Good morning, and welcome to Telstra's half year FY 2021 results presentation. My name is Nathan Burley, Telstra's Head of Investor Relations. I respectfully acknowledge that I am joining today from the lands of the Kulin Nation. On behalf of Telstra, I would like to acknowledge and pay my respects to the traditional custodians of country throughout Australia and recognize their continued connection to land, waters, and culture. We pay our respects to their elders past, present, and emerging. This morning, after presentations from our CEO, Andy Penn, and our CFO, Vicki Brady, we'll be taking questions from analysts and investors, and then media. I will now hand over to Andy Penn.
Well, thanks very much Nathan. Good morning and welcome everybody to Telstra's results announcement for the half year ended December 31st, 2020. This morning, I will make some introductory remarks and take you through our overview of our results. Vicki will step you through the numbers in more detail before we move to Q&A. 2021 is a significant year for Telstra. It represents a turning point for the company in our T22 journey. A turning point in our financial outlook. For the last four years, every year, we have had to face the confronting challenge of the financial headwinds which arise from the transfer of a material part of our business to the NBN. This has meant that we have started each of the last four years with our EBITDA going backwards by up to AUD 800 million.
This has been occurring in a market where competition has led to material reductions in both fixed and mobile ARPUs, as well as technology disruption and significant structural change across the industry. In many ways, it was these dynamics that provoked our T22 strategy that we announced almost three years ago. It was these dynamics in conjunction with a conviction about how technology innovation was going to continue to accelerate that led us to understand that we needed to radically transform. Now, we are substantially through the T22 program, and it is delivering significant benefits leading to a financial turnaround. That is why I say 2021 is a significant year for Telstra, and I will demonstrate this in my presentation shortly. In summary though, today we are reporting underlying EBITDA for the first half of FY21 of AUD 3.3 billion.
Against this, we're also announcing guidance for the second half of the year of AUD 3.3 billion-AUD 3.6 billion. Now, we do not usually provide explicit half year guidance. However, we thought it was helpful to do so this time around to illustrate the turning point that we are at. The turning point is also illustrated by our ambitions for FY22 and FY23. Now, these are not guidance, but they do demonstrate our aspiration for mid to high single digit growth in underlying EBITDA for FY22, and for underlying EBITDA to be in the range of AUD 7.5 billion-AUD 8.5 billion in FY23. This range is important, of course, to support an AUD 0.16 dividend inside our dividend payout ratio. It's also important to deliver a ROIC of around 8%.
We know how important the dividend is to our shareholders. That is why the board expects to pay a total dividend for FY 2021 of AUD 0.16 per share, including an interim dividend of AUD 0.08 per share. We also know how important the group restructure that we announced in November is to our shareholders. We are therefore committed to progressing this this year, including the monetization of towers. You can see the strategies that we have been deploying are paying off. We are performing well in the market and our T22 program is delivering. After a decade of disruption following the creation of the NBN, with its rollout now declared complete, we can clearly see the path to underlying growth ahead.
Our investment in innovation and technology, in digitization and networks, improving customer experience, and being disciplined in our capital management means that Telstra is in a strong position to grow. With that, now let me turn to the results. Total income from the year decreased 10.4% to AUD 1.2 billion on a reported and guidance basis. After adjusting for lease accounting on a like-for-like basis, EBITDA decreased 11.7% to AUD 4 billion. Underlying EBITDA on a guidance basis, which excludes one-off NBN income and restructuring costs, decreased 14.2% to AUD 3.3 billion. Excluding the in-year NBN headwind, underlying EBITDA declined by approximately AUD 180 million. The estimated financial impact of COVID in the half was AUD 170 million. Our underlying EBITDA in the half was consistent with our full year guidance of AUD 6.5 billion-AUD 7 billion.
With our guidance for underlying EBITDA in the second half at AUD 3.3 billion-AUD 3.6 billion, it follows that our guidance range for the full year has narrowed to AUD 6.6 billion-AUD 6.9 billion. Net profit after tax decreased 2.2% to AUD 1.1 billion on a reported basis, while free cash flow was up 88% to AUD 1.9 billion. This was due to working capital improvements related to reduced handset receivables from lower device sales, improved inventory, and creditors. The lower device sales followed the ongoing trend of customers holding onto their devices for longer as they have become more expensive, as well as from lower store traffic from COVID-related restrictions. This was also the main reason for lower income in the period. The board has resolved to pay a fully franked interim dividend of AUD 0.08 per share, which will return approximately AUD 950 million to shareholders.
As I have already mentioned, the board has also confirmed it expects the total dividend for FY 2021 to be AUD 0.16 per share. Turning to the operating highlights for the half. We continued to see strong customer growth in mobiles. We added 80,000 net retail post-paid mobile services, including 58,000 branded and 22,000 from Belong. This is in fact the strongest branded performance in several halves, and it reinforces the benefits of our clear leadership in 5G. In wholesale, we added 163,000 services, while we added a further 456,000 IoT services. Importantly, in mobile, we saw our lead indicator, transacting minimum monthly commitment, or TMMC as we refer to it, increase by AUD 3. Fixed had a more challenging half, where we lost 53,000 net new retail bundles, including 11,000 from Belong.
While we had negative net adds for the first time, encouragingly, bundle and standalone ARPU in Consumer & Small Business has stabilized due to our focus on price, higher speed tiers, new add-ons, improvements to Wi-Fi, and the Telstra Smart Modem. In fact, the Smart Modem is now in over 2.2 million homes, almost 80% of our consumer customer base, and it was key to keeping customers connected when they were working and studying from home last year, and of course, increasingly in this year too. We have also launched our Adaptive Networks product in Enterprise during the period to mitigate downside risk from digital disruption and competition from the NBN, and also to support its return to growth, which we have foreshadowed for FY22.
One sector where we have seen digitization accelerate dramatically is, of course, in healthcare technologies, and Telstra Health is strategically very well positioned in this growing market, with income up 17% in the half and further improvements in EBITDA. In customer experience, some aspects of our service were impacted during 2020 as a result of workforce capacity challenges. These flowed directly from the COVID-related working restrictions that were in force, particularly overseas. Pleasingly, service levels were recovered in the second half of the calendar year, and episode NPS improved by three points in the half, albeit was down one point when compared to the December 31st, 2019. I am conscious of the delays that some customers have experienced in trying to contact us, and I want to apologize for those.
I also know that not all aspects of our customer experience are where they need to be, and we have more work to do in this regard. However, I am confident that the many initiatives we have taken under our T22 program in simplifying the business and in digitization will further improve experience as we go forward into the future. Notwithstanding these specific challenges, strategic NPS has improved five points in the last six months and 11 points in the last 12 months. Finally, on our operating highlights, we have made very strong progress in our productivity program. For the half year, underlying fixed costs were down AUD 201 million, and total operating expenses were down AUD 876 million. This brings the total annualized cost reductions achieved under our productivity program to AUD 2 billion on an annualized basis.
Today, we are upgrading our productivity outlook for FY 2021 to AUD 450 million and for the whole program to AUD 2.7 billion. Before I talk more about the progress we have made on our T22 program in the half, let me come back to the point that I made in my introduction about being at a turning point. The progress that we have achieved in our T22 program shows that we are building momentum and we are building financial momentum. The left-hand side of this slide shows the evolution of underlying EBITDA over the last several halves, with the guidance provided for the second half of FY 2021 of AUD 3.3 billion-AUD 3.6 billion.
The chart on the right-hand side shows the evolution of our full year underlying EBITDA, including guidance for FY21, as well as the aspirations I've referred to for mid to high single-digit growth in FY22, and to be in the range of AUD 7.5 billion-AUD 8.5 billion in FY23. As I mentioned before, the figures for FY22 and FY23 are obviously not guidance. They're aspirations or ambitions, if you like. There are greater risks and uncertainties associated with them compared to our guidance statements. Nonetheless, the charts clearly demonstrate why I say we're at a turning point. I also want to comment on what sits behind this and draw out some of the key underlying operating metrics that we have been focused on improving. The first is postpaid handheld TMMC, which as you know, is the leading indicator for ARPU.
TMMC increased by around AUD 3 in the first half. We expect it to increase by a similar amount in the second compared to its prior corresponding period. We are confident, therefore, that mobile ARPU has reached a turning point and will return to growth in the second half of FY 2021. Secondly, Consumer & Small Business fixed bundle ARPUs have stabilized. We expect this to continue into the second half as we focus on building value and achieving a mid-teens NBN resale EBITDA margin by FY 2023. Thirdly, as I have mentioned, we have continued to make strong progress against our productivity target, reducing underlying fixed costs by AUD 201 million during the half, with another approximately AUD 250 million expected in the second half and AUD 450 million next year.
At the same time, the major headwinds that we have been facing from the migration to the NBN are coming to an end. The in-year NBN headwinds peaked in the second half of the last financial year. They reduced in this half, they will be substantially less in FY 2022. We also expect the COVID impacts to reverse over time. After an estimated AUD 200 million impact in FY 2020, we expect a AUD 400 million impact in FY 2021. This will be weighted to the second half with an estimated AUD 170 million in the first. This is why I am optimistic that we are at a turning point financially. With that, let me now turn back to this year and make some comments on our progress against our T22 strategy and initiatives. Firstly, to say we are on track to deliver more than 80% of our T22 scorecard metrics.
Against the first pillar of T22, we now have more than 7.6 million services on our 20 Consumer & Small Business in-market plans. We also have 2.8 million members signed up to our very successful loyalty program, Telstra Plus, and we are seeing very strong engagement from these customers. For Consumer & Small Business customers, digital sales interactions are up 10 percentage points to 40% compared to FY20, which itself was already up very strongly on the previous year. Overall digital service interactions now account for more than 70% of all service interactions. Under our T22 strategy, our aspiration had been to reduce the number of calls to our call centers by improving our service by 2/3 by FY22. With the acceleration to digital, we have already seen this run rate and achieved it more than a year before the end of the program.
That means that over time, we will need smaller call centers for these customers, and in fact, many more will work from home. We are on track, therefore, to have all inbound calls from our Consumer & Small Business customers answered in Australia within the next 18 months. Last week, we closed our call center in Cebu in the Philippines. Today, we are also announcing changes to another important part of our customer strategy with our intention to transition to full ownership for all of our branded retail stores across Australia. As more customers interact with businesses online as a result of COVID, we think now is the right time to bring back ownership to ensure a consistent and integrated customer experience across all of our online channels and entire store network.
At the height of COVID last year, in fact, we were able to redeploy frontline staff from Telstra-owned stores to assist customers through our digital channels or via the phone. It's exactly this type of flexibility that we'll be able to unlock as more retail branded stores are under Telstra's ownership. Currently, Telstra has more than 60 Telstra-owned and operated stores with another 166 branded stores, which are run by individual licensees and a further 104 that are operated by the Vita Group. Vita Group and individual licensees have been notified of the plan with discussions and transition arrangements expected to progress over the coming months. Importantly, though, we know that in many regional and rural towns, the local Telstra store is a valued part of the community, providing support and connectivity to a range of businesses and industries.
That's not going to change, and neither will our commitment to ensuring current licensee store customers continue to receive the excellent level of service that they have. Turning to InfraCo. We have continued to make progress in the establishment of InfraCo. It is now a fully operational business function with separate accountabilities and reporting. For InfraCo Towers, we are finalizing the intercompany agreements and are now setting up separate IT systems. We are also undertaking the necessary significant due diligence across the asset classes to ensure that we have fully documented inventories of an acceptable standard to third-party investors. Now, one of the reasons that we said for setting up InfraCo was to provide for an increased level of operational and commercial focus on these assets through a dedicated CEO and a management team. Encouragingly, we are starting to see the benefits from this.
In the half, total passive income was up 2.3% to AUD 1.3 billion. EBITDA after leases was AUD 904 million for a 68% margin. Our new Dark Fibre products have been very well received by our customers, with more than 20 services now ordered or complete. We also won two significant government contracts for tower services in Tasmania and New South Wales. We have also started simplifying our field infrastructure and maintenance with a goal of reducing from more than 20 vendors to five and driving operational efficiencies. On the group restructure that we announced in November, we have commenced detailed consultations with key stakeholders, including government, regulators, and the NBN, and we have also appointed advisors and are working with them on a detailed implementation timetable.
We anticipate making a further announcement in March to set out this timetable for the market, including the major steps that we will need to go through to give effect to the restructure. It remains our intention to complete the restructure by the end of the calendar year. On towers, as outlined at our November Investor Day, once established, InfraCo Towers will own and operate the largest mobile tower network in Australia. We have been conducting detailed due diligence and documentation to support launching the monetization process with potential investors in the third quarter of 2021 and with binding offers to be submitted in the fourth quarter. Turning to pillars three and four. Our workforce continues to change significantly. Since June 2018, when we launched T22, we have reduced 22,000 roles, including 6,000 from our direct workforce and 16,000 from our indirect workforce.
This does need to be put in the context of a very significant part of our business and its associated work being transferred progressively to the NBN over the last few years. NBN today is itself a very significant employer. At the same time, we have recruited more than our planned 1,500 new roles with new skills in new areas such as software engineering, data analytics, and cybersecurity. We have also further progressed our journey to introduce agile ways of working. Today, we have around 11,000 people across the business working in agile. We anticipate being fully agile right the way across the business by the end of the calendar year. Last year, we announced a pause on our T22 job reductions to give our people as much certainty as possible during the very challenging time we all experienced through COVID.
We extended that pause to February this year, which is obviously where we are today. Last week, we did announce the next wave of proposed organizational changes. With these, and subject to appropriate consultations, by June, we expect to be more than 90% of the way through our T22 target to reduce our direct workforce by 8,000 roles and to have completed it by the end of the calendar year. In terms of reductions in indirect headcount, it was initially our expectation to reduce by around 25% or 10,000. However, we have already reduced 16,000, and we expect to make further reductions to our indirect workforce due to the significant progress we have made in digitizing the business. The majority of these roles have been offshore.
Needless to say, these changes have been difficult for our people on top of the challenges presented by COVID to them personally, particularly as they directly impact members of our team and their families. I am proud, therefore, that our employee engagement has remained high with a score of 80, reflecting a concerted leadership effort to support everyone during this time. Also under Pillar 4, we have now exceeded our target of monetizing up to AUD 2 billion of assets to further strengthen our balance sheet. Announcements in the half year included the sale of the Clayton data center, the Pitt Street exchange in Sydney, and the Velocity and South Brisbane fibre networks. We also sold the business of our e-commerce subsidiary, Neto, which we acquired six years ago but no longer fits our strategy.
We are currently in the final stages of negotiating contracts for the sale of our remaining stake in Sensis. You will also recall that we restructured Telstra Ventures through a partnership with HarbourVest two years ago. That move has paid off, with Telstra Ventures performing very strongly. Since the inception of Telstra Ventures, the team has completed 73 investments and achieved 25 liquidity events, which have increased the value of the portfolio substantially. In the past half year alone, some of the portfolio highlights include the BigCommerce, CrowdStrike, and Skillz IPOs, as well as the Rancher investment and Cumulus Networks acquisitions. The valuation of investments held by Telstra Ventures is in our accounts recorded in reserves, and we recorded AUD 187 million increase in the valuation of these investments in the half year period. On 5G, we're not just leading in Australia, but we are also among the global leaders.
We have expanded our 5G rollout to selected areas in more than 100 cities and towns across Australia, and the network now provides 5G coverage to more than 50% of the population. It is our intention to increase that to 75% by June. In fact, today, we have around 1 million 5G-enabled devices connected to the Telstra network, and our average combined 4G and 5G mobile speeds are faster than our competitors. In fact, we recently achieved a world-first 5G peak download speed record on a commercial network using millimeter wave spectrum of greater than 5 gigabits per second. Not surprisingly, therefore, we continue to also lead the market in the major mobile industry network performance benchmarks, including umlaut, where we were ranked number one for best in test and best in data. Before I close, a few comments on our T22 scorecard.
Now, as you heard me say earlier, we have delivered or are on track to deliver over 80% of our T22 scorecard metrics. Needless to say, some of the measures, of course, are rated either amber or red, and I want to take one moment to explain why. Firstly, underlying ROIC. As announced at our FY20 results, we will not achieve our T22 ROIC target of greater than 10% in FY23. I know that is disappointing, and I want to reassure shareholders that we are pulling all levers available to us to improve that performance. As outlined at our investor briefing in November, our new target is 8%, and with the ambitions I spoke to earlier, we can see our path to achieving this in FY23. On NPS, we are on track with strategic NPS, but as I explained earlier, we are slightly behind on our episode NPS results.
The building out of our new technology stacks is also very well progressed, but as in any IT project of the magnitude and scale that this one is, there are of course always a few things that right shift. The enterprise stack is now live, as are the agent-facing components and mobile products in Consumer & Small Business. Product launches onto the new stack have been accelerating in the first half of this year, and we expect that to continue in the second, enabling us also to accelerate the migration of customers. On active app users, these have grown by more than 400,000 to 4.4 million. It's below where we had planned it to be, but it reflects good progress on FY20.
We also need to build more momentum in the average services per customer, and we are continuing to target increased multi-product holdings through entertainment, mobile assurance, and gaming add-ons. Our announcement last week with Kayo is a key to doing this. Let me summarize. The last 12 months have seen us navigate the profound disruptions from COVID. COVID has also highlighted that connectivity has never been more important. We have seen a huge acceleration in adoption of digital ways of working and living, and these things are also going to be crucial to a fast economic recovery for the country. It is interesting to reflect on how seamlessly Australians were able to move to working from home, how quickly people adopted digital ways of working and living, and how we were able to support them with the necessary bandwidth, products, and services. None of that happened by accident.
It happened for us because in 2016, we knew we would see a further acceleration in the use of technology, so we invested AUD 3 billion to build the networks of the future and completely rebuild our digital environment. It happened for us because in 2018, we launched our T22 strategy to simplify and further digitize the business. It happened for us because we continue to make excellent progress in implementing this strategy. Those investments are transforming Telstra. We are now less than 18 months from completing T22. We've achieved an extraordinary amount, and Telstra today is a leaner, more responsive, and more agile company than it has ever been before. I said in my opening 2021 is a significant year for Telstra.
I know we have more to do, but we have reached an important turning point financially, and we look to the rest of the year with great confidence in our ability to deliver our strategic ambitions. Our priorities for the next 12 months include, firstly, making sure we drive the key operating metrics that I highlighted earlier, which will be instrumental in delivering the financial turnaround. Secondly, finishing the job on T22, including the final stages of the digitization program and the migration of customers at scale to our new technology platforms. Thirdly, delivering our group restructuring plans, including the reorganization of the company into three separate entities, InfraCo Fixed, InfraCo Towers, and ServeCo, as well as the monetization of towers. Fourthly, further extending our leadership in 5G and rolling out to 75% of the population by June.
Finally, further improving our customer service by bringing our retail experience in-house and onshore. This means meeting our commitment to answer all inbound calls from our consumer and small business customers in Australia within the next 18 months and commencing the process to bring back our licensee stores in-house. With that, we will have truly transformed Telstra through our T22 program and set the company up for success in the digitally driven and very exciting future that lies ahead. We will have built the capabilities to take advantage of the opportunities this presents, and we will announce how we're going to leverage these and what comes after T22 at our Investor Day in November. Can I close by acknowledging that the progress that we have made is due to the combined efforts of our many dedicated employees.
Despite the disruptions and impact on them personally from COVID, every day, they have focused on working for our customers and keeping Australians connected. For that, I want to sincerely thank them. Thank you. With that, I will hand over to Vicki before we open for Q&A.
Thanks Andy. Good morning, everyone. Thank you for joining us. I'd like to begin by recognizing that I am joining you from the land of the Gadigal people of the Eora Nation. I acknowledge their ancient and ongoing connection to this land and their culture. I welcome any Aboriginal and Torres Strait Islanders joining us today. Turning to the details of our financial performance for first half FY 2021, which you can see on slide 13. The numbers on the left of this slide are our reported statutory results. The numbers on the right are reported lease adjusted, which includes depreciation of mobile handset lease expense as OpEx. This provides a like-for-like year-on-year view, given the exit of mobile lease plans. It is the view we use when managing the business and which this presentation will focus on. For first half FY 2021, income was AUD 12 billion, down 10.4%.
On a reported lease adjusted basis, EBITDA declined 11.7% to AUD 4 billion. This decline included a AUD 268 million reduction in net one-off NBN receipts and a AUD 297 million increase in restructuring and other guidance adjustments. I'm pleased to say that our underlying EBITDA during the first half was in line with our expectations and our FY 2021 guidance. We are also pleased today to be upgrading our free cash flow guidance. Underlying EBITDA was down AUD 550 million, or 14.2%. The largest two contributors to this decline were an estimated AUD 370 million of in-year NBN headwind and approximately AUD 170 million of estimated impacts related to COVID-19. Our estimate for COVID-19 is based on international roaming declines, delayed cost out, additional customer support, and deferred NAS professional services. If both the in-year NBN headwind and estimated COVID impacts were excluded, underlying EBITDA was broadly flat.
Our total operating expenses declined 9.8%, including an AUD 201 million, or 6.6% decline in underlying fixed costs. Depreciation and amortization declined 4.6% on a reported lease adjusted basis. This is consistent with the expected full-year decline of around AUD 300 million due to assets associated with NBN completion and legacy IT assets fully depreciating. Net finance costs declined due to lower average borrowing cost, thanks to recent refinancing at lower rates. Income tax expense declined 60% on a low effective tax rate associated with M&A and asset sales, as existing capital losses were used to offset capital gains. Excluding these one-off factors, our underlying effective tax rate was close to 30%. Reported NPAT was AUD 1.1 billion, down 2.2%. Looking now at income by product, which you can see on slide 14, which reflects the new product reporting framework we announced in January. Underlying income declined AUD 1.2 billion, or 9.8%.
Excluding the in-year NBN headwinds and lower international roaming fees due to travel restrictions, income declined 6.5%. Of this decline, over 2/3 were due to a reduction in hardware. There is detail in the appendix on each product. I will touch on the most significant points. Mobile income declined AUD 645 million in first half '21. This was largely due to hardware revenue, which has minimal impact on EBITDA and international roaming declines. Handset and tablet volumes were around 450,000 lower than first half '20. The average price was also lower, helped by the higher Australian dollar. The reasons for the decline in volumes versus the PCP were firstly, impacts of COVID slowing down sales, including foot traffic in retail stores, down around 30%. Secondly, customers holding handsets for longer. Thirdly, higher outright purchases through independent retailers. Finally, a later release date for the iPhone.
Of these reasons, impacts of COVID on sales and the later iPhone release drove outcomes materially different to our estimates when we set guidance. We anticipate these impacts to continue in the second half. Profit-generating domestic mobile service revenue, which has been in decline since FY 2017, was broadly flat, excluding international roaming declines. In postpaid handheld, net adds remained healthy across all segments. Pleasingly, our lead indicator of postpaid handheld ARPU, transacting minimum monthly commitment, or TMMC, continued the positive momentum we have seen since 2019, up by more than AUD 3 in first half 2021 versus PCP. We expect a similar increase in second half versus PCP. The sustained increase in TMMC, as well as pricing changes, are flowing through into ARPU. However, reported postpaid handheld ARPU in first half 2021 declined 8.6%, including a decline of around AUD 140 million in international roaming.
Excluding this, ARPU declined 3.2%, with uplifts from pricing changes more than offset by the continued downward impacts from accounting for new plans, which allocate more revenue to hardware, lower out-of-bundle excess voice and data fees. Finally, dilution from a higher mix of Belong customers, despite Belong TMMC and ARPU growing in first half FY21. These impacts on ARPU will continue, the thing that gives us most confidence in the outlook is the flow-through of pricing changes we are now seeing. In prior periods, ARPU has had a negative drag from pricing trends. This turned positive in first half FY21, we are confident this will flow through even more strongly in second half FY21. Turning to other mobile categories. In prepaid handheld, an increase in unique users and voucher value returned the product to growth. Mobile broadband remained broadly stable sequentially, with declines in prepaid offsetting postpaid growth.
Our wholesale business achieved strong revenue, SIO, and EBITDA growth. In fixed consumer and small business, our focus is on improving the long-term economics and customer experience. Income declined 7.5% in first half 2021, impacted by NBN migration, legacy voice, and Foxtel from Telstra declines. The decline in income from bundles and data was only 0.6%. This reflects ARPU stabilizing as we hit a turning point, thanks to customer migrations to in-market plans no longer being dilutive. Around 70% of our customers are now on in-market plans, and we are almost 90% through migration of customers within the NBN fixed footprint. We remain focused on increasing ARPU through differentiation, add-ons, and plan mix. Turning to fixed enterprise, where income declined 6.4%. Fixed enterprise has two main categories.
Data and connectivity, where revenue was down 7.2% as we transition from providing virtual private corporate networks to integrating over-the-internet technologies such as SD-WAN with Telstra Fibre or NBN access. Total SIOs declined as copper exits were not fully offset by growth in NBN services. We are now around 65% through the migration of enterprise services to the NBN. Importantly, Fibre SIOs were stable and ARPU decline slowed as we focused on retaining and adding higher bandwidth SIOs. NAS, which has been reclassified following the introduction of our new product reporting framework to enhance transparency and align with our strategy. NAS income declined 6%. Single-digit growth in managed services, including security and cloud applications, was insufficient to offset structural declines in calling applications, including ISDN, as well as equipment sales and professional services.
The result in fixed wholesale is attributable to legacy products, NBN headwinds, and commercial works declines. The ongoing portfolio, which accounts for almost half of the revenues and includes passive infrastructure, grew. There is other information on InfraCo consistent with Investor Day included in the appendix. Finally, in other, our health business continued to scale with revenue growing 17%. Turning to our operating expenses, which you can see on slide 15. We are very pleased to have achieved a significant reduction in costs during first half 2021. Total costs declined 9.8%, and underlying costs declined 7.8%. An increase in NBN payments of AUD 136 million was more than offset by the productivity gains we achieved. Other sales costs declined AUD 463 million on lower hardware costs. Underlying fixed costs reduced by AUD 201 million, and we are now tracking to achieve a full year reduction in FY 2021 of around AUD 450 million.
First half 2021 productivity was predominantly enabled by simplification and adoption of digital channels, ongoing focus on vendor costs, and increased workforce efficiency. Our proposal to move ahead with job reductions, which we paused until February this year, will deliver a run rate reduction into FY22. We have now achieved an AUD 2 billion net reduction in underlying fixed costs since 2016. However, to have a world-leading cost base, we have more work to do. Based on our strong progress to date and outlook, we have lifted our net productivity target from AUD 2.5 billion to AUD 2.7 billion by the end of FY22. Further reductions in FY22 are expected to be delivered from IT and network infrastructure costs, realizations of benefits from digitization, including product simplification and customer self-service tools, as well as ongoing labor efficiencies.
We anticipate that delivery of this new target will get us to the top quartile of global benchmark for full service telcos by the end of FY 2022. With the increased cost target and jobs announcements earlier this month, we expect around AUD 180 million of restructuring costs in FY 2021. Moving to EBITDA by product on slide 16. Mobile EBITDA declined to AUD 127 million versus PCP, but would've been broadly flat, excluding the international roaming decline. Mobile EBITDA also returned to growth sequentially in first half 2021. Our mobile business is clearly building positive momentum. We have strong 5G leadership and differentiation. TMMC is up with a higher proportion of customers choosing plans of AUD 65 or higher. Digital engagement is increasing. 2/3 of mass market postpaid customers are on in-market plans, and our loyalty program continues to scale. We see value accretion across our multiple brands.
These factors give us confidence that mobile EBITDA will grow again sequentially in second half 2021. We also expect full year FY 2021 growth on a PCP basis, then further growth that will support our FY 2022 and FY 2023 financial ambitions. Fixed consumer and small business EBITDA declined AUD 230 million. This includes a revenue decline of around AUD 200 million, an increase of around AUD 130 million in network payments to NBN, partially offset by cost reduction. Our NBN reseller EBITDA margin was around 5% in first half 2021, our ambition remains to grow this to mid-teens by FY 2023. Our strategy to achieve this includes a combination of initiatives, targeting improvements in gross margin, such as speed tier mix and add-on services, along with cost to serve reductions and delivery of productivity. Legacy EBITDA has continued to decline with diseconomies of scale.
We remain excited by 5G home internet opportunities to drive future on-net growth, including through millimeter wave spectrum, increasing capacity. Turning to fixed enterprise. Data and connectivity EBITDA declined 15.4% due to reduced revenue on high margin products and largely stable costs. Our adaptive network strategy, launched during the half, is targeted at maximizing long-term economics. NAS EBITDA declined due to reductions in higher margin legacy calling applications and professional services, along with the announced pause on job reductions and one-off costs negating the benefits of cost reductions. This was partly offset by growth in managed services and cloud applications. We expect NAS EBITDA to be broadly flat versus PCP in second half 2021, and then grow in FY 2022. We remain committed to achieving mid-teens margins from FY 2022.
Global EBITDA, excluding one-offs in the PCP and in constant currency, grew 2% as cost initiatives and mix offset the revenue decreases in data and connectivity and NAS. The commitments we've made about the future financial performance of our products are included in a slide in the appendix. Turning to free cash flow, which you can see on slide 17. Free cash flow after operating lease payments increased close to 90%, to AUD 1.9 billion, largely due to working capital improvements. The improvement reflects reduced hardware revenue, with reduced handset receivables and improved inventory on lower handset volumes and average rate. We have also managed our creditors and receivables position to deliver the favorable outcome. Some of the increase also reflects timing. This means we are tracking ahead of our FY21 free cash flow guidance range and are lifting the range to between AUD 3.3 billion and AUD 3.7 billion.
CapEx is tracking consistent with guidance. M&A disposals in the period included the Pitt Street exchange property and Telstra's Velocity network. We look forward to participating in April's millimeter wave spectrum auction and note that payment terms include the option for five annual installments rather than all payment upfront. Moving to dividends. The board has resolved to pay an interim dividend for first half 2021 of AUD 0.08 per share fully franked, including an ordinary dividend of AUD 0.05 per share and a special dividend of AUD 0.03 per share. The first half 2021 ordinary interim dividend represents 125% payout ratio of underlying earnings, but is well supported by cash flow. The total interim dividend represents a 60% payout of free cash flow after operating lease payments, less net finance costs paid.
The board understands the importance of dividend to our shareholders and remains committed to the objectives of the capital management framework. As stated at last year's AGM, the board is prepared to temporarily exceed the framework's principle of paying an ordinary dividend of 70%-90% of underlying earnings in order to maintain the current dividend. The board stated it would consider the following factors in determining if it continues to do so. First, if our ambition for underlying EBITDA of AUD 7.5 billion-AUD 8.5 billion in FY 2023 is achievable. Second, if the full-year free cash flow dividend payout ratio remains supportive and we retain a strong financial position. Third, if there are any other factors that would make the payment of the dividend at that level imprudent. Based on this criteria, the board expects to pay an FY 2021 fully franked dividend of AUD 0.16 per share.
Turning to our capital position, which you can see on slide 19. Under our T22 strategy, we have monetized over AUD 2 billion in assets, and our balance sheet and liquidity position remains strong. Net debt declined around AUD 700 million in first half 2021, and we remain within our comfort ranges for all our credit metrics. We have also updated our capital management framework, consistent with the outlook for CapEx provided at the November 2020 Investor Day. Principle three now states, "Target CapEx to sales ratio of approximately 12%, excluding spectrum, from FY 2023." As discussed at Investor Day, our revised T22 ROIC target is for underlying ROIC of around 8% by FY 2023, with a long-term ambition to grow ROIC. Turning now to our revised FY 2021 guidance, which you can see, along with the assumptions and conditions upon which we have provided them, on Slide 20.
FY 2021 income guidance is now AUD 22.6 billion-AUD 23.2 billion, reducing approximately AUD 1.2 billion at the midpoint from prior guidance. The large majority of the change is due to mobile hardware. I spoke earlier to how differently trading conditions have played out in the first half compared to our expectations when we set guidance. Our outlook is also AUD a few hundred million lower than prior expectations in global, largely due to exchange rate outcomes. These changes have minimal impact on EBITDA. Underlying EBITDA is now expected to be in the range of AUD 3.3 billion-AUD 3.6 billion in second half 2021, which compares to AUD 3.3 billion in first half 2021. We have therefore narrowed FY 2021 underlying EBITDA guidance to be between AUD 6.6 billion and AUD 6.9 billion. Our underlying EBITDA guidance continues to assume an in-year NBN headwind of approximately AUD 700 million.
Due to timing of disconnections, we expect to be at the low end of the net NBN one-off range. The estimated COVID-19 impact in FY21 is unchanged at approximately AUD 400 million. Free cash flow after operating lease payments is now upgraded to AUD 3.3 billion-AUD 3.7 billion, up AUD 450 million at the midpoint. For clarity, any acquisitions from licensees under our strategy announced today to transition to full ownership of branded stores is excluded from guidance free cash flow. To conclude, first half 2021 was an inflection point for the financial performance of our business. Our underlying results remain challenged, including from ongoing NBN headwinds, legacy declines, and financial impacts of the COVID-19 pandemic. However, our continued focus on T22 is delivering simpler, better outcomes for our customers and greater productivity, enabling us to increase our cost out targets. Product margin improvement is also imminent and already occurring in mobile.
We see clear positive indicators of an improved financial trajectory, which we expect will return us to underlying EBITDA growth in FY22 and put us on the path to achieving our FY23 financial ambitions. We therefore look forward with confidence. Finally, I would like to take this opportunity to add my thanks and recognize our dedicated teams right across Telstra. We will now hand over to Nathan to take us through Q&A.
Great. Thanks, Vicki. We'll now take questions from analysts and investors. Our first question will be from Kane Hannan of Goldman Sachs.
Morning, guys. Just three questions from me, please. Firstly, just on that mobile subscriber growth in the half, can you just talk about how you were able to grow your post-paid subs in the half, despite putting up pricing when the industry didn't follow? I suppose, what that tells you about the sticky nature of your customer base? Secondly, just given the increase in productivity to AUD 2.7 billion, obviously nice to see, but are we still thinking about that hundreds of millions of AUD of productivity savings in FY 2023, FY 2024 that was spoken about at the Investor Day, or is that a little bit of a pull forward in some of those savings? Finally, just in terms of the free cash flow outlook. How should we be thinking about the FY 2022, FY 2023 EBITDA targets translating into free cash flow?
Could you just comment on some of the moving parts that we should be thinking about there? Cheers.
Thanks very much Kane. Good morning. I should mention as well that we've got Michael Ackland, who runs our Consumer & Small Business division, David Burns, who runs our Enterprise division, and Brendon Riley, who's the CEO of InfraCo on the line as well today. I think I'm going to let the team respond to some of the questions. I'll get Michael maybe to comment on mobiles, Vicki to talk about productivity and free cash flow outlook as well. My point on mobiles would be, I think it's a function of the continued focus on value and differentiation and 5G leadership. I don't want to take anything away from Michael, so why don't I give Michael the opportunity to make a comment?
Great thanks Andy. Hopefully you can hear me. I think your lead off, Andy, is exactly right. There was less activity in the market and less movement between carriers, which obviously helped. The major drive for us was really around cementing that 5G leadership, and we saw that through the iPhone launch, and we've seen that through now, where all of the major handset vendors are ranging 5G phones almost completely. I think as Andy mentioned, we've now got 1 million 5G phones on the network. We're pretty confident as we move forward now that our coverage is getting to a point where it's a very significant advantage and lead over competitors. It's a meaningful coverage for many, many customers that is influencing their decisions. It has been strong.
I think the other point I would make is we've seen changing dynamics in prepaid with somewhat the removal of travelers and some of the other sort of international student flows that created a lot of noise and excitement in prepaid. We now have a very strong base of prepaid customers that is growing in both value and unique users as well.
Thanks Michael. I'll pass over to Vicki. In the process of doing so, I just did a quick speed test for you, Kane, just got 800 down and 75 up. Over to you, Vicki.
Thanks Andy for that and thanks Kane, for your question. Let me take the last two questions. Firstly, on cost out. Yes, as you said, we have upgraded our cost out to AUD 2.7 billion by the end of FY 2022. What does that mean? We made those broad comments at Investor Day that we would expect future cost out beyond FY 2022. Yes, you remember correctly, I think we did say in the range of hundreds of millions. What we have completed up our last half is we did complete our update of the global cost benchmarking work. Beyond FY 2022, we do still see opportunities. For example, as we complete digitization, we still have the opportunity to decommission future legacy IT systems, and that'll take some future costs out of our business as well.
There are still areas beyond FY22, and so we do believe that cost out and efficiency and that productivity will continue to play a role in the business beyond FY22. Your second question was on free cash flow. How do you think about free cash flow into FY22? I guess a couple of the key components of that. Firstly, we've been clear about our ambition in terms of underlying EBITDA growth, so that mid to high single digit growth. That's an important factor in thinking about free cash flow. Obviously, we've been pretty clear also around CapEx and how we see that evolving from here through to FY23. Two of the major factors. Obviously, working capital, we've seen a significant shift in that in this half.
I would just say hardware volumes do play a really significant part in that, as we've seen in the upgrade to our free cash flow outlook for this year. Working capital's the one obviously will be dependent on what happens with hardware volumes, and obviously there's always a little bit of timing in that. Kane, I would probably point you towards our outlook on underlying EBITDA and our comments around CapEx probably give a good basis to start thinking about free cash into FY 2022.
Thanks, Vicki. Maybe just on the productivity, is it still right to think about AUD hundreds of millions? Or no sort of comment around what we should be putting in there for FY 2023 and FY 2024?
Look, the comments we made at Investor Day right now are still consistent. Obviously, they're ambitions, and as we come and provide future guidance into FY 2022, we can be a little bit clearer. There's no change in what we said at Investor Day.
Perfect. Thanks very much.
Okay thanks Kane. Our next question will be from Eric Choi from UBS. Go ahead, Eric.
Morning guys thanks for the questions and good execution against target so far. Well done. I've got three as well. I might go through them all at once. First question is, just what have you assumed in terms of roaming impacts reverting in your FY22 ambition? My guess is you wouldn't be assuming much. Let's say you do the midpoint of your FY22 EBITDA of, say, AUD 7.2. Sort of infers getting to AUD 7.5 by FY23. Doesn't seem like much of a stretch, given you might have a couple of AUD 200 of roaming coming back and you just confirmed there's more cost out to come. Obviously there's not much NBN drags in FY23. That's kind of the first question. Second question, just on the cost out upgrade.
I guess AUD 200 million of extra cost out should have lifted your ROIC by about 50 basis points, and it looks like we're still saying 8%. It might just be rounding, but I guess a 7.5% ROIC implies 7.5 to an EBITDA, whereas 8% ROIC implies a number above. My question is, have you actually upgraded your internal budgets for that AUD 200 million cost out upgrade, or is it just offsetting a weaker outlook elsewhere? The third question's on postpaid revenues. They're sort of sequentially down AUD 50 million, first half FY21 versus second half FY20. Can you just help me out with how much of a drag briefings out of bundle roaming and new plan accounting impacts were? Because my back of the envelope suggests the three together were easily over AUD 100 million. So your underlying sequential growth was actually positive.
Then going into the second half, is it fair to say those sort of roaming and new plan impacts sequentially become not much of a drag anymore? Sorry for the mouthful. Thanks very much.
No thanks very much Eric. Firstly, thanks for your kind words of support. That's appreciated. Ordinarily, we wouldn't provide those sort of, I guess, as specific an outlook in relation to EBITDA and FY22 and FY23. I made the point that they're not guidance, just obviously because they're going to be subject to more uncertainties and impacts as we look out into that time. Nonetheless, I felt it was an important moment in time as we've sort of reached this turning point that we do try and lay out a little bit of a roadmap as to how we potentially get to that AUD 7.5 billion-AUD 8.5 billion, because I know just how important that is for shareholders, partly because of how it sort of supports an underlying dividend at AUD 0.16 on our current capital management policy.
I'll hand over to Vicki and see if Vicki wants to involve maybe Michael or David in any of her comments, but she can talk through the give some indication what we're doing on or where we sort of see roaming, I guess, and how we're thinking about that, as well as the cost out and how that benefits. The only thing I would say just in relation to ROIC, I know that the market was disappointed when we reduced our ROIC outlook from 10%, and I acknowledge that, and we're pulling every lever. We've set the target at 8%. We can see a path to get there, but I don't want to be overly literal about whether it's precisely 8.2 or 7.9 or whatever the number is.
We're clearly focused on getting to around 8%, and given that's at least a couple of years away yet, two and a half years, I don't want to be sort of too literal on the precise number. Vicki, with those introductory comments, are you happy to try and tackle Eric's three questions?
I sure can. Thanks Andy and thanks Eric, for those questions. Firstly, your first question around assumptions on international roaming in our outlook for FY22. The comment I would make that mid to high single-digit growth in underlying EBITDA is an ambition. The comment I'd make is international roaming could have no recovery from what we're seeing now, and we'd still be in that range. I won't comment any further on where we sit in that range because obviously we haven't given guidance yet on FY22. It is an ambition, I would just, say, make that comment, Eric, that maybe is helpful. No international roaming recovery, and we would still be within that range in how we think about that ambition for FY22.
In terms of the cost out upgrade, I would just echo Andy's comment, remembering our ROIC target that we talked about at Investor Day is approximately 8%. To your point, there can be small movements on that, but it remains approximately 8%. I know you had a detailed question on postpaid handheld and the component parts of it. Those things you mentioned around out-of-bundle revenue, the impact of the accounting changes, the Belong dilution. The first comment I'd make on a PCP basis in the second half, we would expect that to be pretty consistent. I know you were asking a sequential question, and I guess if I just look at what we've said today, firstly, we've said that postpaid handheld reported ARPU, we expect to return to growth in the second half.
I'm pretty sure that will mean sequentially growing in terms of postpaid handheld reported ARPU when you look at what second half was last year. Maybe that helps just frame it up. I know you wanted to get into the details of the sequential movements in all of those component parts, but perhaps that helps a little bit, Eric, with those questions.
It does help. Maybe we can take some of it offline, but appreciate the feedback, Vicki and Andy.
Thanks Eric. If I could ask the analysts just to limit their questions to two, that would be helpful. The next question is from Sameer Chopra from Bank of America Merrill Lynch.
Morning. Yeah, I had just two questions. One is just on, when we think about the sale of the towers business or the divestment of part of the towers business, what's the thinking in terms of capital structure? Do you think that the proceeds will be used for debt reduction? Do you see some way for the proceeds to find their way to shareholders? That's kind of one. The second question is, again, on the asset sales historically. As these assets are sold and leased back, how does that change your operating cost structure? I presume there's leaseback, which means that your OpEx starts to go up over time.
Thanks very much Sameer. Again, I'll get Vicki to add her comments as well. On the sale of towers business, I think you can assume that we will look at the full sort of spectrum of options available to us, partly around strengthening the balance sheet, partly around investing in the future, and partly around returning value to shareholders as well. I would be surprised if there wasn't some element of returning some of the proceeds and value to shareholders. We know that would be an important element of this process. We'll obviously talk about more of the specifics of that when we get into the process in the second half of the year. On the asset sale and leaseback, I'll refer to Vicki, because I know there's probably some complexities in the accounting there as well.
Thanks Andy.
Thanks Andy and thanks Sameer, for that. Just in terms of the asset sales and the leasebacks. Firstly, yes, our outlook's obviously factor in those costs from the leasing back of those properties. Sameer, Andy is right. From an accounting point of view, those lease costs sit below the EBITDA line in D&A now under AASB 16. All of that is factored in as we think about those ambitions into 2022 and 2023.
Great. Thanks Vicki.
Thanks Sameer. The next question is from Craig Wong-Pan from CLSA.
Morning. First question, just on the Belong post-paid net adds. There's only 22,000 added in the period. Just wondering, that's a lot lower than sort of previous periods. Any particular reason for that? Then second question, on the changes to your retail footprint, could you make any comments about what that means for your CapEx or costs? Thanks.
Thank you. Maybe I'll get Michael Ackland to comment on both of those. Michael?
Yeah. Thanks Andy. On the first one on Belong, we removed the AUD 10 price point from market, which represented a significant proportion of net adds in previous periods. We're very happy with our Belong net adds, and as Vicki said, as well as the ongoing impact on ARPU from the changes we've made in market. I might refer Vicki back to you on cost and CapEx impacts of the retail announcement, if that's okay.
Yeah, absolutely. Thanks Michael. Thanks, Craig, for that question. Craig, just to pick up in terms of the retail footprint changes, we obviously are not talking today about the financial aspects of that because the announcement has been obviously that we're not renewing the licensee agreements and looking at that transition of the stores back in. The things I would call out, though, as you start to turn your mind to it, through that transition, we obviously pay commissions to our channels today for the activity that happens in those stores. That will move across and obviously becomes operating costs for us as we transition store leases and staff to be directly employed by us.
Obviously the other factor that we'll come back and talk more about, given we've only just announced it today and there's many discussions and negotiations and things to be worked through, is obviously there will be cash involved in the buyback of those stores. We'll come back and talk more about that at the full year.
Okay. Thank you.
Great t hank you. Our next question is from Entcho Raykovski from Credit Suisse.
Morning, all. My two questions, firstly on TMMC. I'm interested in whether you've seen any significant improvement in TMMC at the end of the first half. I'm just conscious that you were guiding to an AUD 2 increase back in November at the Investor Day. It's now up AUD 3. If you can make comments around what you're seeing into the second half so far, and whether your expectation of the AUD 3 increase into 2H includes some moderation. I guess what I'm trying to get to is, are you being somewhat conservative given there seems to be that acceleration towards the end of the first half? My second question is a broader question on industry structure. Andy, in particular, I'd be interested in your view on whether you think returns in the industry as a whole can improve from here.
You've got your own ROIC target, talking about the industry as a whole, whether you think the NBN means that much lower returns are likely to persist? Thank you.
Thanks very much Entcho. I might refer to Michael and Vicki on the TMMC thing. I think a lot to do with our success in our 5G strategy. Michael and Vicki, do you want to comment on that? I'll come back and make a comment on the industry structure point.
Sure. Why don't I go first and then Vicki can add detail. I think from a market and competitive position, we absolutely saw an acceleration in the back half of the second quarter, post the iPhone launch, particularly, where we saw a strong uptake, as we talked about at Investor Day, that we were expecting around the AUD 65 and above plan, with people taking the 5G plan. We saw that continue. As we look into the second half, there are further device launches, such as the Samsung device, and we're seeing continued strong demand around those 5G plans. I think the strength of the 5G story was definitely an acceleration on that in the second half, and obviously late in that second quarter, given the delayed iPhone launch.
Yeah, and if I just add, Michael Oops.
Sorry. Entcho 's questions, are we being conservative with our AUD 3 outlook for the second half?
Do you want me to answer that one, Andy, or Michael?
Yes, please.
Why don't I talk to that? I think as Michael has highlighted Entcho , there's no doubt TMMC across can move around month to month, just dependent on launches of devices and mix of plans. What I would say is, in terms of that second half, it is another AUD 3 lift PCP. We are confident in that, and it is consistent with what we're seeing in the early part of this half in trading conditions. I wouldn't say we're conservative. I think we're confident, though, that another AUD 3 increase is reasonable and certainly accords with what we're seeing in early trading for this half.
Just, Entcho, to your comment on industry structure. Obviously, it's not just us that's faced the headwind of the migration to the NBN. Others in the industry have also faced it as well. It's been more significant for us partly because of our scale, but also of course, because we were the wholesale provider of fixed broadband services in Australia. We lose the wholesale business and a very different margin on our retail business, compared with the other players. Typically, in the main, it's just a different margin on their retail business. As we sort of talked about, that is still washing through. For us, the NBN headwind, it's obviously lower this year than last year. It will be quite a bit lower again next year, and there's no reason why it wouldn't follow a similar pattern for the rest of the operators in the market.
Of course, enterprise as well. It's probably coming through at the tail end of the rollout of NBN in terms of the impact relative to perhaps consumer and small business. I think there'll continue to be pressure on ROICs for the industry as that happens. Then hopefully from there, certainly we're forecasting that our ROIC will start to improve. That's probably more to do with mobile than it is to do with fixed, although, as you know, we are targeting a mid-teens reseller margin on fixed as well. So we will see, I think, some improvement in ROIC, and I guess others will as well. I do still think, my comments that overall industry economics are challenged, remains. I do think that with the NBN rollout now complete, it is the right time to, as I've spoken about before, to review the pricing structure.
Stephen Rue, I think, and his team have done a good job. I want to acknowledge that in terms of the rollout. We saw NBN's results, obviously yesterday. I think it's important that we get the economics for the whole industry right, because the infrastructure is going to be crucial to obviously underpinning the growth in the digital economy for the whole of Australia.
Thank you. Our next question comes from Fraser McLeish from MST Marquee. Go ahead, Fraser.
Great. Thanks. And thanks for the comments on FY22 sort of aspirations. Very helpful. Two from me. First, Andy, just on InfraCo monetization. You've obviously got a firm commitment on the towers. After that, I know there's no commitments, but is it fair to say there's a kind of broad aspiration to monetize the other parts of InfraCo, or am I reading too much into that at this stage? Then my second one is just maybe for Michael on the point you got on 5G, that AUD 65 price point. Neither Optus or Vodafone are charging a premium for 5G, and they're obviously going to narrow that network gap with their coverage. Do you think that AUD 65 price point is sustainable? Thanks.
Thanks Fraser. I'll let Michael come back to the AUD 65 price point. They're not narrowing it at the moment. I think we're broadening it, and I think our plans are to continue to lead with our superiority in 5G. I'm confident that Michael can continue to sustain that for a while anyway. I'll let him speak specifically on it. On the InfraCo monetization, obviously, as you guys all know because you're very experienced with these things, that the group restructure is a pretty complex exercise. We'll talk about the steps that we need to go through when we come back and talk a bit more about that in March. That's our fundamental focus for now on that. With towers, I think towers is following a parallel path, if you like, and that's a little bit more straightforward.
We're confident we can move forward on the monetization there. Suffice to say, we wouldn't be doing this on InfraCo if we weren't looking to try and maximize value for shareholders. Setting it up in this way really does that, I think gives us the potential to do that as well as create greater transparency and increase optionality. We've talked about the different types of optionality on that. Without going as far as saying, yep, we're committed to do a monetization initiative on InfraCo Fixed when we've set up the restructure, I would say that obviously we'll be looking at ways in which we then crystallize and maximize that value once we've done that. We want to go through the steps of actually setting the thing up first, because there's quite a bit of complexity in doing that.
Michael, anything you want to add on the 5G, our continued leadership over the rest of the industry and the wonders of the Telstra 5G network and our population coverage and all those great things?
Yeah, no. Absolutely. I think we talked about this at Investor Day, which is that we remain confident that we can continue to demand that premium for the 5G experience. Largely driven around coverage and, as Andy pointed out, the speed experience we've got across both our 4G and 5G networks. It's one thing to not charge a premium for 5G if you're our competitors, which I think is probably reasonable from their perspective given we're sort of at a, what will feel like to a consumer, an order of magnitude difference in coverage. I think our target to get to 75% population coverage by the end of the financial year, which we've talked about before, is incredibly significant in terms of doing that.
All of that said, though, we do not rest on our laurels there, and we are continuing to work on the network experiences and the other experiences we give to support that premium. We recognize that continues to be important. I agree with Andy, I don't think the gap is yet narrowing, and coverage in mobile networks is absolutely the critical differentiator. If you can't get access to that very fast 5G network, then there's not much point in getting it. We're focused on it and I think we can maintain it, but we're not resting on our laurels, and we have a pipeline of activity to continue to support that premium.
Great Thank you. Our next question is from Roger Samuel from Jefferies. Go ahead, Roger.
Well hi morning, guys.
Morning.
Thanks for the questions. First one is around fixed Enterprise. I'm concerned that this division keeps on declining over time, and I'm wondering when it will stabilize. It looks like in data and connectivity, there's still a lot of SIOs on copper, and also on NAS, you mentioned there's some legacy calling applications. Yeah. I'm wondering when this division will stabilize in terms of EBITDA. Second question is on tax rates, which is lower in the half just passed. I'm wondering what's the guidance for the tax rate in the second half. Do you still expect to use any of the capital losses?
Thanks Roger. I might get David, who's with us, as I mentioned, who's recently taken over heading up the Enterprise division, to comment on your question on Enterprise. Vicki may have a comment to add, Vicki can definitely take the question on the tax rate. Thanks Roger. David.
Thanks Andy. Look, firstly to acknowledge it was a pretty tough half in the fixed enterprise segment, don't want to shy away from that. We break down your question as you asked, Roger, into two areas, data and IP. At Investor Day in October, we did talk about that this is a space that we will continue to disrupt. The market is being disrupted. We'll disrupt ourselves. We do expect some of this declining activity through to FY 2024. We talked about that we'll have this portfolio back into growth in the FY 2024 period of time. We did, as Andy also mentioned, launch later in the year our Adaptive Networks portfolio of offerings, which is designed around our own Telstra Fibre and NBN offerings, which we're very excited about.
They were late in the half. We're looking forward to the actions and improvements from that area. Secondly, in and around the NAS questions that you had. We did see an acceleration in the decline of the application calling platforms, and that had started pre-COVID, and it continued through that period of time. We also saw some expected milestones, but impacts in costs in and around some of our major contracts, but they're one-offs, and we're confident and comfortable with that. Lastly, as mentioned, the delay in some of the cost activities in supporting our employees and organization. There is also some fantastic green shoots in amongst that NAS performance, which is in line with our strategy around our managed services portfolio and our cloud portfolio, which are very much part of our strategic direction.
We call it our Next Generation Growth or NGG in the NAS portfolio. The other comment I'd give you is, we've announced in literally last week some of the actions around cost and cost management and reorganizing the Enterprise business and the flow on activities that support those two portfolios. As Vicki said, we're reaffirming that we'll have our NAS business back into the mid-teens EBITDA margin in FY 2022. Lastly, I'll reconfirm what we talked about at the Investor Day, was that as a portfolio of the Enterprise business, that's mobility, data and IP, NAS and international. We will get ourself back to growth at both a revenue and an EBITDA line in totality in next fiscal year.
Whilst it was a tough half, and again, acknowledging that, there are the green shoots that support our strategies and actions and even some actions of the last couple of weeks that I think reaffirm where we've said we will hit in the next years to come. I don't know, Vicki, whether you've got anything you add to that.
No thanks David and thanks Roger, for the question. I think David's covered it exceptionally well. Roger, I would just, in the appendix, there is a slide that does summarize all of our product commitments and ambitions. It's updated from the one we showed at Investor Day. It's page, I think, 29 in the slide pack that we've issued called Building Momentum and Confidence. All of those things David just spoke about for the enterprise business plus more covering mobile, fixed wholesale, fixed C&SB, et cetera, is all there summarized. That may be a helpful one, too, just as you're working through your thinking. On your second question in relation to effective tax rate. Yes, in the first half, as you've rightly seen, our effective tax rate was low.
That was purely due to some of those M&A and asset sales that resulted in us using capital losses. Our underlying effective tax rate, excluding those things, was close to 30%. We obviously, for the second half, we don't give guidance on M&A, but I would continue to expect the underlying business to have an effective tax rate around that 30% mark. They might be helpful comments as you just think that through.
Yes, that's great. Thank you.
Great Thank you. Our next question is from Ian Martin from New Street Research.
Good morning. Look, encouraging result. Just a couple of questions. Turning to mobile, I notice on your website you've got an upfront plan, and I've seen a few industry reports that suggest that might become the main form of financial arrangement, particularly with the post-paid market. I wonder if you can comment on that and whether that's got an ARPU implication. Secondly, looking at fixed, particularly fixed consumer, but also fixed enterprise and your expectations of declining headwinds. That seems contrary to what we've heard from NBN yesterday. They want another AUD 800 million in revenue next year, with, I think, about a AUD 4 increase in blended ARPU and a AUD 10 increase over four years. I don't know how realistic that is.
Even the ambition to achieve that must create some uncertainties about what you might achieve in relation to a better outcome once we get through this NBN migration.
Thanks very much Ian. I'll get Michael just to talk about the upfront plan that you referred to in a second. Just a quick comment from me on the sort of the NBN ARPU point and the headwinds. When I'm referring to the headwind, I'm being sort of relatively Well, not relatively, I'm being very literal in the sense of how we have defined the headwind, which is a function of the process of the migration to the NBN, and the previous calculations and estimates that we've provided on that, i.e., up to sort of approximately AUD 3.5 billion ultimately of impact. We try to sort of update that every year or at least provide you the in-year headwind. That sort of reflects the migration.
What you're obviously referring to is the ongoing plans of NBN to lift their ARPU, which is effectively the cost that gets imposed on the rest of the industry and will flow through into our economics. You're absolutely spot on. You know my view on wholesale prices. They are too high, you can see that playing out. That's why I think we believe now is the right time, particularly with the end of the rollout of the NBN, to review that pricing structure. We look forward to consulting with the NBN on that. As far as what it means for us financially, certainly in our mid-teens NBN EBITDA margins, we are looking to achieve that against the background of we've known what NBN's corporate plan is and what their ARPU plans are.
Certainly, I think wherever that was going to get to about, I think AUD 49 or so over the next period of time. I think the challenge will be is that ultimately part of that is probably going to mean that consumers are going to have to pay more. They're certainly going to have to pay more if ARPUs go up AUD 10, that's for sure.
It is of further economic headwind in that sense, but it's not part of the headwind that I was sort of referring to, which occurs in the process of the migration, if that makes sense. We've got line of sight of it. With that said, why don't I hand over to Michael just to take that point on the upfront plan, and I'll see if Vicki has anything to add on what I just said around NBN impact on ARPUs and how that flows through into our outlook, et cetera. Michael, on the upfront plan.
Yeah, sure. As both Andy and Vicki talked about, we've been progressively migrating our customer base onto our in-market plans where we have one of our 20 plans, the no contract plans, and no back book. Once you're on those plans, we don't have a back book, and we're 70% of the way through the base. The next step of that is then moving those plans onto the new technology stack. What you're seeing on the website with the upfront plans is some different constructs around the new technology stack. I'd probably refer to it as the next step in moving to that subscription model, for service plans, and giving customers absolute price certainty. Our view is that that will not have an impact on ARPU, and that the price and inclusions will be exactly the same as our in-market plans as we move forward.
It is the next step of moving to that absolute price certainty, no back book, simplified plans on the new technology stack. What you're seeing is the digitization program really move into absolute scale rollout as we start to talk about those plans. I would think about it as this is the next step on the move to subscription plans and absolute price certainty for customers. No back book.
Thanks Michael.
What's the time frame for that?
We will start as we're piloting now. We'll be piloting and then over the next, I think six, 12 to 18 months, we will be migrating and moving customers across onto the new technology stack as we've talked about previously. From the fourth quarter, we will be transacting on our new technology stack for new and recontracting customers.
Thanks Michael.
Great. Thank you.
Vicki, anything to add on my comments on the NBN ARPUs?
Yeah thanks Andy. Just I'd only add one comment, Ian, to what Andy said. As we look at those NBN headwinds that we estimate, a key input to that is the NBN corporate plan and their assumptions around wholesale pricing. That's certainly an input and factored in as we look at those headwinds. I think Andy's covered it very well. That's all I wanted to add, Andy.
Thanks Vicki. Thanks, Ian.
Our next question is from Brian Han from Morningstar. Go ahead, Brian.
Thanks, Nathan. Andy, given Optus' purchase of Amaysim and the launch of Gomo and Felix by your competitors, how do you see the competitive landscape developing this year in that value end of the mobile space? How do you think that will influence the pricing of all the mobile operators' main brands over the next couple of years? If I could, second question, in terms of the upcoming 5G spectrum auctions this year, have there been any rumblings from the regulators in terms of limits as to how much Telstra can bid for? How do you think the prices will compare this time around versus the 3.5 gigahertz a couple of years ago? Thanks.
Thanks very much for your questions Brian. On the second one first, just to knock that over, I cannot comment on views or pricing or anything else in relation to the spectrum auction, ahead of the auction for governance reasons. I am not aware of any communication in relation to competition limits. That's probably as much as I can say on 5G spectrum. Not wanting to be unhelpful, but I am constrained because of our position. On the value end of the market on the various different activities, what I would say is I think it validated, frankly, our strategy that we employed three years ago when we launched Belong, where we do see a little bit of a bifurcation in the market between the value end and the premium end.
That's why our wholesale business and our Belong business, we're able to compete at that end of the market, and we've done so successfully. Another 160,000 odd subscribers on our MVNOs, which come under Brendon Riley, who's here, and Brendon may want to comment on that in a second. Also obviously with Belong has continued to grow strongly. We took, as Michael said, the AUD 10 plan out of the market, and that's reduced the activity a little bit, but we're doing that to endeavor to lift ARPU both at that end as well as at the premium end as well.
I think providing we continue to offer a premium differentiation, which is exactly what our 5G strategy and all of the other service aspects of the Telstra branded proposition provides, then I think on the one hand, we can compete successfully at the value end for customer numbers and subscribers, but yet continue to grow our branded end of the market as well, which we absolutely did in this half. I think that it validates that strategy playing out. I think what we're seeing is actually our competitors recognizing that, too, and looking to compete at either end of the market from their perspective as well.
How that continues to play out, always hard to predict these things, but I think it's been exactly the right strategy for us, and enables us to protect our premium on the one hand, and offer customers a very differentiated service at a different value point, and yet continue to serve customers at the value end. Brendon, I'm not sure if there's anything else that you wanted to add, what you're seeing sort of in the wholesale side but you had another successful half there.
Thanks Andy. To Brian, look, the MVNO part of the business is performing really well. Our trends over the last couple of years have been very, very consistent. They continue. We're ready to support our MVNO providers in market with whatever changes and adjustments they need to make to respond to all the market activity. I think we're in a good position. We've got good momentum, and I'm confident we can sustain it.
Thanks, Brendon. Obviously, I think you know, that we differentiate our network for the Telstra branded as well as Belong and the MVNO. Not only we're differentiating obviously at the brand level, but we are actually differentiating at the customer experience level as well. Look, thanks very much, Brian. Nathan, back to you.
There's no more questions on the line. With that, I will close the analyst and investor Q&A. We'll pause for a short break, and then my colleague, Nicole McKechnie, will lead a media Q&A. Before we do that, Andy, do you have any closing comments?
Thanks very much Nathan. Thanks to you. Thanks to everybody for hooking in. Thanks to my colleagues, Vicki, David, Michael, and Brendon. Next time we'll be talking to you will be in March when we talk a bit more about the group restructure and we'll have Brendon there. I notice we didn't get as many questions on that today, that's probably because we foreshadowed making further communications on that then. We'll get to hear more from Brendon then as well. Thanks David t hanks Michael. Thanks, Brendon. Thanks, Vicki. Thanks to the investment community for your support.
Good morning and welcome to the media part of our presentation this morning. Thanks very much for joining us. We do have some media on the line. Just a reminder that if anyone does have questions to make sure you register and let us know that you're online and ready to ask. I think we do have one question so far from Lucas Baird from the AFR. Go ahead, Lucas.
Hi, guys. I just wanted to ask about the T22 progress. I think you said there's still 20% that you're still sort of behind the targets at the moment, not really on track to get to at the moment. I was just wondering what's the strategy to make sure you do reach those? Andy, specifically, what happens to you if you don't reach this sort of last 20% of T22? Is 80% of T22 good enough for you? Is your job on the line if you don't achieve 100% of T22? Is there any sort of perspective you can give us on that?
Thanks Lucas thanks for joining us this morning. As you can imagine, with any large, complex transformation program the nature of T22, there's a lot of measures. In fact, on our scorecard, I think we've got something like 60 measures, which basically are tracking our progress in relation to digitization, customer experience, the financial returns, the realization of assets, strengthening the balance sheet, transformation of our workforce, adopting agile ways of working. It covers the whole spectrum. Pragmatically, you're not going to hit 100% of everything. The most important thing, though, is that we're holding ourselves accountable and we're measuring our progress. I would say that being on track or having achieved 80% of the goals that we've set ourselves is pretty good progress. We will continue to focus on the ones where we're not, and I already touched on some of those.
Some of the stuff on digitization is right, shifted a bit. Ultimately, I'm very confident we're going to get the benefits from the digitization program. Again, in a systems build, we are spending AUD 1.7 billion. I challenge anyone to sort of deliver everything that they thought they were going to deliver exactly as on time and to budget. One of the things that we're doing with our T22 scorecard is it is externally audited so that we are very, very sort of literal in terms of our progress. I would say it's a good discipline. It's a good overall measure of how we're tracking. Ultimately, what's going to matter most as we transition to the completion of T22 and we announce what comes after T22, is actually how that translates into results for our customers, for our shareholders, and for our people.
For our customers, we had some challenges with customer experience last year, but we're building momentum there. Again, strategic NPS is up five points for the six months, 11 points for the 12 months. As we complete the digitization program, the migration, that's going to further continue to improve. We've talked about the other areas. We've upgraded our productivity targets. I think in terms of how we're going to deliver for our customers, I can see that coming through. How we're delivering for our shareholders, I think is very much embedded in the aspirations and ambitions that I talked about today, about mid-to-single high digit growth in underlying EBITDA next year and hitting that AUD 7.5 billion-AUD 8.5 billion EBITDA range for the year after. For our people, our engagement scores are at very high levels at 80. I think we're in good shape.
As far as my own plans, as I said, we look to announce at what comes after T22 in November, and I look forward to doing that.
Thanks, Lucas. Anything else from you? You all good?
I think that's it.
Okay, perfect. Moving on to Zoe Samios from The Sydney Morning Herald. Good morning, Zoe.
Good morning. Thanks, Andy and Vicki, for the presentation. Just two questions from me. One is just around the NBN Co headwinds. I know you spoke about that before, but I just wondered if the worst of the NBN impact is now behind for Telstra, just given the rollout's pretty much near complete, or I think there's only a couple of thousand premises that they have to fix things in. The second question was just around that announcement on the retail stores. Does that have anything to do with what occurred with some of the licensees and indigenous customers and the ACCC case last year?
Thanks very much, Zoe. Look, on the NBN headwinds, there's probably a couple of elements to it. The first is that structurally what the creation of the NBN means is that the NBN was created to effectively provide a service that Telstra used to provide. At a practical level, what that means is that we have been progressively transferring over to the NBN that part of our business. Because it's happened transitionally over time as the NBN's been rolled out, the financial consequences of that have obviously been happening as it rolls out. Now, because NBN they're technically complete, that element of the headwind doesn't change until everybody is connected. There's still some flow-through of that to happen, which will go into our next financial year.
That element of the NBN headwind is largely behind us, or we can see the light at the end of the tunnel. By the end of this calendar year, that's going to be largely behind us. Going forward, of course, every operator in the country has to pay NBN for essentially the access to the fixed infrastructure, which sits behind the broadband services that they're providing our customers and providing to us. Then we resell that to our customers. The NBN ARPU continues to increase, what that effectively means is that how much the operators are paying NBN for access to that is increasing. I think you heard from one of the analysts say today that NBN I didn't see this person myself. I haven't actually had a chance to go through yet.
NBN are foreshadowing that ARPU to increase AUD 4 in the next period and up to AUD 10 over the next three to four years. I think that's going to translate directly into cost into the industry. Unless the industry obviously increases retail prices and passes that on to customers, that does create an element of a headwind. That will continue to be the case if that ARPU continues to increase, which is why I say I think now is the right time to very much review that pricing. I know that NBN are considering a pricing consultation. I think they may have even announced it, or certainly it has been referenced yesterday from what I can observe. On the second point about bringing the retail stores back in-house.
Fundamentally, this is about really thinking about what's the next stage of the retail experience and how do we really integrate that with what we're doing digitally as well. It does give us much more control and consistency over the retail experience that we provide. Yes, there's no doubt that it will also help us, I believe, mitigate that type of situation you referred to, where some of our licensee stores, unfortunately, mis-sold post-paid mobile handheld plans to a small number of indigenous customers. It will enable us to have better control of that dynamic within our network as well. It's not the reason we're doing it, but it is one of the benefits that come from it as well.
Great. Thank you very much.
Thanks, Zoe.
Thanks, Zoe.
Thanks, Zoe. We're going to move to David Swan, but just before I do, just a reminder, if you have any questions to press star one and make sure you register them. Dave from The Australian, over to you.
Thanks, Vicki, thanks, guys, for the time. A couple of quick ones from me. Firstly, wanted to ask a bit about the dividend and obviously for you, Andy, it's always a decision that needs to be made. Can you give us some color maybe just in terms of, I guess, the decision to keep it where it is? A bit of sort of what went into that. Secondly, I wanted to ask a bit about just you've had a couple of executives depart over the last sort of six months to a year, Mark Labeed, Håkan Eriksson. I wanted to ask if you're sort of satisfied with the level of talent that you've got around you at the top.
Sure. Thanks very much, Dave. Look, firstly on the dividend, just a couple of comments. I think the first thing to say is that we understand and know and appreciate how important the dividend is to our shareholders, to all shareholders. We have a very large retail shareholder base as well, probably the largest of any company in the country. I think somewhere in the order of 1.3 million shareholders, and we know that the dividend is particularly important to those retail shareholders as well, so we take any decision on it very seriously. We know that to achieve AUD 0.16 purely out of underlying earnings, we need our EBITDA to be in the range of AUD 7.5 billion-AUD 8.5 billion. At the moment, it's not in that range.
We're able to afford the dividend because we've also got the benefit of the one-off payments coming from the NBN, and we committed to provide 75% of those going back to shareholders as well. As those run off, we need our underlying earnings to increase to that AUD 7.5 billion-AUD 8.5 billion range to make sure a AUD 0.16 dividend is sustainable in the context of the current dividend policy. What I outlined today was the pathway for us to get there, which is why the directors were happy to, and believed it was important to, communicate not only the interim dividend today but confirm that we will pay a AUD 0.16 dividend or we expect to pay a AUD 0.16 dividend for FY 2021, even if we're outside of our range. The other thing that we can also actually point to is our strong underlying free cash flow.
Whilst our dividend policy is a function of underlying EBITDA, obviously it needs to be funded out of free cash flow. It's AUD 950 million for an AUD 0.08 dividend, so it's obviously AUD 1.9 billion for an AUD 0.16 all up for the year. We had very strong free cash flow in the first half, up 88% to about AUD 1.9 billion. As you heard, Vicki has increased the guidance for the second half to, I think it was AUD 3.3 billion-AUD 3.8 billion, or roughly in that range. Therefore, whilst we're out of dividend policy in FY 2021, the board has the discretion to make that decision to continue to pay. That's some of the thinking and the factoring in. I think it's unusual for the board to provide an outlook, if you like, on dividend, even for the second half, which is what they did today.
That's just really in recognition of understanding the importance of that as well. In terms of the team, I'm incredibly pleased with my team. Each of my direct reports, particularly those that are running a business unit. You need to put this in the context of Telstra, which is an incredibly large business. Each of the businesses that they run is as big as any other, pretty much most other companies in Australia. They're very, very senior executives. They have very challenging and demanding jobs. Just the enterprise business itself is an AUD 8 billion business going through technology and other disruption with very significant competitive dynamics. Our retail business, likewise, very, very big business, as well as, of course, our wholesale and InfraCo as well. They're all the biggest of their kind in the country, in the sector. We've got very, very significant talent.
Not everybody is successful in those roles and it's not necessarily for everybody all of the time. It's an intense and demanding role, but I'm very pleased with the team. We've got very good talent in the team, and we've got some really good additional and new hires coming in as well. I've never been happier with the team here at Telstra. Thanks, Dave.
Thanks Dave.
Okay.
Okay. Thank you. Next up, we've got Simon Dux from CommsDay. Go ahead, Simon.
Hi there Andy. I just wanted to ask about the TowerCo's potential sale. The key thing around there is there's obviously a lot of institutional investment into tech at the moment. We've got the moves from Macquarie, Myer on Vocus. Aware Super have also had a couple of cracks and have talked about a potentially AUD 200 billion fund they're looking to put somewhere. My question is, are you comfortable with the timing or is there a fact that you could potentially accelerate the timing of bringing TowerCo to market? The reason I'm asking is also there are potential threats around a tech bubble burst with an antitrust case starting to kick something off a bit later in the year. Does that have you in danger of potentially not maximizing the return?
No, look, thanks very much for the question. Firstly, no, I'm not concerned about the timing and I don't necessarily think we should accelerate it. I think what we have seen, I think it's been a trend over a decade or so, but in particular over the last four years, we've seen very, very strong interest in underlying infrastructure assets. As interest rates have come down dramatically globally, and notwithstanding all of the challenges of COVID and the economic consequences of it, the weight of capital that is available that is looking for secure returns, albeit lower returns than maybe historically, is very, very significant. Infrastructure assets stack up really well in the context of that.
Of course, the reason telco infrastructure assets are so in demand, if you like, is because people know that regardless of the fortunes of the various different competitors in the telco industry, the one thing that's not going to change is the demand for growth in connectivity and the ground for growth in coverage and mobile connectivity. As I said, we've got the largest mobile tower business in the country. It's incredibly strategically important to us. It's very much integrally linked with the success of our mobile's business, which is not only the best in Australia, but one of the best globally as well.
We need to set this up in a way that we maximize the value of the towers for our shareholders, but we also ensure that we maintain our strategic competitive advantage and differentiation and all the things that we've invested in from the point of view of Telstra. Commenting on a tech bubble burst or not, I wouldn't make a comment on whether there would be one or isn't one or that sort of thing. There's people that follow this more closely than I do from the investment point of view to be better informed. I don't think that's going to change any of the underlying dynamics regarding ultimately the demand for infrastructure assets and their importance going forward into the future, if that helps. Our timing is we're well progressed.
There's a lot of work involved, a lot of due diligence and stuff needing to get ready. We expect to launch the necessary documentation and the process of the actual engaging with potential investors in the third quarter of this year, which isn't very far away, obviously, and then with binding offers in the final quarter.
Thanks, Simon. Okay. Next up we've got Jen Hewett from the AFR. Hi, Jen.
Good morning. I know that you've talked about this before, Andy, I'm just still perplexed by this kind of seems to be very different approach by NBN and by you. You say the time is right for renegotiating the price structure, that seems to be absolutely not in NBN's repertoire in terms of it raising its own ARPU. Do you think that there is any alternative but really for consumers to be paying more for their services? On the mobile market, again, I can't remember, I think you said 70% or so were on the AUD 65 plan at least. Can you see prices for the mobile market in the premium end ever declining, or staying flat, or are they going to just continue to increase too along with services?
Thanks Jen. Look, on the first one, ultimately you're right. If NBN ARPU continue to increase in line with the pricing structure, consumers will have to pay more. That is the bottom line. If you model it out to the second half of this decade, and they continue to use the same structure, wholesale prices will be higher than current retail prices. It doesn't matter how much the operators reduce their costs, that's just obviously not going to be sustainable. Something has to give. I believe that NBN have foreshadowed that they will enter into a pricing consultation.
They obviously are not committing to where that's going to end up. I hope that through that pricing consultation, we can start to see some reasonable changes that actually lead to an industry where consumers can win. They don't have to face significant increases or increases to access. The operators can win. They can actually make a return on capital that means that they can satisfy their shareholders. Ultimately the NBN, which I accept and I know, and this will be Stephen's point, needs to generate free cash flow to focus on upgrading the network over the long term as well. That pricing consultation discussion's going to be very important. I strongly encourage that to occur.
On the second point on mobiles, in fact, what we call sort of mobile ARPUs, which is a bit of a function of the weighted average of pricing, has in fact been coming down over the last few years, since about 2015, 2016. At the same time, customers have been getting more for their price, if you like, by virtue of getting more data. I mean, if you think about three or four years ago, we were talking about data volumes or data packages, if you like, on some of the higher value plans of about one or two GB of data per month. Now, we're not allowed to call our data plans unlimited, but at a practical level, there are no excess data charges on our new plans in the market for mobile customers. They are getting more, and prices have come down.
I think they've probably come down too much to basically be able to afford the reinvestment in 5G and stuff. We are seeing a little bit of a change and a move to customers paying a bit more at the moment. Also, they're voluntarily going to the higher value plans because it's giving them access to 5G and some of the other services as well.
Okay. Just on that last point, the idea that you'll get back to a margin of mid-teens reasonably quickly, again, I can't quite see. That seems to be overly optimistic, doesn't it, given you're only on 5% now?
Jen, we've got to take more cost out. That's what we've got to do to get there, as well as move customers up some of the higher tier plans and try and monetize some of the other aspects of the service. I mean, things like the Telstra Smart Modem that we provide, which is in 80% of our customers' homes. It's actually quite an expensive device, and it offers a great service because when NBN goes down, which it does, it basically makes sure customers continue to be connected. We've got to find ways to monetize the other value that we provide, the Wi-Fi guarantee that we're providing as well, increasingly, entertainment, Telstra TV. Overall, we can see a path to getting to mid-teens EBITDA margins. You're right, that will continue to be challenged if the NBN ARPUs just relentlessly keep going up every year.
Thank you.
Thanks, Jen.
Thanks, Jen. We don't have any further questions. I will just do one quick call to make sure that no one else has anything burning on star one. If I don't have anything come back in the next couple of seconds or so, then we might call it. No, we're going to call it. All right. Thanks everyone for joining today, and have a great day. Appreciate your time. Cheers.
Don't go. Tell me that the lights won't change.