T-Mobile has a tiny bit more spectrum than Sprint and it's all standard (PCS/AWS) spectrum. T-Mobile has a decently standard GSM/UMTS stack. Sprint is running two non-integrated networks.
On a per-customer basis, T-Mobile isn't so bad on spectrum. 33.7M customers with 50MHz of spectrum. That's actually a better ratio than any other carrier. For comparison
* T-Mobile: 1.48MHz/million customers
* Sprint: 0.92MHz/million customers (excluding 100MHz+ via Clearwire)
* AT&T: 0.93MHz/million customers (including pending 6MHz gain from Qualcomm)
* Verizon: 0.82MHz/million customers*
* The Verizon numbers exclude substantial purchases recently announced which will probably increase Verizon's spectrum depth by nearly a third.
So, T-Mobile isn't in a terrible spectrum position and the spectrum AT&T has to give them as part of the break-up will only help that. Likewise, $3B is a lot of money that could go toward an LTE rollout.
In terms of LTE, Sprint didn't have an LTE plan until mid-October of this year and doesn't expect to have any LTE devices until the third quarter of 2012. If T-Mobile put together an LTE plan today, they wouldn't be behind the times. I mean, they'd be behind Verizon and AT&T, but they could roll out as fast as Sprint (which just raised $4B for its LTE rollout - barely more than the $3B in cash that T-Mobile is getting as a breakup fee). Plus, LTE in America is a bit ahead of the curve. Mostly that's because carriers in other countries can stretch HSPA+ (as T-Mobile is doing) to provide much higher speeds than 3G.*
So, T-Mobile does have the ability to move forward if DT wants to. It has the spectrum, it has $3B in new cash, and it's profitable (which Sprint isn't). I think DT wants to see T-Mobile a bit more successful if it is to stay on in the US, but it is profitable. If T-Mobile can forge a joint venture with Dish Networks (which is interested) it would gain 40MHz of unused spectrum from Dish and might have a partner who would promote the service to its customers. We'll see. I'm definitely not saying you're wrong. T-Mobile needs a bit of energy and momentum to start gaining customers again. However, I think T-Mobile has the assets needed to keep going.
--
The one area where I think you're spot on is in terms of competition with Verizon. When AT&T Wireless and Cingular merged, it increased competition. Verizon had been allowed to merge together greater assets than its competitors previously. Allowing AT&T Wireless and Cingular to merge meant that there was a real, nationwide competitor to Verizon with low-frequency spectrum that provided good coverage. Right now, AT&T and Verizon have decently similar assets (although the Alltel merger gave Verizon a slight edge). AT&T has a greater number of data customers which puts a bit of a strain on their network, but they should be able to manage it now-a-days (and the data shows that AT&T has a better dropped call/blocked call rate than Sprint, although slightly worse than Verizon's, and significantly faster data speeds). If the FCC allows Verizon to gobble up the licenses from Cox, Comcast, Time Warner, Leap, and others, it will again create an imbalance that will see Verizon having superior assets that make it hard for other carriers.
* Before someone says, "but HSPA+ is offering 3-6Mbps while LTE is offering 20Mbps" I want to explain a bit. HSPA+ is backward compatible with HSPA and UMTS. So, when AT&T upgrades its UMTS network to HSPA+, there are tens of millions of voice and data customers on it. By contrast, the LTE networks we've seen have only had a couple million customers on them. As more customers use it, the spectrum will get more crowded and speeds will likely drop as a result. Likewise, there are many different HSPA+ versions. AT&T is running HSPA+ at 21Mbps while T-Mobile has it up to 42Mbps in some areas. However, even HSPA+ 14.4 has been showing average real-world speeds on the iPhone 4S of 3.2Mbps (http://www.macrumors.com/2011/11/18/nationwide-study-shows-a...). That's on a loaded network using 5MHz channels. The LTE speeds we're seeing are on a near-empty network using 10MHz channels (from launch to the end of the third quarter, Verizon has sold less than 3M LTE devices). None of this is to say that HSPA+ is superior, just that we're seeing LTE showing better numbers than it would under the same load HSPA+ networks are experiencing and that means the gap is actually smaller and explains how overseas carriers are using HSPA+ in the interim to offer customers a better experience without the cost of LTE (and it is a cost to the consumer as well as one can see from the price of HSPA+ devices against the cost of LTE devices).
How does Sprint's "Network Vision" impact the MHz/million figures? Can we expect to see improvements there? (Excluding challenges due to backhaul off the towers....)
It doesn't. Sprint isn't getting new spectrum. Depending on how things fare with Clearwire, it might gain some additional capacity, but it looks like Sprint is going to be mostly alone for LTE. Clearwire will help a bit, but it isn't going to be as close as it was in the past. Sprint's current plan is to build its own LTE network and give Clearwire money for the use of Clearwire's network in high-traffic areas.
It's hard to say what Sprint's "network vision" is. Sprint announced it in December 2010 as integrating their SMR, PCS and Clearwire's spectrum and services. Since then, Sprint and Clearwire have gone their own way and come back a bit, but Clearwire will be building its own network. Sprint's efficiency will increase when it gets rid of iDEN and refarms that spectrum for CDMA, but it doesn't change the MHz/million figure. In fact, it mostly means that Sprint is running a less efficient network right now than the MHz/million would suggest.
I like that Sprint has a concrete plan to roll out LTE and has raised the money to do it. However, it's going to be challenging. AT&T and Verizon have a decent head start. If you're wondering, "will Sprint be able to become big like Verizon and AT&T?" the answer is no. Sprint doesn't have the spectrum nor the capital right now.
That said, there's a lot of value in smaller carriers. Sometimes people think that "competition" is about winning. It isn't (at least in terms of economics). Having smaller carriers around can mean better pricing for customers whose usage patterns differ, a different mix of services that match different people in different ways, etc. Millions of people have chosen Sprint and T-Mobile. As long as those companies continue to make their customers happy (and can find a way to be profitable), that works. I think Sprint's future won't be as big as AT&T or Verizon's, but I think their commitment to LTE will help them continue to make their customer base happy if they're able to accomplish it in a timely manner.
EDIT: I should note that Sprint is in around the same range as AT&T and Verizon in terms of MHz/million customers. There are some economies of scale there due to how wideband channels fit into the spectrum, but Sprint isn't unduly spectrum poor. I believe they have 10MHz of PCS adjacent spectrum they aren't using right now that they're targeting for their initial LTE rollout. As customers start getting LTE equipped devices, more spectrum can be allocated to LTE use.
Two years ago, I was a Sprint customer. I left with the Nexus One.
At the time, in CDMA land, because Sprint had a roaming agreement with Verizon, and Sprint did not bill me for roaming on Verizon, my (voice) network was Sprint + Verizon. Which was actually pretty darn great.
Is any of that possible or likely to happen with LTE?
Where Sprint or any carrier, should it sign the right papers, can allow customers to roam off their LTE network onto another one (and the phones will work just fine?)
T-Mobile has a tiny bit more spectrum than Sprint and it's all standard (PCS/AWS) spectrum. T-Mobile has a decently standard GSM/UMTS stack. Sprint is running two non-integrated networks.
On a per-customer basis, T-Mobile isn't so bad on spectrum. 33.7M customers with 50MHz of spectrum. That's actually a better ratio than any other carrier. For comparison
* The Verizon numbers exclude substantial purchases recently announced which will probably increase Verizon's spectrum depth by nearly a third.So, T-Mobile isn't in a terrible spectrum position and the spectrum AT&T has to give them as part of the break-up will only help that. Likewise, $3B is a lot of money that could go toward an LTE rollout.
In terms of LTE, Sprint didn't have an LTE plan until mid-October of this year and doesn't expect to have any LTE devices until the third quarter of 2012. If T-Mobile put together an LTE plan today, they wouldn't be behind the times. I mean, they'd be behind Verizon and AT&T, but they could roll out as fast as Sprint (which just raised $4B for its LTE rollout - barely more than the $3B in cash that T-Mobile is getting as a breakup fee). Plus, LTE in America is a bit ahead of the curve. Mostly that's because carriers in other countries can stretch HSPA+ (as T-Mobile is doing) to provide much higher speeds than 3G.*
So, T-Mobile does have the ability to move forward if DT wants to. It has the spectrum, it has $3B in new cash, and it's profitable (which Sprint isn't). I think DT wants to see T-Mobile a bit more successful if it is to stay on in the US, but it is profitable. If T-Mobile can forge a joint venture with Dish Networks (which is interested) it would gain 40MHz of unused spectrum from Dish and might have a partner who would promote the service to its customers. We'll see. I'm definitely not saying you're wrong. T-Mobile needs a bit of energy and momentum to start gaining customers again. However, I think T-Mobile has the assets needed to keep going.
--
The one area where I think you're spot on is in terms of competition with Verizon. When AT&T Wireless and Cingular merged, it increased competition. Verizon had been allowed to merge together greater assets than its competitors previously. Allowing AT&T Wireless and Cingular to merge meant that there was a real, nationwide competitor to Verizon with low-frequency spectrum that provided good coverage. Right now, AT&T and Verizon have decently similar assets (although the Alltel merger gave Verizon a slight edge). AT&T has a greater number of data customers which puts a bit of a strain on their network, but they should be able to manage it now-a-days (and the data shows that AT&T has a better dropped call/blocked call rate than Sprint, although slightly worse than Verizon's, and significantly faster data speeds). If the FCC allows Verizon to gobble up the licenses from Cox, Comcast, Time Warner, Leap, and others, it will again create an imbalance that will see Verizon having superior assets that make it hard for other carriers.
* Before someone says, "but HSPA+ is offering 3-6Mbps while LTE is offering 20Mbps" I want to explain a bit. HSPA+ is backward compatible with HSPA and UMTS. So, when AT&T upgrades its UMTS network to HSPA+, there are tens of millions of voice and data customers on it. By contrast, the LTE networks we've seen have only had a couple million customers on them. As more customers use it, the spectrum will get more crowded and speeds will likely drop as a result. Likewise, there are many different HSPA+ versions. AT&T is running HSPA+ at 21Mbps while T-Mobile has it up to 42Mbps in some areas. However, even HSPA+ 14.4 has been showing average real-world speeds on the iPhone 4S of 3.2Mbps (http://www.macrumors.com/2011/11/18/nationwide-study-shows-a...). That's on a loaded network using 5MHz channels. The LTE speeds we're seeing are on a near-empty network using 10MHz channels (from launch to the end of the third quarter, Verizon has sold less than 3M LTE devices). None of this is to say that HSPA+ is superior, just that we're seeing LTE showing better numbers than it would under the same load HSPA+ networks are experiencing and that means the gap is actually smaller and explains how overseas carriers are using HSPA+ in the interim to offer customers a better experience without the cost of LTE (and it is a cost to the consumer as well as one can see from the price of HSPA+ devices against the cost of LTE devices).