Two years of admiring the slide; one morning with a calculator. TPG closed down about ~8% on Investor Day – the market having finally totalled a $2.1 billion spectrum bill, a flat premium base, and a return on capital below its cost. The deeper story is a company quietly rewriting its own scoreboard: merging premium and half-price subscribers into one line, redrawing the regional map, and reaching for “intention” wherever a number used to be. The romance of the doubled network has met the spreadsheet. Then, a fortnight later and by way of demonstration, the network went down nationwide. The arithmetic has arrived – and it does not leave early.
TPG Telecom has spent the better part of two years asking investors not to do the maths. It offers, by this masthead’s count, four separate definitions of profit, and gravitates to whichever one is least embarrassing on the day. It reports postpaid and digital-first as a single blended figure so that nobody need notice the premium brand isn’t growing. It declined – twice, on the FY25 call – to give analysts the subscriber and ARPU assumptions behind its own guidance. The whole apparatus is built to keep the audience admiring the slide rather than reaching for a calculator.
On Investor Day 2026, the audience reached for the calculator anyway. The stock closed down about 8%.
This is what happens when the arithmetic finally turns up uninvited. It had been waiting outside for some time.
The Weighing Machine
There is a line of investing wisdom, the sort Warren Buffett has spent sixty years repeating to anyone who will listen, that over the long run a share price tracks the underlying performance of the business beneath it. The market is a voting machine in the short run and a weighing machine in the long.
The operative word, and the one TPG would prefer you skimmed past, is underlying – the real performance, not the headline performance, not the version assembled by the bean-counters in the good suits.
And the gap between those two things, at TPG, has become a chasm. You could park the $2.1 billion spectrum bill in it and never hear it land.
This is a company that reports four separate definitions of profit and reaches for whichever flatters the day. It quietly pulled mobile churn out of its decks for two reporting periods, at precisely the moment churn was the number that mattered most – the corporate equivalent of removing the thermometer because you don’t care for the temperature.
It trumpeted operating free cash flow that had “nearly doubled,” a figure inflated by a one-off $687 million handset-receivables sale to a Macquarie trust, while the underlying free cash flow it didn’t put on the celebratory slide languished around $604 million.
It parks inconvenient costs below the line as “discontinued operations” and outside the statutory result, and continued to reframe Vocus more than a year after selling the thing – a transaction swapped, restated and re-presented so many times the accounts now read like a magician asking you to keep your eye on the wrong hand.
Vocus was sold the better part of two years ago, and yet it went on appearing in the results for reporting periods well after the deal was done – one does wonder how many times a business can be sold and still turn up to work.
And the wrong hand has plenty left to show you: small one-offs sitting quietly outside the statutory result, a larger slug of deal and separation costs resting “outside guidance” as discontinued items, NBN swapbacks still threading through the Vocus numbers, and the handset-trust changes washing through cash flow.
Room, in other words, to manage the EBITDA and NPAT optics in the near term – an unusual number of levers for softening a bad quarter. The question was never really whether the underlying performance shows up. It is how long a sufficiently determined finance team can keep smoothing it before it does.
For two years, all of this kept the headline tidy and the slide admirable.
The share price, in the end, declined to be fooled. It weighed the business, not the wizardry – the spectrum bill, the Fixed losses, the growth that never came – and found about 8% of it wanting in a single afternoon.
The Bill That Was Always Coming
The headline of the day was spectrum. TPG confirmed that renewing its spectrum licences will cost approximately $2.105 billion between 2028 and 2032 – $840 million in 2028 for the 850/1800MHz bands, $570 million in 2029 for 700MHz, $315 million in 2030 for 3.4GHz, and $380 million in 2032 for 2GHz. The payments fall due, with a sense of theatre, two months before each licence lapses. There is no negotiating with that calendar. The spectrum is leased from the government, and the government, unlike a long-suffering shareholder, does not accept payment on an underlying basis.
Set the $2.1 billion against the company’s own cash generation. Free cash flow sits at roughly $604 million before the dividend, before bank interest, before anything else with a claim on it. The single 2028 payment is larger than a full year of that (underlying) cash flow on its own. Which means the money comes from borrowing – the same borrowing TPG spent the past two years proudly unwinding, having sold its fibre assets for $4.7 billion and returned $3.3 billion to shareholders to celebrate a balance sheet it is now preparing to re-leverage.
For scale, because scale is the entire point: Telstra faces a comparable spectrum bill of around $4.1 billion, against roughly $2 billion in net profit. TPG faces $2.1 billion against an underlying pre-tax profit of about $7 million. One of these companies can write the cheque out of earnings. The other will be writing it out of a credit facility, on a business that does not currently earn its cost of capital, with a dividend already running at – and one does have to read this twice – around 644% of statutory profit.
The market did that division on the day. Hence the 8%.
The Composition Problem, In Four Lines
Beneath the spectrum headline sat the 1H26 trading detail, which management presented in the manner of a stable result and which, on inspection, was nothing of the sort.
Vodafone postpaid: flat.
Vodafone prepaid: down 35,000.
The digital-first brands: up 50,000.
The $40 million Ali Wong campaign, the emu, the “98 whatever percent,” all of it, has not moved the postpaid base by a single net customer – the front-book additions, such as they are, washing out almost exactly against a back-book bleeding to price rises and plan rationalisation, and arriving at lower ARPU besides, so that the net position barely stirs.
And the half-price promotion – 50% off, running roughly a hundred days from early March – has, after all that discounting, produced a postpaid result best described as flat.
When you cannot give the product away at half price and still move the needle, the problem is not the price. The problem is the product, and the fifteen years of reputation stapled to it.
Management’s response, per the day’s materials, is to pivot to cost-cutting to defend its EBITDA guidance – to hit the number by subtraction rather than addition.
This from a company whose own TIO complaint volumes remain elevated, whose service is the thing under strain, and whose cost base is the thing it now proposes to strain further.
Defending a growing EBITDA line with the cost ledger rather than the revenue ledger is an ambitious trick to pull off twice, let alone for the years of guidance TPG has staked on it – there is, after all, a finite supply of costs to remove, and an apparently infinite supply of customers willing to leave. One cuts one’s way to a margin. One does not cut one’s way to a customer.
The Arithmetic Behind the Optimism
An investor day is, by design, an exercise in framing. Slides are chosen, metrics selected, and the eye guided gently past the lines that do not flatter. There is nothing sinister in this – every company does it – but it does reward a reader willing to do the translation. So here, in the spirit of public service, is the translation.
Begin with the favourite rhetorical move: the size of the prize.
Whenever the presentation turns to one of the divisions still being built – business mobile, the wholesale platform, premium network services – it turns at once to the market. A three-billion-dollar business-mobile opportunity. A wholesale market worth north of a billion and growing at a double-digit clip. “Over ninety per cent of the market to target.” It is worth saying plainly what that last phrase means, because it is doing a great deal of quiet work.
A company with ninety per cent of a market to target is a company that today holds roughly ten per cent of it.
The total addressable market is the figure a business reaches for precisely when its own share is not the figure it wishes to dwell on. There is no shame in sizing a prize. One notices only that the prize is always quoted in full, and the current takings rather less often.
Then the soft metrics, presented as momentum. Network perception among non-customers, up nine points. Mobile consideration, up three. Awareness rising year on year. These are offered as evidence the strategy is working – and they may, in time, become that. But consideration is not connection, and awareness is not a customer. The company has its own cautionary tale on exactly this point, and it sits a few divisions over in Fixed, where years of perfectly respectable brand consideration have run alongside a long and steady decline in the actual subscriber base.
People considered. They did not, in the main, switch.
A survey answer is a sentiment; a subscriber is a fact; and the distance between the two is the graveyard of a great many telco strategies. Counting tomorrow’s customers today, one is obliged to point out, does not pay this year’s spectrum bill.
Now the growth story the deck tells most enthusiastically – the digital-subscription brands – and the hole in the middle of it. Every new low-cost digital subscriber is counted, here, as a gain. What the charts do not capture is where some of those subscribers come from.
The market has shifted, decisively, to SIM-only: the handset-and-bundle gimmickry that once disguised a plan’s true price has lost its grip, helped along by a public growing wise to the single-price-point trick – and helped, too, by the regulator’s findings against the larger players, which have made the old tricks more expensive to attempt.
This masthead flagged the dynamic in 2025. Growth that eats your own margin is not growth. It is rotation, wearing growth’s clothes.
Beneath the strategy sit the numbers, and the numbers carry a few adjustments the presentation would prefer the eye to glide over.
EBITDA – earnings before interest, tax, and the cost of one’s assets wearing out – is, as ever, the most flattering measure in the book: it tells you what a company might have earned in a world without lenders, the tax office, or depreciation. Useful in its place; load-bearing it is not.
And in a telco it is close to the least honest measure available, because the thing EBITDA politely sets aside – depreciation, the cost of assets wearing out – is, for a telco, not a footnote but the entire game.
This is a business that must continually pour capital into towers, spectrum, core network and equipment simply to stand still; the network depreciates relentlessly, the spectrum expires on a government clock, and the kit needs replacing on a cycle that never stops.
To report earnings before depreciation and capital costs, in an industry whose defining feature is enormous and unavoidable capital costs, is to report the score of a game with the hardest innings removed. It is precisely the number a capital-hungry business reaches for when the capital is the problem.
There is a sharper way to see what that add-back is really doing. A dividend is meant to be paid out of profit – out of what the business actually earned after everything, including the wearing-out of its assets, has been accounted for. TPG’s dividend is not being paid out of that.
On the statutory profit – the real, after-everything number – there is barely a dividend’s worth of earnings to go round; we have already noted the payout running at something north of six hundred per cent of statutory profit. The dividend is instead being funded, in effect, out of the gap – the space that opens up between EBITDA and the bottom line precisely because depreciation and amortisation have been added back.
The company is paying shareholders out of the very cost it has told them to ignore. And when the statutory number will not support the story, the slide reaches instead for NPATA – net profit with the amortisation conveniently added back, and, by the company’s own account, management’s preferred measure – the flattering number a company turns to when net profit alone has stopped being flattering.
It is the same move twice: add back the cost of the assets, then declare the result distributable. A dividend paid from depreciation is not a dividend from earnings. It is a dividend from the bit of the accounts where the earnings used to be.
The net-profit growth the company points to looks rather less muscular once one strips out a research-and-development tax credit of around forty-five million dollars – a one-off kindness of the tax system, not an achievement of the business. Underlying earnings are the ones that recur; a tax credit does not recur, and it does not sell a single SIM.
And the capital expenditure, shown to be falling – a reassuring picture, until one recalls that much of the “reduction” is simply the arithmetic of having sold the assets that used to demand the spending. Something on the order of four hundred million dollars of avoided CAPEX is not the fruit of discipline. It is the fruit of divestment. Spend less because you have grown efficient and spend less because you now own less arrive at the same line of the cash flow statement and mean entirely different things.
Which brings us to the dividend – or rather, to the company’s “intention” to grow it. Intention is a beautifully chosen word: it has the shape of a promise and the substance of a wish, the corporate cousin of “we really must catch up sometime.”
It deserves to be read beside two facts the presentation is less eager to linger on. The first is that the accumulated tax losses currently sheltering the company’s earnings do not last forever; when they run out, a tax bill in the order of ninety million dollars arrives on roughly three hundred million of earnings – a cost that does not exist in today’s figures and very much will in tomorrow’s.
The second is interest: a facility approaching eight hundred million dollars, to be drawn and termed out in the years ahead, at rates that have lately been travelling in one direction, and it is not down. A growing dividend, a looming tax bill, and a rising interest cost, all drawn against a cash flow that is – as we have noted elsewhere – already heavily spoken for. Intention, indeed, may be the only honest word available.
And finally, the pillar the company is right to name, and unwise to spotlight: ROIC above WACC.
Return on invested capital above the cost of that capital. It is the correct metric – in the end, it is the only one that matters, the single line that divides a business creating value for its owners from one quietly consuming it. Which is what makes its prominent appearance here so striking, because on the available numbers the company’s return on capital sits below its cost of capital.
Stated plainly: for now, each dollar the company puts to work returns less than that dollar costs to raise. A business in that position is not compounding its owners’ money – it is eroding it, gently, regardless of how briskly the headline earnings climb.
And the figure that most flatters the present return has been helped along by the very same one-off handset-receivables sale that flattered the cash flow; strip the one-off out, and the gap between what the capital earns and what it costs only widens.
The men who built their reputations on exactly this distinction – Buffett, and closer to home the value investors who follow Montgomery‘s line on return on capital – would not require the slide explained to them. They would need only to see which side of that line the number falls on. It is, at present, the wrong side. Everything else in the presentation is a discussion of how, and how soon, the company intends to cross back over it.
That is the arithmetic behind the optimism. None of it makes the company a bad one, or its managers dishonest; the divestment may prove wise, the digital brands may yet find genuinely new customers, the return on capital may climb back above its cost.
But an investor is entitled to read the slides in their own dialect – to notice that the market is always quoted in full and the share in part, that consideration is offered where customers are scarce, that growth is booked where rotation is occurring, and that the one measure which finally decides the question is, for now, pointing the wrong way. The optimism may even be warranted. It is simply not, on these numbers, yet earned.
The Romance of the MOCN
Every network-sharing deal is sold as a love story before it becomes a spreadsheet. The MOCN was no exception. When TPG struck its arrangement to share most of Optus’ coverage, the promise was expansive: a doubled footprint, a transformed regional proposition, and – the commercial heart of it – churn reduction, customers who would stop leaving because the network finally reached them. There was an emu. There was a celebrity. There was forty million dollars and a campaign, built by the agency before the current one, instructing the nation to behold Double the Network. It was, in its way, romantic.
The spreadsheet has since arrived, and it has stripped the romance out fairly comprehensively.
Start with the churn claim, because it is the one that matters commercially and the one most easily flattered. Churn is, on the recorded numbers, supposedly down – which sounds like vindication until one asks what “churn” is counting. A Vodafone customer who leaves for Telstra is churn.
A Vodafone customer who quietly trades down to one of the group’s own digital brands has not, on one view, been lost at all – he has simply moved from one TPG pocket to another. But a retention figure that counts him as saved tells you nothing about the revenue that left with him: he is now paying roughly half the ARPU he did before. Whatever the churn line shows, the revenue line has taken the hit – and it is the revenue line the company discusses least.
The mechanism here is not transfer pricing, which concerns how a group prices dealings between its own entities; it is plainer than that – it is internal migration, on-net cannibalisation, the group keeping the body and losing the ARPU and booking the result as reduced churn. Retention that is really just downgrading does flatter the slide. It does not flatter the revenue line, which is why the revenue line is discussed less.
Then there is Fixed Wireless, about which one is conspicuously curious. Here was a product handed two gifts in quick succession: a slice of the Optus coverage footprint to sell into, and a new Wi-Fi 7 modem, touted with some fanfare in August last year. Two tailwinds, a larger addressable market, a better piece of hardware – and yet the Fixed Wireless base has conspicuously declined to take off.
When a product is given more coverage and better equipment and still does not grow, the missing ingredient is not coverage or equipment. It is demand, or it is the brand, or it is both – and none of those is fixed by a firmware update.
In fairness, there is a genuine technical wrinkle here, and it is worth naming because it is real.
In some metropolitan pockets Fixed Wireless is capacity constrained – so constrained that, on Vodafone and TPG‘s own sales page, a customer at an affected address cannot sign up at all. The product is, in effect, cease-sold at those addresses: the network is full, and the door is politely closed.
One understands the mechanism, and one even understands the likely cause – a company that has spent two years trumpeting reduced capital expenditure should not be astonished to discover its capacity has not kept pace with its ambitions. Fair enough. But that constraint applies to particular congested pockets.
It does not explain the absence of Fixed Wireless growth across the new MOCN coverage area – the expanded footprint, the freshly available addresses, the very places the deal was supposed to open up. Those addresses were not cease-sold. They were available, advertised, and attached to a better modem. So where are the additions? Did the emu forget to mention them? Or was it simply that, on this product, the bird was less persuasive than billed.
But the romance leaves most visibly in the growth numbers themselves, which after five months of operation – the MOCN nominally launched on 30 January, though some sites were lit from around 9 January – read less like a doubling than a rounding error.
The campaign was national and loud. The movement was neither. Sydney, five million people and the single biggest prize, recorded no growth at all – flat, after the full weight of the campaign.
The next six largest cities, around twelve million people between them, moved up by about half a per cent.
The other towns, perhaps five million people, managed something on the order of one per cent; the rural footprint, under five million, half a per cent.
This is the harvest of the emu, the forty million, and Double the Network: a metropolitan market of five million people moved by precisely nothing, and the regions – the very places a doubled network was sold to transform – moved by amounts a rounding convention could swallow whole. One does not spend that to stand still in Sydney.
The Investor Day’s own state-by-state chart tells the same story in a higher resolution, and tells one further thing by omission.
Melbourne was the best of the bunch, up around one per cent. Brisbane and Perth each crept half a point to about twenty-one per cent share. Adelaide rose six-tenths of a point to roughly twenty-three. Regional and other-metro gained nine-tenths of a point, to about nine per cent. Sydney, held at or near its thirty-per-cent metro ceiling, was always going to be the slow one – that much is conceded.
But one notices which jurisdictions made the chart and which did not: room was found for five mainland capitals, and none for Tasmania or the Northern Territory – the former being, as it happens, where Vodafone keeps a resolutions centre. A company tends to chart the geographies that are cooperating.
Hold the numbers that did make the chart against the targets the company still professes, and the arithmetic turns unkind. The stated metro ambition is something like thirty per cent share across the major cities. Brisbane and Perth sit at twenty-one, Adelaide at twenty-three, and they advanced by roughly half a point in the period.
To travel from twenty-one to thirty on momentum of half a point per period is not a stroll – it is a decade-long march dressed as a quarter’s progress. A three-year break-even was once floated for this venture. One is entitled to take the growth rate actually achieved, lay it beside the distance still to cover, and ask – politely – by which year, precisely, the lines are expected to meet. On the period’s run-rate, the honest answer is “not the one in the plan.”
The company would prefer one did not dwell on break-even at all, and has said as much.
On an earlier call, when an analyst put the break-even question to the chief financial officer – noting the Street’s estimate, as reported in the Australian Financial Review, that the MOCN needs something like 100,000 to 200,000 net incremental postpaid subscribers to wash its face – the response was instructive.
Break-even, the CFO replied, was “definitely not our aspiration”; the company intends to do rather better than that. Note the word. We met its cousin in the dividend policy – the “intention” to grow – and here it is again, dressed as ambition: break-even is beneath us, we aim higher.
It is a fine thing to aim higher than break-even. It would be a finer thing to first arrive at it. Because if the aspiration sits above break-even, the obvious question is where the subscribers to clear even the lower bar are – and the answer, on the company’s own trading detail, is that postpaid is flat and the digital-first brands do not carry the ARPU to fund MOCN economics in the first place.
An aspiration is not a subscriber. And one notices the narrative quietly relocating beneath all this: from “premium postpaid net adds,” the metric originally promised, toward “churn improvement and total subscriber growth” – the softer, blended framing a company reaches for when the original scoreboard has stopped flattering it.
And now there is a newer, shinier toy still: digital-first subscribers, so attractive a metric that the chart itself has been redrawn to fold postpaid and digital-first into a single combined line. It is a remarkable thing to watch in real time – a company merging its premium product and its half-price product into one number precisely because the premium half, counted alone, has stopped moving. Two categories become one the moment one of them stops cooperating; the scoreboard is helpfully rebuilt around whichever metric is still climbing. The target is not being hit. It is being rewritten – and now, it seems, redrawn.
And note, while we are here, that the goalposts have a way of relocating. The regional opportunity, previously broken into two honest buckets – “rural,” sub-five-per-cent share against a four-million-person addressable market, and “other towns,” five-to-fifteen-per-cent share against a five-million-person market – has been quietly consolidated into a single tidier line. It is easier to report progress against a category you have just redrawn. This masthead has remarked before on the company’s fondness for changing the unit of measurement at the precise moment the previous unit stopped cooperating.
There is a subtler tell in the ARPU, too. Consumer ARPU is up – but it is up on price rises, and it is being dragged the other way by a sales mix tilting toward corporate. Corporate wins are real and worth having; they bring scale. But corporate fleet customers are, by their nature, lower incremental ARPU per service – bought in volume, won by competing against the other carriers on price – and if the consumer share of the mix is shrinking while corporate grows, the headline ARPU is flattering a consumer franchise that is in fact softening beneath it.
The implication the consensus has not fully priced is that consumer momentum is weaker than the blended number admits. Scale from corporate, individual SIOs from consumers; and it is the consumer SIOs, not the corporate fleets, that feel the outages, the service complaints, and the TIO volumes that turn up elsewhere in this company’s record. Winning the low-maintenance customer while the high-maintenance one drifts away is a fine quarter and a worrying trend.
None of which was lost on the analysts. On the most recent call, management came across as notably defensive on the postpaid net-add questions and on the MOCN break-even targets now twelve months in – to the point that one analyst characterised the mobile reporting, in a word the company will not enjoy seeing in print, as marked by “opacity.”
When the people paid for a living to model your business reach for that word on your own earnings call, the romance is well and truly over. What remains is the spreadsheet – and the spreadsheet, after the emu has taken its bow, is asking when the network it doubled intends to double anything else.
And Then, On Thursday, the Network Went Off
If Investor Day supplied the theory, Thursday supplied the demonstration.
On 18 June, the Vodafone network went down nationally. Customers in Darwin, Melbourne, Sydney, Brisbane, Perth and Canberra reported no service for several hours from early morning. The cause, the company said, was a power failure at one of its network hubs – which, as an RMIT engineering academic observed in The Conversation, is the kind of thing backup power and batteries exist specifically to prevent, and the kind of cascading national failure that points to a centralised core network with a single point of failure sitting in it somewhere. A decentralised network reroutes around a dead hub. This one fell over.
The company’s network status checker – the page a customer visits to find out whether the network is working – was, with a certain dark elegance, also not working. It was hosted, Vodafone later explained, on systems at the very hub that had failed. The instrument for measuring the outage was inside the outage.
All of which lands with particular weight given that TPG‘s chief marketing officer told The Australian in March, on launching the Ali Wong campaign, that “any previous network issues no longer exist.” The network has a way of auditing that statement on its own schedule.
A Rumour, Handled With Tongs
There is industry chatter – and it will be left precisely at that, chatter, without a source, a figure, or a form of words attached, because some things are better gestured at than quoted – suggesting the outage has already produced an unusually heavy run of customers heading for the exit. Whether that proves out is not yet a matter of public record, and this masthead is not in the business of pinning numbers to a rumour.
But one need not have the figure to see the logic. A telco that has just spent $40 million advertising its reliability, in the same month its network collapses for millions and its outage-checker collapses with it, is not a telco whose customers are feeling especially loyal.
Outages of scale tend to be followed by port-outs of scale; it is the most predictable sequence in the industry. And TPG‘s book is already the one losing Vodafone customers and backfilling with half-price on-net and digital ones. An outage does not start that bleed. It accelerates a bleed already on the page.
The precedents are not encouraging, and they are TPG‘s own. When a major outage strikes, customers leave – and the historical record puts numbers on it.
After Optus‘ 2023 network collapse, analysts estimated the damage would outstrip even its catastrophic data breach a year earlier, on the simple logic that the breach had not cost anyone their service while the outage had; one broker modelled that a 2.5% churn from Optus would be “material” enough to shift nine figures of EBITDA to the rival that caught the departing customers.
And Vodafone needs no analyst to explain the mechanism, because it lived it: the original Vodafail era of 2010-11, an outage-riddled stretch this very masthead is named after, cost the company around 2.5 million customers over three years. That is the ghost at this particular feast.
And it is a ghost that never quite left: more than fifteen years on, the company has never won back that headline number, and remains – on its own historical base – well over a million customers short of where it once stood.
Some companies recover from their annus horribilis. This one has spent a decade and a half failing to.
A brand that spent the better part of a decade rebuilding from a reputation for dropping out has just, in a single morning, reminded several million people of precisely the thing it has spent $40 million and one celebrity emu trying to make them forget. Whether the chatter proves accurate or not, the structural vulnerability is real, documented, and wearing a fifteen-year-old name tag.
The market knew this on the day too. The 8% was not pricing the spectrum bill alone. It was pricing a company being squeezed from every side at once – by the upstarts on value, by the incumbents on quality, by the government on spectrum, and now by its own infrastructure – fifteen years after the original Vodafail crisis it has never quite stopped paying for.
Late August
There is a date in the diary, and TPG cannot move it.
The full-year results land on the ASX in late August. That is when the trading detail becomes audited fact, when the spectrum framework meets the balance sheet in writing, when the post-outage churn – whatever it turns out to be – starts showing in the subscriber lines, and when management is once again obliged to stand in front of analysts and choose which of its four profit definitions to lead with. Not all of them will accept the choice.
On the most recent call, at least two analysts declined to take the framing at face value – pressing management on the gap between the headline figure and what sat beneath it, in the way analysts do when the arithmetic on the page is not quite the arithmetic in the room. The facade, such as it is, holds better in the slide deck than in the Q&A.
The market’s verdict on Investor Day, it should be said, was not really about the spectrum bill alone, much as that supplied the headline. It was about the thing underneath it – the subscriber trajectory and the ARPU that was supposed to climb alongside it and conspicuously did not.
Investors were shown a premium brand growing its customer base by importing cheaper ones, a postpaid line going nowhere at half price, and a forecast of digital-first additions at materially lower revenue per user. They were, in short, shown growth that costs more to acquire than it earns, and they marked the company down accordingly.
The 8% was the sound of a market declining to pay a growth multiple for a business that has quietly stopped growing the only way that matters.
Which is what makes the late-August results worth reading with particular attention. With churn supposedly on the rise since the outage – and we will say no more than supposedly – the full-year numbers are the first opportunity to see whether the trajectory management sketched on Investor Day survives contact with reality.
One will be watching the subscriber lines, certainly, and the ARPU, and whether the postpaid base has finally been coaxed into life or merely discounted into stasis.
This will be the first results since the market stopped admiring the slide and started doing the sum. The slide can say what it likes between now and then. In late August, the company faces the music – and the music, this time, will be played on a calculator.
The arithmetic showed up on Investor Day. It is not the sort of guest that leaves early.
📩 Right of Reply
TPG Telecom Limited, its directors and officers, and any individual or entity who considers themselves referenced in this article are warmly invited to correct, clarify, or add context to anything set out above – including, should they wish, the matters this article has been careful not to assert. If any figure is wrong, any inference unfair, or any of the accounting entirely innocent, the author would genuinely like to know, and to say so.
Verified responses can be sent to vodafailed@gmail.com and will be published in full, without editorial amendment, alongside the original article. This right of reply remains open indefinitely – rather longer, one notes, than the average TPG network hub.
⚖️ Disclosure, Disclaimer & Legal Notice
This article is independent commentary and analysis based on publicly available information, including TPG Telecom‘s 2026 Investor Day disclosures, its FY25 results and ASX announcements, ACMA‘s published spectrum determination, ASX market data, and media reporting.
All financial figures are drawn from public disclosures and analyst commentary and are subject to the limitations of external estimation. All views expressed are the author’s honest opinions, formed on reasonable grounds. This is not financial, investment, or legal advice, and nothing in it constitutes a recommendation to buy, hold, or sell any security; readers should seek their own independent professional advice before making any decision.
References to “industry chatter” regarding post-outage customer movement are expressly identified as unverified market commentary, not statements of fact, and no source, figure, or specific representation is asserted or relied upon. References to the June 2026 outage reflect the company’s own public statements as to cause. Historical figures concerning prior outages and their customer impact (including the 2010-11 period and the 2023 Optus outage) are drawn from contemporaneous reporting and analyst commentary and are presented for comparative and contextual purposes only. No assertion of any breach of any law, regulation, or accounting standard is made.
TPG Telecom, Vodafone, and related names and logos are trademarks of their respective owners and are used in this article for identification, commentary, and analysis only. Their use does not imply any affiliation with, sponsorship by, or endorsement from those entities, and no such association is asserted or implied.
The author has an active dispute with TPG Telecom Limited (ASX: TPG) and has made protected disclosures under Part 9.4AAA of the Corporations Act 2001 (Cth). The author holds a very immaterial shareholding in TPG Telecom Limited. These interests should be considered when evaluating the commentary presented. All entities and individuals retain the presumption of lawful conduct unless determined otherwise by a court, tribunal, or regulator of competent jurisdiction.
Previous posts in this series:
Post #65 – When The Music Stops
Post #66 – The $2B Problem TPG Can’t Afford
Post #67 – The Bonus Year: Thin Earnings, Thick Optics
Post #68 – Buying the Narrative
Post #69 – The Smart Money Just Left the Building
Post #70 – Who’s Watching the Watchers?
Post #71 – Nine Lives: The Ad Agencies Vodafone Burned Through on the Way to Zero Growth
Post #72 – Marked Safe from the Whistleblower Policy
Post #73 – The Story Nobody Will Publish
Post #75 – The Gift That Keeps Giving
Post #76 – The Seat Nobody Wants
Post #77 – Houdini Never Filed a Form 605
Post #78 – Fifteen Years and a Footnote
Post #80 – Two Companies in a Purple Coat
Post #81 – Transformational: A $7 Million Result With a $1.6 Billion Costume