How To Invest

Want to know more?

Find out more arrow-white

Island Investing

Riffs, rants, and the upside of investing from way off Wall Street


Q&A With Clearwire Longs and Shorts

This is the second in a series of additional thoughts about Clearwire, presented in a Q&A format. Here was the first post in this series, and here is the article that kicked it all off.

As a reminder, none of this should be considered investment advice, or anything even close. It’s almost entirely opinion and, in parts below, is pure speculation. But those are the kinds of questions I’ve been getting lately. I’ll return to more objective, quantitative answers in the next post in this series.

And despite a crazy few weeks, I still believe Clearwire’s spectrum, under a set of pessimistic assumptions, is still worth around $6 a share. Implicit in that estimate is that the company finds the short-term funding to be able to ultimately realize that intrinsic value. And while that spectrum value is of less importance to both debt and equity investors right now then the company’s need to find funding, I believe that emphasis will likely change dramatically before the end of the year. Just my opinion, mind you.

On with the Q&A.

Q. Who is going to fund Clearwire?

A. I have no more insight into the answer to that question than any other investor out there.

The reasons I invested in Clearwire prior to knowing the answer to that question, however, were (1) I’ll get a much better price and a bigger margin of safety now then later, (2) I believe the actual probability of the raise happening is significantly higher than what the market appears to be pricing in, and (3) I have faith that somewhere in the parallel universe than can be the telco industry, someone with much bigger brains than mine did the same calculations of Clearwire’s spectrum value as I did – and because that gap between true spectrum value and its current implied pricing in the market is so large, and the TDD technology so potentially disruptive, those brains are going to get off the couch and do something with Clearwire. And, frankly, I also trust that in the negotiations that ensue, John Stanton will defend his own equity interests as strongly as I would defend my own.

So, I like my position in the capital structure as a common equity holder – though I concede Clearwire’s debt could quite be attractive here as well – because I simply don’t believe that my position as a "subordinate" equity holder is going to be relevant for long. I want to be a partner in this business – forever, hopefully – and not a lender to it, and there’s no way to do that without eating some risk. So be it.

That said, it’s also important to be realistic about the hurdles Clearwire could encounter as it seeks additional funding from a strategic partner. Specifically, investors should acknowledge that a funding partner (or partners) might not want to play second fiddle to the first-lien holders (i.e. those debt holders who essentially have first dibs on the spectrum in the event of bankruptcy). That’s a valid issue that folks like the credit analysts at Moody’s have brought up recently. I would argue, however, that the mere act of putting enough "junior capital" into Clearwire will also make the "junior" part of that phrase immaterial, as the company is on a path to self-funding its own operations in the near future. In other words, that equity stake may not be secured, but the ultimate reward for whatever group that funds Clearwire should more than make up for that risk, too, in terms of total return across its investment.

As I explained in my original article, at current prices, the market is pricing in a capital raise of well over than a billion dollars through just a very dilutive equity sale by Clearwire. I believe the actual probability of that happening is very low – less than 10%, I would say – based on a number of factors including public comments the CFO has made as well as the availability of vendor financing. I would guess the financing will be done in stages – and there will unequivocally be some dilution, but my point is that it won’t be nearly as mush as the market seems to believe.

So when shorts assert that Clearwire tried to raise money last year, and apparently couldn’t, or they tried to sell spectrum, but couldn’t, and/or otherwise appear to have no funding options last year, and then try to pass those observations off as being relevant today, I would remind them that: (a) the company only very recently formalized (and pegged the cost of) a switch to LTE Advanced, (b) the entire management team and board has recently been overhauled, and (c) the competitive dynamics throughout the industry have changed quite a bit from a year ago given the T-Mobile acquisition, another year’s worth of network congestion, a shift in emphasis at CLWR to a wholesale business model, and the emergence of Dish’s interest in the space. These things should take a little time.

Whether you’re long or short Clearwire, I’d also suggest you be sure you know the difference between the FDD and TDD flavors of LTE, specifically as they apply to both coverage- and capacity-based models. And not because it will mean you could get a 2 point bounce in the stock in the second Thursday after the full moon, but because it means the actual probability of Clearwire finding strategic funding/getting another wholesale customer/raising capital through vendor financing is considerably higher than most people who don’t understand the difference appear to believe. The technology matters here, period.

Lastly, with regards to specific partners for Clearwire – standby for rank speculation – I am personally intrigued by the potential for the company to do something with Dish. Again, I have no particular insight into either of these companies on this issue, and the odds are that my interest ultimately ends up being little more than wishful thinking, but I would encourage both longs and shorts to read this week’s Bloomberg article on Charlie Ergen’s plans at Dish and ask if there is a better fit out there than Clearwire. And while you’re at it, check out this article, in which Dish’s CEO mentions the possibility of buying or partnering with Sprint or Clearwire.

So, speculate away, I suppose. The point is that Clearwire has other options than Sprint.

Q. Who do you think is shorting Clearwire, and why?

A. This, or some derivation of it, seems to be the second most popular question I’ve gotten. I have had a decent amount of back and forth with some CLWR shorts based on my previous articles, which is valuable in terms of trying to poke further wholes in my own long thesis. Haven’t found any yet – though funding remains the big topic, obviously. That aside, I don’t particularly care who might be shorting the company, although I can certainly understand how frustrating it can be to see something that’s already ridiculously cheap get even cheaper.

Clearwire has been an excellent short if for no other reason than it’s a complicated story with ugly losses to date, a relatively limited float, and an enthusiastic but impatient retail shareholder base – and that makes its daily price subject to all manner of fear, uncertainty and/or doubt sown by the more nefarious shorts. And yeah, that happens – rules be damned, unfortunately. Those rumors show up in all kinds of places, from Yahoo message boards to crazy WiMax iPhone 5 rumors to mysterious, debunked calls to Bloomberg reporters. It’s ridiculous, but it’s also just part of the market these days.

I would characterize Clearwire shorts as follows:

1 – Those simply hedging the 2015 notes sold last fall. These would be hedge funds largely agnostic about most daily events at Clearwire simply covering their bases.

2 – Those seeing an attractive short-term trade. For instance, by selling puts on Sprint (or just going long), buying Clearwire’s 2015 senior secured debt and then shorting Clearwire’s common stock, you can create the cash that pays for the purchase of the 2015 notes, exploit the angst between the two companies, and have plenty of breathing room if there is a debt to equity conversion at Cleawrie. This trade looks good on paper, I suppose, but it (a) implies that conversion and/or major equity dilution, (b) assumes Sprint is better positioned than Clearwire, and (c) I would guess this trade is getting pretty crowded at this point. But those notes are certainly attractive right now.

3 – Those who believe the company will fail – or be massively diluted. These would be shorts who are skeptical of the company’s future based just on its historical financials and past mis-steps – and/or who believe Sprint is actually trying to bankrupt a company it owns most of. I simply disagree that this will be Clearwire’s fate, as per my previous articles. Time will tell who is right. To their credit, the shorts in this group seem to admit they don’t know the technologies here…and to me, that’s kind of the whole point – it is key to the company getting funding. More on this below.

4 – The amateur shorts. If internet message boards are any indication, this group is large.  It includes quantitative and momentum shorts as well.  Ten years ago, shorts were to be respected, but nowadays, any idiot can short anything, and many do. See this post for a whole ‘nuther rant about restoring the uptick rule and reigning in high frequency trading.

It is also important to realize that the shorts – particularly the ones that own Clearwire debt – have large financial incentives to see the company fail. It’s nothing personal, mind you, it’s just math. Should the company go bankrupt, then shorts never have to cover, and debt holders could acquire that valuable spectrum for pennies on the dollar. Fortunately, a decline in stock price alone can’t make a company go bankrupt, and Clearwire’s actual financials have never been better – though we still have a way to go, and they admittedly are taking a back seat to the company’s current funding quest. So the traders will likely have their way with shares until the current fundraising uncertainty is cleared up.

All that said, in my opinion, Clearwire is currently a dangerous short. If Clearwire raises money, signs a new wholesale customer and/or sells excess spectrum, theshort case in 2 through 4 above evaporates instantly. And we know they are currently working hard on the first two efforts there, while the third would be on the table for the right price. Same holds true to a lesser degree if Sprint clarifies on next week’s conference call that it will contract with Clearwire for WiMax beyond 2012. Lot of risk there, too.

So I can’t quite grasp the bravado on the short side here. Clearwire has one significant challenge right now – finding funding – but shorting based on the presumption that they won’t be able to do that seems awfully risky.

Q. Have you considered that Sprint is trying to force Clearwire into bankruptcy?

A. This is a common perception floating around out there – ever since the Pardus Capital letter this past May and the Sprint analyst day disaster of a few weeks ago. Fortunately – and this will disappoint the conspiracy theorists – it is also doesn’t hold any water. Never mind my own beliefs about why this is silly – here is the Chairman of Sprint’s board two weeks ago when asked about Sprint’s partnership with Clearwire:

“No question we want them to do well; it’s in our interest that they do well. Nothing good happens in a restructuring and there’s nothing good in the outcome of that.”

Sprint will also be paying Clearwire a minumum of $1 billion dollars between now and the end of next year as a wholesaler. That is not the act of one company trying to drive another into receivership. That, incidentally, is also why I am not losing sleep over the impact of new iPhone sales at Sprint on Clearwire’s wholesale subscriber trends. By the time we’ll have that billion dollars in hand, we won’t have to guess and can react accordingly then.

Q. What do you say about the fact that Clearwire’s spectrum is of poor quality?

A. In short – hogwash. Let’s step back for radio frequency (RF) engineering 101.

It’s important to understand that the advantages/disadvantages of any given band of spectrum are not absolute – Clearwire’s included.

Of the many trade-offs an RF engineering team has to make when covering an area with a wireless network, the dominant one is the balancing act between the capex budget, cell size and signal strength. It’s really just physics…higher frequency radio frequency waves can carry more traffic, but they peter out quicker and can’t penetrate structures as well.

Lower frequencies, like those owned by AT&T, Verizon, Sprint and T-Mobile travel farther and fade less over time – so they are "stronger" in the sense they’ll travel through walls. Up to now, in telecom, because all those guys were racing to cover the country with signal in order to get and keep as many subscribers as possible, those lower frequencies were an advantage, because the capex costs were lower (specifically, they had to deploy less towers at $150K each to cover the same area then someone at a higher frequency would have to). So in the historical sense, that spectrum has been judged of higher quality – for that capex reason, really.

But if the whole country is already covered with RF waves, and there is more stuff traveling over them, then coverage is irrelevant, and it’s all about making sure your dense areas can handle all the additional stuff. And higher frequencies are better suited to that.

If you’re an RF engineer designing a new network, you’re basically looking at a map, trying to figure out (among other things) where you should stick the pins, or cell phone towers. Around each pin, you draw a circle with a radius of a certain length. That radius depends (among other things) on the frequency you’re working with.

Lower frequencies travel farther, so your circles are bigger. But you’ll have a lot of empty space in the gaps between where the circles come into contact. No coverage in there. And if you overlap the circles too much to reduce that empty space, you can get interference problems – or you’re wasting money covering an area with redundant signals and gear.

But a wide range of frequency in that area is yours to use in your design. You’ve just got to play with all the variables long enough until it all comes together.

So let’s call that entire frequency range in a given area "A to F." You can best cover that whole area if you (a) minimize the gaps and the overlaps in circles and (b) alter the specific frequencies going in and out of any cell sites sitting right next to each other, to avoid interference problems. So, the first cell site will use frequency A, the next one uses B, etc, until you get to a location far enough away from the first one that you can use frequency A again without interfering with anything other A’s around. RF engineers call this "frequency re-use."

In a capacity-based model – one that means you’ve got to get really deep, tight circles – those large radius cells of the traditional lower frequency operators are in that sense less efficient because you don’t always know how closely you can snug the circles together without causing interference problems. Those signals lose strength gradually and inconsistently. So, frequency re-use can be poor.

You can have much more confidence in the spacing of smaller radius circles (like Clearwire) – and better re-use – because it’s much clearer where the signal is going to drop off. And if, say, four of those little circles cover the same general area as one big circle from above, then after enough of them, they’re going to fill in those previous gaps between big circles, too. The problem is you need more towers to do that, which jacks up your cost greatly. Only Clearwire has already built them out – thus the lack of cash flow to date.

Anyway, that’s greatly simplified. For one, the RF guys use hexagons to cover areas, not circles, but never mind. The point is that advantages in a coverage-based approach to designing a network can be disadvantages in a capacity-based model, and vice versa. Clearwire’s higher frequencies give it an advantage in the capacity-based approach – in no small way because its already put up all its towers and has better frequency re-use. It’s just inherently more efficient. For that kind of use.

A dozen years ago, the propagation problem through walls for higher frequency signals was an issue. That will not be anywhere near the same kind of problem for Clearwire under a wholesale model, though, because (1) smartphones contain chips that also can pass traffic over the WiFi that already exists inside many buildings in urban places – and didn’t a dozen years ago – and (2) as long as another carriers’ lower frequency signal can get inside the building, then the initial lift on getting that traffic from a phone to back outside the building is much more the responsibility of that carrier, and Clearwire picks up the slack, so to speak. In a wholesale model, with everyone on LTE.

So that propagation through walls issue for Clearwire is more of a problem in their retail network – but as they’ve announced, they’re no longer trying to compete there. They’re basically going to milk that retail business to fund the wholesale side. So whatever complaints you’ve seen online from retail customers are no doubt valid…but at some point, they’ll be irrelevant, too.

Also – on the differences between TDD and FDD.  The term "disruptive" is easily the most abused adjective in all of tech, but here’s why it might actually apply to Clearwire’s TDD LTE.

Actually, first you should know that the difference in efficiencies between the two technologies can be defined by a number of things – ranging from chipsets to duplexers to power consumption.

What I think is most relevant, though, is that (a) mobile data traffic is very asymmetric…meaning much more stuff is downloaded than uploaded; (b) asymmetric data is difficult/inefficient for FDD systems to handle, because they were originally built for voice, where traffic up and down is roughly the same; and (c) that asymmetry is inconsequential in a TDD system like Clearwire’s because you can so easily change the up/down ratio.

So here’s my lousy metaphor:

TDD (like Clearwire will have) is sort of like a single two-lane road with a clairvoyant yellow line painted down the middle that moves to either side to let more cars go by.

FDD (what all other 4G operators will have) is more like having two one-way roads that are running parallel to each other, with traffic going opposite ways, and a big wall separating the roads from each other.

All things being equal, you can move the same number of cars both ways, but one is going to be a helluva lot more efficient.

The size of Clearwire’s spectrum means all things are most definitely not equal, though…specifically, that spectrum represents a much bigger road with many more lanes. So more cars can use it. And combined with those crazy lines of TDD, it is a very tough combo to beat. Bigger roads, better traffic management.

All that said, if there is an undesirable part of CLWR’s spectrum, it isn’t due to the frequency or technology, but ownership. Specifically, a relatively high percentage of that CLWR spectrum is leased versus owned outright, and based on industry history, the company’s leased spectrum could be valued less than the owned spectrum by the two big dogs in the industry.

To be clear, that is irrelevant to me at these prices, and implicit in my thesis is that the pending spectrum scarcity problem will either outweigh the historical acquisition preferences of AT&T or Verizon, or – perhaps more important – the company’s willingness to opportunistically sell spectrum to anyone, including private equity firms, means the valuation should reflect the perspective of a much broader range of firms than just AT&T and Verizon.

But there is no denying that AT&T and Verizon have shown a historical preference for license ownership in their acquisitions. I just believe the pool of potential buyers and/or ways to monetize the CLWR spectrum asset is much bigger than those two firms, which is why I did my valuations the ways I did. Plus, AT&T also just paid a big premium for the QCOM bands for which nobody even makes gear yet, so they might not be as stingy here as I am presuming, either, in a theoretical buyout. Too hard to say.

Anyway, that’s a minor but legitimate knock on the spectrum – and for some folks it understandably adds even more uncertainty to things. But it’s overblown to me. Again, I think things are cheap enough down here that even if you thought it necessary to attempt to capture that owned/leased issue in a valuation sense, it would still be immaterial in a pragmatic sense for a potential buyer like DISH, or anyone else, really. So, it’s a negative, but it always has been, and the most likely outcome related to it would seem to be that AT&T or Verizon would attempt to claim a lower price on that portion of leased spectrum in the event of a buyout.

I can live with that.

Q. What do you think about the threat from LightSquared?

A. I don’t want this to sound too harsh, but I have yet to talk to anyone I would consider a reputable source in the industry that believes LightSquared will be a credible operating company. That said, it’s possible I haven’t asked around enough – and I am no doubt biased against them, both as a potential competitor to Clearwire and because they appear to be trying to exert political pressure to get what they want, which annoys me to no end.

These guys knowingly acquired slices of spectrum that were absolutely going to interfere with GPS, then bought political favors in the hopes that they could simply shrug their shoulders when it came to solving the very real GPS interference problems that they created. Things just don’t get much more corporatist.  Where are the hipsters in Zuccotti Park on this?

It’s a crapshoot on whether or not L2 gets FCC clearance, but that is hurdle one of many dozen, really, on the path to building a viable company. One of the other immediate challenges for L2 is, well, raising $3 billion – and some portion of that is likely going right out the door to Inmarsat for debt and lease payments they’ll owe.

So, as an investor, you can either write a $3 billion check to L2, see a chunk of that forked over to another company straight away, and then hope to see an undefinable eventual return on that money in about five years…or you can write a $1 billion check to Clearwire, which is guaranteed to bring in at least a billion dollars from Sprint next year, possibly see a cash on cash return within 18 months, and own a piece of the most attractive remaining independent asset in all of telecom on the eve of the Great Spectrum Crunch.

No brainer. And it seems logical to presume that both companies are probably talking to at least some of the same investors.

That said, L2 does have a point in that the GPS manufacturing industry has been sloppy to date about bleedover in their receivers…but those guys also no doubt felt entitled to bleed over given the nature of those high precision uses, and the fact that all RF anywhere near it was supposed to be low-power. I would guess that’s probably why the FCC has let it slide all these years.

Regardless, there is a lot of spin on both sides of the debate, and we still do not yet know how L2 will address any potential latency problems, as per my original article.

But assuming the FCC okayed L2 using Javan’s mysterious filters, and the GPS industry somehow agreed to eat that $400m retrofit cost, interference is unequivocally still going to be a problem in the L-band for high precision receivers. It’s just that L2 says it’s no longer going to be their problem.

And that brings me to this:

It’s hard to see the FCC endorsing the position that no one is responsible for that interference, as approval of L2 would imply.

And I would also doubt that unfunded mandates foist on everyone operating a high precision receiver in an election year will get much traction, politically.

Regardless, I thought Sprint said it best at their analyst day; the capacity that L2 represents still won’t be enough in the long-run. They believe they’re going to need new spectrum in about four years, regardless. And after all the horsetrading L2 is doing now, the usable spectrum they have will be much reduced if they do get through the FCC, at least for the foreseeable future. Specifically, they’ll have a 10Mhz by 10Mhz carrier – or what Verizon has right now for their LTE network. It’s not a small slice by any means, but my point is they’ll all still have the same problem at year five or so – needing more spectrum.

So if L2 makes it through the FCC, that might delay the time frame over which Clearwire can monetize its spectrum, but TDD over bigger swaths still gives them the easy nod in the long-run – again, assuming, as everything does these days – that they can find the funding to get there.

The other interesting angle to all this is that if L2 can’t get FCC clearance by the end of the year, then Sprint can cancel its agreement and move on to other things. So, it’s going to be interesting one way or the other.

There are also those who do not necessarily believe it is a zero sum game between L2 and CLWR, either, depending on how everything shakes out…though you can probably guess my preference for how all this might end.

So take all of that for whatever it’s worth.

Disclosure: Long CLWR. None of the above should be taken as investment advice. Ever.