Multiple reports have appeared in the industry press recently about T-Mobile investing in Clearwire to tap into its spectrum or using it as a wholesale provider for T-Mobile’re retail customers. Other reports mentioned T-Mobile and AT&T participation in a private auction for Clearwire’s spectrum. More recent reports surfaced mentioning that interest in Clearwire’s spectrum dwindling. It appears that earlier reports were the result of wishful thinking. The fundamental reason is Clearwire’s spectrum is not as valuable as it is claimed to be. Let’s analyze why.
First the financial aspect: Last May Business Week published an article claiming that Clearwire spectrum may be worth as much as $20B. There were quotes from industry pundits (in the same article) claiming as much as $50B for the same spectrum. $20B was based on a $0.50 per MHz-POP which was the average price paid by AT&T and Verizon during the 700 MHz auctions. Considering Craig McCaw has promised to pay $5B over 3 decades to the original owners of the spectrum, even $20B sounded a great deal. However, Yahoo Finance computes enterprise value for Clearwire at $2.77B as of January 25th, 2011. This renders the spectrum value to be closer to $4.5B as shown in the 2009 company yearly report. Since the current stock value correlates with the enterprise value given by Yahoo, it seems the market is pricing their spectrum assets to be what company reported in their SEC filings. On the other hand, Clearwire was planning to sell approximately 40 MHz (considered to be around 1/3 – 1/4 of its assets) for $5B, their expectations were more inline with the Business Week figure of $20B.
Now let’s look at the technology situation: Many analysts have already provided assessments of merits and challenges of using the Broadband Radio Service (BRS) and Educational Broadband Service (EBS) spectrum that span the band 2496- 2690 MHz. Comparisons were made between 700 MHz and 2.5-2.7 GHz bands with respect to number of cell sites needed and indoor penetration capabilities. One view (advocated by 700 MHz auction winners Verizon and AT&T) suggests that:
- Deployment at 2.5-2.7 GHz may need 2-4 times more sites for the same coverage.
- Indoor penetration of 2.5-2.7 GHz would be extremely poor compared to 700 MHz.
Even though both assertions are true, they don’t reflect the whole truth. Since technologies such as LTE, WiMax, HSPA+ are all targeting data usage explosion, it is too simplistic to assume that a thin macro coverage layer will be sufficient to handle the demand. Contrary to the original radio network deployments in previous decades, in the future operators will be deploying more and more capacity-limited networks. In such a scenario, there are two basic parameters operators can control:
- Total available bandwidth
- Density of cell sites
Following figure from FCC website describes the details of channelization within the 700 MHz band.
Verizon is at an enviable position where it has close to an average of 20 MHz of spectrum (downlink) in major population centers. However, even this spectrum is not going to be adequate to have a thin macro coverage layer in the long run. Since Verizon’s EV-DO network is rapidly becoming outdated (compared to AT&T and T-Mobile’s HSPA+ networks), it will move its data customers more rapidly to its new LTE network. This will require Verizon to start deploying its 700 MHz LTE network in micro (street-level) and pico (in-building DAS) modes. We think Verizon will start using its AWS spectrum assets as well to cope with capacity hotspots or as femtocell carriers in 2012-2013 timeframe.
Compared to Verizon, AT&T has more limited spectrum within 700 MHz band. Between its auction 73 purchases, and Aloha Partners acquisition, it was sitting on less than 10 MHz (downlink) average spectrum in lower 700 MHz band until recently. However, with the recent purchase of Qualcomm’s unpaired band D, it added another 6 MHz. Even though it is quite difficult to use band D in the downlink direction since it is adjacent to band C uplink, we expect that would be AT&T’s objective. We anticipate that such a terminal may require a separate Rx path for channel D in addition to the typical Tx/Rx used for channels A, B or C instead of sacrificing valuable spectrum as guard band. Similar to Verizon, we anticipate AT&T eventually start deploying LTE in AWS band as well. However, since AT&T still has a lot of mileage left with its HSPA+ technology, deployment of AWS may not be at the same pace as for Verizon. Obviously the precedence of AWS band versus use of lower 700 MHz band D would be dependent on solving technical challenges for device designs and any standardization changes related to band D.
Compared to Verizon and AT&T, Sprint and T-Mobile have significant spectrum limitations. Even though both companies combined have less customers than Verizon or AT&T, their growth is tied to the availability of spectrum. In order to better understand the spectrum limitations of Sprint and T-Mobile, let’s look at the following table showing the spectrum positions of four big carriers along with Clearwire’s spectrum in the most populated five counties in the USA. In this table we include only the downlink bandwidth since mobile networks will continue to be downlink skewed due to the asymmetry in data usage. For Clearwire’s BRS band we assume 80% of the available bandwidth to be used for downlink. For AT&T 700 MHz band, we didn’t include channel D.
Table above clearly shows that for Sprint, T-Mobile and Clearwire BRS band is the only viable option to deploy a technology competing with LTE. So far T-Mobile has clearly indicated that they will be deploying LTE as the next generation technology after fully exploiting HSPA+. On the other hand, Sprint and Clearwire (same network) have already been deploying WiMax in TDD mode for few years. However, it looks quite likely that with the recent changes in Clearwire leadership and Sprint’s announcements they will adopt LTE as the next generation technology to replace WiMax.
In order to better understand the use of 2.5-2.7 GHz band, let’s look at the following figure from FCC website describing the new frequency plan.
- EBS spectrum, originally allocated for educational video purposes, has certain licensing clauses such as 20 hours of educational use per 6 MHz channel and 5% of transmission capacity being dedicated to educational use.
- BRS as the successor of Multichannel Multipoint Distribution System (MMDS) is a dedicated for commercial service.
We expect Clearwire to focus on primarily the BRS spectrum due to license restrictions of EBS as well as limited number (20) of mostly secondary markets for the EBS availability.
Biggest problems for the BRS spectrum are its fragmented nature and lack of paired channels forcing the use of TDD. Even though European countries adopted the same band for LTE use, they decided to use 14 5 MHz paired channels in 2500-2570 MHz and 2620-2690 MHz bands to deploy it in FDD mode. Unfortunately it is too late to impose a new channelization plan to harmonize the US 2.5-2.7 GHz band with the European spectrum plan.
Since we also established the likely plans for Verizon and AT&T to use 700 MHz and AWS bands in FDD mode, we can envision a possible scenario where an iconic device like iPhone may not be available for the 2.5-2.7 GHz band especially if Sprint, T-Mobile and Clearwire end up going their separate ways. In order not to repeat the current situation where T-Mobile ended up as the only AWS operator for a long time, best thing Sprint, T-Mobile and Clearwire can do is to bring their forces together (options ranging from a combined company to sharing a common radio access network) to deploy LTE in the 2.5-2.7 GHz band. Hence, the real price of Clearwire’s spectrum is heavily dependent on the future plans of these three companies and how they can collaborate to exploit this band. Unless they combine forces it is unlikely that its valuation will be of the same order as the 700 MHz band ever. On the other hand, with combined forces among these three companies it is quite possible to achieve the necessary scale to have the next iconic device: iPhone with LTE/HSPA+/UMTS/EDGE/GSM or LTE/EV-DO/IS-95.