Chorus: How Structural Separation Built New Zealand’s Regulated Fibre Utility

Chorus is the clearest example in the developed world of a telecom incumbent being broken in two to get a fibre network built. This article explains why New Zealand insisted on structural separation, how the Ultra-Fast Broadband partnership was financed, how the company earns a regulated return today and what remains to be done once the last copper line is retired. It is part of the New Zealand Company Stories hub.

How did Chorus begin?

Chorus began as the network division of Telecom New Zealand, created by operational separation in 2008. It became an independent listed company on 1 December 2011, when Telecom shareholders received one Chorus share for every five Telecom shares they held.

The assets it took were the physical layer of the old monopoly: the copper lines to almost every building in the country, the exchanges and roadside cabinets, the ducts and poles, and a growing fibre backbone. Telecom, later renamed Spark, kept the customers, the mobile network and the IT services business. The fate of that half is told in the article on Spark New Zealand.

From the first day Chorus was designed to be a different kind of company. Its constitution and the legislation that enabled the split prohibit it from selling to households or businesses directly. It owes open-access obligations to every retailer on equal terms. Its shares were sold to investors as an infrastructure holding with long-lived assets and predictable income, the opposite of the high-margin integrated monopoly that Telecom had been in the 1990s.

Why did New Zealand force structural separation?

New Zealand forced separation because the government wanted to put public money into fibre and refused to hand it to a company that also competed in retail. The Ultra-Fast Broadband tender excluded vertically integrated bidders, so Telecom had to split in order to take part.

The background was a decade of frustration. Light regulation after the 1990 privatisation had produced high profits and slow broadband. Local loop unbundling was ordered in 2006 and operational separation in 2008, but both were behavioural remedies that needed constant policing. Structural separation removed the incentive to discriminate altogether: a network owner with no retail arm earns more when every retailer sells more.

Very few countries have gone this far. Britain kept Openreach inside BT, legally separate but still owned by the group. Australia chose the opposite route and built a state-owned network, the NBN, at far greater public cost. New Zealand’s version was a negotiated bargain. Telecom’s board accepted a break-up it had resisted for years because the alternative was watching a subsidised rival overbuild its copper network across the main cities.

How did the Ultra-Fast Broadband build work?

The UFB programme was a public-private partnership in which the Crown provided long-dated, low-cost financing and private partners built and owned the network. Chorus won the largest share, covering Auckland, Wellington and most other regions, while three smaller local fibre companies took the rest.

The first phase, announced in 2011, aimed to reach 75% of the population by 2019 with NZ$1.35bn of Crown investment, of which Chorus’s share was NZ$929m. The money was not a grant. The Crown subscribed for interest-free debt securities and equity securities in Chorus as each premise was passed, with repayment or conversion due many years later. Chorus signed construction contracts worth about NZ$1bn with Visionstream and Downer in April 2013.

Later extensions, known as UFB2 and UFB2+, lifted the target to 87% of the population. The rollout was completed in December 2022, covering 412 towns and cities across the country. Chorus now passes 1,499,000 addresses, about three-quarters of the national fibre footprint. The other partners, Enable in Christchurch, Tuatahi First Fibre in the central North Island and Northpower in Whangārei, are local monopolies in their own areas and do not compete with Chorus street by street.

The programme is widely regarded as one of the more successful public-private infrastructure projects in the country’s history, finishing roughly on time and at a public cost far below Australia’s. The comparison with less happy experiences is drawn in the article on New Zealand’s infrastructure deficit and PPPs.

How does Chorus make money?

Chorus earns a monthly wholesale charge for every active line. In FY26 it generated NZ$1,029m of revenue from 1.147 million fibre connections and a dwindling 44,000 copper lines, with average fibre revenue per user of NZ$59.51 a month.

Revenue growth comes from three sources. The first is uptake: 75.9% of the addresses Chorus passes now take a fibre service, and the company is targeting 80% by 2030. The second is price: wholesale prices rise each year, and the company has proposed increases from January 2027 that lift the mainstream 500Mbps home plan from NZ$58.73 to NZ$61.08. The third is mix: persuading households to move up to gigabit and multi-gigabit Hyperfibre plans, which had about 9,000 connections at June 2026.

Costs are low and falling. Operating expenses were NZ$303m in FY26, down 2%, and the company employs only about 745 people directly, with field work contracted to service companies. That produces an EBITDA margin of roughly 70%. The price of that margin is capital intensity: gross capital expenditure was NZ$375m in FY26, and the network carries NZ$3.2bn of debt.

How is Chorus regulated?

Chorus is regulated like an electricity lines company. Since 1 January 2022 the Commerce Commission has set a maximum allowable revenue for each regulatory period, calculated from a regulatory asset base, an allowed cost of capital, depreciation and forecast operating and capital spending.

The asset base is the number that matters most. At 31 December 2025 the core figure was NZ$5.1bn, plus a financial loss asset of NZ$0.862bn that compensates Chorus for the losses it incurred building the network before enough customers had connected. Together they give a total of about NZ$6.0bn on which the company may earn a return. The first regulatory period ran from 2022 to 2024; the second covers 2025 to 2028, and in setting it the Commission trimmed the company’s proposed capital programme from about NZ$1.5bn to about NZ$1.3bn.

Two further rules protect consumers. An anchor service, an entry-level fibre plan, must be offered at a capped price so that nobody is forced to pay for speed they do not want. And quality standards cover availability and fault restoration. The model will be familiar to readers of the article on Vector and regulated networks, and it shares its central tension with Auckland Airport’s regulatory settlement: the owner wants a larger asset base, the regulator wants only efficient spending to enter it.

What were the key turning points?

Four moments define the company: the 2011 demerger, a pricing dispute in 2012-2015 that nearly broke its finances, the completion of the fibre build in December 2022, and the move to utility-style regulation in the same year, which finally gave investors a predictable framework.

The crisis came early. In December 2012 the Commerce Commission proposed deep cuts to the regulated price of copper broadband, on which Chorus still depended for nearly all of its cash flow while it was spending heavily on fibre. The company warned of a funding gap of around NZ$1bn, suspended dividends and renegotiated the timing of its Crown financing. The share price fell heavily. Final copper prices were not settled until late 2015, and the episode taught both sides that an unfunded mandate to build a national network was not sustainable.

The recovery was gradual. Fibre uptake ran well ahead of original forecasts, helped by streaming video and by the pandemic, when home broadband became essential. Dividends resumed and grew. Leadership passed from founding chief executive Mark Ratcliffe to Kate McKenzie, then JB Rousselot, and now Mark Aue, with Mark Cross as chair. Aue describes FY26 as the start of a second strategic horizon to FY29, focused on growth, simplicity and efficiency.

Who owns Chorus and how is it financed?

Chorus is listed on the NZX and ASX and is owned by institutional and retail investors, many of them offshore infrastructure funds. No shareholder controls it, and its constitution restricts large holdings without government approval. Its balance sheet relies on bonds, bank debt and Crown financing.

Net debt was NZ$3,174m at June 2026, equal to 4.37 times EBITDA. Lenders agreed in 2026 to relax the senior debt covenant from 5.5 times to 6.75 times, giving more headroom, and the board targets a credit rating of at least BBB; Moody’s rates the company Baa2 and S&P rates it BBB with a positive outlook.

The Crown financing is the unusual element. On 7 August 2026 the Crown agreed to sell the securities it held in Chorus to institutional investors for NZ$702m, turning a state funding arrangement into a market instrument. The obligations themselves do not change: repayments fall due in three tranches at 30 June 2030, 2033 and 2036, with the largest at the end. Chorus must refinance or redeem them as they mature, which is one reason capital expenditure has been held down and why the dividend, 60 cents a share in FY26 with a minimum of 62 cents guided for FY27, is growing only gradually.

Who competes with Chorus?

Chorus has no rival fibre network in its own areas, so its competitors are other technologies: fixed wireless broadband sold by the mobile operators over 4G and 5G, low-earth-orbit satellite services such as Starlink in rural districts, and the legacy copper that it is itself retiring.

Fixed wireless is the most important. Spark, One NZ and 2degrees each earn a far better margin when a household takes broadband over their own mobile network instead of a fibre line rented from Chorus, so they promote it heavily to price-sensitive customers. This places Chorus in the odd position of depending on retailers whose commercial interest is partly to sell something else. Its response has been to raise entry-level speeds, upgrading more than 700,000 homes from 300Mbps to 500Mbps at no extra charge in June 2025, and to market fibre directly to consumers. The company cites research showing 66% of customers prefer fibre as their first choice against 12% for fixed wireless.

Outside the fibre footprint the contest is different. About 43,000 of the remaining copper connections are in areas with no fibre, and for those households the realistic alternatives are wireless or satellite. The mobile operators’ strategies are covered in the article on One NZ and 2degrees.

What do the latest numbers show?

FY26 was a year of steady utility growth. Revenue rose 1.5% to NZ$1,029m, EBITDA rose 3% to NZ$726m and net profit reached NZ$37m, up from NZ$4m, as fibre revenue growth of 6% outweighed the loss of copper income.

Fibre connections rose by 32,000 to 1,147,000. Average monthly data use per connection reached 731 gigabytes, up 9%, with uploads growing faster than downloads. Operating cash flow rose 4% to NZ$740m. Capital expenditure fell from NZ$415m to NZ$375m, split between NZ$205m of sustaining spend and NZ$170m of growth spend.

Copper is nearly gone. Connections fell by 48,000 in the year to 44,000, fewer than 1,000 of them in areas where fibre is available. Chorus has brought its target for full retirement of the copper network forward from 2030 to the end of 2028, and it now expects NZ$50m-70m of cumulative proceeds from recycling the recovered metal, having processed 817 tonnes in FY26. Guidance for FY27 is EBITDA of NZ$730m-760m and capital expenditure of NZ$375m-415m.

What can founders and CFOs learn from Chorus?

Chorus shows that the structure of a deal with government matters more than the headline subsidy. The company nearly failed when it had an obligation to build without certainty over revenue; it prospered once the financing, the regulation and the build obligations were aligned.

1. Never accept a build obligation without revenue certainty. The 2012-2015 copper pricing dispute arose because one arm of the state was funding fibre while another was cutting the cash flow that paid for it.
2. Patient capital changes what is possible. Interest-free Crown securities repayable from 2030 onward allowed a listed company to fund a network with a payback measured in decades.
3. Separation can create value for both halves. A pure infrastructure owner attracts a different, lower-cost investor base than an integrated operator, a point also made by the tower and data centre sales across the sector.
4. Retire legacy assets on a timetable. Running copper and fibre in parallel doubles maintenance costs. Setting a firm switch-off date, and bringing it forward, releases cash and management attention.
5. Match debt maturities to regulatory cycles. With a reset in 2028 and Crown repayments from 2030, treasury planning is as important as engineering.

For finance chiefs in other regulated sectors, the further lesson is to invest in the relationship with the regulator. Chorus’s disclosures on asset values, expenditure and quality are now extensive, and predictability on both sides has lowered its cost of capital.

What happens next for Chorus?

Chorus’s next phase is about finding growth once the build is finished: lifting uptake to 80%, selling faster plans, extending fibre beyond 87% of the population where funding allows, and adding new wholesale infrastructure services on top of the network.

The company has proposed spending about NZ$234m to reach roughly 40,000 additional rural premises, but only with regulatory and policy certainty that the investment will earn a fair return. New products launched in FY26 include data centre interconnection across eight locations, a national transport service and a precision timing service, and a feasibility study is under way for a new subsea cable between the North and South Islands. Trials of battery storage at network sites point to modest opportunities in energy.

The larger questions are financial and political. The 2028 regulatory reset will decide allowed returns for the following period, just before the first Crown-related repayment in 2030. Policymakers are also reviewing whether rules written for a copper monopoly, including the last copper-era obligations, still make sense. And with a stable asset base, a depressed share price relative to the wider market and infrastructure funds hungry for digital assets, a change of ownership cannot be ruled out, although any buyer would need government consent.

Frequently Asked Questions