The approval includes a requirement for AT&T to add 12.5 million fiber to premise locations and 13 million fixed wireless connections.
The fiber to customer deployments are supposed to happen over a four-year period, including about one million by the end of 2016.
Some question whether that is going to happen, and how soon, given fixed network capital investment of about $2 billion a year at the moment.
In addition to boosting capital spending, AT&T might be banking on the ability to build more affordably than was the case when Verizon Communications installed most of its fiber to home networks.
Indeed, Verizon’s unexpected decision to deploy FiOS in at least some neighborhoods in Boston suggests something has changed in the perceived business model.
Perhaps one key element is the difference between cost to “pass a location” and “cost to connect a customer.” AT&T could pass many more homes than it “connects,” as perhaps half the total cost of activating a customer is related to installing drops and network interface units for each paying customer.
The amount of deployed capital therefore includes a fixed element (fiber pass a location) and a variable component (cost to activate a location).
Forced to predict, some of us would argue that many more passings will be covered than “connected,” as initial take rates for consumer optical fiber connections can be as low as 20 percent in the first year.
So, of 100 locations passed, about 20 will require additional capital to activate. Assuming new deployments are targeted neighborhood by neighborhood, where propensity to buy is the highest, the amount of stranded capital is reduced.
Back in 2008, it might have cost $3,800 just to pass a location. Now it might cost $600 or less per passing. But you can see the reason for skepticism in some quarters.
It might cost $600 million to pass a million homes, at $600 per passing. At $500 per passing, it still costs $500 million to pass a million homes.
Using the $500 per passing figure, AT&T would have to invest at least $1 billion in capital to pass two million homes. On an annual fixed network capital budget of $2 billion, that suggests AT&T might be able to add about two million passings a year.
That works out to a total of about eight million over four years, short of the total of 12.5 over four years.
So, yes, one might argue, capital investment would have to climb beyond $2 billion a year to satisfy the FCC requirements.