We are short shares of DISH Network Corporation. Please click here to read full disclosures.
Today we are publishing our report on DISH Network Corporation, which you will find below. We also created a website, makespectrumgreatagain.com, that has a 2-minute video and additional information explaining what DISH has done with our nation’s spectrum.
In our previous reports on Globalstar and Straight Path, we challenged the market’s complacent belief that spectrum prices can only go up and that wireless carriers have a desperate, boundless need for ever more megahertz. The most influential promoter of this bogus notion – with tens of billions of dollars riding on its veracity – is DISH Network. Since 2008, DISH has accumulated a massive portfolio of spectrum licenses and convinced investors that, any day now, it would unveil a brilliant strategy to extract value from these assets, even as evidence mounted that no major counterparty was interested in paying DISH’s price.
Now, with an imminent new spectrum auction promising to drastically reduce benchmark price expectations, DISH is in the weakest position it’s been in for years: sitting on a warehouse full of overpriced inventory, devoid of interested customers, and – with regulatory deadlines looming – running out of time. Meanwhile, DISH’s core pay-TV business is likewise entering dire straits, with declining subscribers, strengthened competitors, and obvious secular problems only beginning to manifest. Overall, we believe that the fair value of DISH’s equity is 58% lower than the current stock price – and, in a reasonable downside scenario, [[more than 80% lower.
Scrappy and entrepreneurial, DISH has always prided itself on its willingness to take outsized risks; as its founder, chairman, and CEO Charlie Ergen – a former professional gambler – said in 2005, “I like to bet a few hands and bet them big.” After achieving great success building the second-largest satellite-TV operator in the US, DISH has watched its earnings flatline for a decade, a victim of consumers’ growing array of entertainment choices and the rising importance of broadband internet connections, which cable and phone companies can offer but DISH largely cannot. Casting about for a way out of this strategic morass, DISH hit upon wireless spectrum.
Initially focused on the concept of a mobile TV service, DISH’s spectrum ambitions have become grander yet vaguer over time, with empty talk of “multiple options” taking the place of a concrete plan. But as DISH has continued to bet big on this single hand – culminating in its widely criticized hijacking of the previous AWS-3 spectrum auction, which earned it the ire of the FCC, Congress, and the wireless industry – it hasn’t noticed that the other players have stepped away from the game. Already armed with large reserves of un- and under-utilized spectrum, combined with better options for cheap capacity enhancement in congested areas, major carriers like AT&T and Verizon now gain less and less from incremental bandwidth – a simple case of diminishing marginal returns. DISH may put on a show of confidence, but carriers are calling its bluff, and DISH shareholders will learn that sometimes, when you bet big, you lose.
I. Investment Highlights
Consensus Valuation vs. Kerrisdale View
Source: company filings, sell-side reports, Kerrisdale analysis
*Long-term debt and capital lease obligations net of cash and marketable investment securities.
†Incorporates (1) cost of AWS-3 auction penalty based on difference between DISH’s original bids and our estimate of the fair value of the spectrum in question, (2) estimated minimum cost of meeting the AWS-4 performance requirements by March 7, 2020, and (3) present value of designated-entity put price.
Carriers have plenty of spectrum already. DISH bulls generally buy into the notion that spectrum is extraordinarily scarce and cellular data consumption is relentlessly skyrocketing. However, even if these beliefs were true, it would not necessarily imply that spectrum prices should be high and rising. After all, the revenue generated by using that spectrum to provide wireless service grew relatively slowly for years, declined in 2014 for the first time ever, and declined again in 2015. Fierce competition has kept a lid on what users pay, while better devices have increased the scope of the demands they place on networks. Thus it might take twice as much spectrum to satisfy a $60-per-month customer today as it did a few years ago – implying that that the profitability of a fixed unit of spectrum is falling, not rising. It takes more input to produce the same economic output, so the input is worth less.
Moreover, the scarcity of spectrum is vastly overstated. Consider, for instance, Verizon – the largest carrier in the US and the most frequently rumored counterparty for DISH. As of the first quarter of 2016, the vast majority of Verizon’s data traffic (92%) ran over its LTE network. But though Verizon holds, on average, 114 MHz of spectrum nationwide, its LTE network uses only half that. The rest of the spectrum is either totally unused (newly purchased AWS-3 spectrum) or still largely devoted to inefficient legacy technologies like CDMA. Over time – indeed, more quickly than previously expected given the rapid uptake of LTE-capable devices – this under-utilized spectrum will be deployed for LTE, effectively doubling the amount of traffic Verizon can handle. A similar story is unfolding for the other carriers as well. As T-Mobile’s chief technology officer put it,
But probably, the biggest thing to think about is – [let me] do rough math for you. Half our network – just over half our network today is LTE, in terms of the spectrum that we own. So half our spectrum is on the LTE technology. And that covers almost 90% of our data.…So when you look at the other half of the spectrum and what it’s doing, it’s not doing that much.
Meanwhile, AT&T, for its part, is beginning to roll out 40 MHz of virgin LTE spectrum nationwide – approximately doubling its capacity even before taking into account its large reserve of under-utilized spectrum, which it’s also gradually shifting to LTE (a process sometimes called “refarming”).
Verizon & AT&T’s Spectrum Holdings:
Large Stores of Untapped Potential
Source: FCC 18th Mobile Wireless Competition Report, Kerrisdale analysis
But more spectrum is certainly not the only way for mobile operators to increase capacity. New generations of technology are also more spectrally efficient – that is, they can transmit greater quantities of data using the same exact bandwidth. (One approach, called higher-order MIMO, is to increase the number of coordinated antennas serving a given cell.) In a report prepared for Ofcom, the British equivalent of the FCC, the telecom consultancy Analysys Mason estimated that, thanks to better technology, LTE spectral efficiency would approximately double over the next five years, thereby doubling network capacity without requiring any new spectrum.
All this fails to even consider the most important way that networks have added capacity in the past: more cell sites. Indeed, American Tower, one of the nation’s largest tower companies, has estimated that, over the last few years, only 20-30% of the gains in cellular capacity have come from more spectrum and higher spectral efficiency, while all the rest has come from new cell sites and more radio equipment. While putting up new, full-fledged towers is difficult in some locations, carriers today can be surgical, deploying sites only where needed to relieve local congestion and making greater use of cheap small cells. Free and low-cost spectrum in the 5GHz and 3.5GHz bands will also play a role in addressing traffic hot spots – without requiring conventional, exclusively licensed frequencies.
Spectrum bulls might contend that even quintupling effective capacity won’t satisfy consumer demand, but they overestimate just how close today’s networks are to their limits. According to the market-research firm NPD Group, the average smartphone user consumes about 3 gigabytes per month (though the distribution is highly skewed – median usage is only ~1 gigabyte). Three gigabytes per month equates to just 0.009 megabits per second – or, assuming that 10% of a day’s usage occurs during the peak hour, just 0.02 megabits per second at peak. By contrast, average LTE throughput in the US is approximately 10 megabits per second, highlighting the large gap – on average – between what users actually do with their devices and what their networks are truly capable of. To be sure, data consumption has grown over time, but some evidence suggests that this growth is already plateauing: Verizon, for example, has seen traffic growth decelerate sharply over the past several quarters (from 75% to 40%), while NDP Group’s detailed analysis (based on tracking individual users’ behavior patterns) suggests that average consumption has been flat in recent months – a far cry from the facile assumption of an endless hockey stick.
Against this backdrop of underutilized spectrum, growing efficiency, improving infrastructure, and potentially plateauing demand, it’s no wonder that wireless-industry insiders regard sensationalistic claims that carriers are on the verge of “running out of spectrum” as utter nonsense. But this dynamic puts DISH in a far worse bargaining position than its supporters appreciate. While they believe that DISH’s spectrum portfolio is a near-term must-have for bandwidth-starved carriers, the reality is that carriers are well-equipped to simply wait DISH out and let it squirm.
Spectrum prices should reset dramatically lower. Much has been written about DISH’s cunning, disruptive bidding in the AWS-3 spectrum auction that ended in 2015; FCC Chairman Tom Wheeler reportedly said from the beginning that the company’s actions “didn’t smell right.” Bidding against itself by way of three separate legal entities, DISH successfully drove prices far above expectations, and investors dutifully applied those distorted prices to the rest of DISH’s portfolio.
However, as the 600MHz auction, beginning at the end of the month, approaches, DISH faces a very different situation. Already highly levered, it lacks the financial firepower to have a large impact on prices, and, after getting into hot water for its complex bidding by proxy, it’s unlikely to try a similar approach again. Meanwhile, carriers’ balance sheets also show the strain of the previous auction, and Sprint – the fourth-largest carrier and arguably the one most in need of low-band spectrum like the 600MHz band – is not even participating. As a result, the 30 MHz of spectrum reserved exclusively for bidding by carriers with modest low-band holdings will likely go in large part to T-Mobile, leaving the rest of the auctioned-off spectrum for AT&T and Verizon. While the total amount of available spectrum is uncertain (subject to participation by TV broadcasters opting to cash out), AT&T has expressed interest in only 20 MHz. If Verizon seeks a similar quantity, then that amounts to around 60 MHz of baseline spectrum demand among the three major carriers – compared to expected supply of 100 MHz from the broadcasters. Thus competition is likely to be subdued, with enough wiggle room to accommodate new entrants as well. No one has a good reason to bid aggressively. Therefore we expect the average auction clearing price to end up close to the reserve level of $1.25 per MHz-pop. Here we assume $1.50 – a 45% decline from the inflated AWS-3 results.
This type of outcome has rapidly become the consensus expectation. Recently Bloomberg reported that the 600MHz auction “may yield a lot less than anticipated,” summarizing analysts’ views that carriers “simply don’t have the war chests to bid up” for bandwidth. Wireless Week echoed these comments. But few have fully considered the ugly consequences for DISH. Not only is the 600MHz auction going to reset all spectrum prices downward; the fallout is also likely to cost DISH billions of dollars in regulatory penalties.
Why? If 600MHz spectrum goes for ~$1.50/MHz-pop, then DISH’s most commercially viable holdings – the AWS-3 licenses purchased in the previous auction – must be worth far less, given their ~3x higher frequency and attendant weaker propagation. While the superior coverage provided by low-band spectrum has become less important over time, especially for carriers like Verizon and AT&T who have already achieved strong coverage, the fact remains that low-band is still more valuable, and recent precedent transactions point to 2-3x higher prices for low-band relative to mid-band (like AWS). But, after partially defaulting on auction-related payments and walking away from many spectrum licenses for which it was the winning bidder, DISH is now on the hook to the FCC for the difference between the pumped-up prices it bid and the proceeds of a future repeat auction. If average prices step down from an inflated ~$2.70 to a more reasonable ~$0.75, DISH’s make-whole liability will balloon.
Meanwhile, DISH’s more “boutique” spectrum holdings have their own problems. Not used by any carrier, they lack an ecosystem of handsets and equipment to give them value – something that likely only Verizon or AT&T have the clout to provide. But when the carrier holds the cards, the carrier extracts most of the value, resulting in very low prices for unusual bands (like the ~$0.30/MHz-pop that AT&T paid for NextWave’s WCS spectrum and Sprint paid for Clearwire’s EBS/BRS spectrum). To varying degrees, DISH’s AWS-4, AWS-3 unpaired uplink, H Block, and 700MHz E Block holdings all suffer from this crucial problem. Even if DISH can ultimately entice a carrier or carriers to buy or lease all of its spectrum, we estimate that its intrinsic value is a small fraction of what the market currently believes it is.
Breakdown of DISH’s Spectrum Portfolio by Band and Kerrisdale-Estimated Value
Source: company filings, Kerrisdale analysis
Time is not on DISH’s side. Making matters worse, DISH can’t simply wait around for carriers to some day exhaust their existing spectral resources; it faces strict build-out deadlines imposed by the FCC. For instance, within four years DISH needs to demonstrate AWS-4 signal coverage and service for 70% of the US population; otherwise, it automatically loses its licenses. With no network of cell sites to provide such service, and with the major carriers busy for the foreseeable future deploying the spectrum they already have (and will win in the 600MHz auction), DISH does not have the luxury of playing “hard to get.” It needs a partner as soon as possible – a fact that every plausible partner is well aware of and won’t hesitate to use against DISH.
Indeed, while we do expect a deal to ultimately materialize, investors must take seriously the risk that DISH can’t find a partner in time and simply loses its non-standard spectrum. After all, while all the carriers will be deploying AWS-3, no one but DISH has any stake in the success of, say, the H Block, and altering existing plans to incorporate such a band takes time, money, and effort. Why not just wait for DISH to lose the spectrum, then pick it up as needed down the road when the FCC puts it back on the market?
DISH’s core TV business also faces grave threats. For years, satellite TV has been a reliable, if stagnant, source of cash for DISH, funding its great spectrum adventure. But with over-the-top entertainment ascendant and satellite-market leader DirecTV gaining strength from its merger with AT&T, DISH is beginning to shed core subscribers. As content costs continue to rise and DISH’s already lean structure makes it difficult to cut expenses, we estimate that a further 10% subscriber loss could reduce EBITDA by 26%, potentially jeopardizing the solvency of DISH’s satellite subsidiary. While some pin their hopes on DISH’s own over-the-top offering, Sling TV, early indications suggest that the product is riddled with bugs and already losing steam, leading to recent headlines like “Is Sling TV a Failure?” In short, DISH’s weak TV business can’t make up for its disastrous spectrum gamble.
Click here to read our full report.
 Source: Kerrisdale analysis of Verizon, AT&T, T-Mobile, and Sprint wireless-segment service revenues.
 See e.g. Verizon’s comments at the UBS Annual Global Media and Communication Conference, December 7, 2015: “And if you look at it today, we only use 40% of our spectrum for LTE.”
 Source: Capital IQ transcript of T-Mobile 2016 Q1 earnings call, April 26, 2016.
 Of course, DISH might choose to bid irrationally, perhaps hoping to drop out at the last minute and thus conserve its cash, but in so doing it would just be digging itself a deeper hole.
 Based on paired spectrum.
AWS-4 Final Build-out Requirement in any EA [economic area], its authorization for each EA in which it
fails to meet the requirement shall terminate automatically without Commission action”).