Sprint (NYSE:S), which is majority owned by Softbank (OTCPK:SFTBY), continues to be what it has been for the better part of the last decade: a conundrum. On one hand, it has all the ingredients necessary to be a first-place wireless provider in the United States - a well-known brand, a wealthy corporate parent to back it, enough low-band spectrum to offer reliable service throughout the country, and a boatload of high-band spectrum to win the new wireless data price wars - more than any other carrier, in fact.
On the other hand, the company never seems to live up to its potential. Losses continue to pile up at the carrier, much as they have for years. And new CEO Marcelo Claure's decision to enter the wireless price wars with T-Mobile (NASDAQ:TMUS) with both feet has only exacerbated the situation. Sprint, which recently fell behind John Legere's Un-carrier for fourth place in the wireless sector, has been hemorrhaging cash in an effort to keep up. In addition to a "limited-time" half-off offer that never seems to end, which has seen the carrier offering four lines for as little as $50 a month, Sprint also offered four lines of unlimited data for a record low $150, as well as brand-new iPhone 6S phones for as little as $1 a month. Even after deducting the value of the required trade-in, Sprint's loss was calculated at several hundred dollars per iPhone. And the king of cash burners was the offer to DirecTV (now part of AT&T (NYSE:T)) subscribers of free service for one whole year. Yes, absolutely NO payments to Sprint for a free year-long ride on its network at 2GB per line. That offer ordinarily would cost over $1,000 per family, but Sprint ate the whole amount. That's how desperate the company is for subscribers.
On the other hand, there had lately appeared to be a glimmer of light on the horizon. Sprint's subscriber count is rising rather nicely - it certainly ought to be, with all this cash going up in flames - and its Network Vision plan, whatever its deficiencies, has seen fewer dropped calls and better data speeds appearing in more and more cities. It is still one of only two carriers to offer a generally available Unlimited Data plan - AT&T only offers it to U-Verse and DIRECTV subscribers - and its most recent earnings call saw net earnings climb from a $2.1 billion loss to "only" $600 million in red ink. With Softbank still there to backstop the company with cash until it gets back on its feet, Sprint seems to be on the way up. The market bid the stock up on the news, back above $3.
So, is the worst behind Sprint? Is this the bottom where traders should get in and make a killing over the next few years as Claure rights the ship, much as 2013 was for T-Mobile and Legere?
Actually, yes. Sprint is a buy. But not for the reasons most people think. There has been precious little improvement at the yellow carrier so far. But you should buy it anyway.
The key to valuing Sprint is not its network, or its subscriber count, which it is raising only by absolutely gutting its cash flow. Cash flow itself isn't the key either, and neither is net earnings. It is the carrier's spectrum. With such perpetual false starts and refreshes of its network and business plans, it is not clear if Sprint, as a company and as a brand, will ever get out of the doghouse with consumers. But the company's hoard of high-band spectrum is such that it doesn't need to for its stock to pay off.
The thing to understand about radio spectrum is that it is a little like a wedding cake. An upside-down wedding cake, with the smallest layer at the bottom. Just as you get through the top of a real cake quickly, but need a lot of mouths to polish off the bottom, wireless spectrum is very quickly exhausted at the bottom of the spectrum, but almost indefatigable at the top. But you have to start with the small section first, just like with a wedding cake.
When AT&T and Verizon (NYSE:VZ) built the first cellular networks decades ago, they built them with mostly low-band spectrum, usually in the 800-900 MHz range. This spectrum was ideal for their purposes at the time, because low-band spectrum can cover vast distances and can also punch through obstacles like walls or roofs without much degradation of signal. What it is not so good for is capacity. Precisely because low-band spectrum travels so well, cell towers have to be placed quite a ways away from each other so they don't interfere with one another. So, the aggregate number of towers built and the amount of bandwidth they can carry is pretty small. But back when cell phones cost four figures just for a unit the size of a purse that could do nothing but make calls, not many people were buying phones anyway, so this wasn't an issue. And anyway, fewer towers meant that it was cheaper to cover the whole country.
Today, of course, it's very different. Apple (NASDAQ:AAPL) invented the iPhone, and Google (GOOG, GOOGL) invented Android. Netflix (NASDAQ:NFLX) invented video streaming, and Facebook's (NASDAQ:FB) Instagram has billions of people uploading and downloading high-resolution photographs. Everyone is sending and receiving gigabytes of data every month. It wasn't long before cell carriers started wanting large amounts of that high-band spectrum they had previously shunned. Yes, it was prone to interference, the signal couldn't propagate far from the tower, and even a single wall was usually enough to kill the transmission. But boy, the bandwidth. 1900 MHz towers could be built so much more closely together, and so the aggregate bandwidth was so much more. Cell companies put up with the other stuff because this spectrum was the only way to carry all the data everyone was asking for and willing to pay for. They could build more than four times as many 1900 MHz towers over the same area as 800 MHz towers, with accompanying increases in capacity.
When building new towers to house the new spectrum, cell carriers did not completely have to start from scratch. Usually, transmissions at different frequencies don't interfere with each other, even if they originate from the same point. So sometimes, they could put the new high-band spectrum on the same towers as the low-band spectrum. But of course, the whole point of the high-band spectrum was to be able to build more towers and have more capacity, so there are still significant new construction costs.
For wireless carriers trying to find the lowest-cost method of providing all the bandwidth their customers demand, this balance between spectrum and towers is the key. Once they have bought a slice of spectrum, they have to pay for it, no matter whether they build it out to its fullest potential or not. So they want to fully utilize the spectrum they have before they buy any more, by building as many towers using that spectrum. But eventually they hit a wall, where bunching the towers any closer together will wreck the network with interference, and that's when they have to go back to the spectrum market and buy more. The only exception to this pattern would be if spectrum somehow was offered cheaper than it should be. Make the next level up of spectrum cheap enough, and it may be cheaper to buy than towers are to build.
This is essentially what is going on in the chief technology officer's office of every major cell carrier every day. Demand is ever-increasing, and the CTO and his team run the numbers and determine what the most optimal way of adding capacity to match it is. How much does spectrum cost right now? How much does tower construction cost? On a per-gigabyte basis, is it cheaper to buy more spectrum to add to existing towers, or cheaper to build more towers to utilize existing spectrum? Is adding more towers even an option, or has the network already hit the limits imposed by physics on its spectrum assets?
What the media has up until now referred to as low-band and high-band spectrum (0-1 Ghz and 1-2 Ghz, respectively) is actually low- and medium-band spectrum. High-band spectrum, 2-3 Ghz, hasn't even really started to be utilized yet. It has taken until now for all the mid-band spectrum to be exhausted. But that point is at last approaching where the middle layer of the cake is all gobbled up. High-bandwidth spectrum will go from being an afterthought to being absolutely essential over the next few years, and probably faster than most people think.
And Sprint owns almost all of it.
As I said, sometimes carriers buy spectrum "early," i.e. before their lower-band spectrum has hit its limit, if it is cheap enough. And high-band spectrum was absurdly cheap when Clearwire bought around 150 MHz of it. And Clearwire was absurdly cheap when Sprint bought it, first half of it and then the other half.
High-band spectrum was cheap when Clearwire bought it, because no one was using it. Remember, carriers start at the small, low-capacity, low-band bottom and work their way up, with each layer bigger, and potentially more valuable than the one before.
When CTOs add capacity by building towers, their costs are spread over all the towers. They pay one company $10 million to build a set of towers in one city, pay another company half that to build a few less towers in another city, etc. But the spectrum that runs on all those towers is the same frequency, the same block, sold in a single batch.
That means the value of that spectrum is intricately bound up with how many towers it can run on. When CTOs are deciding how much to bid in a spectrum auction, they ask how much capacity that spectrum will buy them. And to know that, they have to ask themselves how many accompanying towers they can justify building.
Earlier, when there was hardly any data traffic on the networks, the answer to that question, when it came to high-band spectrum, was "zero." The 900 MHz and 1900 MHz spectrum already had all their needs covered, and it was more reliable and less prone to interference than the higher bands. So, the high-band spectrum was effectively useless. No one was going to build the towers to make purchasing it worthwhile.
But that sword cuts both ways. If spectrum is powerful enough to cover large distances, it becomes worth something sooner, since cell carriers want to buy it earlier in time. But that same characteristic also imposes a ceiling on how much the spectrum can ever be worth. The value of any spectrum is basically the total value of the capacity it produces, minus the cost of the towers it runs on. The maximum capacity, in turn, is a function of how many towers it can possibly run on. Since low-band spectrum cannot run on all that many towers, it can never be worth all that much. High-band spectrum, on the other hand, when carriers finally get around to using it, can run on literally hundreds of thousands of towers. It can be extremely, extremely valuable, if carriers fully exploit it. But they won't even start to exploit it until the lower spectrum is chock-full of traffic.
That is the paradox of spectrum. The spectrum that is worth the least now is the one that will be worth the most later. And "later" is now for US wireless carriers. Data use is increasing exponentially, and is projected to continue to. Low-band spectrum and medium-band spectrum are approaching their physical limits. Almost every carrier is announcing plans to develop a newer, denser cell network with a lot more towers, and those plans only make sense in a high-band context. With the top, big layer of the cake finally about to start filling with data traffic (or get eaten, if you want to stick with the metaphor), and with the number of towers being supported so far above what other bands can support, the sky is quite literally the limit for how high the value of this spectrum can go.
Sprint's Real Gold Mine
Which brings us back, at last, to Sprint - the owner of this high-band spectrum - and my thesis that its own performance doesn't really matter. Losing money, losing subscribers, gaining subscribers. It doesn't matter. What matters is how quickly traffic is growing on the networks of the Big Three (AT&T, Verizon, T-Mobile). The more it grows, the sooner they will need high-band spectrum, and the sooner Sprint will be in the driver's seat of one of the most pivotal negotiations of the wireless industry's history.
Nor would they be the only bidders. Apple, Google, Netflix, Amazon (NASDAQ:AMZN) and Facebook are all companies which depend on Internet connectivity with their customers to run their businesses. At some point, they may decide to it would be worth the cost to have some bandwidth capacity of their own, cutting out the middlemen who currently act as gatekeepers between them and their customers. And not always friendly gatekeepers, either. Meanwhile, Comcast (NASDAQ:CMCSA) and other cable TV providers are probably already thinking of ways to try to match AT&T's "quad play" offer with DirecTV. Sprint's spectrum would appeal to them as well. With this many potential buyers, Sprint can be assured of commanding top dollar.
And top dollar would be higher than most analysts seem to think. Right now, analysts are talking about how much of a discount high-band spectrum should sell for compared to mid-band spectrum. That is absurd. In a high-data traffic world, where capacity is king, the real question is how much of a premium should high-band spectrum go for. In the recent FCC auction, mid-band AWS spectrum went for well over $2.50 per MHz-POP. Even with a 20% discount for uncovered populations (in some low-populated areas, there won't be enough traffic to justify using the new spectrum), close to 150 MHz times America's 320 million people means Sprint's Clearwire spectrum holdings alone are worth almost $100 billion. That's before any additional premium for high-band spectrum, before counting the value of Sprint's low-band and mid-band holdings, and before counting any residual enterprise value in the company's wireless brand and business itself. Even with almost $34 billion in debt, Sprint's 4 billion shares are worth a minimum of $16 apiece. The stock currently trades for around $3. That is in excess of a 400% return potential.
The only knock on Sprint is its management. I have nothing against trying to compete for customers, but the desperation with which Marcelo Claure is going about it staggers me. The one and only way Sprint can fail to achieve stupendous returns over the next few years is if it is forced into a spectrum fire sale in bankruptcy before AT&T, Verizon and T-Mobile are desperate enough to let it name its own price. With that in mind, the last thing Softbank and its chairman and CEO, Masayoshi Son, should want is a sales and marketing expert like Claure, for whom customers are apparently worth buying at any price. Dan Hesse might not have been a genius, but he was a prudent and conservative operator who was careful with Sprint's cash. Right now, the company doesn't need to win the race, it just needs to still be standing when the other three cross the finish line. Once the value of its spectrum sky-rockets, it can try to see if it can use the spectrum itself, confident that it can simply sell if it turns out it can't. Bust will mean simply "slightly less profitable." Right now, bust means bust.
Despite this, I recommend buying Sprint stock. Softbank has seen Claure's style in action for well over a year now. Son made Sprint's acquisition of the other half of Clearwire a priority when buying into the company, as it was then the only other network in the world besides his own utilizing a fully deployed TD-LTE standard. So, it seems clear that he appreciates how important it is. My interpretation of recent events is that he has already decided that he is willing to stick with Sprint come hell or high water, so he might as well let Claure compete now as later, when T-Mobile is even better established as a competitor than it already is. If Son is going to backstop Sprint's losses, relying on the spectrum to make him whole in a few years, then there is no danger of a fire sale. And absent a tragic fire sale, I just don't see how Sprint can lose.
As for the potential bidders, I think many shareholders are underestimating both how expensive it would be for them to acquire spectrum and how valuable it would be once they did. Waiting is only going to make spectrum even more expensive down the road. I would be bullish on anyone who actually manages to pry Sprint's spectrum from it. Meanwhile, failure to acquire spectrum, from Sprint or anyone else, would have to be seen as a bearish factor, though not necessarily deserving of a sell.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.