The New AT&T: It’s about the spectrum

You can make as many arguments as you like about what AT&T (NYSE:T) stands to gain by buying T-Mobile (NYSE:DT), but in my view it boils down to just one: frequencies. Sure, AT&T would gain another 34 million subs, it would have access to a largely built high-speed packet access plus (HSPA+) network, it would leapfrog over Verizon Wireless (NYSE:VZ, NYSE:VOD) in scale and sheer dominance in the U.S. market. But I think none of that justifies the huge integration headaches and costs necessary to knit their two networks and operations together (the Cingular-AT&T Wireless integration is probably still fresh in memory).

However, with a single stroke AT&T would alleviate all of its spectrum worries for the next few years, giving it not only a consistent and large band over which to launch LTE but also more options than it knows what to do with when it comes to future network planning. AT&T certainly isn’t hiding the fact. In multiple interviews and presentations AT&T has stressed how the deal will relieve it of having to ferret out new spectrum in the near term, while its competitors may not be feel so comfortable (CP: AT&T: T-Mobile deal would produce a bigger, better operator).

So how big exactly would AT&T’s spectrum position be post-merger? Conveniently Bernstein Research produced a handy analysis covering just such a scenario (see chart below).

Assuming its acquisition of Qualcomm’s 700 MHz FLO spectrum is approved (Unfiltered: Fight over 700 MHz brewing into proxy battle) as well as the T-Mobile acquisition itself, AT&T would have a whopping 132.4 MHz of spectrum on a population-weighted average per market. In comparison, Verizon Wireless and Sprint (NYSE:S) would have a combined 140.2 MHz. Clearwire (NASDAQ:CLWR), which often brags about its exalted spectral holding has an average of 100 MHz in most markets and maxes out at 120 MHz in its most license-rich markets. Of course, the FCC and Justice Department will most certainly make AT&T divest that spectrum in many markets. But any resulting deal will most certainly still leave AT&T with the premier spectrum position in the U.S., both in terms of quantity and quality.

Let’s start with AT&T’s stated plans. It would deploy a long-term evolution (LTE) network covering 95% of the U.S. population using AT&T and T-Mobile’s combined Advanced Wireless Service (AWS) holdings.

According to Bernstein’s numbers AT&T would have a population-weighted average of 33.4 MHz in AWS per market. That would give AT&T the 40 MHz it needs to deploy LTE in a 20 MHz-by-20 MHz configuration—either as two carriers or a single mega-carrier. That’s twice the bandwidth of Verizon’s current LTE network at 700 MHz, and a third larger than the three 10-MHz unpaired WiMAX channels Clearwire currently uses in most of its markets. Clearwire definitely has room to grow in its 2.5 GHz band, but Verizon is limited in its LTE expansion to areas where it bought extra 700 MHz and AWS licenses.

Furthermore, AT&T would still have 700 MHz spectrum averaging 17 MHz per market, allowing it to deploy an LTE overlay that fills in coverage holes while using its AWS LTE network primarily for capacity. The lower propagation makes 700 MHz ideal for punching through walls and creating big cells at the edges of metro markets. Just how it deploys the network is still an open question since it would require integrating Qualcomm’s unpaired FLO spectrum into a dual-channel framework (Unfiltered: AT&T would have few options for FLO spectrum). Unless AT&T pulled off some radical network engineering feats, it might have to limit itself to a 5 MHz-by-5 MHz carrier at 700 MHz, but that would be more than adequate for a network designed primarily for data coverage rather than capacity.

More: But that's just scratching the surface