I will say, I'm not terribly fond of mergers on principle. However, based on my insight as a former customer and shareholder of the company, Sprint's goose was most definitely cooked to a crisp. If this merger had not happened, I could have seen Sprint file for bankruptcy, with Verizon and AT&T picking the carcass clean.
I think folks forget how dire Sprint's straits were at the time, and this specific merger truly was the least of all evils.
The regulators could have forced the big four carriers (now three) into a common carrier model.
In that setup, all cell providers would effectively be an MVNO, and the three (then, four) physical networks would be operated at arms length from the consumer facing side. Also, the four networks could be structured so that their financial incentives were to improve cell coverage and bandwidth instead of undermining each others' operations.
The best way to resolve these issues is never "force the bad guys to do good things" but rather destroy the walls which enable their rent seeking.
Regulators can work to make it extremely simple for new techs in this space to come up so there is real competition for ATT etc.
We the People should own infrastructure. Want to start a new shipping company to compete with UPS? You don't have to build roads, We the People provide those to you. So, do the same with wireless.
Nothing prevents you from working with liked minded individuals to create your own "infra". The phrase "we the people" sounds cool but in reality what you mean is "your money, my idea".
Roads is a very common argument, but none of the roads are build by "we the people". Government takes your money by force irrespective of how you think it should be spent and then a completely unaccountable red tapy system that employs otherwise unemployable people decides how to spend it. After a massive waste you have some roads which are poorly built even worse maintained.
There is no need for roads to be public infrastructure. It can be fully privatized and people be asked to pay for its use. (While entire compontent of taxes that go towards road building be returned back to the people.)
We will have better roads, less traffic and more money in pocket with that model.
As evidenced by the many successful and popular real-world cases where it's played out exactly this way. For example, um, uh...
The "accountability" is people voting in/out politicians. I agree they need more accounability, but in this case the folly falls on the ignorant for not properly researching who will lead and manage their towns.
Sounds like an awful idea. Land isn't infinite, so it falls into the same problem we have with housing. Lots of inefficiently crafted roads and a nightmare of tolls to manage based on the route you take. Which finals down to a few big boys either battling for the best roads, or worse, colluding with each other to keep tolls high. Navigation now has to incorporate not only for physical distance, but approximate cost as well.
I fail to see an upside here unless your underlying narrative is to force people onto public transportation. Even then that doesn't mean public transportation imroves for those with bad access (also managed by "unaccountable red tapy system")/
Oh good grief.
I call bullshit.
Going 4 to 3 sucked, let’s go all the way to a state monopoly!
It didn't suck for the shareholders of the remaining 3. With a state monopoly, we'd all be shareholders. Win!
Right, because we only have one single shipping company using all the roads...
This is the biggest solution to the problem that many people seem to ignore. If you want your communities to prosper, then the public needs to own the infrastructure that companies provide services for. Full stop.
Anything less is giving companies control over something they should not have control over.
I think there's an argument to be had on 'owned by the public, but operated by private industry.'
Another on 'built by private industry, but owned by the public.'
Government ownership isn't a panacea and has historically faultered when faced with innovative and expert requirements.
But I do think anything that trends towards monopoly makes sense as 'let the public own the simplest level, exposed via standards, and innovation happen above and/or below that.'
That's a terrible idea. If the government owned wireless infrastructure then we'd still be stuck with 1G analog service. Competition between private cellular carriers is the primary factor driving innovation.
There is no "We the People". There is a government, or really layers of government, composed largely of politicians and bureaucrats. Whether those people running the cellular network is good or not ought to be assessed on what actually exists or is reasonably possible, not appeals to vague abstractions.
And yet even with all that infrastructure, the market has coalesced around 3-5 shipping companies for vast majority of consumer shipments.
The walls here are defined by physics. There's only so much spectrum to go around.
You can't have more than 3-4 large cell phone operators working in the same area. Decoupling the radio part and forcing everyone to play as MVNOs is a way to work around this.
That sort of decoupling could be fascinating. IMO, the walls may be as much a function of economics as physics.
Running a small number of powerful radios on low-frequency spectrum is a cost-effective way to cover large areas. Physics constrains how much of that can happen in the same area. But if networks are willing spend cash, they can get a ton of capacity by densely deploying high-frequency radios. Don't think we're anywhere near hitting the walls on what's possible there.
Putting this another way, networks provide enough bandwidth for everyone in Manhattan. It's just expensive.
It won't help much. Radio waves start getting too much attenuation to be useful outside the LoS around 5GHz.
That's still not a lot of spectrum. And you need to share it with other consumers (WiFi).
Then you start getting into the practicalities. You'll need to spam EVERYTHING with your access points. Can you imagine dozens of different wireless providers installing access points on every floor of every building? It's just not going to happen.
Or as I once heard by way of explanation: the capacity of an early analog cellular ~30km AMPS cell was ~60 simultaneous calls.
Modern protocols do magic things with spectrum efficiency, but there's only so much you can do.
The problem is that there is finite spectrum available. The upfront cost of building a national cellular network are also astronomical, regardless of regulatory hurdles.
The FCC was created to manage the scarcity inherent to the radio spectrum. It's not an area where regular free market economics apply. GP's proposal would actually make it easier for startups to horn in on the territory of the big established players, since the underlying infrastructure would effectively be socialized.
Out of curiosity: What are some examples of "regular free markets"?
There are finite amounts of everything. Free market economics solve this problem all the time. There are finite amounts of food, land, energy, shelter, human resources, etc. The problem isn't that spectrum is finite, it's that there is no way to prevent interference.
It's questionable if the FCC is really optimal though. There are huge amounts of spectrum still devoted to dying industries like AM/FM radio and broadcast television.
Regulators cannot magic spectrum into existence.
In USA the CBRS band allows running real LTE or better, real 5G SA on very lightly licensed spectrum.
All kinds of more experimental community networks make use of it. Even better, Magma core is a free 5G core and multiple vendors like Baicells make cheap 5G CBRS base station hardware. CBRS also fully overlaps international 3.5GHz 5G bands, so phones and modems are cheap.
I mean Sprint was strictly CDMA, so that would be an issue on it's own.
How so? Sprint operated 4G LTE, which is a GSM technology (or alternatively, unified the two, depending on how you look at it).
I once got tossed out of a sprint store for pissing off a salesperson with this factoid.
This is a story that's just begging for more detail
Meh, I was a smart-alec kid there with my friend. The salesperson had to go get a sim card "for the LTE to work" and I said "oh right, because LTE is GSM and requires a sim". The salesman insisted Sprint didn't use GSM so I looked up the wikipedia page for LTE on one of their demo phones and started reading out loud "In telecommunications, long-term evolution (LTE) is a standard for wireless broadband communication for mobile devices and data terminals, based on the GSM/EDGE and UMTS/HSPA standards."
That's when the salesman told me to leave.
The reality is, when a CDMA carrier deploys LTE without deploying UMTS, there are usually compatibility layers between CDMA and LTE such as CSFB and eHRPD for when VoLTE is absent. CDMA was never designed to interoperate with LTE as LTE was built around IMEIs and SIM cards but CDMA was built around burned-in ESNs and PRLs, so LTE support was hacked on.
This is why Sprint and Verizon used whitelists: they literally couldn't accept GSM-only devices because you wouldn't be able to make a phone call (the phone would try UMTS but only the non-supported CDMA2000 can be used to actually call, so in turn no phone calls).
Sprint went further by using the CDMA provisioning system on top of LTE instead of just using SIM cards and 3GPP provisioning like most GSM and CDMA carriers. This was a nightmare for custom ROM users like me as custom ROMs were designed for GSM carriers in mind and Sprint was at best an afterthought.
Sprint also had Wimax https://www.rvmobileinternet.com/farewell-wimax-sprints-orig...
Sprint had started as CDMA. They were in the process of building a 5G network but lacked the capital necessary for competitive nationwide coverage.
They didnt start to build their LTE network until 2012-13.
CDMA, from an RF performance perspective regularly outperformed GSM, but as is noted downthread, the provisioning/auth system was inherited from AMPS/D-AMPS, CDMA2000 with a GSM network core would have been amazing.
Which, tbh, is exactly what LTE is.
That's kinda orthogonal to the merger deal, though, no? The government could do that today, if there was the political will to do so. Of course there isn't, though, and the merger (or lack thereof) didn't change that.
The reason their goose was cooked is because they previously were planning to acquire T Mobile, but SoftBank got back-channel info that it would never be approved by the anti-trust regulators. At the time they had Marcelo Claure running Sprint, basically a corporate "fixer" guy for SoftBank. So he ran the company into the dirt in order to make the merge feasible to regulators (e.g. - Sprint purchased a 33% stake in Tidal, the music streaming service. Or how they entered a partnership with bankrupt RadioShack after it got scooped up by PE, and decided it was a good idea to take over all the physical RadioShack locations and turn them into Sprint stores).
"Oh, whatever shall we do, our company is failing, you MUST let us merge with one of our primary competitors or we'll go bankrupt. No company at all is worse for consumers than a merged company."
Sprint went all-in on WiMax as it's 4G network. I owned an early Sprint 4G smartphone (made by HTC) and the 4G never worked. I would go to the Sprint store and ask to show me a signal with 4G turned on, and they would always blame congestion, or weather, or some other made-up excuse.
Was it more about the device than the network? I used mobile WiMax from Clear for years until it ended its service. For what it was, it worked great. It wasn't necessarily a speed demon, but it was reliable. I used it to avoid public WiFi congestion, as a backup when my home network was down, and in 'bring your own infrastructure' situations.
That's possible but it was a Sprint-branded device bought from Sprint that was clearly supposed to be compatible with its new 4G network. And they charged an extra $10 4G access fee every month!
The other bonus was that being a CDMA device, there was no connectivity when traveling internationally, and no option to buy a local SIM card. Wifi only.
Why does that exclude international use?
Because just about every other country is on a GSM network, not CDMA.
Yeah it depends on the market you were in. I helped build the WiMax network but it was built very quickly and in places it was built by people who didn't care very much.
It was all microwave back hauled so rain fade in stormy weather was absolutely a thing. Most of those were FCC licensed or should have been but I know of at least 1 market where they just never filed the paperwork to get the licenses and built it anyway.
The clearwire side of things was wildly oversubscribed on backhaul, often totally saturated 10m circuits.
I had a clear hotspot puck in the DMV area, worked well most of the time. I don't think I had any issues with price paid vs performance given the current state of tech at the time. I think targeting homes was just a hard battle, FIOS and others were really ramping up their initial push into fiber and it was just not going to compete sadly.
in all fairness, the 5G isn't much better these days, at least in my area. T-mobile seem to have inhereted Sprints bad coverage, to the point where they had to send me a booster modem for my own home.
5g is higher frequency, lower wavelength. That means it gets better potential speeds when it does work, but it's going to have more difficulty penetrating barriers, reaching longer distances, do worse in inclement weather, and so on. This is also a fundamental limitation of trying to reach higher speeds with this form of tech, because frequency and wavelength are inversely proportional. The point of this being that 5g is inherently less reliable/robust than 4g, and so it's not too surprising that you'd need a booster, especially if 4g in an area is not the hottest.
At this point telcos are also deploying 5G to old bands, so the assumption that 5G is 3.5GHz and mmW only no longer holds.
In my part of the world I see 5G on 1800MHz and 2100MHz bands in addition to the 3.5GHz one.
Unfortunately that's not done near my home, so I get 1-2bars of 5G, instead of the full bars I see at the office and elsewhere in town.
I think the only place I knew where WiMAX actually worked properly was in Japan.
I don't think that materially really had anything to do with it.
Sprint was dying - with extraordinarily high debt, in 2007, well before Softbank bought them, and indeed they lost money every year from 2008 forward - https://www.statista.com/statistics/481739/sprint-corporatio...
The Merger with Nextel managed to kill what was great about Nextel, and what was good about Sprint, and they lost customers in droves (mostly former Nextel ones). In reality Sprint bought Nextel's OAM equipment and their customers, and moved all the legacy Sprint customers onto the surviving billing and network management platforms (Nextel). The iDEN turndown also lost even more customers, most of whom who realized they didnt need PTToC after all (which is too bad, because on dedicated CDMA hardware, it worked really great).
Then they needed to start rolling out LTE (Network Vision) - and NV didnt start in earnest until 2012/13 - and as someone who was on the field end of it, was very very very poorly managed. Sprint some years prior had outsourced all their engineering expertise to Ericsson, which means they had no one in house with any knowledge. They only realized that 18 months in, and then scrambled to get people back from Ericsson (who I will note, they did not contract any of the deployment management to).
I only know this because I was in the middle of the deployment as a field resource in Seattle.
My guess is only half the sites in the network (in Seattle Market) had enough customers to pay their fixed costs.
I concluded while I worked there that there was no way for four carriers to be viable, there isnt enough spectrum allocated, and you pay the same fixed costs over and over again.
I'll go further, Sprint had a massive switch facility for the LD operations with room for like 4 DMS250's in Tacoma, but that's not where they put the SPCS 5ESS, that was in Kirkland in a rented building (and interestingly enough, it's still part of T-Mobiles operations today), there was also another Motorola iDEN switch also in Kirkland.
Post merger they never really made any effort to reduce their fixed costs (sites, switching centers, et al), because that would have cost money - they also got bled dry by having to foot the entire bill for rebanding the SMR band, which was on the order of 2.5 billion dollars. They did close stores (and RS was a major outlet for Sprint Sales, before it went belly up) which contributed to problems later.
So I don't know where you got your info, but I think its hooey - before Softbank bought Sprint, they didn't have the capital to upgrade their 2G/3G network to LTE, much less consider a merger with T-Mobile.
and then there was also the Wimax debacle. Sprint had invested pretty heavily in Wimax being their future network before they realized this wasnt going to work.
I got bit by the wimax nonsense as a consumer. I had a 4G HTC flagship phone on sprint and it was awesome. Then when I went and upgraded my phone a few years later to another HTC flagship phone, also with 4G, I was very confused why I could only use 3G.
The first phone was labeled 4G and second one was 4G LTE, which sprint didn't have in my area. I had been using wimax.
I switched to Verizon not too long after that so I could have 4G again
Wimax actually did work great in practical terms even if the clearwire network was made out of compressed spit.
All of them - all the CW sites, were under provisioned for backhaul.
Yeah network vision was a disaster. I think in the market I helped built our average number of site touches was like 21 or 22 from construction complete to on-air. All the funds gathered from the high interest junk bond sale to fund NV was used very inefficiency because sprint lacked the expertise to not get taken for a ride by their vendors.
It was a shitshow, you'd have GD tell us a site was complete and show up and find no equipment or equipment on pallets. GD rather than the equipment vendors were the villains, it was subcontractor-o-rama, and no one was responsible for anything. There was one site I went to like six times to inventory non-existent equipment - never mind when they decided to preload the inventory but used the barcodes on the samsung gear that didnt match anything.. I escalated that to Ericsson and then Sprint directly (it mattered for ericsson, because they'd have failed ATLAS audits), about two weeks later all that vanished, and we got a tranche of tickets to go reinventory everything.
I also saw at least half a dozen sites, cut and in service still sitting on their pallets, such a mess.
Imagine for a moment you go off to the crusades, leave the barn to one group, the fields to another group, and leave the house to a third group, then gleave no one in charge and give them no way to communicate beyond the most informal means - and you're surprised everything is on fire upon your return?
In the end it did work once it was done, but it was only hell for the customer during deployment because it was deployed and cut in a hopscotch fashion, which would be fine, if you could roam from new back to old - there was a one way roam, from old to new, once on new, there was no path to roam back.
I think both of us can be right at the same time though. Just because they had problems before the SoftBank acquisition as well doesn't necessarily make what I'm saying unreasonable. There was still sentiment in 2013 when SB closed the deal that regulators would not have approved of Sprint acquiring T Mobile [0], despite the struggles going on at Sprint at the time (that you describe). Sprint was definitely putting together a bid to acquire T Mobile, WSJ reported on it [1].
As you yourself said, Sprint was _dying_ at the time of the SB acquisition, but as far as large firms go, they were far from bankrupt yet. SoftBank simply twisted the dagger and then presented the corpse to congress instead of the dying patient.
[0] https://www.theverge.com/2014/2/4/5376824/fcc-chief-reported...
[1] https://www.wsj.com/articles/SB10001424052702303293604579256...
I remember saying at one point to my operations manager at Ericsson, "how does anyone make money in this business?" he laughed and said "I have no idea".
I cannot explain how poorly managed Sprint was, it'd take me an essay to just explain the various dysfunctions I saw there.
That said, it did improve some once Softbank bought them.
I'm a former engineer at Sprint and I strongly disagree with this characterization. Sprint's goose was cooked but it was due to debt from selling junk bonds to build Network Vision at the time of the original LTE rollout. Their credit was ruined by that point from 30+ years of absolutely terrible and corrupt c-suite executives.
Marcello has a lot of faults but he didn't run Sprint into the ground. He is actually pretty smart and at that time we cut over a billion dollars out of the operating budget circa 2016/2017 iirc. It was an impossible position and it's really sad because it was a great old company in my estimation. T-Mobile is just the worst.
Yeah, I always thought the two primary factors here were the lost bet on WiMax (which probably cost a lot to build infrastructure for) and the Nextel aqui-merger causing a lot more friction than synergy.
I'd be a horrible businessman, because I really can't imagine keeping so much debt and simply being okay with it until its too late. I'd be considered a fool if I managed my personal finances like that, but that's the normal operation when managing millions or especially billions at a time.
WiMAX was the answer for build 4G for cheap, and also how to use a bunch of spectrum that was not conventionally useful.
It did work quite well in practice, I can assure you.
Why is T-Mobile the worst?
He finally reduced the fixed costs that should have been done post Nextel merger.
I remember Nextel and iDEN sites co-sited on the adjacent towers.. but with different shelters (sprint was often outdoor cabinets).
How is ruining and destroying the value of Sprint, that SoftBank owns, possibly good for SoftBank?
What is SoftBank to gain here from enriching TMobile?
I don't know if the OP post is true-- or partly true-- but to explain how it might be good for SoftBank: This was an all share deal, so sandbagging the value of Sprint shares so you can merge with TMobile could be good in the long run because you own shares in the new entity which has much less competition.
I'm not sure if SoftBank deserves the credit of assuming that their actions are based on sound logical reasoning.
They want to spend whatever billions it costs to consolidate the industry and then reap the monopolized profits down the road.
Radio Shack was a major retail channel for Sprint since the 90s. By the mid-00s, I would wager most of Radio Shack's gross profit came from wireless retail. They were already more or less Sprint stores with some overpriced PCs and stereos by then.
Why would it be a prudent move to take on the corporate real estate costs associated with RS if it was already a profitable retail channel? What does that change other than increasing overhead for Sprint? And if RS was working, why would converting them to solely cell phone stores make things any better?
Because the choice presumably wasn't {Radio Shack business-as-usual} vs {acquire Radio Shack stores}, but rather {Radio Shack disappears as a retail entity and channel for Sprint} vs {acquire Radio Shack stores}.
Let's not forget the amazing Palm Pre that was released 18 months too late. If it could have been released on time, it might have done much to save Sprint. But by the time it was released it was merely competitive instead of compelling.
Tell me about it. I bought my Palm Pre shortly after launch and, to this day, I can't think of another smartphone experience on the software end that was as good as webOS, save for Sailfish OS perhaps. If only Palm had been able to get their device on Verizon earlier, and perhaps started off with what the Pre 2 hardware was like.
What makes you think someone else wouldn’t have picked up the pieces? Whether that be dish/comcast/google, or just some private equity.
Being acquired by a competitor wasn’t their only exit option, and pretty much anything else would’ve resulted in more competition.
In an infinite universe of possibilities, you're right: a failing brand with a household name being acquired by a competitor is not the only possible outcome.
But it is historically far and away the most likely one.
Isn't that circular reasoning? Regulators have to allow the merger because other outcomes are unlikely, but they are unlikely because regulators always allow the merger. If the merger is blocked, the probability of that outcome falls to 0, and the others' increase, no?
You're assuming anybody else would have wanted to acquire a debt-laden company with an enormous infrastructure deficit as they mismanaged the move to 4/5g. Sprint essentially was only valuable to a company with a better LTE/5G network already in existence that users could me moved to and then Sprint's old spectrum repurposed. Anyone else would have been left with many tens of billions in network upgrade costs.
If the merger was blocked and no one else wanted to buy it whole, then their assets would get sold off and the proceeds divied up by creditors (and if anything was left after that, to shareholders), same as any other company. Why is this an unthinkable scenario? There would have been some market-clearing price for the Sprint brand name and CDMA network (possibly separately, or even the network itself parted out) and even if that price was zero, then it was the shareholders (and possibly creditors) who should have taken the bath, not the entire phone-using public.
Allowing an anti-competitive merger simply because the alternatives for Sprint shareholders were bad is a bailout by any other name.
The sprint shareholders already took a bath going from over $80/share in 2000 to $5. Hard bankruptcy and selling off its network would have been disruptive to their remaining customers, more complex/expensive to unwind (causing more of the money to go lawyers), and would still result in 1 less major carrier, resulting in essentially the same thing. The results would have most likely been most Sprint customers eventually being part of the other carriers anyways. If there was a market clearing price for sprint, a private equity firm would have snapped it up - that didn't happen because there was too much debt to go on as its own entity.
Could things have been done differently? Sure. A condition of the merger could have been guaranteeing MVNO access or selling off a portion of the spectrum (maybe that happened).
But T-Mobile, combined with Sprint, went from being a distant competitor in subscriber numbers competitive. If sprint was "sold off" separately, T-mobile would most likely eventually run out of steam and end up like Sprint. They just wouldn't have the number of subscribers to amortize costs down the way AT&T and Verizon could.
Keeping struggling, small players going somehow would likely of only delayed the inevitable.
I have to say I don't actually care or know very much about the merger, but the fact remains that there were other possible outcomes than the merger going through as it did. If regulators had blocked the merger, one of those would have happened instead. AFAICT there is no way to refute this argument. Possible futures at that point in time were are F_1 through F_n. If we had eliminated F_1, we would instead have F_2 through F_n. Maybe F_2 through F_n all entail a complete collapse of the US and world economy, or maybe they would have better outcomes. I don't know, but one of them WOULD have happened.
Yes, obviously "something" else would have happened had regulators blocked the merger - I don't know why you're hyper-focusing on that. My point is that it would almost certainly have been an even worse outcome for everybody involved.
Unless you can show me how a cellular company that and backed itself into a technological dead end (and built up enormous debt during said process that sapped most of the money it earned to service) and needed quite possibly multi-tens of billions to even catch up to the others, you're not really adding to any conversation. If it was possible, business-people with more money and brains than either of us would have leapt at the opportunity.
The only alternative I can possibly think of is taxpayers ponying up - but that would in the end mean that consumers would "pay" indirectly, to say nothing of the moral hazard costs of government bailing out private businesses (works so well in the financial industry, doesn't it?).
Sorry, I thought we were just talking about what we were taking about. Have a good day.
Sprint had outdated technology w.r.t their infrastructure and cell towers, so an outside acquirer didn’t make as much sense. It needed an existing player with more advanced infrastructure to take on the customer base, or a large investment in upgrading it’s own infrastructure.
Pretty much this. Sprint bet the farm several times on ultimately dead-end tech (CDMA, WiMAX) and were too focused on other things to get ahead of technical evolution. They ended up saddled with debt by the end of it, which hampered any ability to actually upgrade their networks (another poster already pointed out the technical issues with running LTE without UMTS). By the time of the merger, the only things of any value they had was the spectrum and customer base to be moved over.
An external acquirer would almost have to build from scratch after absorbing all the legacy costs and run the risk of inheriting Sprint's bad business decision culture or spending an enormous sum building out a new team. Of course, they'd still have to support the old and new setups at the same time for awhile as the new stuff was built out. The ROI would have been decades at best.
With a merger with another telecom, people can be migrated over already existing infrastructure (with some upgrades to deal with new traffic) and have sprint's old spectrum slowly merged into the existing infra.
I'm oversimplifying of course, but one gets the idea.
By the time they merged with T-Mobile, they had a fully modern LTE network - I know because I helped to deploy it.
That never would have happened without Softbank buying them however.
To be honest, I truly do not know all possible ends, as I'm not an oracle. Based on the information I had at the time, I felt that this merger was the best option out of all of the ones being explored at the time. I'm happy to have been wrong about this if they went a different direction and it worked out better, but history is history.
A Sprint bankruptcy may have been inevitable, but Verizon and AT&T would have been forced to steer clear (a failed merger having made that inevitable).
The private market would have provided a bounty of suitors for Sprint if it couldn't recover from bankruptcy. It may have emerged in a far weaker fourth place, but it would still be around.
The only other possible dark horse that could have emerged was either in US Cellular or Dish Network. However, I believe there were concerns with issues like market capitalization (in the case of US Cellular) or having different priorities (like Dish) that would have jeopardized a proper fourth option for the US.
I'm not suggesting the merger was "good" or anything like that. Just that the other options seemed quite unlikely.
Dish is already making a hash out of its mobile customers (Ting, etc)
What's wrong with Ting?
Have you been a customer lately?
Yes, and haven't had issues with it recently... Was there some bad news I missed?
The reddit forum for Ting has a lot of complaints about billing issues, lack of eSIM support, and poor customer service (for the TMobile/DISH customers, of which I was one)
I'm a current Ting customer for the last ~5 years and have had no issues with them.
There is already a burgeoning dark horse competitor to the wireless companies: SpaceX. I believe within a few years Starlink will be a viable global wireless network.
They are using local partners
For now. But this is where the tech is heading:
https://www.pcmag.com/news/att-ast-spacemobile-promise-true-...
https://www.starlink.com/business/direct-to-cell
As a Canadian (so possibly biased here), my own hypothesis for what would happen if a sufficiently-large long-tail power vacuum emerged in the US cellular data market — either back then or today — is that one of the major Canadian carriers would try to move in, beginning by serving cities just across the border from major Canadian cities.
You could easily do cell-tower-maintenance truck-rolls from offices in Vancouver BC to towers in Seattle or Portland; from Toronto to Buffalo (or, less plausibly, to Chicago); or from Montreal to Boston. And that's only if they even bother to operate towers — if they tried today, they could just as well operate as pure MVNOs.
In fact, flagship plans on Canadian carriers today, already usually build in no-cost full-speed US roaming data access through partnership with US carriers operating on the same frequency bands. It's a very short distance from there to operating an MVNO atop the same carrier's network.
(I would say that I'm surprised they haven't tried to do this already; but until recently, Canadian carriers were addicted to the extremely-high-profit-margin rate plans they built up through oligopolist price fixing. Our current government has seemingly broken that up for now, with much cheaper plans finally appearing — so they might finally decide it's time to expand their TAM to stay profitable.)
US Cellular is actually pretty probably, same tech stack, same hardware, similar customer bases, but they didnt have acres of capital. The issue was the albatross of the debt from the Nextel merger.
Were they legitimately in financial straights? or was it the thing where they want to force regulators to allow the deal by blowing up their own company?
For example, Albertsons is blatantly doing this so Kroger can acquire them.
Is that why they are absurdly expensive compared to all of their competitors?
They're paying out an excessive dividend that completely empties their cash reserves and makes operating the company unsustainable
https://www.opb.org/article/2022/11/12/oregon-ag-files-court...
Corporate bankruptcy is a scam. Any merger on auction should acquire existing debts, and execs should be a lien against wealth and future income.
What a ridiculous proposal. Eliminating corporate bankruptcy and making employees personally liable for business debts would wreck the US economy. If debtors take a haircut then it's their own fault for lending in the first place; no one is forced to buy corporate bonds and vendors always have the option of requiring cash on delivery.
Letting employees loot the company for their own profit isn't any better.
Shareholders and lenders already have the necessary legal tools to prevent employee looting, if they choose to use them.
Counter-argument, moneylenders shouldn't loan money to companies doing this. And secured lenders have their protections.
Anyways, they knew the game. Bankruptcy auctions are essentially debtors recouping as many of their costs as possible before writing off the rest (further minimizing future taxes).
I don’t have a real strong view on this specific situation one way or the other but it’s relevant that the appeals courts pretty quickly threw out this suit and allowed Albertsons to pay the dividend.
Yes, they consistently lost money from 2008-2019.
Sprint was cooked for one specific reason: parent company QWEST refused to do the prism-split of fiber cables into secret NSA colocation rooms like AT&T did.
Qwest never owned Sprint.
Ok young'n
In March 1998 Qwest announced it would acquire long-distance carrier LCI International Inc. for $4.4 billion. The deal created the fourth-largest long-distance carrier in the United States behind AT&T, MCI Worldcom, and Sprint Corp. The combined companies had about 5,800 employees and revenue of $2.3 billion. The acquisition gave Qwest 2 million long-distance customers and a well-established sales force.
Which has what to do with Sprint?
Oh nothing. Move along, nothing to see here.
Hmmm - interesting!
https://en.wikipedia.org/wiki/Joseph_Nacchio
EDIT: more info here: https://www.eff.org/deeplinks/2007/10/qwest-ceo-nsa-punished...
I don't follow this industry at all, so I'm out of the loop. Are you saying Sprint could not have possibly reversed course? (E.g. through pricing or even new leadership)
Sometimes when companies are trying to merge, executives from both sides will come out and say it's "necessary" and push the narrative that one or both will go bankrupt without the merger. In cases where that strategy doesn't work it, unsurprisingly, turns out to be a lie. The companies will just keep competing and figure out a way to operating as usual. (I remember reading not-too-long-ago about an example of this exact thing, but don't remember what it was)
So personally, I don't trust anything a business says when a merger is on the table.
if memory serves, Sprint had huge debts racked up from buying Nextel, as well as buying massive amounts of spectrum for WiMAX. They also purchased Clearwire. But WiMax never actually happened, everyone went to LTE, so sprint did as well.
The main reason they were attractive to T-Mobile is that large amount of spectrum they owned, which was very valuble for 5G.
https://www.rcrwireless.com/20160401/featured/worst-week-bol...
A huge costly mistake for Sprint trying to grow subscribers. Nextel's network was incompatible with Sprints, just a ton of business and technical issues. Sprint was always trying to be on the forefront of technology whether wireless or fiber. They spent a ton money on a project called ION that was last mile fiber to businesses, ahead of there time but a ton of costs actually laying down fiber doomed it. Poor timing for a lot of ideas and projects I saw. Source: ex-sprint emp.
Sprint could not have possibly reversed course because they had already wasted too much capital on a failed WiMax network and couldn't afford to build a competitive nationwide 5G network. Their options for raising more capital weren't looking good. New leadership or a different pricing model wouldn't have changed that reality.
To be frank, I thought the merger was necessary at the time as a lowly shareholder, only a few dozen shares at best, based on what I could tell on financial reports. I really wanted Sprint to not be a complete mess and be strong again, but they made way too many boneheaded decisions back in the day and the chickens were coming home to roost.
Dish network the supposed new 4th wireless carrier's goose is being cooked now and i bet will file for bankruptcy in the next year or so. Satellite pay TV is a dying to dead industry.
This situation seems extremely similar to the failed Spirit/Jetblue acquisition [1].
[1] https://www.bloomberg.com/news/articles/2024-03-04/jetblue-a...
Exactly. If you map out all the MNOs around the world in each and every country the trend is very clear where they merge to around 3 - 4 MNOs depending on population density. It is both economical and technical decision, I wouldn't even put it in the category of "least of all evils".
That's what's going to happen with US Cellular: https://arstechnica.com/tech-policy/2024/05/report-t-mobile-...
I believe you but I also think there was a third path available… wasn’t Dish going to buy them at some point, or something like that? Sprint were in dire straits but plenty of their core businesses seemed viable if placed in competent hands.
You're spot on, I posted a long comment downthread elsewhere, but you're 100% correct here.
Sprint lost money basically every year after they merged with Nextel (having to pay for rebanding was part of it).