If I’m right and DISH is determined to win a significant AWS-3 spectrum position at the end of the auction, then it seems highly likely that one or both of AT&T and Verizon will leave the auction with a significant shortfall in AWS spectrum in major cities including New York, Los Angeles and potentially several other markets.
Then it seems Ergen’s calculation is that he will have significant leverage to force AT&T and Verizon to deal with him and lease spectrum on his terms (including supporting interoperability for his AWS-4 spectrum holdings). However, one way for AT&T and Verizon to freeze Ergen out and avoid having to make a deal would be for them to instead purchase 2.5GHz spectrum from Sprint. Its plausible that Sprint could raise as much as $10B relatively easily from selling say 30MHz to each of AT&T and Verizon, leaving Ergen holding an asset with no clear route to monetization and a buildout deadline which will start to become a pressing concern within a year or two (especially if DISH has not yet standardized the AWS-4 band).
So does Masa Son want to boost DISH’s position at the expense of AT&T and Verizon, or would he like to get revenge for DISH’s actions in the Sprint & Clearwire bidding wars last year? If DISH is stuck with billions of dollars of spectrum it can’t lease, then DISH will be disadvantaged in mounting a competing T-Mobile bid, when Sprint renews its attempts after the 2016 Presidential election, because DISH will struggle to raise as much cash and DT will be reluctant to accept shares whose value is based primarily on spectrum assets with limited utility (remember that T-Mobile isn’t in a position to create an ecosystem for AWS-4, unlike AT&T and Verizon).
In fact, Sprint could point to DISH’s reserves of spectrum as providing the basis of a new competitor in the wireless market, and could even gain the tacit endorsement of AT&T and Verizon for a purchase of T-Mobile. In addition, by selling some spectrum now, Sprint raises money to participate in the 600MHz incentive auction (where DISH may not have the resources to compete) and gets out from under the spectrum screen limitation. So it might well make sense for Masa to make a choice which boosts AT&T and Verizon, rather than cooperating with DISH.
Incidentally, another side-effect of the AWS-3 auction prices is that Phil Falcone is now scrambling to get back into the LightSquared reorganization plan, as his argument that LightSquared’s spectrum should be valued at more than the debt gains support from these price benchmarks. For example, the unpaired uplink 10MHz B1 block (1700-1710MHz), currently valued at almost $1.3B, will be used to argue that LightSquared’s two 10MHz uplink blocks alone are worth double this sum. So the obvious counterstrike from Ergen is likely to be to try and blow up the reorganization plan and force LightSquared into liquidation.
I understand conversion to Chapter 7 would invalidate the Inmarsat Cooperation Agreement, and thereby make it much harder for anyone to take on the risk of buying LightSquared’s assets. Of course, that is unlikely to worry Ergen (he would be expected to take a hard line with Inmarsat in any case), and would provide an opportunity to potentially buy LightSquared’s satellite assets for considerably less than the value of the LP debt and boost Ergen’s attempts to corner the spectrum market. As one person close to the case told me, such an outcome would literally make Judge Chapman cry.
UPDATE (11/26): Another interesting question is the status of the 650M MHzPOPs of EBS spectrum (38MHz) that NextWave holdco controls in New York City. I would expect hectic bidding to secure access to that spectrum, if DISH turns out to be the winner of much of the AWS-3 spectrum in New York. Of course, Ergen has likely already thought of that, and I’d speculate that he might even have locked up an agreement to buy that spectrum block in advance of the AWS-3 auction, making it harder for Verizon and AT&T to address their potential spectrum shortfall in the New York market.
That seems to be a pretty good summary of what Charlie Ergen has told SNR and NorthStar, the Designated Entities (DEs) which appear to be doing a lot of the bidding in the AWS-3 auction. We’re still seeing multiple bids on the New York and Los Angeles J block licenses (which are now priced at over $3.6B for just these two licenses), and so its possible that DISH’s DEs may now hold in excess of $10B of Provisionally Winning Bids (PWBs) between them.
Its interesting to note that because Ergen is a designated bidder for American AWS Wireless I LLC (the DISH subsidiary), he wouldn’t be allowed to communicate with either SNR or NorthStar during the auction. So they must be bidding in line with instructions he gave them previously, and he can’t change course in the middle of the auction, if bidding has gone well beyond what even he expected. Ergen could have set a limit on the $/MHzPOP that he is prepared to pay for a single license and/or an overall dollar cap on bidding, but it looks like neither of those factors have been affecting the bidding to date. That suggests to me that Ergen is more likely to have imposed an overall dollar cap, which could perhaps be as much as $8B to $10B each, before the 25% discount on the PWBs that each DEs is expected to receive.
A plausible bidding strategy for each DE, which requires no communication after the auction starts (and seems to roughly accord with what happened after DISH likely switched to the unpaired spectrum in Round 17), could then be something like:
Bid in every round until you see high activity in the unpaired spectrum (a sign that DISH has switched away from the AWS-3 blocks), then defend what you have, no matter what the price of each license. When you reach the overall dollar cap, drop the G block licenses first, then the H and I blocks, but keep the J blocks.”
So now the question is whether, with a weekend to think about it, AT&T and Verizon decide to leave Ergen holding the J-block licenses in major cities. Is there a point at which DISH becomes so financially stressed by the burden of spending perhaps $15B on spectrum, that it is in a weaker position after the auction than it was before (e.g. with insufficient resources to bid for T-Mobile or build out its spectrum)?
Does it matter if AT&T and Verizon are unable to deploy as much AWS-3 spectrum in major cities as they wanted? How much leverage can they exert at the FCC if DISH fails to build out these licenses? And should AT&T or Verizon force up the price of the unpaired spectrum to a level which puts the Ergen-led LightSquared reorganization plan in jeopardy? We’ll see in the next few days whether AT&T or Verizon are able to play Ergen at his own game.
UPDATE (11/23): Thinking further about the end-game, it looks to me like the mysterious spin-off of DISH’s spectrum, mentioned briefly by Ergen in the Q3 results call, is the likely way forward if one of the DEs is successful in acquiring a large block of spectrum. DISH could inject its existing spectrum assets into this entity and raise debt against the combined spectrum portfolio. Then (given the FCC rules against selling DE spectrum for at least 5 years), DISH would have a standalone entity it could spin-off to shareholders, and could lease or all of its spectrum holdings to the major operators, just like Grain Management.
UPDATE (12/1): Round 38 saw another apparent “tell” from DISH with the bidding on the New York I block: 2 new bids were submitted for a price of $1.316B, when the identical H block could have been secured for only $1.235B. Its implausible that either Verizon or AT&T would have bid for the I block at this price unless they already held the H block and wanted to double up. But equally well, its impossible for both of them to already be holding the H block, and only the holder of the H block would want to bid a higher price than necessary for the I block. So the only logical scenario under which 2 bids could have been submitted for the I block is if DISH holds the H block and both its DEs were bidding to add the I block, potentially sending a “signal” to AT&T and Verizon that DISH is determined to capture both blocks (and preserving eligibility for both DEs for when the FCC moves to the second stage of the auction). Note also that a similar single bid in Round 40 for the New York H block (after several rounds with no bids) was followed by a double bid in Round 41 (albeit at a lower price than the I block).
If we go back a few rounds then we can see many instances of these paired bids in New York, Los Angeles and a few other major cities. Until now I had assumed this simply to be a sign that DISH was competing with both AT&T and Verizon. However, now I think that quite possibly only one of AT&T and Verizon is competing very actively for the paired spectrum blocks. Most likely that would be Verizon (who can put AWS-3 to use more quickly as supplementary downlink for its AWS-1 holdings), with AT&T perhaps switching to the unpaired uplink B1 block, where DISH appears to be facing unexpectedly intense competition.
Notably, we can also see that there were two bids for the New York J block in Round 36 (after 1 bid in Round 35 and 2 bids in Round 34), followed by a single bid in Round 37 and no other bids thereafter. That sequence may indicate that Verizon finally forced DISH to relinquish the J block in Round 37. However, the fact that DISH’s DEs both appear to have enough eligibility to still be bidding for such a large license (when only one could be the winning bidder in any round) at that stage also suggests that DISH must hold a very significant share of the major city licenses. So the question now is just how far Verizon will need (and is prepared) to go in order to capture the remaining licenses it wants to secure.
After the AWS-3 auction reached its reserve price of $10.066B for the 50MHz of AWS-3 spectrum (1755-1780MHz uplink/2155-2180MHz downlink) in Round 12, the most notably feature of the auction so far has been in the high level of bidding activity for the largest cities, such as New York and Los Angeles. As shown in the table below, in the first 12 rounds, there were up to 8 bids per round for the three main (BEA) licenses in these cities, implying that multiple players other than the major operators (AT&T, Verizon and T-Mobile) were bidding.
By now it would have been expected that the three operators would sort themselves out and bid on a self-selected subset of licenses (e.g. AT&T for 20MHz in H+I, Verizon for 20MHz in J and T-Mobile moving to the smaller CMA G-block license as it did successfully in the AWS-1 auction in 2006). However, it appears that this has all been disrupted by DISH’s desire to drive up the price. DISH has presumably concluded, logically, that the major operators will have to buy the New York and LA licenses (plus a few other places such as Chicago, Washington, Boston, San Francisco and Dallas) whatever the cost, so has been bidding simultaneously across all of the main licenses in these cities.
Uniquely amongst the participants, DISH is also part of three separate bidding consortia: American AWS-3 Wireless I LLC, Northstar Wireless, LLC and SNR Wireless LicenseCo, LLC. Given Ergen’s interest in pushing up the price that the operators have to pay, it would not be in the least surprising if all three have been bidding simultaneously against one another for all of these licenses. The resulting higher level of bidding activity would potentially sway the decisions of AT&T, Verizon and T-Mobile, making them think they face more competition from each other than is actually the case, and thereby persuading them to bid more than they originally expected. After all, if AT&T thinks Verizon wants all 50MHz and Verizon thinks the same about AT&T, they are both likely to bid more aggressively, since both will think they had underestimated how valuable the spectrum is.
What happens next? DISH could potentially have instructed each bidding consortium to cease bidding at the same level, minimizing the risk that it would be stuck with licenses it didn’t actually want. That wouldn’t require any coordination, merely setting a near identical budget/price limit for each of the three consortia. Then I’d expect a sharp drop in bidding activity, probably later today or tomorrow, when DISH reaches its desired price point (to provide a high comparable for its other spectrum holdings) and probably switches to buy the 1695-1710MHz unpaired uplink block, where to date the price has lagged significantly due to a lack of competitive bidding. Recall that this unpaired uplink block is one of the best comparables for LightSquared’s valuation (where Harbinger is arguing that the 20MHz of uplink should be worth up to $5B), and so its clearly in DISH’s interests for the end price to be little more than the $580M reserve price (for 15MHz of spectrum).
Of course, DISH is in a win-win position: if Verizon, AT&T and T-Mobile bid up the price of AWS-3 then Ergen can claim his AWS-4 spectrum is worth even more, but if they call his bluff and leave DISH owning the key licenses in major cities (that will be needed as part of any AWS-3 rollout), then Ergen can demand interoperability for his current AWS-4 spectrum as a condition of selling the spectrum to those operators.
If this is actually what is happening, then I’d expect criticism of how the FCC has enabled DISH to game the system (by participating in multiple consortia), just like there was criticism of their decision to auction off the H-block earlier this year in what some likened to a “retail sale” to DISH. However, the result will be the opposite to the H-block, which only just reached the reserve price, because in this case the auction revenue will be significantly higher than expected. Nevertheless, I’d expect mobile operators to be even less enamored of DISH than they were already, because Ergen will have just cost them billions of dollars they didn’t want to spend, just like he cost Masa Son billions of dollars by forcing him to raise the price that Softbank paid for Sprint and Clearwire last year.
UPDATE (11/20): It seems that I may have been right, because the FCC announced on Tuesday afternoon that the reserve price of $10.066B was met for the paired spectrum blocks at the end of Round 13:
At the conclusion of Round 13, the provisionally winning bid amounts, net of any applicable bidding credit discounts, for the paired 1755-1780/2155-2180 MHz licenses (the licenses in Blocks G, H, I, and J) exceeded $10,0660,326,600, thus meeting the aggregate reserve price for these licenses in Auction 97.
However, at the end of Round 12, the total of Provisionally Winning Bids (PWBs) in these bands was already $10.375B. The PWB is calculated before bidding credits (of 25% for a small business) are applied. Thus holders of bidding credits must have held sufficient licenses for the total of net bids to still be below the reserve price (i.e. had bids with the 25% discount valued at more than $311M). That means these small bidders held licenses valued at more than $1250M at the end of Round 12.
Two of the three DISH entities (SNR and Northstar) both sought 25% bidding credits (despite apparently having credit agreements with DISH to fund their bids) so it appears likely that they were bidding aggressively throughout this period. Then we apparently saw DISH move its own bids over to the unpaired spectrum in Round 17, leading to a reduction in the number of bids for the large metro licenses. Given their use of bidding credits, it would not be at all surprising if SNR and Northstar emerge still holding billions of dollars of spectrum, funded by DISH, and it would be logical to take this approach (if DISH buys the unpaired spectrum at close to the reserve price, then it would have gained no benefit from using bidding credits, since the reserve price has to be met by the net bids). Of course the availability of bidding credits will probably also be a cause for further criticism of DISH’s apparent moves to game the system by pushing up the prices paid in the auction.
I’ve just released my new 69 page Globalstar profile, analyzing both Globalstar’s MSS business and their spectrum valuation and potential partnerships. Over the last month its been interesting to observe the rather hyperbolic comments from both supporters and opponents of the company, and unfortunately those on both sides appear poorly informed about the potential value of the TLPS spectrum and the growth of the MSS opportunity.
In the MSS business, its crystal clear that growth has fallen far short of expectations in the business plan presented to COFACE in summer 2013, and it is expected that Globalstar will need to make equity cures to address an EBITDA shortfall in 2015, and perhaps even at the end of this year. However, those amounts will be modest, nothing like the $200M raised by Iridium earlier this year, and so there is no need for concern that MSS challenges will disrupt Globalstar’s attempts to monetize its spectrum in the next year or two.
Nevertheless, it seems unlikely that the new Hughes-based devices will dramatically change this picture: MSS terminals offering WiFi links to smartphones but no standalone functionality have generally been fairly unsuccessful in comparison to self-contained communications devices (compare SPOT Connect vs SPOT or the original inReach vs inReach SE/Explorer), and low price terminals/consumer distribution channels have not altered the dynamics of the handheld market very much (for example, the SPOT Global Phone has not changed Globalstar’s business prospects materially). Its also little use selling a new device for $100 if the customer still has to choose from the existing handheld airtime plans (which have a four times higher ARPU than SPOT). So overall, we don’t expect Globalstar’s MSS business to generate enough value to match the $1B invested, or even the current COFACE debt.
While the MSS picture may not be encouraging, I’m more positive about Globalstar’s TLPS opportunity. It’s clear that Globalstar’s spectrum does have potential value to partners, since Globalstar came close to a deal with Google in early 2013, and we suspect a deal with a different partner was almost reached earlier this year. I also expect the FCC to approve TLPS without material concerns, although it seems likely to come in Q1 not Q4, and may involve giving up some L-band spectrum to Iridium, as happened in the past when Globalstar was seeking ATC authority.
The key question is therefore whether a partner can now be secured who will pay a substantial sum for access to the spectrum, from a limited universe of possibilities in the service provider category. Equipment and infrastructure providers are more likely to want to make money from selling equipment to Globalstar (and its service provider partner), than to pay Globalstar for access to the spectrum. The second question is then what the appropriate valuation would be if a deal can be struck: based on comparable valuations for high band spectrum and the alternative sources of spectrum out there (including AWS-3) it is hard to believe that anyone could seriously envisage a $6B or $10B valuation for TLPS. However, the debate should be about what Globalstar’s spectrum is worth, not whether it is worthless.
All of these issues are discussed in more detail in the report: we give specific forecasts for the MSS business by product through 2018 and explain what we believe to be the most appropriate valuation benchmarks for TLPS and who is now the most likely partner. Contact us if you’re interested in more information.