Wednesday, June 29, 2011

NBNCo. Don't say you weren't warned.

Wednesday column

You know a business case is hopeless when the company that’s drawn it up has to bribe its competitors not to compete against it.

The NBNCo business case has a tenuous relationship with reality.

The publicly-released corporate plan detailing how the company will spend $27 billion of public money and hit its targets reads like a cry for help, and also an exercise in laying down a paper-trail so its executives can say “we told you so” when their targets are nowhere near met.

Optus cable passes 1.4 million homes in Sydney, Melbourne and Brisbane. Fast already, it is capable of being upgraded to a top speed of 240 megabits per second, far faster than anyone needs for conceivable uses; way faster than NBNCo’s entry level speed of 12 Mbps and on the way to the maximum NBNCo speed of 1000 Mbps.

Optus was in the process of upgrading its cable from 30 to 80 Mbps. Instead it will now “decommission” it in return for payments from NBNCo worth $800 million.

Telstra is already offering 100 Mbps on its Foxtel cables in Melbourne. As a condition of taking $9 billion from NBNCo over time it will “disconnect” its broadband customers from the cable although in its case it will continue to use it to deliver pay TV.

These ‘Do Not Compete’ agreements are not mere icing on the cake, enhancing a business model that already makes sense - they are necessary for the business model to make sense. NBNCo will never make a return on the cost of its capital or meet its customer targets if it faces competition. Its corporate plan says so, at point 1: “The plan assumes effective regulatory protection to prevent opportunistic cherry picking... the viability of the project is dependent upon this protection.”

Cherry picking means competing for good customers on price.

And cherry picking is impossible to prevent...

One of the reasons we don’t have a very fast train between Sydney and Melbourne is because as soon as the money was spent the airlines would discount the flights that competed with it to cream off its valuable customers.

In the case of the National Broadband Network as soon as it is built someone will come along and offer something cheaper to its most valuable city customers. The corporate plan tries to ensure against this, noting wistfully that “the government will consider the introduction of a levy, if necessary to prevent opportunistic cherry picking”.

The deal with Telstra requires Australia’s biggest supplier of wireless broadband to “not promote wireless services as a substitute for fibre based services for 20 years.”

Even if Telstra complied, and it has already indicated it won’t let the clause prevent it promoting wireless broadband, there’s nothing to stop a firm which hasn’t signed an agreement with NBNCo aggressively promoting a cheaper and more convenient product.

In the OECD fixed broadband connections are growing at 6 per cent per year; wireless connections are growing at 10 per cent per six months. South Korea, said to show the way for Australia because of its lightening-fast fibre to the home connections, has just 34 fibre or cable connections per 100 residents. It has 90 such wireless connections.

To some extent the new Korean figures are good for NBNCo - they show people buy wireless connections in addition to fixed lines. To a larger extent they are bad news. They suggest that even with fast fixed broadband on offer an awful lot of households forsake it for a still-fast, cheaper and more convenient alternative.

Research reported in NBNCo’s corporate plan finds that in Australia around 1 in 3 wireless broadband users have no fixed lines.

Acknowledging there is no real need for the very fast speeds it will be offering the plan says “the main limiting factor in the early years of the NBN will be the availability of applications that require high bandwidth”.

Further down the track it is expecting applications such as remote hosting and 3D imaging to become mainstream, pushing bandwidth demand towards 100 Mbps - which Telstra cable already provides. In the long-term it says ultra high definition video will require speeds of more than 250 Mbps - which would be encouraging had not Australian viewers already voted with their remotes and as good as closed down Australian HD TV broadcasts, preferring instead reruns of sixties and seventies sitcoms on low-definition channels such as GEM and 7Mate.

It’s a shaky foundation on which to build a business case.

Just as it will remain legal for anyone to offer a cheaper and lower-bandwith product competing with NBNCo, despite the NBNCo’s wishes, it will also be remain legal for large CBD-based organisations to take their high-bandwidth traffic “off-grid”. Why pay NBNCo a large sum to communicate internally when you can do it yourself for a fraction of the cost?

As the customers NBNCo is counting on are leached away it will have to continue to service what will eventually be $50 billion of capital and debt.

Its corporate plan says it won’t be to stay ahead servicing the debt. Its cost of capital will be 10 per cent, its internal rate of return will be 7.04 per cent.

And that’s assuming no corporate customers leach away to wireless, assuming the bleed to wireless-only households stops at 16.4 per cent, assuming no non-wireless competition, assuming Australians take up ultra high definition television in a way they have shown no inclination to, assuming it isn’t required by the government to deploy the superior single rather than split fibre technology the government will force it to test... in short, assuming every one of its judgement calls turns out right.

Including population growth. The NBNCo plan assumes housing growth of 177,000 premises per year to 2025. Downwardly-revised ABS projections released last week have just 151,000 households per year added over the coming two decades.

Maybe this will come right too. More likely, NBNCo will look sick. Those of us who have flicked through its corporate plan can’t say we weren’t warned.

Published in today's SMH and Age

Related Posts

. Overselling fibre. Ouch.

. Just build the NBN and bugger the expense

. The NBN is a slowly unfolding disaster

. We're spending a fortune on new wires, we'll disconnect the ones you have

. Questions for the NBN cheer-squad


Bernard said...

I’ve had 100Mbit Optus cable since April 20, 2010 – a service that is available most (but admittedly not all) Optus cable customers.

It is a shame your article didn’t point out that both Optus & Telstra cable customers have been able to get 100Mbit over cable for more than a year already; it would have strengthened your point.

My comment on hearing the $800M price for the Optus buy in to NBN was that every man, woman and child in Australia has each paid $40 to NBN to replace existing 100Mbit connections over cable with brand new 100Mbit connections over fibre for Optus cable alone.

Rob Thomas said...

There are a couple of incorrect base assumptions you have made here. Firstly, you're assuming that Telstra and Optus want to keep their cable (and in Telstra's case, copper) networks.

They're both jumping for joy about the ability to throw the albatross of their HFC networks away. Telstra have a nice profitable deal with Foxtel for their cable, so they're keeping it for that, but Optus are all too glad to be rid of it.
But you honestly expect a privately owned company to happily throw away their insanely expensive infrastructure without demanding some form of compensation?
Optus are going to remove their overhead cables. That's going to cost them money all by itself.

Telstra are still keeping their Pay TV, so they aren't removing their overhead cables, and don't get as much money per customer as Optus do, in case you were wondering why the prices per customer were different.

The 'Do Not Compete' clause you keep talking about, somewhat confuses me. At first you say NBNco will never make a return on its capital, because of wireless. This kinda doesn't make sense - NBN is using the latest Ericsson 3G LTE Australia-wide which is upgradable to LTE-Advanced, which is a contender for the 4G title. So the wireless that they're not using? Yes, they are using it.

Then you say that someone will offer something cheaper to its City customers.. What? More fibre? How?

Sure, there's existing fibre in some city buildings - PIPE have 6 buildings in Brisbane connected with Fibre, and other minor players have a couple of other buildings cabled up, but who, exactly, do you expect to dig up city streets and run new fibre? They can't use Telstra Conduits, becuase NBN have exclusive use of them now.

It's not like NBNco are going to be going 'Sure, hey, use our pits and conduits. We don't mind'.

The other major point you're missing is that Mobile Broadband (3G data that you use on your iPhone or Blackberry) is _not_ what the NBN is about. You can continue to pay Telstra or Optus (or Vodaphone if you're unlucky) for the privilege of using their data. Nothing will change. That's mobile broadband.

Rob Thomas said...

(Part 2) The NBN is about FIXED wireless broadband. Telstra offer that, but not in 3G LTE. Check the Telstra Business website.

Now, continuing on, you talk about speed. DOCIS3 can do about 300mbit down and 100mbit up. That'll probably be fine for.. ooh, about 10 years. And then in 10 years, we'd need to replace it with something. Probably fibre.

Hey, remind me - how long is the NBN rollout going to take? A reasonably long number of years. About 10, wasn't it? Yeah. 10. What? Those numbers match up? Don't tell me that someone is thinking about the future, and being proactive rather than reactive?

You need to pay attention to the difference between 'Wireless Broadband' and 'Fixed Broadband'. Wireless Broadband is someone with a Smartphone. Can you even BUY phones that don't have internet access on them these days? I don't think you can. So lets assume that every new phone is a wireless broadband consumer. And as Australia has an insanely high mobile penetration rate (I think it's like 1.4 phones per person), we can reasonably happily take it for granted that anyone who wants mobile broadband already has it.

So why are fixed broadband connections growing at all? Why aren't people turning off their home ADSL and HFC connections?

Because they're a totally seperate thing, that's why. And while you continue to assume they're the same thing, everything else you assume will be based on the same faulty assumption. Ass(u)(me). Except, in this case, it's just you.

Rob Thomas said...

(Part 3) My 19yo son is a prime candidate of the 1-in-3 broadband consumers who don't have a fixed line. He has an iPhone. He doesn't have a home phone at all - he doesn't need it. Why would he? He uses facebook on his iPhone, and does his banking on it. But when he wants to use the Internet, to download something, or watch some stuff on Youtube, he doesn't plug his laptop into his phone and Tether it. He drags his laptop over to a friends place, or our place, and uses our ADSL2.

I could keep going, but as the rest of your article is trying to figure out some way of saying that people won't use the NBN over other things that don't exist that somehow will be cheaper, and until you can actually offer some real alternatives (and note that Telstra's 3G LTE service is not available, at all, to fixed networks - you can only connect to Telstra's 3G LTE network with a USB dongle) to the NBN, the article is just wishing for something that doesn't, and won't exist.

I don't know why you're wishing for it either. It's not like people are going to go 'No thanks, I'll stick with my more expensive and slower ADSL2 connection!' Sure, there's the lucky 10% (that percentage I doubt, but I'm yet to find a more accurate figure) of broadband consumers that have DOCSIS connections, but the NBN isn't about that. The NBN is about giving EVERYONE the same footing. And if we're going to give them the same footing, why not give them a reasonably future-proof one? Not one that, even the people who manufacture it say, is going to need to be replaced in 10 years with fibre.

Anonymous said...

Lovely stuff, Rob Thomas.

I don't know why, but I have this picture of Peter sitting in front of his computer with his eyes scrunched shut, his fingers in his ears yelling "I can't hear you, la la la, I can't hear you"...


Marek said...

has just 34 fibre or cable connections per 100 residents. It has 90 such wireless connections.

I would have thought that since wired connections are made to homes and wireless connections are made to people that these 2 numbers are not directly comparable(unless of course you have one person per home)

Anonymous said...

Is Peter just counting customers of the big Telco in South Korea? The one that is losing customers to the other two telcos there? I think he is. Just like Turnbull claimed South Koreans were turning away from 100mbps fibre to slower wireless—no, they were turning to the other two telcos.

Peter just doesn’t like the NBN. Hey, he is alright, has fast internet at work!

Peter Martin said...


The OECD numbers are "Fixed and wireless broadband subscriptions per 100 inhabitants."

I included a link to the spreadsheet.

You are right. The number would never reach anything like 100 for fixed connections because there are fewer households than residents.


As far as I know the OECD figures are complete.

Peter Martin said...


Big firms such as banks, universities and broadcasting organisations are already setting up or have set up their own optic fibre connections to escape Telstra and now NBNCo. Don't doubt it.

You ask:

So why are fixed broadband connections growing at all? Why aren't people turning off their home ADSL and HFC connections?

The real question is why demand for HFC services is not exploding where Teltra is offering 100 Mbps.

I don't know why, but it doesn't bode well for NBNCo.

Rob Thomas said...

The reason why 100mbit isn't 'exploding', is because most people don't need 100mbit YET.

Everyone (and, honestly, I mean EVERYONE - even Turnbull) admits that 100mbit will be a requirement in a number of years. Turnbull says 15. The rest of the world says 10.

The entire reason that ANYONE is getting 100mbit now means that some people (but not all) want or need 'fast' bandwidth now.

The difference is, is that 'fast' is a moving target. 5 years ago, 'fast' was 8mbit. 5 years before that, 'fast' was 256k, and 'average' was 56k.

Now we're at a point where 'fast' is 100mbit, and 'average' is 24.

Trying to draw a line in the sand and say 'Fast is 100mbit, and fast will always be 100mbit' is provably false, and honestly, a pretty silly assumption to make.

So, to answer your question, "I don't know why 100mbit isn't exploding" is because you haven't actually thought about WHY it isn't exploding.

10 years ago, most office computers were connected with 10mbit networks and cables. Now, most computers come with 1Gbit NICs and a 8 port home gigabit switch costs $35.

Rob Thomas said...

Oh, sorry, I missed a point. Companies are building their own Fibre networks.

Well yes. Because Telstra do not, for any amount of money, let other people use their fibre networks. Well, not until now. And it cost us $11b to make them release their perilous clutch on their fibre networks.

I work at the largest power station in QLD. We _desperately_ need faster bandwidth, and Telstra have refused to even QUOTE on providing us a piddly 10mbit data connection. We'll be happily paying NBNco pretty much whatever they want for their 100mbit connection, because it'll actually exist. And it'll work.

How many other big companies are in that same situation? Lots. Which is why they've run their own fibre.

Marek said...

The number would never reach anything like 100 for fixed connections because there are fewer households than residents.

so the figures then don't support the follow up statement

They suggest that even with fast fixed broadband on offer an awful lot of households forsake

Peter Martin said...

They suggest it, they don't confirm it. NBNCo expects it. Its corporate plan expects 16.4 per cent of households to foresake fixed lines for wireless.

Michael P said...


You are assuming mobile and fixed broadband are substitutes. A lot of current industry thinking is that they are more complements.

Demand for capacity on wireless broadband networks is growing very fast. If this continues, it is likely not to be able to be serviced by technology improvements, more available spectrum or a greater density of base stations. Demand will be price-limited unless there is another way out.

The other way out on current industry thinking is that everyone has a wirless broadband device that connects to the public network outside your house or place of work. At home (or work) it will connect to your own private base station (called a pico cell) that connects to the network via fibre connection (supplied by NBN or someone else). People will connect through their pico cell when they can because it will be faster and cheaper than though a common base station.

No one in he past has been able to predict the future, so I'm sure I can't, but this seems like a reasonable way for technology to develop.

Peter Martin said...

I can't predict the future either Michael P, but I did explicitly make the point they are both substitutes and complements:

To some extent the new Korean figures are good for NBNCo - they show people buy wireless connections in addition to fixed lines. To a larger extent they are bad news. They suggest that even with fast fixed broadband on offer an awful lot of households forsake it for a still-fast, cheaper and more convenient alternative.

Dong said...

Peter, there are already overload problems with wireless. At a large conference recently when everyone was using their smartphones and laptops speed dropped and there were frequent dropouts.
When you consider upload needs then NBN is a no brainer. A business in Scotsdale (photos) that was taking two days to send its files to the mainland can now send them in an hour making the business viable. In a few years enterprising business people will have new ways to use the bandwidth to make money. The first telephone lines did not have enough users to make them economical but increasing use of them built up. It couldn't have done so unless it was built.
Check out how much it would cost to power the number of towers needed to provide the coverage anticipated by the NBN, not to mention the carbon footprint.

Anonymous said...

The critical infrastructure health providers that I work with *need* very fast broadband now (100mbps would be OK) but they can't get it. Their IT systems won't work effectively until they do.

Anyone who has information heavy business running across multiple sites will benefit from this type of infrastructure, and they can't all be in highly urban areas. And fibre is future proof and much lower maintenance than metal cables.

Mostly I like Peter Martin's work, but he's blinded by an attack of the Dunning-Kreuger effect (lack of awareness of one's own incompetence) on this matter.


Peter Martin said...

Health providers do not need 100 Mbps to the home. They do need 100 Mbps or more between facilities.

Information-heavy business running across multiple sites are installing their own infrastructure right now or have already. NBNCo won't help them and won't get their money.

Anonymous said...

Rob Thomas,

TPG ADSL2+ is very fast and is much cheaper per month than the $60 - $70 minimum that the NBN is expected to cost.

Oh, and it doesn't cost the taxpayers $40 Billion

Anonymous said...

Peter - All of the above is very interesting but, at the end of the day, I don't want to live in a society where people like Rupert Murdoch (via telstra/foxtel) control the major information infrastructure in this society. I'm happy to pay an absolute bomb to stop that happening.

Isn't it more important that australian consumers (citizens) get more competition concerning the information content they receive, which is something the NBN should facilitate in spades. Why do you think Foxtel can charge more than $100 a month for what it provides? It's not because there is too much competition.

Anonymous said...

You don't need to be a technologist to understand why non-compete clauses are bad for consumers.

It's like we are paying to be punched in the face.

Don't get me started on "cherry picking". Another phrase coined to mask anti-competitive behavior, like "dumping".

I'm also sick to death of hearing about Rupert Murdoch. People who keep on banging about this bogey-man, seem to think that he's successfully brainwashed everyone except them.

Yeah, we are all zombies, except you. You took the red pill.

Alas, your thoughtful analysis has brought out the usual mob of vociferous internet fanbois.

Anonymous said...

Fear not, you can keep reading the Australian and watching Sky news until your brain disintegrates. But some of us want more diversity of views, which the NBN should provide.

Further, the options for niche programming should be immense. It is easy to see the day when you will only pay to receive the game your local football plays each week and so won't have to pay for an immense amount of dross. The clubs will benefit and consumers will benefit because there will be real competition.

Anonymous said...

NBNCo are not in the content business.

Spending public money to shut down pre-existing infrastructure does nothing to improve the situation.

It hurts us as tax payers and consumers.

Also your subjective assessment of sports programming doesn't interest me, I don't watch sports of pay for my television.

In fact, I'm debating this this issue with you over Optus infrastructure.

Anonymous Zombie

badm0f0 said...

The real question is why demand for HFC services is not exploding where Teltra is offering 100 Mbps.

Probably because nobody wants to pay extra for speeds they will never actually receive (unless they the only premise connected to the node). HFC is subject to contention & when the node reaches a degree of saturation speed tumbles.

We switched from our alleged 30Mbps cable connection to ADSL2 at Telstra's suggestion because the actual speed we were receiving during peak dropped as low as 1Mbps. We were paying a premium for that speed over packages with the same download allowance but lower theoretical speeds (not sure how it could actually have been lower in reality unless they were sending us data by carrier pigeon).

You also refer to businesses & institutions having set up their set up their own fibre networks to escape Telstra & now NBNCo. That is fine for the small % of businesses of a size who can afford to extend fibre to existing premises or the small % willing to pay a premium to set up or relocate to premises that can connect to existing fibre. This is not the majority of businesses however.

The current cost & geographical constraints of private fibre acts as an effective barrier to entry in many data-oriented industries which are strongly dependent upon high availability, low latency services. The great advantage of the NBN is less maximum speed than it is the combination of reliable high speed & ubiquity.

Peter Martin said...

You hope. If those advantages are not there at a good price you will have to (try to) shop around.

Anonymous said...

Badm0f0 speaks of contention. Fiber optic has the major advantage of permitting multiplexing.

Badm0f0, no one disputes that Fiber is superior to copper.

The criticism here is two fold.

One, that copper is capable of giving Fiber a run for it's money.

And two, that public money is being used to deny consumers access to this existing infrastructure.

At the very least, this should give NBNCo's adherents pause.

Anonymous Zombie

Anonymous said...

"Health providers do not need 100 Mbps to the home. They do need 100 Mbps or more between facilities."

Yeah, you're right. But now you're starting to tell them where they can have these facilities. Seeing as the old copper needs to be renewed anyway, seeing as Howard stuffed up the privatisation/regulation of Telstra and seeing as fiber is pretty much future proof, the NBN solves a range of problems in one go. Wireless is a distraction.


Peter Martin said...

I am sure you have considered these things carefully kd, but I'd like the Productivity Commission to take a look a as well.

It is an enormous expenditure. The PC has run the ruler over lesser projects.

Anonymous said...

I'm in agreement with Peter on this.

I would love for the Productivity Commission to evaluate this.

This would tell us just how many lives would be saved by the NBN, effect upon life expectancy, infant mortality and so on.

That way we can weigh this up against the alternatives.

After all, the 800 million dollars is a lot of money. You can almost build a Fiona Stanley hospital for that.

Consider that vast sums of public money will be spent scrapping productive assets half way into their useful life.

What if we simply deferred the NBN by ten years?

What if put that 40 billion dollars in ING Direct? The real interest alone would be enough to purchase the same Fiona Stanley hospital.

These questions and many more need to be asked.

I don't know what the biggest problems facing the medical community are.

But I doubt they will solved by having patients Skype their GP.

Unless of coarse, you also are proposing installing MRI machines in peoples livings rooms.

Anonymous Zombie

Anonymous said...

The whole point about the NBN is that it's NOT in the content business. Therefore, it will be happy to allow anyone with content to, for a reasonable fee, use its network (unlike the present players). As a consequence, you won't be beholden to the present TV/PayTV organisations for news, entertainment, etc etc. That is real competition. The most important marketplace in any democracy is the marketplace of ideas and information. You can't put a price on that, but it's worth a fortune. The NBN should help that enormously.

Anonymous said...

I still don't see how using public money to dismantle functioning infrastructure will do anything for the democracy of, or market for ideas.

Your argument is basically that people won't be free to think for themselves until Telstra and Optus are eliminated from this marketplace.

Will you people, please try to stay on point.

The article is about non-compete clauses, who pays for them and why the NBNCo's business plan is predicated on them.

Anonymous Zombie

rog said...

peter, you claim that the reason that there isnt a national very fast train is that the airlines would apply discounting to cream off valuable customers.

Maybe, but in a recent interview Tim Fisher let slip that the lobbyists in Canberra ran the show and Max the Axe was influenced by those lobbyists. Because of its high efficiencies development in trains has been put well and truly on the back burner.

Post a Comment