Chorus Limited (CNU) Earnings Call Transcript & Summary

Good morning, everyone, and welcome to our full year results announcement for FY '20. I'm JB Rousselot, the Chorus CEO, and I'm joined by David Collins, our CFO. Because of the Level 3 restriction in Auckland, I'm coming to you via Zoom today, and David is joining me from our Wellington office. So today's webcast is also about showcasing the benefits of good quality broadband in these difficult times. I'll start this morning with an overview of the key numbers and trends for FY '20. David will then take you through the financials and the guidance for FY '21 as well as a regulatory update. I'll then talk about some of the new dynamics we're seeing in the broadband market and our strategic focus for FY '21, and then we'll go to Q&A. So if we look at the highlights for the year, first and foremost is that despite the operational restrictions and the financial impact of COVID-19, we achieved our goal of modest EBITDA growth. This reflected our ongoing focus on reducing costs and on reshaping our business to a fiber-centric future. We've reported EBITDA of $648 million for FY '20, up from $636 million last year. That's despite a $12 million EBITDA impact from COVID-19, which David will talk through in more detail shortly. Net profit after tax was $52 million compared to $53 million last year. Those results mean that we've confirmed a fully imputed final dividend of $0.14, bringing the total dividend for the year to $0.24 per share. We delivered another strong year of broadband connection growth, with an increase of 10,000 lines. We finished the year with 1,415,000 fixed line connections, down just 35,000 lines from the year before. And this is a marked improvement on the level of line reduction experienced in prior years. We've continued to see strong growth in fiber uptake during the year, up from 53% to 63% across the completed rollout footprint. And just as importantly, customer satisfaction with fiber installation rose from 7.7 to 8.1 out of 10, which exceeded our target of 7.9. Our staff number reduced from 918 to 870 as we moved from a rollout to a more operational focus. But despite that shift, employee engagement was 8.5 in June 2020, up from 7.6 out of 10 in FY '19. And our employee Net Promoter Score rose from 28 to 67, putting us in the top 5% of our international technology company benchmark. Now this time a year ago, we were investing significant time and money into adding to our network capacity in preparation for online streaming of the Rugby World Cup. We thought that, that event would be our big test of FY '20, and our network cope very well with the record bandwidth demand that we saw during some of the matches, though that we know that it was just a dress rehearsal for COVID-19. In late March, just like other countries, we saw New Zealand daily Internet behavior changed radically, with everyone forced to work, learn and connect from home for about a month, unlimited bandwidth suddenly became essential for consumers. Our daytime traffic more than doubled, and we saw evening peak bandwidth traffic grow by about 35%. And this increase would have been approximately 10% higher if global streaming services such as Netflix and YouTube hadn't limited the bandwidth requirements of their streaming services. And to be clear, these measures from streaming services were to assist other operators with capacity constraints on network small point to congestion. They were no way needed for our network because of the significant extra capacity that we have available. As a result of lockdown, average monthly data usage grew from 303 gigabytes in February to 406 gigabytes in April. Behind the scenes, our people did a fantastic job. They kept our essential infrastructure going, all while embracing working from home in an extremely short time frame. They also helped implement a range of COVID-19 initiatives to support our broader industry and the community. Our immediate focus was on supporting retailers to provision additional capacity across their networks and on maintaining the network while providing essential connections throughout the lockdown. Workers' welfare was a priority for us. We work closely with our service company partners to develop clear guidelines for work that could be undertaken and kept customers informed of our focus on social distancing practices throughout the different alert levels. We also sourced and provided additional personal protection kits to service companies. The rapid suspension of nonessential field activity during lockdown stopped our UFB build and placed significant financial strain on our service companies, and we chose to provide $5 million in financial support to them and their subcontractors. This help assist with reduced workflow and retained our workforce so that we could rapidly restart activity when alert levels were relaxed. A relief fund of $2 million was also made available to retailers to help address the expected increases in debt from consumers and small businesses that were unable to pay their bills during the lockdown. We removed some of our fees to enable temporary business disconnections. And when the Ministry of Education identified critical households needing a broadband connection for students to keep up with their schooling, we made our existing network connections available free of charge for a 6 months period. And by the end of FY '20, we had connected about 10,000 homes under this initiative. We introduced price reductions on our 1-gig product and on our small business plans from 1 July, and we also delayed our annual CPI increase until 1 October. And of course, we continued to invest in broadband capacity as network demands continued to grow in this environment. We're very pleased in the way we've been able to play a part in supporting the wider community. The health and well-being of our people has also been a big focus. During lockdown, we provided a range of initiatives, including online webinars on remote working, financial learning and mental well-being. And the later included providing the Mentemia app fronted by Sir John Kirwan. Our team really appreciated the support, and they rated Chorus 9.3 out of 10 for supporting their physical and mental well-being swept the pandemic. As I said, the lockdown period had a substantial impact on field activity. That's clearly shown in this chart with fiber installation plummeting during that time. The UFB2 rollout was also paused. In consumer-facing activity like our managed migrations program was impacted by social distancing requirements throughout lockdown and throughout Alert Level 3. This meant that fiber connection growth was affected, down from 32,000 in Q3 to 27,000 in Q4. And we revised our CapEx guidance downward based on what we were seeing at the start of the lockdown. Happily, though, the UFB2 rollout was able to restart faster than we'd expected in the second part of Q4 and so did our volumes of installations and connections. As the slide shows, we saw a strong V-shaped rebound in installation activity, and July installation levels were back at around 16,000. By comparison, the recent return to Alert Level 3 in Auckland has not restricted our build, our connections or our maintenance activities. Despite COVID-19 restriction and the ongoing reduction in field cruise, we completed 167,000 fiber installation in the year, which was slightly down from 186,000 in FY '19. The investment we made in prior years to streamline our processes and introduce digital tools for fiber in a day have really paid off. We lifted customer satisfaction for all fiber installations from 7.7 to 8.1 out of 10, which was above our target of 7.9. And customer satisfaction with technicians has also risen from 8.5 to 8.7 out of 10. Customers are getting connected quickly with lead times in June this year, down to 8 business days compared to 13 days last year. Our managed migration campaigns proved very successful in stimulating additional fiber demand throughout the year. We had 32,000 customers take-up of fiber installation and we'll continue to see about half of managed migration customers activate a fiber service within just 6 months. With the UFB1 rollout completed in the first half and UFB2 fast approaching the halfway mark, we're about 88% of the way through the UFB build program. This means that there are approximately 150,000 homes and businesses remaining to pass in the UFB rollout area by the end of 2022. The socioeconomic benefits of the rollouts are huge, especially when you look at the size and location of the communities we're now connecting. And as this slide shows, uptake in UFB2 areas where we've completed the rollout is already at 37%, while in UFB1 areas, it is at 63%. We saw a significant shift in fiber product -- sorry, in fiber product mix in the FY '20, with uptake of 1 gigabit plans almost doubling from 58,000 to 115,000. So gigabit connections grow from 10% to 16% of mass market plans. The shift will continue to be helped by our drop in gigabit pricing from $60 to $56 per month from July and the market incentives that we put in place for retailers. The demand for 1 gig connections now represents about 25% of our recent fiber orders. And as I mentioned before, we've lowered our -- some of our prices from July and delayed our CPI increases until October, and some of those price changes are shown on the chart. When we look at connection movements across our 3 network zones, you can see that we've had strong broadband growth within our planned UFB footprint. Broadband connections there were up by 34,000, while copper voice-only connections reduced by 29,000. Wireless competition has driven some of that line offs. And in the last month, we've seen mobile networks operators seeking to reduce their own costs by pushing more of their low-usage copper line customers to fixed wireless. In the rural zone, where UFB fiber is not planned buyers or other fiber companies, connections have been largely stable. Wireless competition is increasing in the zone as a result of government-funded wireless networks. We've seen some reduction in copper voice and ADSL lines, some of it to wireless networks, but we've also seen growth in VDSL and fiber connections. And that fiber growth is mostly greenfield's development, just outside the UFB zone, and we now have about 20,000 fiber connection in this zone. Finally, the local fiber company, or LFC zone, is where we continue to see consistent copper voice and broadband line loss. As you can see, there was some moderation in the rate of line loss through Q4 as a result of COVID-19 restrictions, but that has since returned to the pre-lockdown level. This last slide gives you a sense of the progress we've made in migrating consumers from copper to broadband -- or copper broadband to fiber throughout the year. The chart on the left shows the broadband mix within our planned UFB zone. Of the 980,000 broadband connections, fiber now represents 73%, up from 61% at the start of FY '20. In our biggest market, Auckland, fiber now makes about 82% of our broadband connections, and that's up from 73% in June last year. I'll tell you more shortly about our plans to really drive fiber migration in FY '21. But first, I'll let David talk through the FY '20 numbers in more detail.

Thank you, JB. Good morning, everybody. It's great to be with you today. Starting with our income statement. We're very pleased to report EBITDA of $648 million, a growth of $12 million against the prior year. As JB mentioned, we were impacted by COVID by $12 million split across expenses and revenue. And you can see the split on the slide. In terms of depreciation, we see depreciation continue to grow in line with the CapEx growth in our assets, slightly offset by a reduction in amortization of software assets. Looking at our interest line. Interest costs have grown year-on-year. The main driver of that is the growth in our crown funding and, therefore, CIP interest. With regard to external debt, we've seen a lower rate during the year following the refinance of our GBP bond in April, but average external debt level is slightly higher over the year, which has mostly impacted or offset the rate impact. And lastly, on the income tax expense, we've seen a favorable impact in the year due to the reinstatement of building depreciation, which IID has announced, which provided us with a one-off benefit on the income tax line. Moving to revenue. Revenue is down $11 million year-on-year to $959 million. The COVID impact, as I mentioned on the previous slide, was $3 million. Other than that, the key trends that you can see are consistent with prior years and that is an ongoing growth in fiber revenue. Voice lines continuing to decline. And copper broadband as a result of migrations to fiber continuing to decline also. Field services revenue is also with calling out. We have seen lower network relocations across the period, which has impacted the field revenue line a 0 impact on EBITDA due to the cost offset. Moving on to costs. Our cost outcome for the year was very pleasing. We've come in at $311 million, being a $23 million reduction on the prior year. A couple of the key elements to call out. Firstly, labor. Our FTEs are actually down 5% year-on-year and that's mostly as a result of the customer network operations reorganization, which occurred in late calendar 2019, which we spoke about at the half year. The key thing to note there is, most of those employees are actually capitalized. So the impact of that is more a free cash flow benefit rather than a benefit on the OpEx line. It's also true that labor was impacted by COVID. During the period, we were able to capitalize a lower level of our employee costs as a result of the COVID shutdowns. And we have also invested in our regulatory capability. I've more to say about that a little later, which has impacted our labor cost line also. Maintenance is another one that's worth calling out. We do see a continuation of the reduction in maintenance costs, which you've seen also in prior years. It's mostly driven by a continuing reduction in our copper lines across our business, which we think is a very pleasing trend for us. Two other cost items to call out. On the regulatory line, we have seen a significant saving year-on-year, which was a result of the reduction in the telecommunications development levy announced by the government earlier in the year. That has benefited us by some $8 million versus the prior year. And lastly, our IT costs continue to fall as we look to exit legacy and shared systems and look to invest in new systems going forward. So a very pleasing outcome in terms of expenses for the year. Looking at reactive maintenance, in particular. This is the traditional slide that we've shown now for a few of the last few results, releases. The key trends to see, and it shows half-on-half for reactive spend and both fiber and copper on the top right-hand side and then just copper across geographies down the bottom right-hand side. The 2 key trends here are the ongoing reduction in our copper variable maintenance cost, which we expect to continue. And secondly, within the UFB area, specifically, our maintenance cost for copper, in particular, does continue to fall. There was a slight COVID impact in terms of the lockdown period, but the overall trends are consistent with what we've seen in prior periods. Moving on to CapEx. We've reported gross CapEx at $663 million for the year, marginally above the range that we had announced earlier this calendar year of $610 million to $650 million. It's a good news story. The reason we're above range is the faster than expected ramp-up post the lockdown period, particularly for UFB2 communal, but also for connections. JB mentioned the V-shaped recovery, which we saw that was very pleasing and that faster ramp-up did result in a higher CapEx outcome for the year than we had anticipated when we set the guidance. In terms of the cost per premises passed for the communal build, that's come in at $1,558, right in the middle of our range of $1,500 to $1,600. The other key item to call out on the page is the communal spend. Communal now is all UFB2 and 2 plus looking forward. In the current year, full year '20 UFB1 finished in December of 2019. So that line will continue to reduce as UFB2 and 2 plus complete by December of 2022. Moving on and looking at connections in a little more detail. As I mentioned before, we spent $282 million on connections across our business. Our cost per premise is connected, which is a key metric that we monitor for the rate at which we pay for connections has come in at the bottom end of our range of $1,000 to $1,100 coming in at $1,022 per connection. We've invested a little more in capacity in layer 2, which is the $31 million. And backbone builds, you can see at the bottom of the slide at $68 million over the period. Moving on to common and copper CapEx. Copper continues to decline, which is a trend that we will see in the future and is a continuation of previous trends. Copper spend has come in at $55 million. As connections continue to decline, the flow-on impact is a reduction in copper CapEx spend. In terms of common CapEx, that's fairly stable. It's at $60 million for full year 2020, mostly IT spend around modernization of our systems and also looking to improve the provisioning and assure experience of our customers, which is always a key focus area for us. Moving on to guidance. Starting with EBITDA guidance, we are providing guidance for full year '21 at a range of $640 million to $660 million. It is important to note that we have made an allowance for COVID in setting this range of $10 million. The reason we've done this is we can see uncertainty in 2 main areas: firstly, at a macro level, the ongoing economic impacts of COVID on the broader New Zealand economy are difficult to predict, and we see a lot of uncertainty there. And secondly, at a more granular level, the question of how many more lockdowns will there be? How long will they go? And what level will they be? Nobody knows the answer to that. And of course, as we sit here now, we can see Level 3 lockdown in Auckland and Level 2 where I'm sitting in Wellington. In terms of the dissection of the $10 million, it's very difficult to be specific about it. To give you a generic idea of where that might eventuate, looking at full year '21, on the revenue line, if the economic impacts are severe, we could see lower business revenue or lower greenfields revenue. We could also see greater price elasticity for end consumers. That is a greater concern about price for the products that they pay. On the cost front, what could we see? We could see similar to what we've seen in full year '20, higher level of inefficient labor, we could also see higher network or circa costs if there were longer lockdown periods. So $10 million, we think is an appropriate allowance, and we will monitor that as the year goes by. In terms of CapEx guidance, we provide guidance at a range of $630 million to $670 million for full year '21. As JB mentioned, we are doubling down on our efforts to accelerate uptake of fiber connections across our business. And we, therefore, are guiding for connection CapEx to a range of $275 million to $295 million. It's important when you look at our overall CapEx guidance that it does include circa $30 million for the West Coast fiber build. The West Coast Fiber build is a government-funded or grant-funded project to roll out some 270 kilometers of fiber cable on the West Coast of the South Island of New Zealand to provide better connectivity experience for New Zealanders. It's partly a resilience based project, and there are some new connections that are related to it. The important point to note is that it is predominantly grant funded, and there is approximately $30 million in relation to that project as part of our CapEx guidance for full year '21. The last important point to call out on CapEx guidance is copper. As you would expect, we see copper CapEx continuing to fall looking forward into full year '21 and, hence, we have guided to a range of $35 million to $55 million. A couple of comments on UFB2 communal. We are now a fair way through the build program. It's going very well. We thought it would be appropriate to provide an update on the total program level guidance. We had previous guidance in market of $505 million to $565 million. We have narrowed that range now to $548 million to $568 million. It reflects a growth in scope of the project. And most of that increase in scope is supported by Crown funding, which I'll come to a little bit later in the discussion. In terms of cost per premises connected for UFB2, we have now moved to providing annual guidance. We had previously given a program level view or guidance for cost per premises connected. Given that it was based in 2017 dollars and had a few caveats attached to it, we felt that it would be more useful for our investors and end users to provide annual guidance for CPPC, which is what we'll be doing for UFB2 going forward. Moving on to dividend. Starting with full year '20, our final dividend, as JB mentioned, we are announcing $0.14 per share fully imputed, which will lift the full year dividend to $0.24, in line with the guidance we had previously provided to market. We will have the dividend reinvestment plan in place, and it will have a 2% discount attached to it. In terms of full year '21 dividend guidance, we are guiding to $0.25 per share, which is a continuation of the modest dividend growth, which we have had in place now for a few years and is the last year before we move to our new dividend policy, which I'll talk about in just a moment's time. So moving on to our new dividend policy. We announced in February of this year that we would transition to a free cash flow based policy from full year '22. I'm pleased to report today that there is no change to that plan, and our intention remains exactly the same. We've defined what the policy is on the slide, which is consistent. On the right-hand side of the slide, you can see our forward trajectory for both UFB2 communal and for connections CapEx. And you can see the trajectory is quite clear looking past full year '21 in terms of the downward trajectory. Two comments to reiterate on our new dividend policy, which I spoke a lot about at the half year. But it's worth reiterating again just in this conversation. We are very focused by our credit rating and we are committed, as we have said previously, to maintaining our existing rating at BBB, Baa2. What that does mean over the next couple of years is that we will be constrained in the full implementation of our new dividend policy, whilst we have elevated levels of communal CapEx, which runs until December of '22 and connections CapEx, particularly, in full year '21 and '22. So we do talk about a transition into our new policy as we did at the half and that's the reason that we talk about transition due to communal and connections CapEx. Moving on to sustaining CapEx. We thought it would be useful for investors and users to give an illustration of what sustaining CapEx would look like in the current year just gone based on our future dividend policy. And the outcome is that $186 million of our full year '20 CapEx spend would have met the definition of sustaining CapEx under our future dividend policy. A couple of points to note about this. We do expect looking forward that our fiber sustaining CapEx will grow over time. That's natural. The asset is quite new at present. It will age over time. So we do expect a slight increase in fiber-sustaining capital. But as we've said to market previously, we think that a reasonable midpoint estimate for sustaining CapEx going forward would be around about $200 million per annum. Moving on to our capital allocation framework. We thought it would be helpful for our investors, given that we've announced what our future dividend policy will be from full year '22, just to give a couple of comments around how we think about capital allocation in that context and in that framework. The answer is it's very simple and easy to understand allocation framework. We have a shareholder value lens when we look at how we allocate capital. So when we have surplus capital in the future, which would come after our dividend distribution in accordance with our new policy and after sustaining CapEx, our decision process around how we use that capital would be through a shareholder value lens. So whether that led to share buybacks in the future, additional dividends or discretionary CapEx and we've defined what discretionary CapEx might be through the asterisk at the bottom of the slide, the key determinant would be shareholder value. So to describe that differently, we would look at discretionary CapEx in the future only where it demonstrably provided additional shareholder value over capital management activities. A comment about credit ratings, specifically. I mentioned before, we're committed to our existing rating. We are, however, also talking to our rating agencies around potentially lifting our thresholds to reflect our transition into our new regulated regime and the change in the underlying risk profile of the business as a result of the completion of the build and the resulting cash flow growth that will flow from that. Moody's specifically have said in their most recent public announcement that they will lift our thresholds once they have certainty over the price quality regime outcome, which I would expect to be mid- to late calendar 2022 -- sorry, 2021, reflecting the transition to the new regulatory regime. Moving on to net debt to EBITDA. This is the usual slide providing details of our debt levels. Our ratio with S&P is at 4.12 -- sorry, 4.14x against a metric threshold of 4.25. A reminder, our covenants are at 4.75x. And we also note Moody's threshold on the slide at 4.2x. Moving on to Crown financing. The same slide that we've shown you in previous presentations. We have a capacity or a facility limit of $1.33 billion. We are drawn now to $1.067 billion, and we'll continue to draw down against Crown funding as we complete the communal build for UFB2 and UFB2+. You can see on the right-hand side of the screen, our financing facilities that are in place. A couple of key points to note there, we did have a significant refinancing during full year '20, which was the repayment of the GBP bond in April of this year and we had a significant financing raise in December of 2019, which is a EUR 300 million bond, which you can see at 2026 on the screen. Our key refinancing requirements looking forward are, firstly, a NZD 400 million bond in May of 2021, which we will look to refinance in the debt capital markets well ahead of that date. One other comment to note in terms of our Crown funding, we have included a note on this issue in our financial statements, we have agreed an extension with the Crown for the dividend dates, which apply to future dated Crown equity securities. So circa $90 million of future equity securities. We've agreed an extension of the dividend dates that attach to those via weighted average of 4 years. And that was the result of Chorus having funded nonstandard installations over the period from 2016 and the outcome or the result of the delay in the implementation of the RAB regime until 1 January of '22. So there is a note that gives a little more detail in the financial statements. All right. Lastly, for me, moving on to regulation. Always plenty to talk about on the regulatory front. We've got a couple of items to call out on the next slide. Firstly, Brett, the next slide, please. In terms of the implementation methodologies, we've had a couple of key dates over the last months. Firstly, in terms of the input methodologies draft decision, that was made in November of last year, and we've had a few announcements over the last couple of weeks. We wanted to call out a few of those in our discussion now. Firstly, in terms of the financial loss asset, the commission has announced a change in approach from a building block model to a discounted cash flow approach. We see that as value neutral, and that is the inputs into a discounted cash flow approach. If they are the same as the BBM approach, they should give the same outcome. In terms of the pre-2022 WACC, the commission has proposed moving to a 5-year risk-free rate calculated each year within the loss asset period to calculate the average WACC over that loss period. Our view in terms of the risk-free rate is that we took on risk at the commencement of the UFB build in 2011. And like most infrastructure projects, that is the point at which financing decisions were made and interest and costs were locked in, and I'll talk a little bit more about that in a couple of moments' time. In terms of Crown financing, the commission has reverted to their previous view, which was that Crown financing is, in essence, costless. The reality for us is that Crown financing was far from costless. It attracted onerous terms and conditions under the terms of the UFB funding agreement, and we saw the outcome of some of those risks over the UFB build period, which I'll talk about in a couple of moments' time. We built ahead of demand. There was no certainty of the demand for our asset when we built it. And the reality is that our investors took on significant risk upfront, which we believe should be better reflected through the regulatory regime. Moving on to a view of what our cost of debt is versus some of the implied or the implied WACC outcome from what the commission has said to date. The reality for us is that there's a significant gap between what the commission's current views are and what the Chorus' actual costs have been over the period of the loss asset. You look at the chart on the right-hand side of the screen, you can see Chorus' actual cost of debt, which are the black bar charts compared with, firstly, what the commission is proposing as our cost of debt allowance, which is the orange line, quite significantly below our actual cost of debt. And secondly, the implied vanilla WACC within what the commission has said to date. And when you look at the slide, you'll see that the implied vanilla WACC is basically in line with our actual cost of debt. We don't think that's an appropriate outcome. Clearly, the cost of equity would be higher than our cost of debt. Hence, the implied vanilla WACC should be higher. Like any infrastructure program, when we took on risk at the commencement of the program, that's when we locked in our financing as best we could. Any commercial organization would look to do that and would also look to look -- lock in debt from as longer-term perspective or as long a tenure as is possible. The outcome of some of the risks we faced over the loss asset period, you can see in the red box in the middle of the chart. For a 2-year period, Chorus was unable to pay dividends and was unable to raise debt as a result of the regulatory volatility we saw during those 2 years. That's a great example of the risks that our investors actually faced during the loss asset period. Our investment-grade credit rating was under threat back in 2014 and 2015. Suggestions within the commission's current views that BBB+ is an appropriate credit rating ignores both our current rating, which is BBB, ignores the minimum required rating under the UFB funding agreement, which is BBB- and ignores the fact that our rating was at threat over 2014 and 2015. We also look at comparators when we think about what's the appropriate WACC for Chorus over this period of time. Ofcom in the U.K. is a very relevant comparator. And they've provided a view on a proposed WACC of 6% for fiber to the home for Openreach, which we think is a much more appropriate and relevant comparator for our business as we look at our returns going forward. So in short, the current views that the commission has, we think, removes even the small margin that government had acknowledged upfront should be appropriate for this project. And we don't think that's a fair bet for our business. We will continue to make submissions on these issues and our investors, we're also encouraged to do so. The last slide for me, we thought it would be useful just to outline some of the key regulatory milestones or steps as we look forward. The next key date is in October, which is the input methodologies' final decision. That's the 13th. In November, we'll have a final decision on the financial loss asset consultation. That paper having been released just a few weeks ago on August 13. And then importantly, we move into price quality through calendar 2021, where the commission will give views on RAB, WACC, maximum allowable revenue. So all of the key outputs, both through a -- firstly, a draft decision in mid '21 and a final decision in late '21. So thank you. I'll hand now back to JB.

Thank you very much, David. Now let's talk about trends and strategic focus. COVID-19 has accelerated some of the positive underlying trends that support our business. International telecommunication providers talk of seeing a bit of a fixed line renaissance with consumers now placing even greater value on the reliability and the unlimited capacity of fixed line networks and especially on fiber networks when they can access them. In recent months, we've seen more and more European network operators committing to rolling out more fiber. In the U.K., for example, BT has said that it will take fiber to 20 million customers by the late to 2020s. Now let's move to New Zealand. New Zealanders made the most of their fiber during the lockdown in April with fiber usage averaging almost 500 gigabytes. We've seen that again last week when Auckland moved to Level 3 with daytime traffic in the Auckland area surging by about 85%. In the past, we've tended to focus on downstream usage, as shown in the chart on the left. Although overall traffic had reverted to some kind of normal in June, we're still clearly on the trajectory of strong long-term growth for download traffic. But one of the significant changes from COVID-19 is that consumers now have a greater appreciation for the value of upstream capacity and upstream performance. And the chart on the right shows national data usage with upstream traffic included. In April, during the nationwide lockdown, upstream traffic peaked about 60% higher than January levels. And that was driven by the rapid adoption of video conferencing applications like Zoom and Microsoft Teams, which generate more two-way traffic. And even after the lockdown ended, the upstream traffic had been about 10% higher than pre-lockdown levels because of the shift to people spending more days working from home. This is one of our favorite charts because it shows so clearly the way our lives and upstream traffics have changed as a result of COVID-19. The blue lines show pre-COVID-19 upstream traffic throughout the course of the day. Then when we went into lockdown, the green lines, daytime traffic increased significantly and we see what our people call the Godzilla or what I call the Zoom effect. And that's the staggered increase and decreasing traffic roughly under half hour that coincides with people starting video conferencing services for meetings. That trend was still evident post lockdown, and it has returned with Level 3 conditionings in Auckland in the last few weeks. Now our research shows that post lockdown, people estimate that they'll spend about 43% of their time working remotely. And that's about double what we had seen pre-lockdown levels. So we see the rapid acceptance of working from home as a very positive trend for our business. And as well, it is driving greater data usage and it means there is a greater need for reliable broadband at home. That's both from the perspective of employees, because they need to carry out their work, and employers who want their people to be just as productive at home as they are in the office. And our research shows that almost 3/4 of employees consider high speed broadband a must have when they're working from home or remotely. We also saw a substantial increase in the proportion of people planning to upgrade their broadband. In April, during the lockdown, just 15% planned to upgrade. By June, that had increased to 39%. And then finally, although about 70% of businesses say that they already have policies to support remote working, only 14% say that they currently provide an Internet connection or a plan for their employees. We expect this to increase as more employers encourage remote working options. At the same time, as consumers are placing higher value on reliable broadband, the Commerce Commission continues to highlight the strong performance of fixed line broadband relative to other technologies. They note the reliability of fixed line copper and fiber services throughout the lockdown period, while average download speed of fixed wireless decreased by about 25%. The commission's report also points out the comparative performance of fixed wireless when it comes to key features like latency. And for the consumers, latency turns up as web page loading very slowly or bit by bit or buffering-type effects when you're streaming or you're video conferencing. Our own consumer perception survey picked up on this trend over the last 9 months. About 70% of respondents are rating fiber at an 8, 9 or 10 out of 10 for Internet satisfaction. By comparison, we've seen fixed line wireless respondents show a strong decline in satisfaction. And this is all consistent with the communications of the Minister's announcement several weeks ago that fiber was the gold standard for broadband. Now our own wording for that is a bit more modest. Our latest campaigning words, it's slightly differently, and we call fiber the way we internet now. Because for us, if you want reliability, capacity and speed performance, there is no question that fiber is the best technology, and that's why international networks operators are following the New Zealand UFB example. Now with the UFB rollout running down, our focus is all about connecting more New Zealanders to fiber. This is our strategy on the page, and we've put it at the top of our strategic focus with a target of 1 million fiber connections to the Chorus networks in 2022. It's a substantial step-up from the 750,000 connections that we have today. But we know that the socioeconomic benefits of fiber connective -- connectivity make it all really worthwhile. Now the purpose of our business remains to make New Zealand better by bridging digital device and by enabling work, education and creativity through better broadband. As you can see from the slide, the 4 key strategic priorities underpinning this purpose are the same as the ones that I talked about at the half year results. First and foremost is winning in our core fiber business and optimizing our nonfiber assets, growing new revenues and developing the long-term future of our business. As I said in February, winning in fiber means getting more of the homes and businesses that haven't connected on the fiber and higher speed plans. We got that number down from 44% in February to 40% in June, and we're going to push even harder in the coming months. As the fiber rollout starts to slow down, we no longer are getting as much of a natural demand tailwind from just the release of new addresses. And so that means that we need to reach the connections intensity across our existing footprint to meet our FY '21 installation forecast of 145,000 to 165,000. Now to do that, we're focusing on 4 broadband residential markets. First, there are roughly 400,000 addresses in the fiber footprint where fiber hasn't yet been installed. Second, there are roughly 80,000 addresses where fiber is installed, but not activated. About 50% of fiber orders now come from these intact addresses. Others who are campaigns or from people moving homes or providers. Third, we're going to pass another 80,000 UFB addresses in FY '21. We've historically seen about a 30% uptake within 6 months of fiber being made available. But in some of these smaller communities, we're seeing exceptional uptake through door-to-door marketing and through word of mouth. In Whatawhata, for example, which is a suburb near Hamilton, we've already achieved 80%. And fourth and finally, greenfield means that we have an existing contracted pipeline of about 25,000 properties, and about 20% of that growth is outside of the UFB analytics results. One of the key ways we can help lift connection intensity is by making the most of our active wholesaling strategy. As an open access wholesaler, we provide a level playing field for almost 100 retailers. Each has their own marketing proposition and target market. The smaller retailers, in particular, are keen to grow their market share, and you can see that in the evolution of the broadband market on this slide. The IDC data shows how the smaller retailers are growing their share of the broadband market with the transition to fiber. Electricity retailers, for example, have been a key part of this growth. Trustpower has become the fifth largest broadband retailer and Contact Energy has reported that it has grown to almost 30,000 customers in a fairly short time frame. We've also seen Vocus start to bundle power with broadband. And the change continues. With Sky TV confirming a few months ago that it is entering the broadband market and their combination of popular content, but also a large existing customer base, could be very powerful in converting some of those late adopters to fiber. As part of our FY '21 initiatives, we've just launched a new line of retailer incentives. This continue to have volume hurdles relative to each retailer's past performance, and we provide credits based on the specific customer segment and plan types. At a high level, this can provide up to $300 in credits for copper late adopters, for example, or up to $500 for the win back of an off-net connection. We also have our own advertising and migration campaigns to drive further uptake. Our managed migration programs had a huge month of July. It achieved more than 5,000 installations, and that's well above our previous average, which was about 3,000 a month. By visiting neighborhoods, we're helping consumers get over the perceived hurdle of having to organizing installations with their retailer. And not only does this activity clearly lift uptakes compared to areas where we haven't door knocked. It also is helping us win back connections from other networks. The proportion of installation orders from off-net customers was around 50% in June, which is up from 35% a year ago. A week ago, we launched our new advertising campaign, providing fiber broadband as the solution to what we're calling bad net. This campaign highlights some of the reliability issues that I've talked about earlier, like slow downloading of web pages or as you can see on the slide, the issue of blitz face when you're video conferencing or poor upstreaming or upstream connection. And I'm sure you've all seen that in terms of the video conferencing that you've attended. Wellington also remains an obvious area of focus for us. Legacy cable networks in Wellington and on the Kapiti Coast means that we have a lower historic market share on copper broadband. The good news is that, as this slide shows, we continue to make good progress on gaining customers in those geographies. Now we're not just focused on fiber for the retail market. The past year has seen some significant changes in our business portfolio and some significant customer wins. As you know, we launched our 2 and 4 gigabit Hyperfibre plans earlier this year. This was initially in a limited number of centers, and we're looking to expand this footprint further. Uptake is growing gradually, but that's no surprise given that we started with a limited initial footprint and also given the retail pricing around $200 in a month. But what's interesting is that even at that level, there are some residential customers who see good value in that service. In February, we talked about our small business plans where we offer enhanced restoration times. And the good thing is that we started to see good pickup in the rate of adoption of these plans with uptake doubling to 3,000 connections in the second half. A reduction in monthly pricing for this product from $55 to $52 from July should give these plans an added boost with more retailers starting to promote them. We've also enhanced our premium business services with a new proactive fault response where we flagged and troubleshoot host with retailers as we see them out here. And at a broader network level, we're continuing to strengthen our role as a neutral network host. During the year, we reached an agreement with the Rural Connectivity Group, the RCG, to provide backhaul for their new rural wireless sites. We're expanding our edge center colocation site with a new site plan in one of our Tauranga exchanges. We developed a national Tel extension service. So it's easier for retailers to expand their fiber services nationwide. This means that they only need 1 network handover point if they want, greatly reducing costs between multiple UFB areas and enhancing their ability to grow at geographically dispersed customer base. Related to this, smaller retailers have complained about barriers to interconnection for some point. We're planning to launch a new peering service in the first half of FY '21, in conjunction with the New Zealand Internet exchange. With many retailers already in our exchanges, we see greater interconnectivity between service providers as a really important stepping stone to increase cloud and other content traffic over time. And the good news is that, that will, in turn, complement our edge center and backhaul services. Now we're continuing to explore opportunities to grow new revenues. We're starting to see significant interest for smart location connectivity from organizations like Auckland Transport. Smart locations are things like CCTV cameras, traffic lights and digital advertising sites. Often, these will be switching from existing copper or mobile connections to take advantage of fiber's higher and more resistant bandwidth. We've refined our processes for handling these connections as they often require complex installs. We've also introduced a new smaller ONT, the size of the flash drive, that enables connection on places like lamppost and traffic lights, and our symmetrical 50 megabit plan for these services wholesales at $50 a month. Our innovation program helped identify and develop the new Wi-Fi service that we are launching soon using our latest Wi-Fi ONTs, like the one on this picture, the new service will have a monthly fee of $1.30. The consumer benefits of this new service are significant. First, this could reduce the retail cost of broadband for short-time customer connections because retailers would no longer need to recover the cost of a router on a short time frame. And secondly, if our Wi-Fi device is already in a home, it means that customers can get their broadband up and running almost straight away. When they move premises or when they switch provider, they won't have to wait for a retailer to deliver the router. Now to help retailers transition their systems to this new service, we're going to provide credits for initial sign-ups. So this is unlikely to generate significant meaningful revenues in the short term, but will in the longer term. We have more ideas in the pipeline, including exploring the opportunities created by the latest Wi-Fi standard, Wi-Fi 6. Wi-Fi 6 is seen internationally as a great potential alternative to 5G in enterprise and other private environments where cost-effective capacity and support for a large number of devices is important. An example of this would be places like airports, stadiums or campus environment. Now as our network becomes increasingly fiber centric, we're looking very closely at the assets and the maintenance programs we need to deliver services into the future. We've already began to dispose of nonintangible network sites. For example, we sold a site in New Plymouth township, which is another fiber company's area, and we've exited some rural buildings where we -- that we no longer require. One of those was the Whangarei Heads site pictures on this slide. Now that site was hit by a car. So we chose to shift the network equipment into a cabinet rather than repair the building. And changes like these will help us realize ongoing maintenance and incidental cost reductions for those buildings. We're also running the ruler over the way we approach productive maintenance on the copper network in areas where we have higher fiber penetration. And with the Commission due to release its couple withdrawal code in September, we've been thinking about our approach to starting to switch some of our copper network off. For example, today, we have about 20 copper broadband cabinets with no customers on them. And another 30 or so have less than 10 customers in the cabin. Now as the heading under this slide suggests, we don't see the withdrawal of copper service being a quick big bang, like the switchover from analogue to digital TV broadcast, for example. It will be a much more gradual cabinet by cabinet process. There'll be lengthy notice requirements that we'll need to follow. The Commission's drive code currently suggests a 6 months' notice requirement. And the chart on the right shows that we have exchanges in Wellington and Queenstown, where fiber uptake is now more than 90% of our broadband connection. Now we know some retailers are seeking to switch consumer to alternative wireless networks immediately because of what tends to be portrayed as our pending shutdown of copper. And that's not necessarily correct, given our own migration time frame. We're going to take a very careful, considered and consumer-centric approach to any copper shutdown. Now as I've mentioned already, there are a range of providers who continue to use our network to provide services over both copper and fiber. And everybody who's worked in this industry knows that when you wake up a customer with a force change, they also tend to shop around. And we will strongly suggest that they should consider moving to fiber at that time. Now the changes to the way we operate our network and the shift to become a regulated utility in 2022 are also shaping the way we think about our future operating model. For example, we need to develop core utility functions, such as compliance and asset management that are subtly different from the way we previously operated. David also mentioned that we need to beef up some of the regulatory capabilities that we have in the business. Our organization and our service company partners will also need to keep evolving as the fiber rollout runs down and we no longer have to operate a copper and a fiber network side by side. We started this process in FY '20 with the merger of 2 business units within Chorus into 1 single Customer and Network Operations team, or CNO. In the near term, COVID-19 is reshaping the way we operate on a day-to-day basis. We already had a flexible working policy, and more than half of our people can see themselves spending more than half their working hours at home rather than in the office. We're now thinking more about how we can make remote working deliver better outcomes, not only for our people, but for Chorus. For example, we have been running a hot-desking approach to several of our office floors for more than a year now. And it's only logical that we'll look to do more of this with people spending less time in the office. As events of the last few weeks in Auckland have shown, the outlook for FY '21 is far from certain. Although we weathered FY '20 well and produced a strong financial and operational results, we don't yet know how much economic impact New Zealand will see from COVID-19. Our role as essential infrastructure provides some protection, but we're not necessarily immune. We also know that there will be increased competitive intensity as mobile network operators seek to address their own cost pressures by promoting fixed wireless to some of their existing customers. And the release of short-term 5G spectrum means that we expect more campaigns promoting 5G in the coming months. But we follow international development closely. And based on what we've seen of the real-world performance of 5G, we continue to have strong confidence in the superiority and in the economics of fiber. Recent COVID-19 experiences and the shift to more home-based work means key value unlimited data and rock-solid broadband like never before. If consumers fully understand the network choices that they are being asked to make and shop around, we're confident that they'll choose fiber. And finally, in the next few months, the Commerce Commission will provide its final determination on key aspects of the new regulatory regime, like David has mentioned. We believe there is still a significant gap between retrospective economic assumptions and commercial reality. These decisions provide the Commission with an opportunity to ensure that investors get a return that represent the risk that they took to build the fiber network. It can do so while knowing that end users are well protected by the price cap, 100 megabit per second anchor service. So there is a huge potential for a win-win here. Under the current proposed setting, we believe that investors will not get a return that reflects the risk that they took, instead their commissioning approach banks the successes of the build and with the benefit of hindsight, effectively derisks it. If this occurs, we run by risk that investors will not consider UFB as a model for a successful transformation of more New Zealand infrastructure. We and our investors will, therefore, continue to advocate for a fair return that respects the risk taken in the first decade of our partnership with government. And we'll keep doing what we can to promote gigabit uptake and bridge digital device, where the regulatory settings support those goals. COVID-19 has proven the capabilities of fiber. Now we need to help New Zealanders realize the true value of the network that is already on their doorstep, and we're committed to do so. Now on that final note, let's move to Q&A.

[Operator Instructions] First question is from Arie Dekker.

You're just starting with guidance. I mean, I think you mentioned on the call that you've seen a little bit of an increase more recently in hours fees and pushing fixed wireless for the lower usage corporate customers. Obviously, in FY '20, the impact on fixed wireless is reasonably moderate. What are you sort of -- what allowance are you making in FY '21 guidance for line loss to fixed wireless?

Sure. Arie, happy to make a couple of comments on that. As I mentioned on the call, the allowance of $10 million is -- we haven't dissected that into specific increments. So it's possible, as you say, that there could be an impact in terms of fixed wireless. That would happen if there was greater price sensitivity in our end consumer base. We don't expect the trend to be any more significant than what we have seen presently and in the past. So we do acknowledge that fixed wireless has a role to play in the market. It has an attraction to a certain demographic but we don't see, as JB mentioned earlier on, significant increased risk in that area, we believe in the value and performance of fiber.

And if I can add a little bit to this -- sorry, Arie I'll just add a little bit to this. I think for us, what's really encouraging is to see the continued growth in people moving to fiber. And we're quite confident in the numbers that we've put out there in the targets that we have for ourselves to generate installations and then, ultimately, connections on fiber. What's slightly more difficult for us to forecast is the rate of copper disconnect. We have an influence on the uptake of fiber. We can do a lot of things with our active wholesale strategy. If some of the retailers want to move voice-only customers to wireless, that's something that we don't control. So it's harder for us to kind of predict how big that will be.

Okay. Just on the $10 million allowance, so what I sort of assumed was COVID in your FY '21 guidance. Just in terms of Auckland at Level 3, which I guess is sort of, hopefully, just a couple of weeks, are you incurring any additional costs at this point in FY '21 associated with COVID that you might sort of put into that allowance or nothing yet? And just in terms of the $2 million of support you provided with regards to hardship in retailers in FY '20, is there -- has that set a preset? And is there any sort of commitment, I guess, on your part in FY '21 to make more support for RSP bad debts?

Okay. Let me handle the 2 questions you put. So regarding the latest Auckland Level 3 restriction, we haven't seen as big an impact. So first of all, what happened in the first wave is we moved into Level 4, which is very, very restrictive. Then it got relaxed into Level 3. But when it got relaxed into Level 3, we think that a lot of consumer were still very nervous about potentially letting people coming into their home. So while in the Level 3 we were allowed to do connections, the uptake of those connections took up quite slowly. It's only when we go to Level 2 that we really saw a return to the volumes that we had previously. This time around, I think people are more comfortable with the growth that we have in the Level 3. So we haven't seen a big impact in terms of the rate of connections that we have. We've continued to do build work. We continued to do connection works. And the only thing that's been impacted was our door-to-door managed migration. So we had to move it from a physical door knocking to a call center type of activity. So that took about a week to reformat. That was the only little impact that we saw currently from the Level 3 restriction in Auckland. Now regarding your other questions, the $2 million fund that we made available was really something that reflected the very unique and unforeseen impact on the total level for long term that we experienced for more than a month. The industry took our thought the right position in saying that during that time, people should not be forced to disconnect because the broadband connectivity [indiscernible] their lifeline. The RSPs committed not to disconnect people. They removed data caps when they could. And because of that, we -- in post RARE decided to support them with this $2 million fund. We see it as a one-off commitment that was done to reflect the unique circumstances of Level 4 lockdown. And there are no ongoing commitments besides the fact that we'll do the right thing to support the community.

Okay. And just a look with regards to Slide 26 and the framework for capital allocation, could you just clarify for me where and that the connection CapEx associated with UFB1 and 2 as such?

Sure, Arie, very much. So the fourth bullet point down that talks about the transition period, the reality is over full year '22 in particular. And then in a -- from a declining perspective from full year '23, we still have communal CapEx spend and we still have connection CapEx spend. So they are part of the transition period. So they are -- this is an illustration of post the completion of communal and post the elevated connection CapEx.

But just in terms of going into sort of '24, '25, you'll presumably still be connecting customers to the UFB network. Can you just clarify where in the framework those connect -- where that connections CapEx as such?

Yes. Sure, Arie. So that would be picked up in the range that we set the dividend payout ratio at. We'll set it at a point that allows a certain level of CapEx over and above sustaining. By full year '24 or '25, the connection CapEx number will be materially lower. So we will make allowance within the range.

Yes. Okay. So it's not discretionary, but it's offset in your sustaining CapEx.

For connections, correct.

Okay. Great. And then just final question from me. Just in terms of the incentives, I think in terms of the breakdown you provided, it looks like incentives cost around FY -- sorry, cost around $14 million in FY '20. Can -- you're making a concerted push in FY '21 to maintain very high connections, which makes sense. And you've outlined sort of the extent of the incentives. What should we sort of be allowing for incentives cost in FY '21?

Sure. Happy to pick that up, Arie. So we -- and you're right to call out the difference in customer retention costs. There is an incentive component, which is not sustaining capital. And then there is a provisioning component, which is sustaining. So your question on the incentive part, I would expect that to be slightly higher. In full year '21, given the expanded managed migration and incentive programs that JB spoke about, it's not a material number, but I would expect it to be a little higher in full year '21.

And just in terms of the $300 and $500 incentive maximum, how weighted are they to hitting targets on upselling customers to the 1 gig product?

JB, would you like to respond to that one?

We target them more towards the 1 gig product, but there are also some that cover all of our products, including the entry-level product. What's really important for us is to make sure that we get off-net network people onto fiber. If they can get to the 1 gig, great, but we also will reward customers for bringing off-net customers onto fiber even if it's not on the 1 gig.

Okay. So there's not like a tilting of the incentives to the higher spec products, that's for volume achievement overall.

It's a mixed thing. We do drive -- we want to drive both outcomes. One is getting people on fiber and also, if possible, getting them onto the high-speed plan. So it's a mixed approach for those incentives.

The next question comes from the line of Phil Campbell.

Just a couple of questions for me. Just maybe, David, if you could give us a little bit more of an update in terms of the copper withdrawal. I think listening to JB, I think it was going to be more of a gradual process and it's done on a cabinet by cabinet basis. But is there any kind of guidance in terms of how we should think about potential OpEx or CapEx cost of that? And I suppose the other question was just whether or not some of the CapEx cost is actually included within the RAB or not? I don't know if you've got any guidance for me yet.

Sure, Phil, happy to talk about that. In terms of copper CapEx, generally, we do expect it to continue to fall. We've guided at $35 million to $55 million in full year '21, which compares with a spend of $55 million in full year '20. So it will continue to decline. That does reflect our plans around withdrawal. As JB mentioned, a couple withdrawal is not a big bang approach. It will be done carefully and always through a consumer lens so the trajectory will continue to be a reduction, quite significant when you look at the range that we've given for CapEx. On the OpEx front, the main impact is around maintenance. So the slide that gives dissection of reactive maintenance, you'll see that copper variable cost has continued to fall in recent years. That is where we will see most of the benefit in the coming years for copper. There is a significant portion of our cost base, which is shared. So it's a little more murky to be able to dissect the impact on our shared cost base. We will do that as part of our regulatory submission, but the key direct link is around our maintenance cost for copper.

Great. Just another one, just following up. I don't know if you've given this number before, but just in terms of your current thinking around what you think like the terminal fiber penetration would be in, say, 5 years' time?

Sure. I'm happy to give a view on that one. JB, if you would like? We've put…

No, no, you go ahead. I'll top up at the end.

Okay. So we put a target out of 1 million fiber connections by full year '22. I think in terms of how far we would like to get -- the reality for us is we'll continue to push to get the penetration rate as high as we can. We haven't actually put a percentage in market. So I probably wouldn't want to actually give a percentage, but it's certainly -- we don't stop at 1 million, if I could put it that way. So you can do the math on what 1 million is, and then you can assume we will be targeting something north of that number.

Okay. Awesome. The other one was just, you talked a little bit about greenfields in the note. Obviously, there's been a number of shovel-ready projects that have been announced. I was wondering if any of that fringe build, whether you've applied for shovel-ready funding on that or whether that's just going to be funded internally?

So at this stage, we did a submission to the fiber ready for specific UFB rollout projects. These were not taken as part of the government program. Some of the real estate development programs that are part of the shovel-ready programs will potentially trigger some fiber connections and fiber work, but that will have to be funded as the normal engagement with developers as we do for every greenfield. Greenfield is a little tricky for us to pick right now. There is clearly a lot of projects that were put on hold during the lockdown that are now back and trying to get completed as fast as possible. So this has put a lot of pressure on our capacity to continue to support these programs. The longer-term outlook for greenfield is a bit more uncertain. So that's something that we'll focusing on and following very closely.

And sorry, just wondering, final one for me, David, in terms of the dividend policy transition period. I think at the interim result, you kind of were talking kind of '22 to '24. I just wanted to make sure that was still what you're thinking then post kind of '24, you can actually step up to more of the typical cash flow payout type method?

I think, Phil, certainly, our views are consistent with what I said at the half. The transition period is constrained more heavily in full year '22 due to the communal build completion for UFB2 and also the elevated connection CapEx. The constraint does start to release from full year '23, but there will be a level of connection CapEx for at least a year or 2. So the way I've described it is not unreasonable, and our view is consistent with what we described at the half year. And Phil, just on -- I'd just add a comment also on imputation. I made some comments on this at the half year, which I thought I should just reiterate again. We do expect to be in a low to 0 tax paying position for the next couple of years, which does mean that our imputation balance will run down from its current levels to close to 0 by the end of full year '22 or thereabouts. We do expect to commence paying tax again within the short term. So we'd expect from full year '22, the imputation level would start low, but then start to build back again over the ensuing years. So I wanted to just give a little bit of clarity, given we're talking about dividend on our imputation profile also. That's all consistent with what I said at half year also.

Our final question comes from Brian Han.

Further related to the copper decline issue, the fall in network maintenance costs obviously accelerated in the second half. Is it possible that cost line could fall to around $50 million this year or next?

Sure. I can give a response to that. I wouldn't like to put a specific number on maintenance. We haven't given guidance on any of our specific expense lines. I do expect maintenance will be lower again in full year '21 due to the continuation of the trends that we've talked about. I think the number you've just quoted is -- would probably be a little bit too low, Brian, I would suggest.

Okay. And sorry for the silly question, but does -- is a permanent step-up in upstreaming traffic have any implications on the cost side, whether OpEx or CapEx?

No. The beautiful thing for us is that our network is really completely transparent from that perspective, while some of the setting on traditional networks like copper network do have a significant impact in terms of delivering symmetrical when you're on fiber. It's really a configuration thing. So no, it does not. And we see that, in fact, as a good trend for us. Because fiber networks are particularly good at handling that upstream traffic.

Following question is from Ian Martin.

Just 2 questions for me. First, you mentioned 1 million fiber subscribers target and currently a 63% penetration rate. Are you able to tell us how that penetration rate varies from area to area? What the peak rate you're seeing is in the U.S. area and how that -- looking at UFB2 and penetration at UFB2, how that compares to a similar period with UFB1? And secondly, just going to that question of risk and perceptions of risk and regulatory opportunities and so on, can you comment on how the risks with UFB2 compared to those in UFB1? Because clearly, the CapEx seems more volatile and you've increased the CapEx recently. And penetration rates seem to be different and the whole range of differences that might affect the risk and return profile in UFB2 compared to UFB1. And that kind of begs the question that the marginal cost of capital as you expand into 2 areas is greater than the allowed cost of capital, which I don't know whether it's got implications for your rate of spending in those areas, but surely, those discussions around risk and return and allowed cost of capital be done before the capital commitment rather than lift the regulatory opportunism after the event.

Thanks, Ian, and I'll handle the first one and then I'll let David handle the risk one. On the take-up rates, we've mentioned there are a number of areas where we're now well above the 90% take-up rates. So in terms of how high can we get, we have areas that are in that 90% range. In terms of comparing UFB2 to UFB1 areas, what we've tended to see is a slightly faster take-up rate in UFB2 areas. And you can kind of rationalize that by the fact that these tend to be areas that would have been in poor delivery area in terms of quality of the broadband they were getting. They will probably still either on VDSL on long -- sorry, an ADSL or on long-line VDSL. So the moment we bring fiber to those communities, we tend to see a very quick uptake. Those are close-net communities. They tend to rely a lot on word of mouth. And I did mention during the presentation, we do have one particular area, which is a UFB2 area. Whatawhata, next to Hamilton, already 80%, which would be higher than some of the store at UFB1 area. So we tend to see a faster uptake in UFB2 areas. And it can be explained by the fact that those people have been waiting for it for a long time. I'll hand over to David on the more risk calculation question.

Thanks, JB. Ian, thank you for the question. In terms of risk, I think the point you're making is very valid. I'd reverse or repeat the comments I made during my presentation that from our perspective, the risk we took was at the date of signing the UFB agreements. So UFB1 and then UFB2 and 2+ and, therefore, we think that the assessment of the appropriate rate of return should relate to the time period that we have committed to through that contract. So there's that conceptual point, which goes to how you set WACC and risk-free rate over the period of the loss asset. In terms of the respective risk profiles of UFB2 versus UFB1, it's true that UFB2 is a little more challenging. The towns are a little more diverse. It's a little more expensive. That's, of course, all reflected in the contractual arrangements we make with [ Circos ]. But the underlying risk profile around construction risk. Could you argue that's higher? Perhaps you could. So it just goes back to my comments about reflecting the real commercial risk that we face when setting the appropriate regulatory return.

Could I also ask about the new revenues? Presumably they're outside the regulatory framework. But when you look at things like smart locations, for instance, I imagine government has a lot of interest in connecting up locations. So I wonder whether there's potential for a kind of UFB3 for smart locations. Is government showing any interest in accelerating that rollout?

We haven't seen that yet, but that's a good suggestion. What we typically do is a lot of those smart locations tend to be with local government. So we've talked about Auckland Transport being one of the potential customers. So they tend to be local government agencies. But yes, I like the idea of potentially having a UFB special tool for smart locations. Thanks for the idea.

But that's outside the current RAB arrangements, isn't it?

David?

We -- it's a little bit unclear, Ian, but potentially. The reality is that we don't yet have a final view on input methodologies from the Commission. So I can't be totally black and white on that at present.

We have a follow-up from Arie Dekker.

Just on the regulatory side of things. I mean you've talked about the issues you sort of see with the proposed settings and the risk this has taken. Can you just comment in terms of the other parts of the framework, are you sort of more comfortable with what you're seeing on the sizing of the RAB and OpEx allowances in the rules?

Sure. Thanks, Arie. I can talk about that. In terms of the sizing of the RAB, we don't yet have a view from the Commission on their perception of the size of the RAB, either the loss asset itself or the base build. We probably won't have that, and it probably won't be public until the draft price quality decision, which is mid- next calendar year. So I don't really have a view on a number from the Commission's perspective. In terms of what the rules say for RAB, it's pretty clear in the legislation around the spend on assets and how that will be measured. The loss asset, I spoke about in my comments, the move to a discounted cash flow approach. We don't have a concern with that compared to BBM, if the inputs are the same, where our concerns lie, as I talked about earlier, are how you assess the risk-free rate over the loss asset period and the treatment of Crown financing, in particular. So they're the key issues from a RAB perspective. When I look in the market at what views there are on the level of the RAB, and we've been very careful not to put views into the marketplace because we don't think it's appropriate yet. It's too early. There's too many moving parts. There's a lot of judgment that's needed. But market views across the analyst community seem to be around about $6 billion as a midpoint, in a range. So that's the information we have on the RAB. On the cost allowances, we are preparing our initial regulatory submission. It's due in December of this year. You'll see when that goes in, what we're proposing or what our view is on those allowances. The rules seem to be okay, Arie, at present. The big judgment is around cost allocations. We have a lot of shared assets. Our costs are shared across fiber and copper. So that's where the judgment is for us. I think there'll be more detail on that from the Commission in the coming months and into next year. But at the moment, it seems reasonable.

We have no further questions at this time. I'd like to hand it back for any closing remarks.

Okay. Well, thank you again for joining us on this occasion. These are complex times. The good news from our perspective is that we continue to see the underlying trend for growth in fiber demand to continue. And in some cases, some of the things that are triggered by the COVID-19 pandemics are things that are underlying some of the strengths or accelerating some of them. We believe that we have a good plan for next year. It's one that we're going to be very closely implementing. We'll keep very close attention to what's happening in the broader telco sector, but also how the pandemic is impacting this market. We look forward to briefing some of you in more details over the next couple of weeks. And thanks again for joining.