700 MHz: Breaking the C Block Package

I apologise for the hiatus between this and my last 700 MHz Auction update, but with 36,419 bids over 261 rounds, analysing the data set is taking a bit of time.

Among the several controversies arising the from now-completed auction has been ATT’s claim that bidders were deterred from bidding on C Block because of the open access rules imposed on the block. I can say with confidence that this is a bald-faced lie.

Twenty-six companies bid on C Block spectrum: Alltel Corporation, AST Telecom, LLC, AT&T Mobility Spectrum, LLC, Bluewater Wireless, L.P., Cellco Partnership d/b/a Verizon Wireless, Cellular South Licenses, Inc., CHEVRON USA INC., Choice Phone LLC, Club 42 CM Limited Partnership, Copper Valley Wireless, Inc., Cox Wireless, Inc., Cricket Licensee 2007, LLC, Google Airwaves Inc., King Street Wireless, L.P., Thomas K. Kurian, MetroPCS 700 MHz, LLC, NatTel, LLC, PTI Pacifica, Inc., Pulse Mobile LLC, QUALCOMM Incorporated, SAL Spectrum, LLC, SeaBytes, L.L.C., Small Ventures USA, L.P., Triad 700, LLC, Vulcan Spectrum LLC, and Xanadoo 700 MHz DE, LLC.

Note that the lying buggers at ATT bid on REAGs 2 and 4. They were deterred, but only by Verizon’s deeper pockets.

The interesting dynamic in C Block is the effect of combinatorial bidding on the outcome. Under the combinatorial bidding rules three packages of REAGs were available (the 50 state package, the Atlantic package, and the Pacific package) as well as the individual REAGs. The rules provided that so long as the bid on a package exceeded the total amount of the bids on all the individual REAGs in that package, the package bidder would win (assuming that the package bid reached the reserve price). If the total amount bid on the individual REAGs exceeded the package bid in a round, then the package was “broken” and the package bidder wouldn’t be required to take any REAGs if it couldn’t have the whole package (this was to prevent a bidder who wanted a national footprint from getting stuck with less if another bidder outbid on one or two crucial components of the package).

Echostar was a strong proponent of combinatorial bidding, insisting that they wouldn’t show up and bid if the C Block did have a combinatorial bidding rule. Oddly enough, they got the rule and then their bidding entity, Frontier Wireless, didn’t even show in C Block bidding (they bid mainly in E Block without combinatorial bidding). But what they inadvertently did was screw at least one major bidder with the combinatorial bidding rules they insisted on.

More below…

Google bid exclusively on the 50 state package (REAGs 1-8), gradually pushing the bid up until they exceeded the reserve price in round 17 with a bid of $4,713,823,000. That meant that the open access rules would kick in. They then sat on that bid. There is speculation that Google’s board had a failure of will (and a serious loss of stock value), and decided to go no further in hopes the Verizon would outbid them and have to build the network under the open access rules.

However, the combinatorial bidding rules put MetroPCS in a real bind. MetroPCS really wanted REAGs 2 and 5, but the bidding was lagging on most of the individual REAGs, so they would have to bid far more than the value of those two REAGs to bring the total amount bid on individual REAGs high enough to break Google’s 50 state package. The situation sat that way until round 26 with bids on some individual REAGs creeping up. In round 26 QUALCOMM broke Google’s package with a bid of $1,405,293,000 on REAG 4, which sent the total amount bid on individual REAGs over Google’s package bid, $4,755,048,000 to $4,713,823,000. The dynamics from rounds 18 through 26 can be seen in this table; the bid amounts are in millions of dollars.

Seeing that the package had been broken, Verizon, which reasonably assumed that the package bid was Google (remember that all people knew was the high bid in each round, not who had made it, because of the anonymous bidding rules), knew that there would now be a scramble for the individual REAGs. Between rounds 27 and 30 Verizon swooped in with large bids and secured the vast majority of the individual REAGs.

Now how does this screw MetroPCS? According to a reliable source, by the time the package was broken MetroPCS had spent enough eligibility on bids in other blocks that it didn’t have the resources to counter Verizon on the two REAGs it wanted. If the package hadn’t had to be broken for MetroPCS’s high bids on REAGs 2 and 5 to prevail, they’d have known that they were in a fight with Verizon early enough to redeploy resources to counter Verizon. I haven’t had the time to run all the numbers, but preliminary indications are that this scenario is very plausible.

Thanks, Echostar. And they didn’t even show up to play by their own rules in C Block.

I predicted, when the combinatorial bidding rules were proposed, that all they would do was advantage bidders with the deepest pockets, i.e., the biggest incumbents.

Anonymous bidding helped competition, the open access rule at the least didn’t hurt it, but combinatorial bidding was a competition-killer.