FCC tries to tie down crucial elements of underlying auction design, including definitions of “clearing targets”, “spectrum impairments”
While the schedule may have slipped some, the Incentive Auction is still on its way. And even if the Auction may not start until sometime in 2016 (at least according to the current thinking), the Commission is facing – now – the monumental task of working out the myriad details that will govern the auction process. To that end, last month the FCC invited comment on a mind-numbing array of highly technical questions about both the reverse and forward components of the Incentive Auction. We’ll summarize a few of the highlights concerning the reverse auction. (And let’s be clear, this is just a summary of about 80 pages – i.e., half of the 160+-page request for comments – of densely-packed Commission-ese.) We’ll address forward auction highlights in a separate post.
Setting the Clearing Target and Impairment, Categories of Licenses
The principal goal of the reverse auction is to clear UHF spectrum of current TV broadcast licensees in order to make that spectrum available to wireless operators. A couple of years ago, the Commission was hoping to be able to clear a nationwide “clean swath” of spectrum amounting to as much as 126 MHz. The thought was that the reverse auction software could be set up to use that level as a starting point from which the on-going reverse auction calculations would be based. In other words, unless a set of particular reverse auction deals would clear that pre-identified amount of spectrum, those deals would be non-starters.
The FCC now seems to be accepting the reality that using a fixed 126 MHz starting point (or even some lower level, like 108 MHz) might be a tad ambitious because of various practical constraints (for example, border interference considerations). As a result, it is now proposing an approach that would rely on a dynamic, rather than static, clearing target concept. The target would be the highest clearing target possible from among the available options depending on broadcaster participation in the reverse auction. But the Commission makes clear that it would like to hear from commenters relative to the notion of omitting any initial clearing targets.
“Impairment” and the Definition of “Cleared” Spectrum
Regardless of the starting point, however, it will in any event be essential for the Commission to determine precisely how the notion of “cleared” spectrum will be defined. And that gives rise to additional problems.
In some markets, certain TV stations will need to be relocated somewhere within the reallocated 600 MHz band, meaning stations so relocated will find themselves in a band populated by wireless providers.
Such proximity threatens interference (both to wireless and to broadcast service), which will be calculated using the “inter-service interference” (ISIX) methodology.
Wireless licenses subject to such interference from a TV station would be considered “impaired”. (Wireless licensees would not be permitted to cause interference to TV broadcasters, but they could agree to operate in areas where they might suffer interference from TV service.) But just how much potential interference will cause a wireless license to be deemed “impaired”? And should “impaired” spectrum be included in the calculation of cleared spectrum for clearing target purposes?
The Commission suggests that the interference threshold should be in the range of 10-20% of the downlink, and it’s looking for comment on whether that range would be appropriate. Comments are sought as to whether a threshold for impairment to uplink service could be significantly higher – possibly up to 50% – than the downlink threshold. Whatever the threshold ultimately adopted, a county subject to impairment over that threshold would be deemed to be “impaired”, which would then be taken into account in calculating the relevant clearing target. In addressing these questions, commenters also might consider why the proposed methodology for determining impairment is based on county-level data rather than the partial economic areas (PEAs) that will be used to apportion wireless licenses for auction.
In any event, once the Commission has resolved these preliminary questions, it proposes that the initial clearing target will be determined as one in which impairments, on an aggregate nationwide basis, affect less than 20% of the total U.S. population, weighted to reflect geographic variation in license prices as determined in prior auctions.
Price Break for Impaired Wireless Blocks
In addition to its importance in determining the clearing threshold that will be crucial to the operation of the reverse auction, the definition of “impairment” will be similarly crucial in the forward auction. The extent of impaired spectrum in a given block of wireless spectrum available for auction will affect that block’s price. The Commission proposes to place in the forward auction two categories of generic license blocks:
- Category 1 blocks, which would include no more than 15% (and as little as zero) impaired population; and
- Category 2 blocks, which would include spectrum subject to more than 15% but not more than 50% impairments.
Prices for both categories of blocks would receive a 1% discount for each percentage of impaired pops. Blocks with more than 50 percent impaired pops would not be offered.
Reverse auction bids methodology
The overwhelming success of the recently closed $45 billion AWS-3 auction demonstrates without question that demand exists for wireless spectrum. Presumably, then, the success of the Incentive Auction will depend on the ability of the reverse auction to make enough wireless spectrum available. And that, in turn, will depend on the willingness of broadcasters to clear the spectrum to be repurposed for wireless operators.
Because of the need to encourage broadcaster participation, opening prices offered to broadcasters in the reverse auction must be high enough to draw broadcasters into the process. BUT they can’t be so high as to lead to infinite rounds of reverse bidding or, worse still, result in an aggregate financial obligation to broadcasters so high that the forward auction of the clear spectrum won’t generate enough proceeds to satisfy the auction’s final stage rule.
How to walk this delicate line? The Commission has developed a methodology for determining each station’s opening price.
The FCC proposes that a station’s opening price first be weighted based on a number of factors that will affect the station’s impact on the repacking process. The Commission will look at (1) the “interference” for which the station’s continued use of its current channel would be responsible and (2) the amount of population it serves. Taken together according to the following formula, these factors would be called the station’s “Volume”:
Station Volume = (Interference)0.5 * (Population)0.5
(This concept has also been referred to colloquially in the broadcast community as “scoring”.) “Population” here means the number of people residing within the station’s interference-free service area. “Interference” is a far more complicated notion which, for our purposes here, may be summarized as “the number of co-and adjacent channel constraints a station would impose on repacking on a pairwise basis.”
Once the Volume has been calculated, the opening price will be figured by multiplying the station’s Volume by a “base clock price”. What’s a “base clock price”? We probably couldn’t do it justice, so we’ll just quote the FCC’s own description:
The base clock price is a constant amount per unit of volume. Based on our work to date on the design of the incentive auction, we expect that a base predicated on an opening bid price of $900 million for the station with the highest volume will achieve robust participation by stations across multiple markets. We therefore propose to set the base clock price so as to yield an opening bid of $900 million for this station. To do this, we will calculate volume for all stations and then rescale so that the maximum station volume is one million. Dividing the $900 million opening bid price for the highest volume station by one million results in a base clock price of 900. The base clock price will drop in each round of the reverse auction, while a station’s volume will remain constant. The price offered to a bidder to go off air in a given round will be the product of the base clock price in that round and the station’s volume. The markets and stations needed in the reverse auction will depend on which stations choose to participate, and actual compensation to stations will be determined by the auction.
So Volume * Base Clock Price = Station Clearing Bid. The base clock price will decrease in each subsequent round; the Volume will stay constant.
There are some further possible quirks. For example, if a UHF station opts to move down to a low VHF band, its opening price would be discounted to a percentage (in the 67%-80% range) of the opening price it would get for fully relinquishing its license. If the station opted to take a high VHF channel, the discount would reduce the opening bid further (to perhaps 33%-50% of the full relinquishment price). Other similar discounts could kick in in later bidding rounds.
Stations will be required at the outset to identify all the options they might accept. At each auction round, the broadcaster will be shown bid prices for the various options available to it at that point – i.e., the price offered for full relinquishment; the price for a move to high UHF (if available in a proposed reallocation analysis); and the price for a move to low-VHF, if available. Broadcasters will also get to see at each stage whether repacking is feasible and what that will entail. So long as there are competing stations still accepting the descending bids and the clearing target has not been met.
The bids for a specific station will stop descending and a provisional bid will be accepted – meaning that that station will be “frozen” in terms of further auction activity – when the repacking feasibility analysis determines that there is no way the station could be assigned a new channel in its pre-auction band. In other words, the bid is accepted at the point where the station must go off the air. Keep in mind that one station’s bids might be “frozen” while the bids continue for other stations. It is also possible that a station might become “un-frozen” if the landscape changes due to stations nearby dropping out of the auction.
The Commission is looking for input as to whether these proposed formulae are appropriate and whether the opening bids are likely to result in broad participation leading to a successful auction. Commenters are specifically asked for input on the proper discounting of UHF-to-VHF moves. In addition, the Commission asks whether intra-round bidding should be permitted – i.e., that a broadcaster should be able to indicate within a round the price at which he or she would relinquish the station’s spectrum rather than accepting the offered lower price in the round. Finally, broadcasters should consider the ramifications of possibly being frozen, then unfrozen, then frozen again, during the course of the auction.
If this is not complicated enough, there is another wrinkle: dynamic reserve pricing (DRP). Because (obviously) there are a lot of moving parts here, it’s possible, if not likely, that in the early rounds some bids to broadcasters could be unduly decreased by an apparent – but not necessarily actual – lack of alternate available channel reassignments. In other words, the FCC is concerned that the auction would fail in Round One if opening bids were predicated on the notion that alternate UHF channels are available for relocation of all participating stations. To put this in context, in a congested market like Los Angeles, there are few, if any, open UHF channels, and certainly there won’t be any if the band is shrunk down by a minimum of 13 channels.
But the FCC is confident that, if it keeps the auction rolling for a while, things will fall into place. To keep the bids high (and the broadcasters in the game), the Commission to use DRP to goose the system. The DRP mechanism would have broadcasters accept artificially lowered bids or risk being reallocated to the 600 MHz band. DRP, in theory, will promote enough spectrum to be relinquished so that repacking does become feasible at a later stage. However, it is unclear whether a station which drops out during DRP would then be then “frozen” into the 600 MHz reallocation. The FCC asserts that this DRP procedure will be discontinued once the remaining UHF stations could be repacked into the UHF band without unduly impairing the spectrum to be offered in the forward auction.
Commenters are asked to weigh on whether the DRP plan is reasonable, and if so, how long into the auction process the Commission should continue to be able to use DRP as a thumb on the scale. The FCC also asks whether the auction process should include a full channel reassignment optimization analysis – a time-consuming chore – during the use of DRP, or whether a limited version of the analysis would be acceptable. Commenters may also want to consider whether DRP could artificially lower auction bids, thus causing diminished participation and/or the reallocation of too many stations back into the 600MHz band so as to make the forward auction unworkable in many places.
There are plenty more proposed auction details in the invitation for comments, but these should give you the idea. Again, it’s a very dense, highly technical document which underscores the incredibly complex interrelation of the forward and reverse auctions. It reminds us all of the importance of getting each and every detail right, because a misstep with respect to any detail here or there could send the entire process off-track. Readers inclined to immerse themselves in such things would do well to read the entire request of comments and provide input on the FCC’s suggestions. The deadline for comments is February 20, 2015, and for replies it’s March 13, 2015. Comments can be submitted through the FCC’s ECFS online filing system; refer to Proceeding Numbers 14-252 and 12-268.