Wednesday, May 30, 2012

Build the NBN if you must, but please don't destroy working assets

Wednesday column

The Optus high speed cable internet network is a national asset. Comprising 25,000 kilometres of coaxial cable strung across 550,000 poles in Sydney, Melbourne and Brisbane supported by 7000 kilometres of fibre the provided Telstra with its first genuine competition, putting its own wires direct into half a million homes.

Telstra fought back with FoxTel, Optus got burnt and has probably never recovered the cost of stringing the cables. But from an economic point of view what’s important is that the asset exists. Its costs have been sunk. What Optus has now is an asset that costs relatively little to operate and can deliver peak download speeds of 100 Mbps - far faster anything on Telstra’s copper wires.

Right now it has 496,000 customers. It is within connecting distance of another one million meaning that for very little cost Optus or a buyer of the network could provide a very fast very cheap internet service to as many as 1.4 million households - a service far faster than ADSL.

Only a vandal would destroy such an asset. Only a seriously confused regulator would allow it happen.

NBN Co has asked the Australian Competition and Consumer Commission to let it pay Optus a reported $800 million to shut the network down. The only precedent for the destruction of infrastructure on such a massive scale is the $4 billion NBN Co is to pay Telstra to rip out its copper network, transfer its customers to the national broadband network and remove the internet from the cables it uses to deliver FoxTel.

In no other industry would the ACCC approve such an agreement not to compete. In no other industry would it countenance a bribe to decommission working infrastructure.

Mid last year it approved the Telstra deal. This week it published a draft decision approving the Optus deal.

In terms of national assets the Optus decision is arguably worse than the Telstra decision...
Telstra will decommission its copper phone lines street by street as the NBN cables are switched on. While there will be a loss of competition, copper probably isn’t able to compete with fibre over the long-term. By contrast the Telstra and Optus coaxial cables are as good as new. They were strung up in the last half of the 1990s. They are already fast and capable of being made faster. They cost almost nothing to maintain. They are something that should not be destroyed wantonly.

The Telstra coaxial cables won’t be. It has only agreed to disconnect the internet from them. It could put it back as soon as changed legal or political circumstances allow. The Optus coaxial cables are scheduled for destruction. The ACCC’s decision will have physical consequences. It should not be taken lightly.

And yet the ACCC gives every indication that its decision could have gone either way. It was “finely balanced”, according to chairman Rod Sims in Monday’s statement.

Its decision to approve the agreement was based on weighing carefully “clear public benefits” against “a potentially large but less clear detriment”.

The main “clear benefit” is odd. The Commission says the agreement will “avoid the cost of operating the Optus network to provide a service the NBN is also able to provide”. Would we apply it elsewhere? Should Virgin shut down its airline network to avoid the cost of operating a service Qantas is also able to provide? Should Woolworths shut down its network to avoid the cost of operating a service Woolworths is also able to provide? Of course they shouldn’t. We normally value competition.

Some will argue that wiring up houses is different. Frontier Economics, the consultant used by Optus in its submission to the ACCC says fixed broadband services are a natural monopoly - they shouldn’t be provided twice. But Optus cable network is already in place. It costs next to nothing to keep it in place. It is NBN Co which is planning to duplicate it. Given its plans and the rate at which it is duplicating infrastructure right now, its complaint against “inefficient infrastructure duplication” is simply strange.

In truth is competition that worries NBN Co, not inefficiency. It is paying $800 million to remove a competitor, not out of a public concern about inefficiency.

With its last big fixed line competitor out of the way the only market restraint on its prices and quality of service will be wireless internet, and it’s on to that as well.

The Optus agreement given a preliminary tick by the ACCC prevents Optus from advertising wireless data services within the area served by its existing cables in a way which is “expressly critical of or makes any express adverse statement about the performance or functionality of the NBN where such a statement is misleading or deceptive or involves the making of a false or misleading representation in contravention of the Australian consumer law”.

The ACCC waived it through because it essentially meaningless, requiring Optus to do no more than obey the law. But it indicates how deeply concerned NBN Co is at the prospect of competition, or as it puts it “cherry picking”. Competition would force it to provide value, and its national pricing structure would force it to provide it to all Australians. Its $36 billion cost base won’t allow it. That’s why it needs to destroy perfectly good working infrastructure. It’s why I think the ACCC needs to think again.

In today's Sydney Morning Herald and Age


ACCC Authorises NBN CO Optus HFC Subscriber Agreement


"Optus is being paid $800m to shut down a tax-avoidance scheme"

The ACCC has authorised the NBN-Optus deal. So where do the public benefits come from? As background, remember that the ACCC can only authorise an otherwise anticompetitive agreement if the public benefits outweigh the public detriments. I have previously expressed my skepticism about this.

So where are the benefits? The Box 1 on page 27 of the decision has the answer. And it illustrates the problem of the hidden cross subsidies built into the NBN.

Suppose that the cost of providing broadband services to Optus’ broadband customers is lower on the NBN (formally, incremental costs are lower) but for some reason the NBN charges Optus (and other broadband retailers) a wholesale price that exceeds Optus’ own incremental cost. Then, in the retail marketplace, Optus will undercut the NBN and keep its customers despite it being cheaper to service them on the NBN. Hence, the customers must be forced to the NBN to achieve the incremental cost savings (albeit at the cost of creating an allocative loss as the end users face higher prices).

So the question is – why is the NBN charging a wholesale price that is so high that it can be undercut by an inefficient competitor? The ACCC give two reasons. First, that the NBN must recover its fixed costs (which presumably Optus have written off). Unfortunately, if this argument is correct it undermines the benefit test. If, taking new investment into account, it is less costly to continue using the Optus network then the best outcome would be to not role out the NBN in areas covered by the Optus network. Of course, this is what the NBN is threatening, so taking the NBN at their word, if the fixed costs of the NBN rollout eliminates any incremental cost advantage, then the agreement should not be authorised and Optus can service that part of Australia. Of course the ACCC may not believe the NBNCo …

The second argument is more convincing. Excessive pricing by the NBN. It is cheaper to provide services to the Optus customers using the NBN but the wholesale prices that the NBN will charge are so over inflated that this cost gain will be more than eaten up by the NBN monopoly pricing in the Optus service regions. In this case, the Optus customers must be forced onto the NBN to save the relevant costs, albeit to then be exploited by the NBN monopoly pricing.

Now normally the ACCC would roll around the floor laughing if someone put this sort of argument to them. However, in the case of the NBN, monopoly pricing in city areas is necessary to create a uniform national price. The city will subsidise the bush. So the monopoly pricing is really just a hidden tax and, of course, the customers must be forced to pay this tax.

So in brief the ACCC has authorised the NBN-Optus deal because (a) it believes that the cost of servicing the Optus customers is cheaper on the NBN and (b) it can’t be left to the market to sort this out because the NBN prices are so inflated by a hidden tax that the Optus customers would probably stay with Optus.

Hmmm. So Optus is being paid $800m to shut down a tax-avoidance scheme?

Well, that is what happens when you get bad policy. I have no trouble with a uniform national broadband price. But have the NBN set prices to reflect cost. And then provide a transparent subsidy to those consumers who the government believes should receive a lower price. Don’t hide the tax – it just leads to problems like that faced by the ACCC.

RELATED READING: Why the ACCC should NOT authorise the NBN-Optus arrangement
Former ACCC Commissioner Stephen King

Related Posts

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

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

. Just build it and bugger the expense


Chris Grealy said...

Build the NBN we must, especially here in Capalaba, 30 minutes from the Brisbane CBD. The copper is rotting away and is on its last legs. ADSL is limping along at 1.2Mbps when it works.Can't get cable. No mobile coverage. Abbott tells me I don't need the NBN, but offers no working alternative. Our local (Liberal) MP doesn't want to know.
And they whinge when people call them negative. I wonder why?

Geoff U said...

The HFC infrastructure is perfectly good for providing Internet services but is no longer adequate for future needs.

Morse code was perfectly good for sending messages in the late 1800's but was no longer adequate for future needs once we were able to transmit voice.

The horse and cart was perfectly good for long distance transportation but was no longer adequate for future needs once the automobile was invented.

Travelling between US/Europe and Australia by boat was perfectly good but was no longer adequate for future needs once air travel was made available.

Anonymous said...

Please Geoff U, I don't think you are speaking from the requisite knowledge base to make such sweeping and frankly incorrect statements.

I work for the worlds second largest cable TV and HFC operator. For the last five years in all markets that we operate in, we have significantly outcompeted the telcos and in some markets have been offering 120-160 Mbps for more than 3-4 years now. We outcompete them through network superiority and aggressive price points and bundling.

The capabilities of HFC is improving along a similar development curve as wireless and fibre. Cable operators, and their vendors, aren't going to sit and stagnate on 120-160 Mbps, and you are seriously deluded if you think they will. Competition demands it.

But don't take my word for it. Have a look at the collective share prices and market capitalizations of the worlds biggest cable TV companies in the US and Europe. They have collectively hundreds of billions of market cap between them, and most are nearing all time share price highs. (all in the face of increasing 'fibreisation' from they key telco competitors)

Who am I going to believe: the market, or Mr Geoff U with his simplistic relativitities?

I'm going to go with the market on this one.

Will said...

Is $800m not a fair price for the customers? What's the problem?

The whole point of the NBN is to provide a universal access CAN model. You can't do that cost effectively through HFC because it involves multiple private parties with proprietary control on key bottlenecks. Moreover, the technology is not suitably advanced to tolerate such a premium and the harm to the fibre model that goes with it.

Continuing to run HFC in parallel just hurts the economies of scale of the fibre system. The argument that it is already a sunk cost is true but largely misses the forest for the trees in trying to preserve the lemon tree when you really need to re-scape the whole garden. This is about industry structure - something Peter completely fails to understand.

Peter Martin said...

You ask:

Is $800m not a fair price for the customers?

It is an outrageous price for taxpayers to pay to shut down something that works.

It's your money Will, your taxes.

Wouldn't you rather spend it building something up?

Marek said...

Hmm interesting situation, i understand my the NBN want to buy the customers, why optus wants to sell them and also why Peter thinks its stupid! I agree with u on this one Peter, because i believe that optus would shift it's cable customers over to the nbn anyway

Marek said...

why* not my

Letter to the Editor said...

David Epstein, Vice President, Corporate and Regulatory Affairs, Optus

April 30, 2012

Peter Martin misses the point in his criticism of the ACCC’s draft decision to approve an agreement
to migrate Optus’ HFC customers to the new National Broadband Network (NBN) [‘ACCC in a twist to
banish network’s rivals’].

Mr Martin has misunderstood the nature of competition on the NBN, which will focus on investment
in services and applications, not investment in pits, ducts and cables.

He argues that Optus’ HFC network could, hypothetically, provide real competition to the NBN in the
future and uses the analogy of airlines to make his point.

It may come as a surprise to Mr Martin but the airline industry is actually a useful example of where
the ACCC has found there are clear benefits from not duplicating infrastructure.

In recent years the ACCC has given several approvals to Virgin Australia, allowing it to rationalise its
own airline infrastructure and partner with others to expand its destination footprint.

Such transactions removed the potential for infrastructure based competition to occur if Virgin had
been forced to build or enhance its own airline infrastructure to compete with its competitors.

But in each case the partnership was justified by clear increases in product choice and service
competition at the retail level and funded by the capital Virgin reallocated to those priorities rather
than it into duplicate infrastructure.

Classic examples of this are the Skywest and Etihad authorisations. The ACCC recognised that
retail innovations like bundling, more extensive coverage and greater product choice were more
important to consumers than duplicating underlying aviation infrastructure.

The net results of this have been widely publicised competition between Qantas and Virgin on price,
destination and increased customer offerings. In telecommunications such benefits are likely to be
magnified even further.

The ACCC’s draft decision on the NBN Co/Optus HFC transaction recognises that similar factors
are likely to play out in the advent of a wholesale NBN, which will unlock a level playing field for
nationwide service-based competition.

However, the potential for increasing retail competition and choice between carriers in an NBN
environment might well be diminished if the theorists had their way. Unlike the dominant player
and smaller providers, Optus would be forced to maintain infrastructure that is confined to a
regional footprint and which cannot be enhanced cheaply.

The ACCC’s decision to approve this agreement recognises that Optus can become a stronger retail
competitor because it will be free to reallocate capital to things customers want – product and
service innovation rather than pits, ducts and cables.

This can also benefit NBN Co because the revenue it receives from early access to Optus’ HFC
customers will get it to a lower point on its cost curve earlier. This will in turn help NBN Co to deliver
on its commitment to keep access prices low and ensure that all Australians can enjoy access to a
superfast broadband network at affordable retail prices.

Anonymous said...

Wow, when you explain it like that Mr David Epstein it really does Sound like a fantastic deal for taxpayers, I guess it is true after all that what's good for Optus really is good for all!!

Whilst Mr Epstein you are in the mood for clearing up some misconceptions for us all, perhaps you'd be so kind as to clarify what the CASH value of this supposed "$800m" deal really is.

$800m is post tax NPV. That is a misnomer. A 1st year commerce student can figure out pretty quickly that grossing that figure up to pre tax, and then assuming a 10 year spread of those migration payments, then you get to a CASH value of $2.1 billion or thereabouts. That is a substantially different number, and frankly where the discussion should be focused. After all, the "post tax NPV" formula was dreamed up by the Govt in their Telstra deal to keep the number in the press as low as possible.

Let's not forget this Optus deal was a condition precedent of the Telstra NBN deal. Hence why the ACCC is approving

How much sense does this deal make now?

Post a Comment