isen.blog

I've been reading a May 2004 study by Telcordia and Sanford Bernstein (the investment house) called Fiber: Revolutionizing the Bells' Telecom Networks. The study paints a scenario in which Fiber to the Premises (FTTP) built by Bell South, SBC and Verizon would become profitable in eight years. The scenario envisions a five-year buildout in 20 states to 50% of residences and 80% of businesses. Astoundingly, the cost of this build would be a mere $27 billion; this would be spread over the five year period. The capital expenditure (capex) would be higher -- $48 b -- but significant savings from the newly-built fiber would begin almost immediately, mostly in reduced operational expenses (opex). New revenues from fiber-based services would also help, as would savings from reduced copper investment.

[Aside: I visited the NJ Bell 5ESS in Murray Hill NJ, just down the street from Bell Labs, in 1996. Most of the interior space -- one large-ish room -- was taken up by the "wire frame." This is the huge patch panel where local loop wires come in from the neighborhoods to be connected to the switch. The supervisor giving the tour then showed us the brand-new OC-12 box. This was about the size of an under-desk PC, and it sat all alone in a frame. It handled about 1000 twisted pairs -- it had one fiber going into it and a couple going out. The corresponding copper lines took about 8 feet of rack space. In the case of fiber, all the "provisioning" was done in software. In the case of copper, it is done in software.]

The Telcordia-Bernstein report quantifies what I witnessed in 1996. It shows wireline opex as 43% of total current incumbent telco expense structure. 43%!!! Most of this is in Central Office (CO) expense, and it says that the ILECs would save 100% of CO opex by switching to fiber. Other opex would be reduced by 30-90%. At the end of the buildout, 57% of the fiber's benefit would be from opex savings. Repeat: the case for FTTP is driven by opex savings more than by new service revenues.

The study looks at a comparable Fiber to the Neighborhood (FTTN) buildout, where the last few hundred meters are via DSL. It says, "We came away disappointed." While FTTN is but 5% of the cost, it would deliver 8% of the revenue and none of the cost savings of FTTP.

The authors write that the prospect of this FTTP buildout would be met by "investor derision." They conclude that the Bells are "damned if they do, damned if they don't." However, if they don't, the telcos face "long-term earnings stagnation worse than anything they have experienced." This "dour outlook" is "much gloomier" than the prevailing consensus. The report shows wireline voice revenues falling by 4.5% per year over the next five years while wireless revenues flatten. Meanwhile, the cost of keeping a copper access line has been growing, and this is expected to continue to grow.

Studies like these always have problems. This study does not anticipate the effects of future regulatory changes. It does not explicitly cover the WiFi phenomenon, or additional unlicensed technologies that are likely to burst into the marketplace, or innovative new architectures that might suddenly arise in unlicensed spaces. It is not clear whether the study's estimate of VoIP's effects takes end-to-end (aka "free") VoIP into account. But the study rings truer than many I've seen.

Will the ILECs actually do something like what the study envisions? Today I'd bet they won't. And if they tried, I'd bet they wouldn't be able to execute, or that they'd take years longer than planned, or have cost overruns, or Cheney-up in some other way. On the other hand, Telcordia knows ILECs, and ILECs listen to Telcordia. So if I were a betting man, I'd be hedging those bets. The ILECs can't afford not to do it.