Interested in further auction analysis and wireless strategy? Contact us at or 703-506-0041.

Friday, February 8, 2008: Bidding at the end of Round 51

License MHz Aggregate Total PWBs Aggregate Reserve Delta % of Reserve Total Pops Current Cost per MHz-pop
A 12 $3,882,309,900 $1,807,380,000 $2,074,929,900 215% 285,620,445 $1.13
B 12 $9,089,970,400 $1,374,426,000 $7,715,544,400 661% 284,636,891 $2.66
C 22 $4,748,146,000 $4,637,854,000 $110,292,000 102% 285,620,445 $0.76
D 10 $472,042,000 $1,330,000,000 $(857,958,000) 35% 285,620,445 $0.17
E 6 $1,036,700,400 $903,690,000 $133,010,400 115% 285,480,684 $0.61
Total 62 $19,229,168,700 $10,053,350,000
Graph of FCC Auction 73 New Bidding Units at the end of Round 51

(Click to enlarge)

Graph of FCC Auction 73 Regional Cost Per MHz-Pop at the end of Round 51

(Click to enlarge)

As the activity in the A and B blocks slowly ramped down today, and C and D lay idle, the bidding war in the E block raged on, and pushed the average cost per MHz-pop to $0.61, and increased the total PWBs by $116M on the day. The E block is now 15% over reserve, and has been the big money winner for the FCC the past three days. Interestingly, the E block battle expanded into the New York license in Round 46, a license which hadn't seen any action since Round 9. It appears that the Miami and DC licenses were spared to make room for NYC's 7.5M bidding units, so the attacker would not exceed eligibility limits, and Detroit dropped out a couple rounds later. But in Round 50, the war for the biggest licenses subsided a bit when the bidder who waged the original attack in Round 37 placed bids only on NY, Boston, Miami, LA, and SF from the original list, but added Denver, Phoenix, Kansas City and several other smaller licenses to keep the overall activity level consistent with previous rounds. All of these were promptly taken back by the victim (who is most likely Qualcomm) in Round 51. On Monday morning it will be the attacker's turn to bid in Round 52, so we will see then if the battle continues..

As we pointed out yesterday, the war in the E block started in Round 37 when someone placed bids on several of the largest licenses, many of which had not seen any action since Round 8 or 9. Eleven of the top twelve most populous Economic Areas have been involved in this war, with only the Chicago license remaining unscathed since Round 36. The aggregate PWBs for the E block has increased $213M since the war started, and assuming the victim is Qualcomm, this will likely come out of their pocket. We assume the attacker took a waiver in Round 36 (i.e., the first round of Stage 2) and then placed bids on eleven of the biggest E block licenses, amounting to nearly 25 million bidding units. Looking at today's graph, one can see that the same pattern appears in the B block as appears in E between Rounds 35 and 37, so it's likely that the attacker was active in both blocks. This strongly suggests that the attacker is one of the biggest players in the auction, since only bidders with very high eligibility could have the headroom to make such bids in their first active round after the transition. So is it AT&T? Or is Verizon still using their power of eligibility to drain money from the pockets of their competitors?

The activity in the E block is important because it implies that C block bidding is not necessarily over. Mathematically it may still be possible for the attacker in E to switch back to C and place a bid on the Mississippi Valley license, assuming that the same bidder doesn't already hold the PWB on Mississippi Valley. But even so this is unlikely, since anyone with sufficient eligibility will be discouraged by the price tag: As shown in the second chart, at $2.36 per MHz-pop, the Mississippi Valley C license is the most expensive C block license, and the minimum bid is in excess of $1.6B. It appears that it was a strategic move by the C block mastermind back in Rounds 22 to 27 to intentionally bid up the Mississippi Valley license, since the bidder knew that it would be the most vulnerable of the six CONUS licenses because it has the smallest number of bidding units. Based on late auction bidding eligibility limits, Mississippi Valley is the C block license that would be easiest for a competitive bidder to target if that bidder wanted to break up a nationwide win. The other five C block licenses in CONUS are mathematically out of reach, but temptingly cheap (probably by design), and Mississippi Valley will be out of reach shortly if bidding in E subsides. This is why we very likely won't see any more bidding in the C block.

So why is there a sudden interest in the E block, a 6 MHz unpaired block of spectrum in the middle of the band? It's not clear if there is genuine interest or if it is an anti-competitive move, but there are at least three potential uses for the spectrum that bidders may be considering: 1) Mobile Broadcast, which is why Qualcomm is interested; 2) A TDD technology such as WiMAX, although there would be considerable interference to the TDD uplink from Qualcomm's MediaFLO broadcasts in the adjacent D block; and 3) Supplemental FDD downlink when combined with the A band. This last option would require some development work, but it is not impossible for a technology such as LTE to support an asymmetrical FDD configuration of 12 MHz downlink and 6 MHz uplink, and the recent activity could be an attempt to create A-E pairs in top markets.

Thursday, February 7, 2008: Bidding at the end of Round 45

License MHz Aggregate Total PWBs Aggregate Reserve Delta % of Reserve Total Pops Current Cost per MHz-pop
A 12 $3,869,574,900 $1,807,380,000 $2,062,194,900 214% 285,620,445 $1.13
B 12 $9,083,905,400 $1,374,426,000 $7,709,479,400 661% 284,636,891 $2.66
C 22 $4,748,146,000 $4,637,854,000 $110,292,000 102% 285,620,445 $0.76
D 10 $472,042,000 $1,330,000,000 $(857,958,000) 35% 285,620,445 $0.17
E 6 $920,345,400 $903,690,000 $16,655,400 102% 285,480,684 $0.54
Total 62 $19,094,013,700 $10,053,350,000
Graph of FCC Auction 73 E Block Reserve Price at the end of Round 45

(Click to enlarge)

With no activity in C or D, the story of the day was the E block, which finally surpassed its reserve price in Round 44. Today's graph shows the E block's slow path to its reserve price since the block's bidding pace dramatically slowed down in the Round 9 time frame. Today, after a brief lull in activity in Round 43, in which back-to-back $14.5M gains in Rounds 41 and 42 were followed by a mere $0.5M gain in Round 43, the reserve price plateau looked as though it may again be unreachable. But this scare lasted only for a brief moment, as activity roared back in Round 44 and pushed the aggregate PWBs more than $4M over the reserve. We are guessing that there are several bidders still involved in the E block, but there seem to be two that are dominating, one of which is likely Qualcomm. These two bidders are involved in an alternating round bidding war for about ten of the top licenses (Boston, Philly, DC, Orlando, Atlanta, Detroit, Dallas, Houston, LA and SF) which appears to have started back in Round 37. The pause in upward movement in Round 43 was due to lack of activity on these ten licenses and was likely due to one of two players using a waiver to catch their breath and reevaluate the situation. Bidding on nine of these ten licenses continued in Round 45, with Miami replacing DC in the mix, and by the end of the day the total value of the E block was $16M over its reserve, or 102%. So it seems there is still some stiff competition in the E block, and so far the end is not in sight.

Bidding on the B block is steadily declining round over round, both in terms of the number of bids placed and the bidding unit value of those new bids. As one might expect, most of the activity is focused on the smaller, under-valued licenses that are left. It could take several more days for the bidding to subside to zero, but it is on a steady slope in that direction, and at this pace, the projected end-point will be sometime next week. The FCC has increased the number of rounds per day from five to six starting tomorrow, so that may help to speed things along.

A block activity was extremely steady today, with an apparent bidding war going on among about 30 of the smaller licenses, the largest of which in terms of bidding units is the Norfolk, VA license. This A block bidding war caused an increase of about $13M today, and considering that most of the licenses involved are still relatively inexpensive compared to the average A block price of $1.13, with many that are sub-$0.10 per MHz-pop, the war may rage on for several more rounds/days.

We don't mean to sound like a proverbial broken record, but once again no news is bad news for the D block. If the bidder who placed the Round 1 minimum bid is waiting for something else to happen before making a move, we cannot imagine what it might be. Since C block bidding is over, bidding on A and B is limited to secondary and tertiary markets, and the E block has surpassed its reserve price, it's unclear what else could be worth waiting for. Therefore we'll say it for the last time: the D block is unlikely to be sold in Auction 73.

Wednesday, February 6, 2008: Bidding at the end of Round 40

License MHz Aggregate Total PWBs Aggregate Reserve Delta % of Reserve Total Pops Current Cost per MHz-pop
A 12 $3,856,457,900 $1,807,380,000 $2,049,077,900 213% 285,620,445 $1.13
B 12 $9,075,034,100 $1,374,426,000 $7,700,608,100 660% 284,636,891 $2.66
C 22 $4,748,146,000 $4,637,854,000 $110,292,000 102% 285,620,445 $0.76
D 10 $472,042,000 $1,330,000,000 $(857,958,000) 35% 285,620,445 $0.17
E 6 $868,217,400 $903,690,000 $(35,472,600) 96% 285,480,684 $0.51
Total 62 $19,019,897,400 $10,053,350,000
Graph of FCC Auction 73 New Bidding Units at the end of Round 40

(Click to enlarge)

The auction has transitioned into Stage 2 and has now topped $19B. Stage 2 means that bidders must be active on 95% of their total eligibility measured in bidding units, compared to 80% in Stage 1. This has forced a bit more activity in A, B, and E as shown in the graph to the right, although there was no further bidding on the prime C licenses other than a few bids on the Alaska license today. We still feel that Verizon is holding the lion's share of C, but AT&T and maybe Alltel are other possibilities and they may be holding one or more of the six licenses required for Verizon to establish a nationwide footprint in CONUS.

The transition to Stage 2 forced more bidding on B block licenses, with an average of 80 bids per round today compared to 50 bids per round yesterday, and pushed the average cost per MHz-pop up another penny to $2.66. However, the day's monetary increase on the B licenses was relatively modest at $16M, compared to $65M yesterday, indicating that the increased bidding was focused mainly on the smaller properties. In the category of "out-of-whack" stats, it's interesting to note that of the top 100 most expensive licenses on a cost per MHz-pop basis, 93 are in the B block, 6 are in the A block, and just 1 is in the C block (i.e., Mississippi Valley at $2.36 currently ranks 92nd overall). Of the 93 B block licenses, 18 are RSA licenses, of which the most expensive is Wisconsin-5 Pierce (outside of Minneapolis), which currently stands at $6.51 per MHz-pop, and is the fourth most expensive license overall.

Bidding activity on the A block also increased as a result of the stage transition bringing an additional $19M and another penny to the average cost per MHz-pop. Almost none of the bargains mentioned yesterday were touched today, indicating that the block is dominated by regional players, and is likely split amongst a large variety of local bidders. There are still only four of the 176 A block licenses that are more expensive than the average B block price (Tampa, Boston, Philly, and LA).

E block bidding was back up today as a result of the stage transition, much more so than the other blocks when considering bidding units. Today's graph shows this clearly, as it plots bidding units associated with new bids in each round, rather than just new bids. The graph indicates that the E block bids were focused on the larger properties, as the total number of new bids on B block and E block licenses was about the same on the day. This is even more dramatic when you consider that the bidding units for E licenses are half of the bidding units for the same license in A, since E is half as much spectrum. The E block is now only $35M short of its reserve price, having gained a whopping $47M today. At this rate the reserve price will be met very shortly.

Once again there was no action on the D block today, and with the transition to Stage 2, it has become most certainly impossible that someone other than the bidder holding the Round 1 bid could have eligibility to bid on it. Therefore, the fate of the D block rests in the hands of that Round 1 bidder, who can either choose to bid it up to the reserve, or leave it on the table. There may be reasons to wait until the last minute, but we are not aware of them, so we are still doubtful that the block will be won in this auction.

Tuesday, February 5, 2008: Bidding at the end of Round 35

License MHz Aggregate Total PWBs Aggregate Reserve Delta % of Reserve Total Pops Current Cost per MHz-pop
A 12 $3,837,740,900 $1,807,380,000 $2,030,360,900 212% 285,620,445 $1.12
B 12 $9,059,125,200 $1,374,426,000 $7,684,699,200 659% 284,636,891 $2.65
C 22 $4,748,031,000 $4,637,854,000 $110,177,000 102% 285,620,445 $0.76
D 10 $472,042,000 $1,330,000,000 $(857,958,000) 35% 285,620,445 $0.17
E 6 $821,223,400 $903,690,000 $(82,466,600) 90% 282,958,804 $0.48
Total 62 $18,938,162,500 $10,053,350,000
Graph of FCC Auction 73 New Bids at the end of Round 35

(Click to enlarge)

The auction is approaching $19B, the B block has topped $9B, and it seems the bidding on C has ended, as there were no bids on any of the 50 States C licenses today. We feel that Verizon has come away with most if not all of the coveted licenses by maintaining enough bidding eligibility throughout the auction to be able to bid on the individual C licenses near the end of the auction. The graph at the right shows bidding activity and may provide some clues as to how they accomplished this: Notice the small bubble of activity in the A block between Wednesday and Friday of last week. We believe this was probably Verizon supplementing their activity in B with activity in A to ensure the level of eligibility required to switch back to C. The sharp drop in activity in B in Round 22 corresponds with the first bids in C, and marks the point at which Verizon shifted focus back to C. Incidentally, it's very likely that Verizon's current PWBs exceed their current eligibility, since a few older bids became provisionally winning in Round 30 when the package bid was beaten. This could explain the lack of bidding in today's rounds, since if this were the case, Verizon could not place any new bids, and AT&T may not have had sufficient eligibility to challenge them. If Verizon pulled this off, it was a truely masterful bidding strategy and they deserve to be commended.

An alternate scenario to the above is the following: AT&T got frustrated that they had been played in B and late last week decided to make a run at C. Verizon would have quickly noticed this move and joined in, although with AT&T suddenly decreasing their bidding activity in B first, Verizon was left with more PWBs than they had intended, and had less eligibility than they needed to be completely effective in C. The bidding patterns don't preclude this from possibility, but they don't readily support it either. Therefore we place less probability on this scenario.

Bidding activity on the A block is nearly over, and the average price of the licenses has settled at $1.12 per MHz-pop. Several regional bargains are still available, assuming there are any eligible bidders left: Cleveland and Pittsburgh are currently $0.33 and $0.23 per MHz-pop; Charlotte and Greensboro, NC are at $0.32 and $0.21; Louisville and Lexington, KY are $0.20 and $0.14; and Syracuse, Rochester, and Buffalo are sitting in the bargain basement at $0.05, $0.07, and $0.05, respectively.

E block bidding slowed again today, gaining only $10M toward its reserve price (vs. $30M yesterday and $32M on Friday). The block is still more than $82M short, and with new bids slowing to an average of less than 40 per round, compared to an average 79 per round yesterday, the block may be a candidate for reauction in Auction 76.

The D block is still an open issue, but as every day passes, hope of a bidder stepping forward fades further. With the C block almost certainly finished, we feel that if nothing happens on the D block tomorrow, it will never happen in this auction, and the block will be re-evaluated and sent to reauction.

Monday, February 4, 2008: Bidding at the end of Round 30

License MHz Aggregate Total PWBs Aggregate Reserve Delta % of Reserve Total Pops Current Cost per MHz-pop
A 12 $3,799,550,900 $1,807,380,000 $1,992,170,900 210% 285,620,445 $1.11
B 12 $8,994,146,400 $1,374,426,000 $7,619,720,400 654% 284,332,201 $2.64
C 22 $4,748,030,000 $4,637,854,000 $110,176,000 102% 285,620,445 $0.76
D 10 $472,042,000 $1,330,000,000 $(857,958,000) 35% 285,620,445 $0.17
E 6 $811,158,400 $903,690,000 $(92,531,600) 90% 282,958,804 $0.48
Total 62 $18,824,927,700 $10,053,350,000
Graph of FCC Auction 73 Cost per MHz Pop at the end of Round 30

(Click to enlarge)

Well, a lot has happened since Friday: First, over the weekend we realized that it must have been Google holding the winning bid from Round 17, since if Verizon had won the entire bidding sequence would have had Google dropping out of the auction before open access was triggered, and we believe that this must have been at least part of their objective. Then, interesting things started happening with the individual C block licenses, and this culminated in Round 30 with the Round 17 package bid being beaten by the sum of the eight individual licenses by about $30M. If Google held the Round 17 bid, it's clear that Verizon holds at least part of the Round 30 bids. Finally Round 31 was postponed till tomorrow due to a data glitch, so we will have to wait another day to see how Google and other remaining bidders react. But we wouldn't be surprised if witnesses on the street in Mountain View, CA reported hearing a huge sighing sound around 12:30 pm PT.

We maintain the position stated last week that Verizon was largely responsible for the B block prices climbing so high, and it would be a brilliant move on Verizon's part if they were able to maintain bidding eligibility in C while bidding up B. It appears as though Verizon is satisfied with the damage done in B, and has turned their focus back to C. But we doubt that Verizon holds all eight PWBs on the C block, and this recent development has raised some questions as to what their true objectives are: Will Verizon be satisfied with only a subset of the eight coveted C block licenses, so long as they've made the statement that encumbered spectrum will be significantly discounted compared to unencumbered spectrum? Are they looking only to complement their AWS winnings? More interestingly, will AT&T retaliate, and do they have enough bidding eligibility to make Verizon pay for what they've done? We'll know more tomorrow...

Bidding activity on the A block has been dwindling throughout the day, but total PWBs increased by about $80M on the day. We will see tomorrow if the new activity in the C block spills over to A, but as bidding eligibility decreases, we find this doubtful.

Bidding on the 6 MHz unpaired E block remained slow today, but got yet another $30M closer to the aggregate reserve price today. The block is now just $92M short, and three more days of activity like we've seen today and Friday will bring the block very close to its reserve price. We feel this is a separate ring in the circus with respect to what is happening in C and B, so the next few days will tell whether Qualcomm will seal the deal now, or will have to wait for the reauction.

The new developments in C have brought up new questions about D. We still feel that it is very unlikely that anyone will step up and bid the reserve, but given today's events, it's safe to say that stranger things have happened. So is D a backup plan for Verizon to gain nationwide spectrum if it does not win all of C? We are still skeptical that Verizon wants D, but only time will tell.

Next Week | Previous Week