Yes, it could. And bear with me, please: I'm about to attempt some FiOS math.
The company today said it plans to pass 18 million premises with its fiber network by the end of 2010. It also says it expects to invest $18 billion in net capital from 2004 through 2010 in deploying that fiber-fed network. (See Verizon to Pump $18B Into FiOS by 2010.)
I’m on my second bourbon and coffee, but it seems like Verizon has just copped to spending about $1,000 a home, on average, to pass 18 million homes, over a six-year stretch. Earlier, Verizon said it cost around $900 to pass a home, so let's go with that.
Verizon also says it will cost only about $650 to connect a "passed" home to its network by 2010.
So what does it cost to hook up a neighborhood? These aren't absolute figures, mind you, but let's assume that Verizon passes each home in a 400-home neighborhood, then nabs 10 percent of the homes (40 homes) as customers.
Take $900 and multiply it by 400 homes. That's $360,000.
Now let's hook up those 40 homes. That's 40 multiplied by $650. That's $26,000 added back to the cost to pass the homes, which was $360,000.
So now we have a figure of $386,000 spent in just one neighborhood. But what has Verizon spent per customer? Take $386,000 and divide it by the 40 homes and you get $9,650.
Millennium Marketing principal Kermit Ross worked out some very similar figures for me on a notebook after a session at Optical Expo 2006 last week. I recall his numbers/assumptions were also in the ballpark of $9,000 per subscriber.
Okay, now I'm on my third bourbon and coffee. But I think the larger point here is that even with all the cost savings Verizon has managed to achieve, this stuff is still really expensive.
"Obviously, in the early stages of a network, the cost to connect each home is astronomical and there's really nothing you can do about that," says Graham Finnie, an analyst at Heavy Reading.
So slice it anyway you like -- fiber to the home is damned expensive. And the payback takes years, maybe decades. But without a next-generation access network, carriers simply won't have a business.
Are you sure that your wireless if good to go? When I had FiOS installed, I was getting less than half of the performance I paid for 20Mbps down/5 Mbps up. The installer, a nice guy, couldn't get it fixed. One of the local reps wanted me to "tune" my TCP stack (a process that hasn't been necessary since Windows 98).
I finally fired up a wireless analyzer and determine that there were a few other wireless devices in the 'hood competing for the same channel. I reconfigured my Actiontec for only 802.11G and fixed the channel to a non-overlapping one. There is a decent chart at http://www.unixwiz.net/techtips/wireless-guide.html.
Those two changes doubled my performance to an acceptable limit. To be honest, if I take another wireless hit, I am giving up and running Cat 5 everywhere.
I can't imagine the problem FiOS customers are going to have in higher density neighborhoods where there will be more compettion for spectrum. Verizon need to spend some money training it's field staff in 802.11 wireless network and stop turing to DSL Reports forums for tech advice.
The heck with the dollars and cents, as a consumer i just switched over to FiOS and the service stinks. My Verizin DSL was more dependable and faster. Verizon's attempt at fixing is talking to someone who can barely speak english wanting to go though a checklist ... I've already checked the connections dude and the wireless LAN in my home just fine with DSL, so give me back my dependable DSL !
I listened to the webcast, though the part with the Fios demo was obviously useless...
Doreen explained that the STB was totally separate from the cost-to-connect numbers because it is a separate line item (ie: it is leased by the customer on a monthly basis). See slide 37 where it lists equipment pricing.
A couple of extra notes on the slide package:
- Slide 40 basically says that Phil's $9k is actually $2535 on average through 2010. This assumes a 39% take rate, 18M houses passed (which apparently takes $4.4B out of their copper-based capex in that interval), $500M in support cost reductions, and that they meet their cost reduction targets for cost-to-pass & cost-to-connect.
- It seems that they are assuming that anyone who connects, whether they sign up for video or not, are paying for the video head-ends, etc.
- The 25k CO over 5 years case seems a bit inconsistant though... They assume the cost to connect is $733 yet they also say that 70% of the customers come on in the first 3 years. The point is that the cost to connect is $1200 in 2004, $1163 in 2005, hopefully $880 by YE 2006, and maybe $650 by 2010. It appears that the cost numbers are backend weighted but the customer numbers are frontend weighted...?
If they can meet the 39% "success" rate by 2010 (assuming 18M houses passed. If they don't pass this many the average costs/house change) they may do OK. Up until those conditions occur I'd say their stock will be hanging by a thread.
Regarding the STB CAPex, found the answer in the transcript at: http://investor.verizon.com/news/20060927/20060927_transcript.pdf . I quote:
Unidentified Audience Member Can you give us a general sense of the cost per home for just the set-top box piece of the equation, what we should add to that $2,500? Doreen Toben - Verizon - EVP, CFO I think I would just tell you industry normal, what you think a normal set-top box is.
From rjmcmahon >Costs didn't include set top boxes. Customers >will demand them to be included. TIVO showed us >there is little consumer demand for customer >owned STB equipment, even if it's novel.
But then VZ's FAQ page for FIOS says the STB is leased by the customer. See http://www22.verizon.com/content/fiostv/customer+support/faqs/faqs.htm#q10 . In which case, why doesn't VZ include the STB in their costs? Are they maybe doing something creative, like a lease arrangement through another party so that it's off of VZ's books?
re: "Real humans don't care if their email MySpace access and porn are delivered by magical elves riding flying squirrels, they just care about it being fast, affordable and relatively reliable."
Everyone please rate that last message highly so I can use the above quote in next week's newsletter. Seems like that should be crocheted on a pillow or something.
These numbers seem both realistic and surprisingly pessimistic. Why would a company like Verizon pay $9000 per home served? It's not rational. The OpEx savings, if real, just aren't worth more than a small fraction of that. But then there may be more going on.
FiOS rollouts began shortly after Mike Powell granted "relief" for unbundling. To hell with the black letter law of 47CFR Sections 251 and 271; Powell decided that unbundling obligations are basically limited to facilities that were in place the day the Telecom Act passed. That's not what the law says but who cares? But Powell (and Martin) needed a quid pro quo. So in exchange for allowing them to cut off competitive LEC access to fiber-served homes, and competitive DSL access to lots more homes, VZ had to pull some FTTH. This is classic "industrial policy". "Fiber" is good, so whatever creates it is good. Forget why fiber was supposed to be good, but get it even if it's done badly.
Remember that back in 1993, the ILECs, especially the Bell Atlantic ones, promised to roll out FTTH over the following decade, on an open access basis, 45 Mbps, in exchange for a removal of rate of return regulation. So they got their relief, and have raised their typical returns above 30% (vs. around 12% under regulation). Yet they didn't pull the fiber. See http://www.teletruth.org/ for more details. So now they're repeating the trick, promising fiber ine exchange for killing their competition. They only have to do it long enough to be credible; they can slow down drastically next year, blaming economic conditions, investors, whatever. Martin will buy absolutely anything they say.
The key to this type of business is take rate, of course. They can theoretically migrate their POTS business to glass, but why bother? They could also gradually replace copper with glass, a sort of routine upgrade, as we sort of envisioned in the 1980s would happen in the 1990s. But DSL saved copper's butt. FiOS also fills in some holes where DSL doesn't reach, but there are cheaper ways to do that (e.g., Pronto, Lightspeed). How much incremental revenue will they get? In some areas they're overbuilding where RCN has already overbuilt cable. As a business, this doesn't make a lot of sense.
kbg_lem posts >Anybody knows what it really means to pass a home?
I think you're mostly on target. Chart 38 from VZ's deck http://investor.verizon.com/news/20060927/20060927.pdf seems to give the best detail.
It looks like they have the Network & Support (incl. video-head-end) and STB broken out separately from cost-to-pass and cost-to-connect. So I think CPE and install costs inside the home are not included in cost-to-connect.
But then their Gross CAPex calculation on chart 40 doesn't show STB. Which makes me wonder whether STB (and related costs) is indeed buried in the cost-to-connect numbers.
I'm not sure about your neighborhood, but here in "the Reddest of Red States" the take rate is probably about 20%. I'm not sure about the logic behind people signing up for broadband without TV. It would cost me more and if the FIOS quality is better than Comcast (which it likely is) I would switch on principle. My ILEC is SBC, the ultimate evildoer, therefore it behoves me to consolidate. That's the point of bundling. If all things are equal and the price is lower, and there is no switching costs, I'll go for it. Your not taking into account actual human behavior. Real humans don't care if their email MySpace access and porn are delivered by magical elves riding flying squirrels, they just care about it being fast, affordable and relatively reliable. Although there seems to be some elasticity in that last point.
As for the provider's CPU being cycled out. That stuff is ungoldy cheap. Right now you're looking at $200 per unit on the high side. Within a year or two that could drop below a c-note.
Is the 9K estimate valid though? You're assuming that the provider installs an OLT and populates it right off. If each blade supports Nx32 homes, they can control the offering by rolling it out in certain neighborhoods and increasing take rates through special offers. The OLT, ONU and splittera are cheap. The labor is what really represents the bulk of the cost. I my neigborhood you'll have to trench. There are no aerial drops. For greenfield apps this doesn't present a problem. You need a ton of prgram management to coordinate installs and builds. Turning up OLT's isn't easy. How do you test it if there are no ONU's hung. I'm not sure where folks get the idea that OPEX is reduced, I'd say the cost is just shifted in product cycle.
I'm not sure how FiOS addresses the union issue. It is still installed, owned and maintained by the Verizon Telecom organization, which means it remains in the union shop. So, maybe the fiber network needs fewer union people around to maintain it, so that helps a little. Except that it doesn't. The union isn't stupid. They see where this is going, and there is going to be one heck of a donnybrook when the Verizon union contract comes up for renegotiation. I think the current contract is up in early 2008, which means the ugliness starts happening in the middle of 2007. If your FiOS isn't installed by then, don't count on it being installed until mid-2008 when the whole mess gets sorted out.
The blogs and comments are the opinions only of the writers and do not reflect the views of Light Reading. They are no substitute for your own research and should not be relied upon for trading or any other purpose.
To save this item to your list of favorite Light Reading content so you can find it later in your Profile page, click the "Save It" button next to the item.
If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.
Digg
Del.icio.us
Reddit
