This week, the FCC’s Broadcast Incentive Auction brought two big surprises. On Monday, the FCC held a webinar (slides here) reviewing its post-incentive auction transition plans and it showed that the webs of interference dependencies were hugely more complex than at least some of us thought. Then yesterday, the forward auction component (where wireless companies bid on spectrum that broadcasters are willing to sell) of the auction’s second stage opened and closed some two hours later after the first round. I’m wondering if the two might be somehow related.
In one pre-auction scholarly paper I read, one daisy chain (a string of dependencies with a beginning and an end) was given as an example that extended from Miami into North Carolina. However, the webinar slides showed one daisy chain of 29 stations that extends from central Florida into Canada!
Then there are sets of dependencies called cycles. Cycles are effectively closed daisy chains – they have no beginning or end. For example, station A must transition before station B, station B before station C, and station C before station A. Here’s a cycle of 196 stations that covers most of eastern North America:
The FCC has clearly thought out the problem in great detail and is asking for comments on the strategies it presented. Remember, it has to accomplish this repack in 39 months after the auction closes. Typically, the broadcast industry has expressed skepticism that this is possible, while the FCC and wireless companies (especially T-Mobile) have expressed optimism.
With dependency webs like those above, what could possibly go wrong?
Phil Kurz of TVNewsCheck (see “Repack Plan Draws Tepid Industry Reaction”) asked various industry leaders a version of that question, and their responses are sobering. There are a lot of things that can and will go wrong, from supply chain to tower issues to weather. Also see, "Destination Repack: Not Enough Time, Money" by Tom Butts in TVTechnology. The mathematical probability of delays for stations in these complex webs of dependencies seems extremely high and there doesn’t seem to be much if any margin for error.
So, back now to yesterday’s super quick close of the stage 2 forward auction. My first thought when this happened was that they must have lost one of the big dogs out of the 66 registered bidders. There is some evidence for this. In RCRWireless News yesterday, Dan Meyer has this quote from PwC [emphasis added]:
“The rapid close of the second stage of the forward auction shows just how delicate the balance of supply and demand is for the 600 MHz spectrum in question,” explained Dan Hays, principal at PwC’s Strategy& division. “Unlike the first stage of the auction, which saw a gradual decrease in demand over multiple rounds, the demand profile in the second stage’s single, forward round fell off quite rapidly. Based on the pattern, this could well indicate the sudden exit of one nationwide license being sought, as demand in nearly every major market fell simultaneously from 10 licenses to nine. … Today’s results have further increased the odds that stage three of the auction, which we expect to commence toward the end of October or early November, will be followed by a fourth stage as well. As a result, we could well see auction activities trail into early 2017.”
If this turns out to be correct, then a possible tie to the repack transition plan might be increased anxiety on the part of one of the potential buyers that the 600 MHz band won’t be ready for them when they need it. Major wireless carriers have other options for getting the spectrum they desire, not all carriers are as needy as others for spectrum in this band, and the clearance process introduces risk into their build-out plans that at least one carrier might have found unacceptable.
P.S.: At FierceWireless, Colin Gibbs relays speculation that it might be Comcast who bowed out, though for a different reason than my speculation above. See Tim Farrar's @TMFAssociates Twitter feed. And a friend passed on the thought it could be EchoStar. Lots of candidates.