web analytics

Bill Bennett

Menu

Tag: fibre

New Zealand is building a fibre to the premise (FTTP) network. I’m excited about that.

UFB fibre unbundling fades into background

The Commerce Commission’s final guidance, published this week, may put the UFB fibre unbundling question to bed.

Unbundling is the idea that internet service providers can buy a raw physical connection from a wholesale fibre company, attach their own electronics and sell services on to customers.

At present New Zealand’s ISPs have to buy the physical connection bundled with the electronics at each end to send and receive data.

Unbundling deals on the table

Since the start of the year, fibre wholesalers have offered unbundled fibre to retail service providers.

Chorus, Enable, UFF and Northpower have all published their proposed offers. Few service providers have shown any interest. But two, Vodafone and Vocus, formed a partnership to work on unbundling fibre. That project hasn’t got far for reasons that will be obvious when you read on.

I’m simplifying here. There’s a technical rabbit hole you can go down that discusses matters like Layer 1 and Layer 2. You can find an easy to read summary of unbundling on this site.

You don’t need to understand the details to grasp the basic economics.

The Commerce Commission regulates basic fibre prices. Fibre companies can sell wholesale at a price determined by the Commerce Commission. This price takes into account the cost of building the network. It is set so that fibre companies get a return on their investment.

Fibre expensive to build

As you’d expect, building fibre networks is expensive. The bulk of the costs pay for civil engineering and the physical part of the network. This is Layer 1.

Fibre companies turn Layer 1 into Layer 2 by installing electronic hardware at each end of a line. This moves the data through the fibre.

Installing electronics is a small fraction of the cost of building a fibre network. A few percent at most.

Given fibre companies charge a cost-based fee, the price difference between bundled and unbundled is tiny.

Which means there’s no value in UFB unbundling to a service provider.

While the Commerce Commission regulates UFB wholesale prices, it opted not to regulate unbundled fibre prices. It left that to commercial negotiation and said it may step in later if necessary.

As with other fibre regulation there are two key ideas in last week’s Commerce Commission guidance. These are: Equivalence and non-discrimination.

Equivalence means the terms a fibre company offers must be the same as it offers its own business operations. Non-discrimination means they can’t play favourites with retail service providers.

The big telcos don’t like non-discrimination. It means they can’t use their clout to get better prices than small ISPs.

Room to move

This week’s guidance might not be the last we hear of UFB unbundling, but it doesn’t leave ISPs with much room to move.

It’s worth pointing out that consumers won’t benefit from unbundling. If there are cost savings, the ISPs will want them to rebuild margins. There’s nothing an unbundler can add to a fibre connection, almost all their options include taking things away.

Now the Commerce Commission is investigating whether the non-price terms attached to unbundled fibre offers meet regulatory requirements. This can be trickier. Take the fibre companies reluctance to let external engineers cut and splice their fibre lines. On one level it sounds reasonable, but the regulator may yet decide otherwise.

In its guidance the Commission explains how it will monitor and enforce obligations.

Telecommunications Commissioner Tristan Gilbertson says: “We strongly encourage fibre providers to review their product offerings against the guidance we have issued and to make any changes necessary to bring them into compliance as quickly as possible. We also encourage retailers to raise issues with the LFCs where they believe the product offerings are not meeting the required obligations.”

Which is a way of saying: “As long as no-one is unreasonable, we’re going to leave this with the market for now.”

Data traffic hits new peak once a Fortnite

Chorus says traffic on its UFB network reached a record high on Thursday evening. It says data traffic peaked at 3.15Tbps.

This was at around 8pm. Traffic always peaks in the evening when people stream television services like Netflix.

Yet that wasn’t the reason for Thurday’s surge. That was when developers released the most recent version of Fortnite, the popular online game.

Once a fortnite

Chorus says the Fortnite update beat the previous network peak. That was earlier in the month when there was an update to the Call of Duty game. A pattern is emerging with big game downloads.

The network coped with the load without congestion. We can’t take this for granted.

Earlier in the year traffic surged when New Zealand first went into lockdown. This was a heady mix of people working from home during the day and using video conferencing. They then stayed in at night and streamed movies or TV shows.

Volume all the way up to 11

In March traffic hit 2.8 Tbps. At the time Chorus pointed out this was well within its theoretical maximum capacity of 3.5 Tbps.

Thursday’s peak came close to this level. Chorus says today’s network capacity is around 4.5 Tbps.

New Zealand got lucky when the pandemic triggered a surge in data demand. Only months earlier telcos upgraded networks to cope with anticipated Rugby World Cup demand.

We’re not so lucky when it comes to blockbuster game downloads. Games software companies time worldwide launches to co-incide with international demand lulls. That happens to be New Zealand’s demand peak. So we push our networks to the limit.

Away from games and streaming, Auckland’s level 3 lockdown saw higher daytime traffic. Although it is up on recent weeks, it hasn’t returned to the levels seen during the first lockdown.

Chorus says traffic was at 1.49 Tbps at noon on Friday. That’s around 40 percent above the level earlier in the month when there were no lockdowns. The company says Auckland traffic was up 69 percent, the rest of the country was up only eight percent.

The last days of urban copper

Copper phone networks have served New Zealand for over 100 years. They won’t be around much longer, at least in cities and towns.

Spark has started to call time on urban copper. It says it plans to retire PSTN in Devonport and Miramar by Christmas. That move will affect around 1000 customers.

It is the first step in Spark’s plan to move customers from copper on to either fixed wireless connections or fibre. In the companies words, these are “the modern alternatives”.

Plain old telephone service

PSTN is the public switched telephone network. That’s the engineers name for the old copper-based telephone system. At times people in the business call it Pots: plain old telephone service.

The key word in that phrase is old.

Spark says PSTN is now at the end of its life and needs replacing. It is now 17 years since equipment makers stopped making PSTN hardware. Spark says it is getting harder to find people with the skills to maintain the technology.

Old folk reluctant

Older customers continue to love Pots. There’s a reluctance to move from a technology they have known for decades.

It is no long essential.

It is possible to deliver services resembling PSTN voice calling on fixed wireless and fibre connections.

The alternatives do not work in a power outage. You can’t, say, make a 111 call if there’s no power. This worries planners and officials, but there are few homes without mobile phones.

Alarms

There are difficulties with security or medical alarms designed for copper network technology. Again, there are modern alternatives in almost every case.

Nevertheless Spark says it will not move anyone who needs special hardware until it has found a replacement.

Getting rid of urban copper will simplify telecommunications and lower costs. We won’t have to run two networks in parallel. Companies like Spark can streamline support operations and reduce costs.

No longer controversial

A move from copper networks may have been controversial a few years ago. Today he majority of customers in cities and towns now use fibre or fixed wireless instead.

Fibre uptake is now 60 percent or better in urban areas. There are 180,000 homes or business using fixed wireless connections. Meanwhile there are more mobile phones in use than there are people.

There are problems in areas not yet covered by fibre. Removing urban copper will put further pressure on the government and fibre companies to extend the reach of the UFB network to these places.

Fibre and beyond

Fibre is now scheduled to reach 87 percent of New Zealand by 2022. It is realistic to stretch fibre coverage up to around 90 percent or a little beyond.

Where that isn’t possible, mobile coverage and Wireless Internet Service Providers can fill in the gaps.

Compared with fibre or wireless, copper is expensive to maintain. It is more expensive to maintain in the rural areas where it is likely to remain.

As the number of lines falls, the support cost per line will rocket. There will come a point where the remaining copper network is economically unsustainable.

The flip side of this is that reducing those costs should free up money to pay for rural network upgrades.

Spark not hanging around

Spark is moving early. The Commerce Commission has published a code for dealing with 111 calls after copper is switched off. This has to be in force by the end of next year. Once that is done Chorus will be allowed to stop offering copper services where fibre is available.

Chorus continues to own and operate the copper networks. They will remain in the ground for now even in the areas where Spark has withdrawn services.

By running pilot programmes in Devonport and Miramar Spark will be able to better understand how decommissioning PSTN might work. The company expects to spend years moving off the services across the rest of the country.

This story has been updated (24-07-20) to reflect out-of-date numbers and timings.

When telcos don’t like regulated industry structure

Telcos are waging a public relations war on Ultrafast Broadband. Reading between the lines of their public statements, they don’t like the regulated market model.

This was set-up over a decade ago by John Key’s first government. The government restructured telecommunications. It tilted the playing field in favour of customers. New Zealand businesses and consumers got a great deal.

Regulated separation

Reforms separated the market into two parts. Regulated fibre companies would build, own and operate the UFB network. They are regional monopolies. They can only sell wholesale services.

Retail service providers can sell broadband without geographic boundaries. RSPs all buy wholesale services on identical terms.

This model promotes competition. No single player can dominate in the way companies could before the restructure.

Chorus, Northpower, Enable and UFF are the fibre companies. There are 90 retail service providers. The biggest and best known are Spark, Vodafone, Vocus, Trustpower and 2degrees.

Consumers are happy with how the market operates. At least those in areas that can get fibre are. The fact that people in areas without fibre are grumpy about it speaks volumes about the model’s success.

Best broadband

Thanks to UFB New Zealand has one of the best and most affordable broadband networks in the world.

It is great for consumers. It is less wonderful for big RSPs. The biggest ones are not happy.

There are two main reasons they don’t like the UFB model.

First, from their point of view, separation does too good a job of promoting competition. RSPs fail to compete on anything other than price. They discount broadband to the point where, as Vodafone CEO Jason Paris says; “They compete away all the margins”.

Thin margins

Retail broadband margins are wafer thin.

A consumer might pay $80 to $100 or so each month for a fibre broadband account. Roughly half of that goes to the wholesale fibre company. RSPs have overheads and costs. The margin is often less than 10 percent of the monthly fee. It can be lower, some only make five percent.

They get that money 12 times a year. Yet regulated UFB is not as lucrative as the old ways of selling telecommunications.

This explains why Spark and Vodafone are keen on fixed wireless broadband. It’s an inferior product, but they get to keep a larger slice of the cake.

The second reason the bigger telcos don’t like the UFB model is they are not in control of their own destiny. This bothers them. They have few options, little room to manoeuvre.

Equivalence

Another, less obvious grievance is the UFB idea known as equivalence.

A supermarket chain like Countdown pays less for products than the wholesale price paid by a corner dairy. They get economies of scale. Countdown buys tins of baked beans from a wholesaler at a lower price than dairies pay.

Equivalence means the largest telco pays the same as the smallest RSP for a customer hook-up.

There are economies of scale when it comes to support, back-end services and marketing. Yet the aristocratic telcos resent paying the same price as the peasant RSPs.

Phoney war

All these aspects of the UFB model come into play as telcos wage a phoney war. It is a war that is being fought on a few fronts.

Last month there was fuss from big telcos about something known as the wall of bad debt. This sounds like something from the Pilgrim’s Progress.

Spark, Vodafone and 2degrees have been the most vocal. They say a recession is coming and they face large unpaid broadband bills.

They want Chorus, it’s always Chorus 1, to share the cost.

This is an unusual argument on two counts. Wholesalers don’t shoulder the risks of their retailers in other sectors.

What’s more, the debt risk facing the large RSPs is of their own making. When lockdown began the telcos said they would not cut off customers for non-payment.

Decent

It was a good and decent thing to do. Or at least it is a good and decent thing to do with your own money.

It’s not such a good and decent thing to be generous, then turn around and ask the fibre companies to pay half. It would be fine if they had gone to the fibre companies first and agreed something along these lines. But they didn’t.

This doesn’t mean their argument is unjustified. They have a case. It is not as clear cut as they argue.

It is not as if fibre companies don’t face their own post-lockdown financial risks.

First they had to keep contractors ticking over when there wasn’t much work. What’s more, they face their own potential wall of bad debt if RSPs go bust in the looming recession. Having an RSP fail to pay would be far more serious than an individual consumer missing a payment.

Moral hazard

There is a moral hazard aspect to the wall of bad debt. If telcos get wholesalers to carry half of the risk, they have an incentive to take more risks.

Companies have a tendency to be reckless when protected from consequences. Among other things, they might not chase bad debts as hard if they know they face half the loss. They might be less fussy about who they give credit to.

It doesn’t help Chorus’ case that it is now enjoying the financial light at the end of the UFB-build tunnel.

That said, Chorus agreed to pay $2 million towards the bad debts. In round numbers that’s about half the cost if 7000 New Zealanders fail to pay for six months of broadband. Around one million New Zealanders connect to the UFB. So if seven percent default on payment, Chorus pays RSPs half the cost.

Telcos don’t think that’s enough. Remember here that Chorus has no legal obligation to pay anything.

Minor regulatory risk

There is a small danger here to the idea of regulated separation. If the fortunes of wholesale fibre companies depend on RSP performance, they could play favourites. A bigger danger comes from another battle: fibre unbundling.

To understand how fibre unbundling threatens UFB, let’s go back to the original plan. The government decided fibre companies would sell layer 2 services. It left open an option for layer 1 services at a later date.

In this context layer 1 is an unbundled service.

That later date is now. Or, to be more precise, it was January this year.

Layers

You don’t need to know the technical nuance about layer 1 and layer 2 to understand what is at stake. Here’s the simple version.

In effect layer 2 means fibre companies sell RSPs a complete service from a home or office to a network node. This includes the network connection hardware at each end of the link. Fibre companies wrap all the parts needed to do this into a bundle and sell it as a whole.

If they sold layer one, RSPs would get a fibre connecting the home to the node. They pay for the hardware on the ends of the line. Hence unbundled.

The problem with unbundled fibre is that all the costs in a network are in the civil engineering. Stringing fibre around the country is expensive. The hardware on the ends of the fibre cost peanuts in comparison.

Wholesale UFB prices are regulated. The price depends on the cost of building and supporting the network. The bundled hardware turning layer 1 into layer 2 is a small percent of the total. The input cost difference between bundled and unbundled fibre is tiny.

In other words, in the UFB model unbundled fibre doesn’t make economic sense.

Economics

Big telcos know how the cost structure works, but don’t accept the economics. They continue to argue and lobby for lower unbundled prices. If they get their way, fibre companies would sell connections at below cost.

The entire regulated UFB model would collapse. And future governments would struggle to raise private capital for infrastructure projects.

This connects to another way telcos are waging war on UFB. Last week, a Vocus press release said the latest round of Chorus fibre price rises is: “cynical, money grabbing and unwarranted”.

It’s an opinion. UFB regulations say fibre companies can raise prices in line with inflation. That’s a step away from “cynical, money grabbing and unwarranted”.

Inflation is tiny, around one percent, so the rise is small. It reflects the increased costs fibre companies pay for maintenance. Without it, a fibre company’s margins would ratchet down each year.

Futureproof

Government choose this price rise as part of the regulated UFB model when it feared investors would not want to fund fibre. It needed this clause to attract investors. It still needs this clause to attract private investors for future infrastructure projects.

Vocus, like the other telcos can choose to pass this cost on to customers. As we’ve seen, competition is tight, if a telco raises prices customers may move elsewhere. But that’s how market economies work.

And that’s the story here. Price increases, unbundling and, to a lesser degree, the coming wall of debt are all hardwired into the UFB. Government designed market regulations that way for a reason.

Regulated competition

UFB has been a success, in part, because there is a competitive market. Unpick the regulations and the whole UFB fabric unravels.

The challenge facing telcos is that the market is too competitive. As Paris says, they have competed away their profits. Corporations might pay lip service to free markets, but they don’t like them.

Mobile means Spark, Vodafone and 2degrees have alternative paths to profit. Enterprise services, adding value through content deals or power billing are other paths. Being better at what you do is always an option. Market consolidation might help.

Attempting to repair margins by chipping away at the foundations of UFB is not a wise strategy. There was a time when Telecom thought it could defy government policy and regulation. Look how that ended.

Disclosure: I do freelance writing and editing for Chorus, but the company doesn’t tell me what to write. 


  1. Telcos have figured out it is easier to bash the biggest fibre company. Criticising Chorus gets more media attention and more sympathy, than disrespecting “fibre companies”. Yet almost every accusation made against Chorus applies to the other fibre companies. If, say, Northpower was the target of anger, that would look like bullying. ↩︎

Huawei’s Gigabit Gap: NZ broadband ahead of Australia

Australian title Business IT covers a report released by Huawei: Australia suffering gigabit gap despite spending A$51B.

The report says it cost A$4,500 for each NBN connection network, a total of A$51 billion, but the network still only reaches 28 percent of premises. In comparison New Zealand’s UFB network now reaches 75 percent of premises. That figure will rise to around 85 percent when the second UFB stage finishes at the end of 2022.

 

 Gigabit capability in selected countries
Gigabit capability in selected countries (percent of premises)

Ten years after starting the NBN project, less than a third of homes can get gigabit fibre. Meanwhile New Zealand fibre companies are starting to offer speeds of up to 10Gbps. Australia has no plan to extend its fibre footprint.

The report also shows Australia has the world’s third most expensive gigabit broadband. The only people who pay more are in Norway and Canada.

OMDIA, formerly known as Ovuum carried out the research for Huawei.