The Upcoming Spectrum Auction Might Disappoint

Mark Lowenstein / October 19th, 2015

On March 29, 2016, the FCC is scheduled to kick off the 600 MHz spectrum auction. The stakes are high because a significant chunk of bandwidth (50-80 MHz) is being made available and this is likely the last major swath of spectrum that will be auctioned for the foreseeable future. Additionally, the 600 MHz band is considered ‘beachfront’ spectrum, because it has better propagation characteristics than higher band spectrum. It makes it particularly attractive for operators who need to improve their coverage outside of cities and inside of buildings.

Many analysts believe this auction will be a bonanza, raising upwards of $60 billion, similar to the 2014-15 AWS-3 auction, which raised $45 billion (a record $2.53/MHz-pop), largely from AT&T, Verizon, T-Mobile, and DISH. Data traffic over the wireless network is expected to grow by nearly 50% per year through 2020, driven by video and other rich media. The two largest wireless carriers, AT&T and Verizon, hold two-thirds of wireless subscribers but less than one-half of industry capacity and T-Mobile needs more low-band spectrum in order to be competitive with the “big boys”. Some believe that outside players, like cable companies or perhaps even Google or Apple, might participate in this auction in order to become masters of their own wireless domain.

I have a contrarian view. I don’t believe the auction will be a dud, but it could well disappoint. The markets could go for less money (on a MHz per pop basis) and fewer companies may participate, in a serious way, than anticipated.

There are three main reasons why I am more conservative on the prospects for this auction. First, the 600 MHz auction is extraordinarily complicated, with uncertainty as to exactly how and when the spectrum will be available to be put to commercial use. Without going too deeply into the details, this is a pioneering ‘incentive’ auction, in which the incumbent TV broadcasters share their spectrum or move channels in return for receiving part of the auction proceeds. There is a ‘reverse’ process and a ‘forward process’ (more details here). There is also contention about exactly what the rules are and how it will all work. And there is an uncertainty as to how, and when, the incumbent broadcasters will move off that spectrum – especially in cities. The best-case scenario is this spectrum will be available for commercial use in 2019-2020, but that timeframe could easily move out a couple of years, especially in the larger urban markets. Auction prices could be depressed if some operators believe there are better, alternate mechanisms for obtaining spectrum and putting it to work sooner.

Second, circumstances in the wireless market have changed over the past 2-3 years. Yes, all of the major operators will need more capacity in the coming years to address the expected growth in data consumption. However, the financial resources they need to bid for this spectrum have become somewhat more constrained. We are coming out of a fairly vigorous wireless price war in the U.S., which has affected operator margins. The operators are also quite highly leveraged: AT&T acquired DirecTV, is venturing into Mexico very aggressively (and building out an LTE network), and is upping its spend on fixed broadband in the U.S.; Verizon, whose balance sheet is still feeling the effects of the Vodafone buyout, acquired AOL and is saving some of its powder for other potential acquisitions (in my view); T-Mobile has a conservative corporate parent in the form of Deutsche Telekom which will, at some point, want to wring some profits from its profligate progeny; DISH is in the FCC’s financial penalty box due to the AWS-3 Designated Entity debacle and does not particularly need this spectrum; and Sprint has said it will not participate in the 600 MHz auction (although some have speculated that SoftBank, Sprint’s owner, might). Operator CFOs are going to be more disciplined about this auction than the last time around.

Third, there is more spectrum available right now than many people think and it could creatively be put to work sooner than the 600 MHz spectrum. Examples:

• Unlicensed LTE. Progress is being made to allow wireless carriers to use channels in the unlicensed 5 GHz band (also used for Wi-Fi) to add capacity to their LTE networks. Verizon and T-Mobile are pushing hardest on this.

• Sprint 2.5 GHz spectrum. Given Sprint’s financial challenges, it is possible Sprint could more aggressively court a wholesale MVNO partner, such as Google or DISH, or sell some of this spectrum outright to raise funds.

• DISH. I believe there will be some action involving DISH prior to the 600 MHz auction. It could do a deal with Sprint, or possibly make a bid for T-Mobile, as has been rumored. Verizon could also make a play for some of DISH’s spectrum.

• Spectrum sharing possibilities. The FCC is pushing a spectrum sharing proposal for the 3.5 GHz band, which could be used for small cells; there are spectrum sharing possibilities as well, using “TV White Spaces” in the 600 MHz band; and possibilities for some excess spectrum being made available as part of the FirstNet build.

Putting unlicensed band and spectrum sharing possibilities to actual use will require unique levels of intra- and inter-industry cooperation, as well as a daft FCC (in an election year).

Finally, I am less optimistic than some that non-incumbent players will be major players in the 600 MHz auction. For the wireless historians among us, Cable is already 0 for 3 in the facilities-based wireless business. The TWC-Charter-Brighthouse triumvirate is caught up in merger approvals so the timing is not right for them. Comcast is the cable elephant in the room and, despite their aggressive investment in Wi-Fi, I don’t believe they are convinced they have to buy spectrum if they want to make a cellular play. As for Google, Apple, Facebook, and other potential auction participants, I think there will be more attractive potential wholesale opportunities than there have been in the past. Plus, I don’t see a compelling business case for an additional facilities-based wireless provider, given four existing national LTE networks, plus DISH, out there.

It is not necessarily a bad thing if the 600 MHz auctions are not as lucrative for the Treasury Department as hoped for. After all, the purpose of spectrum auctions is to make more capacity available for wireless use, not to raise money for the federal government. And if wireless players spend a little less than planned to buy the spectrum, conceivably they’ll have more resources to build more network more quickly.

Mark Lowenstein

Mark Lowenstein is Managing Director of Mobile Ecosystem, an advisory services firm focused on mobile and digital media. He founded and led the Yankee Group's global wireless practices and was also VP, Market Strategy at Verizon Wireless. You can follow him on Twitter at @marklowenstein and sign up for his free Lens on Wireless newsletter here.
  • Aaron

    Hey Mark, thanks for the good read. What do you mean by saying AT&T and Verizon have “less than one-half of industry capacity”? Do you mean spectrum or built out data capacity?

    • Aaron,
      Thanks for reading. AT&T and Verizon each have in the low-20% range of industry capacity, by which I mean the total amount of spectrum they own. Not built out, or data used, etc. Just actual spectrum capacity. I realize this is not a complete apples-to-apples comparison (low-mid-high band spectrum, etc.) but it is one of the reasons Sprint and TMO are able to be more aggressive with data plans, as they have more spectrum per active subscriber than AT&T and Verizon.

  • jfutral

    From your lips to God’s ears. My industry (the performing arts) is tired of needing to spend money (most of the industry doesn’t have, especially the non-profit sector) to change all our wireless equipment (wireless microphones and communications) because what used to be available gets auctioned off by the government. We’ve already been lied to once about needing to change our frequencies. Maybe now that will die down for a while.


  • Google Fi and Republic Wireless have shown that you can be a wireless carrier with a low CapEx by leveraging WiFi to the max, and buying wholesale cellular for a sub 20% portion of your user’s traffic. This reality alone lowers the value of spectrum.

    Based on the above, cable operators would almost be nuts to aspire to be facilities-based cellular carriers. Their competitive advantages are: existing branding, existing telco experience, existing retail, existing service trucks and CSR, and Wi-Fi footprint. Today’s technology trends are obviously towards all-IP based networks and devices, voice included. It would be a nutty investment to take those trends and build a 1995-type cellular competitor. Especially when most national “experiments” have shown that having over 3 infrastructure-based cellular carriers erodes overall market profitability.

    Spectrum has never been a bad investment. Seems like you and I think that may become less certain.

    “After all, the purpose of spectrum auctions is to make more capacity available for wireless use, not to raise money for the federal government.” So true. The value spectrum creates is only loosely related to the price paid, and that’s ALWAYS been the case.

Protected by Gerben Law