth more than the entire Verizon Communications is now.
Paying down the $62bn in debt is the other rationale given for a sell-off. Verizon has lost about 10% of its value on the deal and its share price has not yet recovered on an improved outlook. Investors look at its new balance sheet, say that it is over-geared and suggest that Verizon sell something – and the only thing they can find big enough to bring in that amount of cash is the fixed line network.
The truth is, that if Verizon wanted to sell of some component parts, it would sell the unvectored DSL broadband lines, of which it has 3.2m, since these very much in the descent.
But the truth is that these fall into at least two camps, those which can be upgraded to FiOS, and those which cannot because FiOS broadband does not pass the door of these homes. Verizon has 5.7m FiOS internet subscribers and 5m TV subs, and these bring in some 71% of consumer retail revenue ‘ which amounted to $3.6bn last quarter. Now 3.2m broadband lines that are tough to push onto higher speeds are not worth very much at all, so selling even a small part of these is not going to bring in very much cash ‘ nothing like $62bn. Selling the whole thing, which would probably mean around $14bn plus of retail revenue – is also not going to get near denting this debt. And anyway the non-upgraded homes are worth more to Verizon’s FiOS than anyone else.
Another reason given is that Verizon doesn’t plan to build out FiOS any further, and that it is not making back sufficient returns on what’s there. It is perhaps true that the FiOS service ‘ which relies on fiber that goes all the way to the home ‘ is making far more profit for the company in broadband than it is in TV additions and that perhaps Verizon does not want to take on more capex simply to upgrade its older systems to FiOS. It can offer vectoring to move these other broadband lines up, and it can hold what it has always held for the past seven years or so, about 8.9m broadband lines.
We have to remember that it now has partners in cable companies reaching other parts of its footprint that it doesn’t go to, and it could move to reselling DOCSIS 3.0 at some point, if the FCC would ever let it. Building out FiOS is, in its management’s mind, like overbuilding its partners’ networks. Why do it?
The other major issue here is what is on the horizon in heterogeneous wireless networks and small cells. Cellular operators are expected to have as much fixed line bandwidth on the floor, for the 10x growth in backhaul points it will need to pursue the strategy into LTE-Advanced. It will also need as much Wi-Fi as it can handle in the field ‘ whether this is hotspot build-out or its own homespots, in order to go down a path of Wi-Fi offload, the favored first step in upgrading cellular dual-mode device bandwidth.
So in a nutshell, its internal planning processes will make it abundantly clear that Verizon without wireline suffers horribly against AT&T and against its current partners in cable, who could crucify it with backhaul and Wi-Fi offload charges if it did not have a huge chunk of lines of its own, to barter with. What will happen over the next few years is that a flow of service from wireline to wireless will happen within Verizon, and it can reinvent the profitability of its wireline business, as its wireless unit buys more fixed line capability. And that will make that wireline business worth a lot more. So why on earth sell it off before that happens?
Operators round the world have seen wireline networks, for so long a legacy drag on their finances, becoming critical to the quad play strategies which are the only way for many to increase ARPU ‘ as prices of individual services, such as mobile, broadband and TV, fall, it is essential to capture more of a household’s total spend. This has driven Vodafone, for instance, to invest in wireline operators in Italy and Germany, and for a full alliance with cable giant Liberty Global to be mooted by some analysts. Similarly, Deutsche Telekom is investing in fixed line partners for its T-Mobile arm in some countries where the telco does not have wires in the ground, especially in the growth area of eastern Europe.
Even for operators ‘ of which Verizon is arguably one ‘ which have powerful partnerships with cable and fiber providers, there will still be advantages to owning wires, particularly to backhaul the rising tide of traffic on the carriers’ mobile and Wi-Fi systems, and to extract fees from wireless-only rivals in need of such capacity.
Carriers will certainly start make plans to retire DSL lines which have not been vectored (the technology being used to up the speed and capacity of qualifying copper wires where it is impractical or uneconomic to run fiber yet). And there will always be those areas where fiber is inappropriate, and where LTE will gradually replace copper as its own reliability and speed improves.
AT&T has made it pretty clear that DSL’s days are numbered in its own US strategy (though its huge Project VIP infrastructure plan enhances investment in fiber-to-the-home as well as in LTE, with no idea it might sell off one). In August, it launched its own response to HomeFusion, a
a home phone and internet service for selected markets, which runs over LTE or even HSPA. This aims to blur the distinction between wired and wireline, delivering a full suite of services over the cellular connection, especially in areas where AT&T has no wired lines, or only DSL.
Such services will help AT&T phase out DSL, or penetrate further into markets where it does not have fixed coverage, on the back of its cellular systems. The latest offering is being sold initially in Baltimore, Eastern Pennsylvania, Southern New Jersey, Virginia, Washington DC and other east coast locations (mainly Verizon wireline territories), which will be the proving grounds for a nationwide launch. The service works through a modem (free with two-year contract) which backhauls up to 10 Wi-Fi devices, and requires only a power outlet and AT&T 3G or 4G signal. It then offers unlimited national calls for $20 a month and internet service starting at 10Gbytes for $60 a month, up to 30Gbytes for $120 a month, with overage charges of $10 per Gbyte. More interestingly, users can include the service in existing Mobile Share data plans, with the home modem counting as one of the devices which can tap into a pooled data allowance. The product can be added to a Mobile Share contract for $30 a month.
AT&T CEO Randall Stephenson recently called on the FCC to accelerate the process of reviewing the carrier’s request to move to an all-IP, all wireless network nationwide, which would involve shutting down legacy wirelines. AT&T is seeking FCC guidance on the procedures needed to discontinue traditional wired TDM services and replace these with more efficient and profitable wireless links. He told a conference: “It’s a big darn deal, and will take investment, but the amount of cost it removes from the legacy [system] is dramatic. The biggest restraint is regulatory because we can’t just retire certain technology. Our all-wireless, all-IP, all-cloud is moving at a blazer pace.”
Project VIP effectively means the end of real investment in DSL and indeed, where users still have copper lines, AT&T increasingly recommends the use of LTE for fixed access if fiber is not available. (It does use DSL for the last mile for its U-verse fiber system in most areas, but is defocusing on standalone copper lines). It says wireless, fiber-to-the-node and business networks deliver 81% of its revenue and are collectively growing at 6% a year, so it is clear where the investment priority must lie. AT&T last month petitioned the FCC to update its rules so that wired or wireless IP systems can be considered as a replacement for the traditional copper phone-line system.
But at the start of 2015, one-quarter of AT&T’s wireline base will still not have U-verse, which points to another strategic significance for LTE investment, to stop defection of DSL customers. While it may be efficient for AT&T to move to an IP-only system, critics say it will be a blow to rural users and to local CLECs which rely on the larger firm’s network ‘ factors which may attract FCC scrutiny or even encourage AT&T to sell off its DSL business as Verizon did. But selling off fiber and the chance to market a quad play over the most efficient combination of wired/wireless lines, plus backhaul a new generation of wireless technologies? That would certainly be the road to disaster for telcos fortunate enough to have the infrastructure and deep pockets to push out an all-IP, fiber/LTE platform, the ideal framework for a next generation business model, and one which wireless-only players like Sprint cannot emulate.
Multipath broadband?
As carriers weigh the either/ors of wireless and wireline, an Irish start-up believes users should be able to combine any connections they have to produce a better broadband speed. Multipath Networks has launched a campaign on the crowdfunding site Indiegogo for a multipath router, which can aggregate the bandwidth of, for instance, a DSL line and an LTE connection.
Founder and CEO Justin Collery says the device could also be used to share a connection with a neighbour. The router retails for about $200 and can combine up to four connections to boost connection speed and reliability, There is also a $5 monthly subscription fee after a free trial period. This covers the cloud-based intelligence that makes the product work.
While combining links is not a new idea, the speed of modern wireless, cable and broadband technologies makes the prospect more disruptive, and the start-up has some technology of its own driving the platform. It uses Multipath TCP, which is also being harnessed by Apple to allow the iPhone’s virtual assistant, Siri, to combine 4G and Wi-Fi networks.
Unlike previous aggregation attempts, the start-up’s router does not actually send packets in order to test how fast a connection is, which is data hungry, but instead measures the difference in time between a request for a packet and the receipt, and gauges the connection characteristics using the cloud-based algorithms. The router also fills up the lowest cost connection first to keep costs down, a feature which could help justify the monthly fee and upfront cost.
Eventually, like most innovations which initially rely on consumers forking out, Multipath aims to build out an ‘as a service platform’. It would buy connections from operators and resell them so that a single router could support several carrier, Wi-Fi and broadband networks and offer the best possible combination of speed and cost to the end user.