A draft order for the AWS-3 auction is circulating for a vote at the FCC, and while I have not seen the item, early reports of its recommendations suggest to me that the Commission staff has got this one exactly right.
Why?
The item proposes a band plan that includes two 10 x 2 MHz EA blocks and one 5 x 2 MHz CMA block. Incorporation of the large block and license sizes will not only ensure that the FCC drives the greatest efficiencies out of this spectrum, but also that it attracts the most revenue at auction. A quick study of the AWS-1 auction is instructive on these points.
Take yourself back to 2006 – the iPhone has not yet been introduced and most of us were sporting Blackberries and pushing maybe 70 MBs of data a month, mostly doing email but little else. No Facebook postings, no Twitter feeds, no app stores chock full of data-hungry apps. Carriers generally entered Auction 66 with very little awareness of the data tsunami headed their way.
Yet the bidding activity even back then clearly favored the larger license sizes and the larger blocks — both from an activity ratio perspective and from revenue received.
Let’s look first at the bidding activity ratios — a figure that attempts to capture first round bidding (or demand) against MHz offered (or supply). For example, if you are selling 100 units and have opening demand for only 100 units your activity ratio is flat at 1 and bidding competition is virtually non-existent. On the other hand, if opening bids are for 700 units against your same 100 units available, your activity ratio is 7 – which demonstrates high interest in the units and corresponding high bidding competition.
In the first round of the AWS-1 auction, the highest activity ratio by a long-shot (6.99) was for the 10 x 2 REAG licenses. The large block coupled with the large license size simply attracted the most interest (and dollars) out of the gate. On the other side of the auction, the 5 x 2 CMA licenses had little first round interest with an activity ratio of only 0.68 (which represents more supply than demand in the first round).
Final MHz/POPs revenue figures are also instructive. Again, the 10 x 2 REAG licenses topped the revenue charts at $0.73 MHz/POP, followed by the two smaller REAGs blocks ($.60 and $.58). The CMA-based 5 x 2 licenses anchored the auction at only $0.39 MHz/POP. The value of the larger block size is also apparent in the winning bids for the REAG licenses. In three regions, the 10 x 2 block sold for more than the two 5 x 2 blocks added together.
Given incumbent relocation demands (and the still unfunded FirstNet pot), a robust revenue stream from this auction is as essential as it will be in the 600 MHz auction.
The mid-band EA blocks in Auction 66 are interesting. They started with lower bidding activity than for the CMA licenses but ultimately sold for more, with the higher value attaching to the 5 x 2 block adjacent to the REAG licenses. I suspect as values went higher for the REAG licenses the bidding migrated down band with the greater value placed on the licenses adjacent to the REAG licenses. Contiguity has value. And we do know that some of the biggest winners of these blocks (SpectrumCo, MetroPCS and T-Mobile) did not bid for them in the opening round.
The only thing that surprises me about any of this is that these patterns emerged even in a pre-iPhone era where the average customer was accommodated with well under 100 MBs/month. Average usage profiles are quickly pushing toward 1 GB and one of AT&T’s most popular offers right now is a 10 GB shared bucket for four devices.
And in the broadband world in which we now live, 10 x 2 is now table stakes for an efficient LTE deployment, as a recent Peter Rysavy chart illustrates.
For these reasons, we think the FCC staff got this band plan exactly right. And the proposed 10 x 2 MHz EA licenses may be the last big opportunity for the Commission to drive spectral efficiencies through auction structure, as the 600 MHz auction structure demands 5 MHz fungible blocks.
If I had to quibble with anything, we would have preferred to see package bidding for at least one of the 10 x 2 blocks. We continue to believe package bidding addresses important exposure risks and can drive incremental revenue.
But, hey, life is not a box of chocolates, or something like that.