Conan Kudo

Are there any opex changes then? Because wasn't part of Network Vision supposed to reduce opex by simplifying network structures?

Once the main Network Vision program is done, Sprint will ease off capex and reduce it accordingly, which naturally changes the balance sheet. After this year, I'd expect significant cuts in capex to shift Sprint back to profitability.

We don't know about the long-term health of either of these two companies because all financials right now are influenced by the heavy spending on network upgrades right now. You can't just subtract it, either, because there's always a degree of capex spending involved. You can look at previous trends and develop a model of what it might be like if Network Vision and the modernization program were completed. Here's what I see it being like: Sprint would still be losing money, but the loss would be narrower. However, the lack of the headline loss would cause more to peek into the deeper problems, such as poor execution ability. There are some serious fundamental issues with Sprint that need to be addressed, but hopefully Masa is working on that as we speak. T-Mobile would be turning a profit of about $1 billion, because it would ordinarily ease off into $2.9-3.2 billion of capex if modernization wasn't accelerated. Because T-Mobile has been firing on all cylinders and is quite ahead of schedule, it went ahead and accelerated its program, which pushed it down to lose money for the quarter.

I think that AT&T and Verizon are OK with the Sprint/T-Mobile merger. T-Mobile/DT definitely are. I think that if Sprint does a spectrum swap with Dish and then hosts Dish's spectrum in a joint venture that will concentrate on fixed broadband, then the merger will be acceptable to the feds. Stable is bad, in this case. Right now, we don't have a price war. We have healthy competition on value propositions. T-Mobile competes based on features for price, Sprint competes with its unique plan structure, AT&T competes on bundled value and multi-layer mobile broadband, and Verizon competes on coverage of LTE. In the event Sprint and T-Mobile were to merge, the effect would result in three "stable players." This is similar to Japan and Korea, where after consolidating to three players and they reached relative equilibrium in terms of what they wanted, they stopped truly competing. We won't get a price war. Right now, Japan's telecom market is stable as a rock, enabling SoftBank to invest in other businesses and NTT to do research into new telecom technology. Three stable players means that competition would level off, rather than increase. Also, there would be no incentive to compete as long as you can hold nearly a third of all U.S. customers. As long as that balance is held, then there's no point in trying. McAdam isn't stupid. He's keenly aware that the scenario that took place in Japan is the ultimate destination in the U.S. if more consolidation among major players is permitted.

Charlie doesn't negotiate. That's why AT&T/DirecTV instead of AT&T/Dish. SoftBank won't buy Dish either, for the same reason.

Why? DirecTV owns no satellite assets directly. EchoStar, controlled by Ergen, acquired all of Hughes satellite systems. DirecTV and AT&T don't compete at all. In fact, the two have been strong partners since the SBC/BellSouth days. The only wireless spectrum assets that DirecTV owns are for Brazil and Colombia. Obviously, since BellSouth divested its Latin American units to Telefónica nearly a decade ago, AT&T doesn't have anything there. And AT&T is divesting SBC's 10% stake in América Móvil to prevent arguments with Latin American regulators. The situations are completely different. Sprint and T-Mobile compete in every market and they compete for the same customers. So unless you are completely blind to how M&A works, you'd realize that DirecTV/AT&T and Comcast/TWC have no effects on changing the regulators' mind on T-Mobile/Sprint.

Still no. While LTEiRA doesn't prevent roaming deals, it definitely prevents SRA-style deals. That's why Pioneer Telephone left the SRA for LTEiRA not too long ago.

It's generally the same as prepaid, though the limits that trigger the throttling differ wildly based on the agreement. And of course, there are some agreements that don't have them at all, like Ting's (because access is fully metered).

Now all your WCS are belong to AT&T...

No. U-verse will likely be expanded by DirecTV programming agreements, and DirecTV will continue to be available for DSL subscriber bundles.

Heh, no. I know lots of stuff, but I'm no super-machine. You can't assume that everyone who leaves are bandwidth hogs. You make a great deal of assumptions without data to back it up. I get that this is a Sprint-focused site, but we have to be realistic here. The current trajectory is not good, and even putting its network back together may not be enough.

M2M revenue is also pretty high margin (~90% or greater). And subscribership isn't a zero-sum game, which is why you can have >100% penetration in the market. You sound so confident that the gains are temporary. But the data collected by most analyst groups seems to contradict that. In fact, they believe the subscribers are stickier with T-Mobile than they were with their previous mobile network operators. So I will have to disagree.

Yes, but T-Mobile still needs to support several generations of AWS UMTS devices that lack support for PCS. While the number of users who have AWS-only UMTS devices continues to dwindle, it's still significant enough to warrant maintaining it, especially in markets like Cincinnati, where many users come to T-Mobile from Cincinnati Bell Wireless.

That is dumb on so many levels. For one, GSM/UMTS earn T-Mobile more than 70% of its revenue. Within that, UMTS earns T-Mobile greater than 65% of its revenue. GSM, while steadily earning less revenue, is still profitable to maintain due to domestic and multinational roaming and M2M services. As UMTS is shifted from AWS to PCS, it can increasingly take over the role of GSM, allowing GSM to fade from the network. AT&T, ironically, is helping this by infusing a massive boost into the UMTS ecosystem by driving costs down to integrate UMTS into devices. It's now just as cheap as GSM to put into a device, in large part thanks to AT&T. As for CDMA, the reason it's not worth maintaining is due to the increasing costs in the ecosystem. As operators steadily convert from CDMA to UMTS and cut off orders for CDMA-enabled devices, the cost of supplying those devices and network gear goes up. About a decade ago, a CDMA device would probably cost roughly the same as a UMTS one, because nearly all of the Americas maintained CDMA networks, as did several countries in Africa and Asia. This is definitely no longer the case. As a result, a CDMA device costs many times more than its UMTS counterpart. And more CDMA operators are disappearing every month. A few months ago, S-Fone in Vietnam declared bankruptcy and completely shut down. Bangladesh's CityCell is in the process of shutting down its CDMA network for GSM, pending approval to convert its mobile license to GSM/UMTS and turn on the replacement network. CityCell no longer offers CDMA roaming services, and has been preparing for two years for the switch. If its request isn't approved, CityCell will likely shut down. Bell and Telus in Canada are jointly shutting down CDMA across the country throughout the year. Movistar (owned by Telefónica) completed shutting down its last CDMA network in Venezuela about a month ago. Iusacell in Mexico has successfully migrated nearly all of its subscribers to UMTS and is repurposing CDMA spectrum for LTE service, in partnership with Nextel Mexico and Movistar. China Telecom, the biggest CDMA operator in the world, will likely fully switch back to GSM through a network sharing agreement being hashed out by all three Chinese mobile network operator companies to speed up LTE deployment. China Telecom already provides to its customers access to the China Unicom/China Mobile GSM network through its dual-mode CDMA/GSM devices that use GSM1X for the CDMA part. Sprint's problems are compounded by the fact that no one outside the US use PCS for CDMA. And Sprint is the only one in the world using ESMR for CDMA. This makes CDMA device procurement exceptionally expensive, which is why Sprint has trouble with devices from time to time. KDDI elected to use it for LTE to avoid the cost issue, and Nextel Japan remains in limbo (though I expect it to have been fully shut down for the last few years, and they may not even have SMR licenses anymore). We don't live in a vacuum, as much as many would like to think.

I've mentioned it multiple times. I immensely prefer a network sharing agreement where Sprint and T-Mobile share GSM/UMTS/LTE networks, maintaining a common platform while executing on developing broad coverage and high capacity networks using the broad portfolio of spectrum the two companies have. In such a scenario, Sprint would still need to retire the CDMA network, as network sharing in CDMA is not feasible. It's somewhat possible, but it's so ugly, difficult, fragile, and not worth it. 3GPP networks have been designed from the beginning to support network sharing very easily, which is why ACS and GCI in Alaska transitioned to GSM/UMTS/LTE after merging their networks into a NetCo called AWN and now use the shared network. Why should they skip AWS-3? AWS-3 offers T-Mobile the opportunity to get AWS LTE up and running in several markets where it can't now, such as Cincinnati. And the AWS ecosystem is much larger than the PCS one for LTE, making it much more cost effective for both companies.

I doubt that the PLMN ID of 310-120 is blocked, because most modems treat GSM/UMTS/LTE as one big thing (3GPP authentication). While Sprint has disabled 3GPP authentication in US carriers, by its very nature, it had to allow its own. With something like MOCN, that won't be a problem. I seriously doubt DT wants to sell, since it would make very little sense. Post internal reorganization, this is the contribution to DT's revenues. EE (UK JV with Orange, S.A.) is under "Group Headquarters & Group Services", not "Europe". Note that DT made a profit this quarter. While TMUS had an operating loss of ~$100m this quarter, the losses are temporary, since they are largely attributed to the huge influx of EIP usage and waiting for value to come in from traded-in devices being resold. Despite that, DT's revenue went up 8% because of the US unit, and profit increased as well.

A switch to UMTS wouldn't be a disaster for Sprint or its subscribers. >60% of currently sold phones support UMTS on PCS, and the overwhelming majority of phones sold by Sprint are UMTS capable. This is not an issue for Sprint. In fact, the opposite. Most likely, there'd be a massive capacity gain I suspect that the T-Mobile branding may be an issue, though. I doubt Masa would like to kill the Sprint brand so quickly, but he may believe the brand is too tarnished to keep. However, you're wrong about why Masa switched the Vodafone brand to SoftBank. He switched it not because of licensing fees, but because the Vodafone brand itself was bad in Japan. Japan is insular and xenophobic by nature, and when J-PHONE rebranded to Vodafone in 2004, the attractiveness of the mobile network operator to Japanese consumers dropped like a rock because Vodafone is a "foreign brand" that sold phones that "weren't Japanese enough". When SoftBank rebranded Vodafone to SoftBank Mobile for mobile, and fixed assets to Yahoo! Japan, prospects immediately began improving. It wouldn't be until 2010 with the iPhone 4 (the first iPhone that supported docomo's UMTS band 6 that SoftBank used for roaming at the time) that Japanese consumers began reconsidering foreign brands and foreign-designed devices. In any case, I think a network+spectrum sharing agreement to share GSM/UMTS/LTE networks would be much more palatable than a full-on merger. And indeed, it would preserve the four-player market, while providing both with the effective scale in the market they need. The independent company that manages the network would have access to BUYIN, the Deutsche Telekom+Orange S.A. global handset and network gear procurement joint venture that offers highly discounted GSM/UMTS/LTE hardware to DT/Orange companies. This would naturally extend to Sprint indirectly, since T-Mobile US currently participates in BUYIN.

Unfortunately, Netflix's adaptive bitrate mechanism is designed for throughput throttling, not latency throttling. There's not a lot of things you can do to prevent the disruption that latency throttling causes.

So there is one particular use case where latency throttling will be noticed, no matter what: streaming. Your streaming will break. You'll have buffering and lag like nobody's business on video streams, and in some (very rare) circumstances, picture and sound will desynchronize. With audio streams, you'll buffer more often. If you're streaming a Flash-based presentation or a PDF data object, you'll have major corruption issues, because if the packets don't come in fast enough, an EOF byte is injected by the client in the middle (ironically to prevent zombie connections that cause dead capacity in networks), causing the file to corrupt. This is why T-Mobile tries really hard not to invoke latency throttling, or why others have a stable throttle (like AT&T with Aio/Cricket). Latency throttling (which is what Sprint calls "network prioritization") is the kiss of death for any type of streaming. That's why I really hope Sprint has some form of load rebalancing algorithm in place that kicks in before it tries to do this. If it goes straight to the throttling mechanism, there could be serious repercussions. I've been checking, and I've yet to turn up any information that proves that Sprint has load rebalancing supported for this circumstance.

Only because no operator signed up for it. In this case, Sprint/T-Mobile will be big enough to do it. But like I said, VoLTE is close at hand, so I am not sure they will gain much by it. Actually, one did implement GSM1X. China Unicom did and when it transferred those CDMA assets to China Telecom, China Telecom continued to use it. Historically, all three Chinese operators used GSM. It was later that WCDMA and CDMA2000 and TD-SCDMA were deployed. The only reason CDMA2000 was deployed there was because of GSM1X. That's also why SIM cards are required to connect to China Telecom's CDMA network. It uses GSM authentication. And also, China Telecom customers typically have dual-mode CDMA/GSM devices, so that they can access the China Mobile/Unicom GSM network, too.