When a company generating $7 million in pre-tax profit from continuing operations faces a $2 billion spectrum bill, the question isn’t whether it can pay. It’s what breaks first. TPG’s FY25 result was packaged as transformation – but the stock fell on results day, brokers slashed earnings by up to 73%, and UBS now forecasts negative free cash flow in FY27. The headlines are propped up by a one-off securitisation, a dividend exceeding earnings six times over, and a “premium” mobile brand that added zero postpaid subscribers. The CEO collected a $250,000 bonus and fresh equity while ACMA investigated two customer deaths from Triple Zero failures. Barely months after returning $3.3 billion to shareholders as proof of a de-risked balance sheet, TPG may now need to go back cap in hand – for capital it just gave away. The bull case requires everything to go right. Nothing here suggests it will.
TPG’s FY25 result was presented as “transformational.”
The stock fell 2.72%.
Somewhere between the LinkedIn celebration post and the market’s verdict lies the truth – and the truth has a $2 billion problem attached to it.
The Bull Case – and Why It Doesn’t Hold Up
Before pulling the numbers apart, it’s worth understanding why reasonable people may own this stock – and why the arguments, on closer inspection, are insufficient.
Owner earnings yield. TPG generates $600-700 million in owner earnings – net profit plus depreciation and amortisation, minus sustainable maintenance CAPEX – against a market capitalisation of roughly $7.3 billion. That’s an 8–10% owner earnings yield, which is attractive in a market where defensive yields are scarce. But owner earnings are only meaningful if the business earns above its cost of capital. At 5.42% ROIC – below any reasonable estimate of WACC for a leveraged telco with regulatory risk and spectrum uncertainty – those earnings are not creating value. A high yield on a value-destroying business is not cheap. It’s a trap.
Balance sheet transformation. Bank debt has been reduced from $4.1 billion to $1.4 billion. Interest costs are expected to halve from ~$160 million to ~$80 million in FY26. Net leverage sits comfortably below the 2x EBITDA ceiling. But the transformation was funded by selling the EGW assets for $4.7 billion – a one-time event. And if spectrum costs of $2 billion arrive over FY27-30 and can’t be funded from operating cash flow, debt has to go back up. The balance sheet was de-risked. The spectrum wall may re-risk it.
MOCN churn improvement. The arrangement with Optus doubled TPG’s national coverage footprint, was hypothesised to provide some churn improvement and a credible path to regional competitiveness for the first time. But MOCN was supposed to drive postpaid growth, not just churn reduction. Postpaid is flat at 2,846k. Zero net adds. Both Telstra and Optus grew postpaid in the same period. Churn improvement without subscriber acquisition means the MOCN investment is generating defensive value, not growth – a fundamentally different economic proposition than the one the market was sized on.
Cost discipline. Operating costs grew just 0.5% against 3.3% trimmed mean CPI. Management has committed to $100 million in real-terms savings over four years. This is structural, not cosmetic. But cost-out in a service business has a floor. TIO complaint data trending sharply against TPG brands while Telstra and Optus decline suggests the floor may already be in view. And marketing spend rose ~$40 million for MOCN launch without delivering premium postpaid conversion – a structural cost increase with diminishing returns.
Oligopoly structure. TPG operates in a three-player market with high barriers to entry. Spectrum scarcity and network investment thresholds make new entry effectively impossible. The industry structure should support rational pricing over time. But being the weakest player in an oligopoly is not the same as being protected by one. Telstra has brand and network leadership. Optus has a deep-pocketed parent in Singtel. TPG has neither. In oligopolistic markets, the weakest participant typically earns the worst returns, faces the most competitive pressure, and has the least pricing power. TPG’s FY25 result is entirely consistent with that structural reality.
Dividend yield. 18 cents in FY25, guided to 19 cents in FY26, offers north of 4.5% at current prices. But the dividend exceeds earnings by 6x. It is funded from the cash flow gap between D&A and CAPEX, not from profit. And all four covering brokers forecast 19-20 cents through FY27 against EPS estimates of 6-19 cents. The dividend is a signalling device, not a reflection of earnings capacity. Whether the signal survives the spectrum wall is the question the market hasn’t priced.
These are real attributes. They explain why brokers like Macquarie and Morgans rate the stock positively. They are not sufficient to absorb what’s coming.
The Number They Didn’t Want to Talk About
On the FY25 earnings call, UBS analyst Lucy Huang asked a direct question about spectrum renewal costs. She put the estimate at $1 billion to $1.5 billion over three to five years.
CEO Iñaki Berroeta’s reply: “It is more than that.”
More than $1.5 billion. Over a longer period.
That was the extent of management’s disclosure on the single largest capital obligation facing the business.
Days later, UBS published their post-result note. The number had doubled.
UBS now estimates TPG’s spectrum renewal costs at approximately $2 billion for FY27–30. That’s double their previous estimate of $1 billion.
And UBS is forecasting negative free cash flow in FY27 as a direct result.
Let that settle.
The company that just told analysts it would generate $600 million in free cash flow by FY27 may instead be cash-flow negative.
The entire FY27 narrative – the one that underpins the dividend growth story, the capital return framework, the de-leveraging trajectory – runs through a spectrum wall that management acknowledges exceeds $1.5 billion, that its own covering broker now sizes at $2 billion, and that nobody on the call was willing to quantify.
The Spectrum Wall: $7.3 Billion for the Industry, $2 Billion for TPG
In December 2025, the Australian Communications and Media Authority released its preferred pricing position for the renewal of expiring spectrum licences. The total market value assessed was $7.3 billion – significantly higher than ACMA’s earlier preliminary range of $5.0–6.2 billion.
Around 80% of existing mobile spectrum licences are due to expire between 2028 and 2032. The renewals cover the 700MHz, 850MHz, 1800MHz, 2GHz, 2.3GHz, 2.5GHz and 3.4GHz bands – the frequencies that underpin virtually all mobile services in Australia.
This is an industry-wide challenge – all three carriers are affected. But the burden is not distributed equally, and that’s where the analysis becomes specific to TPG.
Spectrum Costs Against Earnings: The Peer Comparison
The $2 billion figure for TPG lands differently depending on the earnings base absorbing it.
Telstra faces a $2.7 billion spectrum renewal bill under ACMA’s preferred pricing. Telstra argues the fair value should be $1.4 billion. Against Telstra’s FY25 underlying NPAT of $2.3 billion and EBITDA of $8.6 billion, even the $2.7 billion figure represents roughly 1.2 years of net profit or 31% of EBITDA. Telstra’s ROIC sits at 8.5% – above its cost of capital. The company is running $750 million share buybacks and generating $2.1 billion in annual shareholder returns. The spectrum bill is a significant cost but absorbable against the earnings base.
Optus sits within Singtel, whose scale gives it balance-sheet flexibility that TPG lacks on a standalone basis. With H1 FY26 EBIT up 27% and an annualised underlying net profit of approximately $200 million, Optus also has operational momentum running in the right direction heading into the renewal cycle.
TPG faces approximately $2 billion against a pre-tax profit from continuing operations of $7 million and statutory NPAT of $52 million (of which $45 million was a non-recurring tax benefit). The spectrum bill represents roughly 285 times pre-tax earnings. Even on the most generous underlying NPATA measure of $69 million, spectrum costs represent 29 years of current underlying earnings.
The absolute number matters less than the ratio. Telstra can absorb its share from earnings over a reasonable period. TPG cannot. That is the structural difference this analysis is built on.
TPG itself acknowledged the pressure. In its submission to the ACMA consultation, TPG cautioned against treating spectrum as “a cash cow,” arguing Australians would ultimately pay through higher mobile bills and slower improvements. The industry lobby group AMTA has proposed that ACMA allow operators to spread payments over the licence term. If granted, this would ease near-term pressure – but it doesn’t eliminate the cost. It spreads it across a longer period of compressed returns.
The $7 Million Earnings Base
To understand why $2 billion is a balance-sheet-shaping number for TPG, you first need to understand what the business actually earns.
Strip out the $45 million non-recurring income tax benefit – from R&D credits and recognition of previously unrecognised tax losses – and TPG generated approximately $7 million in pre-tax profit from continuing operations in FY25.
$7 million. On $4.179 billion in revenue. A 0.17% margin.
That’s not a turnaround. That’s a rounding error. And it’s being presented like a victory lap.
The statutory NPAT of $52 million only exists because of that tax tailwind. It won’t repeat at the same magnitude.
Management’s preferred metric – underlying NPATA of $69 million – is pro forma, excludes material one-offs, and strips out customer base amortisation. Even on that most generous basis, it is a thin earnings base for a business generating $4-5 billion in total revenue.
Return on invested capital was 5.42%. On any reasonable estimate of WACC for a leveraged telco carrying regulatory risk and spectrum uncertainty, the business is still destroying value on incremental invested capital – even after a “transformational” year.
For context: Telstra’s ROIC is 8.5%, above its cost of capital – meaning every dollar Telstra retains creates value. Every dollar TPG retains, on the available evidence, does not.
This is the earnings base against which a $2 billion spectrum bill arrives.
The Cash Flow Illusion
The headline numbers suggest a company swimming in cash. Operating free cash flow “nearly doubled” to $1,291 million.
It didn’t.
$687 million of that came from the sale of handset receivables to a Macquarie-led securitisation trust in October 2025. That’s a one-off balance sheet monetisation – future handset repayment streams sold for upfront cash.
It is not operational improvement.
Strip it out and the picture changes completely:
- Gross adjusted OFCF: ~$604 million – actually below FY24’s $649 million
- Net adjusted (after subordinated note): $594 million impact, giving normalised FCF to equity of $396 million per the company’s own slide
The company acknowledged this. CFO John Boniciolli himself separated out the receivables impact and said the year-to-year effect “should not be material” going forward.
So the honest baseline is $396 million in normalised free cash flow. Not $1.3 billion. Not “nearly doubled.” $396 million.
Now subtract the dividend.
The Dividend That Exceeds Earnings by 6x
TPG declared ordinary dividends of 18 cents per share for FY25. On approximately 1.86 billion shares, that is roughly $335 million in distributions.
Against statutory NPAT of $52 million (tax-benefit-assisted), the payout ratio is approximately 640%.
Against the $7 million pre-tax profit, the dividend is nearly 50 times earnings.
This is not funded from profit. It is funded from the cash flow surplus created by depreciation and amortisation (~$1.5 billion) materially exceeding CAPEX ($771 million in additions).
That gap creates real cash even when bottom-line profit is negligible. And with bank debt down from $4.1 billion to $1.4 billion, the interest savings flow directly to cash available for dividends.
But it raises a question that any quality-focused investor would ask:
If the business doesn’t earn enough to cover its dividend, and it faces a $2 billion spectrum bill, is the dividend sustainable – or is it a signalling device that may need to be revisited when the spectrum bill crystallises?
All four covering brokers forecast 19 cents in FY26 and 20 cents in FY27, suggesting TPG has effectively pre-committed to a growing dividend trajectory. Against broker EPS estimates ranging from 6 cents (Morgans) to 19 cents (Macquarie), that implies payout ratios of 100% to 300%+.
The mechanism carries a latent assumption: that the asset base doesn’t need replacing at the rate it’s being depreciated. If CAPEX restraint means under-investment relative to the true maintenance requirement – a question that becomes more pointed as 5G densification demands increase and spectrum bills arrive – then the dividend is effectively funded by consuming the asset base.
The dividend is being maintained as a signalling device, not a reflection of current earnings capacity.
The FY27 Inversion
Here is how the forward cash flow story was supposed to work:
- FY26: ~$400–430 million FCF to equity, from EBITDA growth, flat CAPEX, lower borrowing costs
- FY27: ~$600 million FCF to equity, from further EBITDA growth and CAPEX declining to $650 million
- FY28+: Continued expansion as cost-outs, capital efficiency, and revenue growth compound
This was the framework John Boniciolli validated on the call when an analyst put the $600 million FY27 estimate to him directly.
Then UBS published their note.
With spectrum costs now estimated at $2 billion for FY27–30, the FY27 free cash flow story inverts. UBS forecasts negative FCF in FY27 once spectrum payments are included.
Even excluding spectrum, UBS estimates FCF grows to $660 million by FY28 from $430 million in FY26. But that gross number must absorb annual spectrum payments of $500-670 million depending on the phasing. It doesn’t work.
Management’s position is that spectrum renewals are “later in FY27” and “very manageable.” On the call, John said they are “absolutely comfortable in being able to absorb whatever spectrum outcomes will be finalised with the ACMA.”
Comfortable. On a $7 million pre-tax earnings base.
The Funding Dilemma: Every Option Is Bad
If spectrum costs are $2 billion and operating cash flow cannot absorb them (which UBS’s negative FCF forecast confirms), the funding has to come from somewhere.
The menu:
1. Draw on debt facilities they just paid down. TPG reduced bank debt from $4.1 billion to $1.4 billion using Vocus proceeds. Re-leveraging to fund spectrum would undo the balance sheet transformation narrative that management spent the entire FY25 results presentation celebrating.
2. Raise equity. Dilutive. On a stock that dropped on results day. With a register that includes concentrated strategic holders – CK Hutchison and Vodafone Group PLC jointly holding ~50%, the Teoh family at ~14%, and Washington H. Soul Pattinson at ~13% – who may or may not participate.
3. Cut deeper into CAPEX. CAPEX is already declining from $771 million to a guided $650 million from FY27. Cutting further risks network quality, competitive positioning, and 5G densification – precisely when Telstra has committed $800 million to mobile network investment and Optus’s EBIT is growing 27%.
4. Sell spectrum. Strategically weakening in a three-player market where network coverage is the only differentiator. And selling spectrum to fund spectrum renewals is circular logic.
5. Cut or freeze the dividend. The nuclear option. Would destroy the income narrative, spook the register, and likely trigger a re-rating.
Every option is bad. The question is which combination of bad options management chooses, and how they present it.
Did They Return Capital They’re Going to Need Again?
This is the question that quality-focused investors should be asking.
In FY25, TPG returned more than $3.3 billion to shareholders – a $1.52 per share capital reduction plus a 9 cent special dividend – funded by the $4.7 billion Vocus sale proceeds.
That was presented as evidence of a transformed, de-risked financial position.
But if spectrum costs of $2 billion arrive over FY27–30 and can’t be funded from operating cash flow, then the capital return was potentially premature.
The sequence would be:
- Sell fibre assets for $4.7 billion ✓
- Return $3.0 billion to shareholders (partially recycled via $438m reinvestment plan) ✓
- Pay down $2.7 billion in debt (using $1.7b of Vocus proceeds + reinvestment plan proceeds + operating cash) ✓
- Face $2 billion spectrum bill ✗
- Need to raise debt or equity to fund it ✗
In that scenario, TPG gave money back to shareholders that the business is going to need. That’s not financial transformation. That’s a capital allocation timing error.
The Mobile Growth That Isn’t
The spectrum bill doesn’t arrive in isolation. It arrives against a mobile business that management describes as its growth engine – but whose engine room is stalling.
Postpaid: Zero. 2,846k at FY24. 2,846k at FY25. Flat. Despite doubling the national coverage footprint through MOCN. Despite ~$20 million in go-to-market spend. Despite the Optus outage being gift-wrapped as a switching opportunity.
Both Telstra (+106k adjusted postpaid) and Optus (+30k postpaid) grew the segment. TPG didn’t.
Digital First: Growing but dilutive. The 228k total mobile net adds came overwhelmingly from digital-first brands (Felix, TPG, iiNet) at $25.56 ARPU – roughly half the postpaid ARPU of ~$50. Every digital-first add that cannibalises a Vodafone postpaid customer compresses group economics.
The Combined Reporting Trick. TPG created a new reporting category – “Combined Postpaid and Digital First” – showing +113k/+3.3% growth. The chart draws your eye to the growth callout. The underlying data shows postpaid at 2,846k at both year ends. Zero growth hidden inside a blended category.
On the earnings call, Iñaki criticised competitors for blending digital brands into postpaid reporting: “some of our competitors will report that under postpaid… it’s getting a bit more difficult to do comparative analysis.” TPG did the same thing on the same day. On the same chart.
MOCN breakeven is drifting. The CFO acknowledged analyst breakeven estimates of “100,000 to 200,000 net incremental subs” then immediately reframed: “break even is definitely not our aspiration.” If the aspiration is higher than breakeven, where are the subs? Postpaid is flat. Digital-first doesn’t carry the ARPU to fund MOCN economics. The breakeven target is drifting, and the narrative is being quietly rewritten from “premium postpaid net adds” to “churn improvement and total subs growth.”
The Prepaid Problem: Why Volume Doesn’t Equal MOCN Payback
There’s an important distinction the combined reporting obscures.
MOCN breakeven was always premised on durable, high-ARPU postpaid subscribers – customers generating stable monthly revenue at margins that justify the network access costs. That’s the economic engine that makes a multi-carrier network arrangement pay for itself.
Prepaid and digital-first growth doesn’t do that.
The prepaid base is structurally low-margin and low-usage. A significant proportion of it is transient by nature: tourists on short-stay SIMs, promotional port-ins who chase a deal and disappear within months, serial SIM-hoppers cycling between providers for the latest introductory offer. These customers cover their variable costs and keep the subscriber count warm, but they contribute almost nothing to the fixed cost recovery that underpins MOCN economics. Their lifetime value is short, their usage patterns are unpredictable, and their revenue contribution is a fraction of what a contracted postpaid customer delivers.
Once you strip out the transient volume, the churn recycling, and the promotional cohorts with poor retention, what remains is a segment that earns its keep but doesn’t shift the needle. Prepaid, in its current form, sustains the base – it doesn’t fund a network-sharing arrangement.
When management frames “228,000 net adds” or “total subscriber momentum” as evidence of MOCN success, the question is not whether the number is real – it is – but whether the type of growth has any bearing on the original investment thesis. A customer on a $50 monthly postpaid plan with a 24-month handset commitment attached is a fundamentally different commercial outcome from a $25 digital-first SIM that churns within a quarter. The first generates the stable, high-margin cash flow that services MOCN economics. The second does not – not at scale, not at that ARPU, and not with that tenure profile.
So when growth is predominantly low-ARPU, high-churn, and short-tenure, the MOCN payback equation hasn’t been answered. It’s been deferred – behind a different set of numbers that look healthy in aggregate but don’t actually address the economics that matter.
If the original breakeven required 100,000 to 200,000 premium postpaid adds and TPG delivered none, then redefining success around digital-first volume doesn’t close the gap
it just moves the goalposts to a pitch where the economics are worse.
And if management’s implicit new target is total subscriber momentum rather than postpaid conversion, what does breakeven actually look like on blended ARPU of ~$35? Are we now talking half a million net adds? Closer to a million?
Nobody on the call was willing to put a number on it. Which is itself the answer.
What Management Said – and Didn’t Say – on the Call
The earnings call transcript is instructive not just for the answers but for the pattern of evasion.
On postpaid stagnation: An analyst from Evans & Partners asked directly whether flat postpaid reflected a smaller addressable market and whether there was pricing power. The response pivoted to a “96,000 year-on-year improvement” framing – emphasising the rate of change rather than the absolute outcome. Management then immediately redirected to “the primary market growth opportunity we see is in our digital subscription brands.”
That’s a rehearsed exit from an uncomfortable topic: the premium brand isn’t growing, so pivot to the sub-brand that is, even though it earns half the ARPU.
On ARPU and subscriber assumptions: The same analyst pressed for the building blocks behind EBITDA guidance – what subscriber growth and ARPU assumptions were embedded in the range.
The CFO declined. Twice.
The analyst said on a public call: “Sounds like you don’t really want to go into specific assumptions around subs and ARPU.” The CFO recovered with generalities about multi-brand strategy. An analyst politely calling out a non-answer, in public, is not something that happens on calls where management is comfortable with the numbers.
On MOCN breakeven: Jarden’s Liam Robertson asked about MOCN payback phasing. The CFO acknowledged analysts have modelled “100,000 or 200,000” incremental subs as a breakeven range – then said “break even is definitely not our aspiration” and wouldn’t give an actual aspiration or cumulative incremental subs target.
That’s a goalpost shift. The floor has been quietly lowered from “200–250k postpaid at ~$45 ARPU” to a vaguer, untestable standard.
On spectrum costs: Lucy Huang put $1-1.5 billion on the table. Iñaki confirmed “more than that” but wouldn’t quantify.
For the single largest known future capital obligation, that’s remarkably thin disclosure.
The CFO said they are “absolutely comfortable” absorbing whatever ACMA determines. Material unquantified future obligations without a framework for how they’ll be funded is exactly the kind of opacity that forces conservative assumptions.
What wasn’t asked at all: In 44 minutes of Q&A with eight brokerages, not a single question on Triple Zero deaths, ACMA investigation announced, the compliance uplift program, governance, executive remuneration, or the whistleblower matter on foot. The annual report discloses all of these. The sell-side ignored every one.
Fixed: Still Shrinking
NBN lost ~116k subscribers in FY25, down 6.9%. Fixed Service Revenue grew just 0.7%.
Fixed Wireless – positioned as the growth offset – added only 17k net subs year on year. That’s an inadequate offset to 116k NBN losses, despite the “new modem launch” and 5G Standalone expansion narrative.
Nearly a quarter of group EBITDA sits on this structurally declining base.
Management says the market is “structurally challenged” with “intense competition from volume-focused NBN resellers and non-telco providers.” That’s true. It’s also an admission that the fixed business has no credible path back to growth. Aussie Broadband, Superloop, and other challenger brands continue to record SIO wins half after half.
The Cost Discipline That Can’t Last Forever
The strongest part of the result was cost management. Operating costs grew just 0.5% against 3.3% trimmed mean CPI. Management has committed to $100 million in real-terms savings over four years to FY29.
This is genuine. But it has limits.
The savings come from lower network maintenance post-MOCN, reduced headcount from business simplification, IT modernisation, and tighter third-party procurement. These are structural, not cosmetic. Credit where it’s due.
But cost-out in a service business has a floor. If it degrades resolution quality, it feeds complaints, churn, remediation cost, and regulatory attention. And the TIO complaint data – trending sharply against TPG brands while Telstra and Optus decline – suggests the floor may already be in view.
The EPS Carnage Nobody Is Talking About
The broker reactions tell the story the presentation couldn’t.
Macquarie maintained Outperform but cut FY26–FY29 adjusted EPS by 73%, 62%, 58%, and 55% respectively. Read those numbers again. A 73% EPS cut for FY26. That is not a reclassification. It is a fundamental reassessment of what the business earns.
Morgan Stanley rates TPG Underweight at $3.50 – implying 11% downside. They prefer Telstra as “higher quality defensive” and Aussie Broadband for capital growth. A broker explicitly telling clients to buy both of TPG’s competitors instead.
UBS rates Neutral at $3.95 – essentially zero upside. Spectrum costs doubled. Negative FCF in FY27.
Morgans is the most bullish at $4.40, but even they acknowledge FCF was “assisted by asset and mobile handset sales.”
Broker EPS estimates for FY26 range from 6 cents to 19 cents. That’s a 3x dispersion across four professional analysts covering the same result on the same day. That level of disagreement is itself evidence of the disclosure visibility problem:
when a company uses statutory, pro forma, underlying, and guidance basis metrics interchangeably, even the sell-side can’t agree on what earnings are.
One analyst on the call said it plainly:
“Sounds like you don’t really want to go into specific assumptions around subs and ARPU.”
Exactly. And markets punish that. They price precision, not vibes.
The Tax Shield Is Depleting
TPG currently pays minimal cash tax, shielded by accumulated tax losses from the VHA merger era and prior-year write-downs.
The Vocus transaction consumed a significant portion – the $473 million pre-tax gain on sale would have absorbed substantial losses. The $45 million tax benefit in FY25 came from recognising some of what remains.
The CFO confirmed cash tax in FY26 will be “similar to FY25” – meaning minimal. But the losses are finite and dwindling.
When they run out – likely around FY28–FY29 on current trajectory – TPG transitions from a non-taxpaying entity to a full 30% corporate taxpayer. On $300 million of hypothetical future pre-tax profit, that’s a $90 million annual cash tax bill that doesn’t currently exist.
Nobody on the analyst call raised this.
The $600 million FY27 FCF estimate may not fully account for when this transition occurs.
And it arrives at precisely the same time as the spectrum bills.
Regulatory Risk: Sized by Precedent
Rather than speculate on outcomes, we can look at precedent. ACMA’s largest ever fine – $12 million – was levied against Optus in November 2024 following the nationwide Triple Zero failure during its 2023 network outage. Telstra paid a $3 million infringement notice for a 90-minute Triple Zero call centre outage. CBA was fined $7.5 million for spam breaches involving 170 million marketing messages.
These are the current benchmarks. But the penalty regime is expanding. Proposed legislative reforms would increase the maximum civil penalty for breaching industry codes and standards from $250,000 to $10 million, with penalties potentially reaching 30% of adjusted turnover for the most serious breaches. The enhanced enforcement framework would also remove the current two-step process (direction to comply first, then enforcement), enabling ACMA to take direct and immediate action.
For TPG, the realistic near-term financial exposure from current ACMA investigations is likely in the tens of millions – material against a $7 million pre-tax profit base, though not in isolation a balance-sheet event. The more significant cost is likely the ongoing compliance uplift program that TPG has acknowledged in its statutory filings, the remediation and governance overhead, and the incremental cost pressure from TIO complaint volumes that are trending adversely against TPG brands while Telstra and Optus trend down.
What could the ACMA investigation actually be touching? The annual report is vague, but there are several plausible threads.
The TIO’s Systemics function is known to have received reporting on systemic billing issues and complaint-handling failures across TPG brands – the kind of pattern-based escalation that can trigger ACMA interest when individual cases are symptomatic of broader non-compliance.
Separately, dealer channel conduct remains a persistent risk vector across the industry: coverage representations that don’t match on-the-ground reality (particularly post-MOCN, where marketing has at times outpaced actual network experience in regional areas – hello: Dalton, NSW), number cycling to generate acquisition commissions, promotional port-in behaviour that inflates gross adds while destroying cohort lifetime value, and hardship or collections practices that intersect with vulnerable customer protections.
Any of these could form the basis for an ACMA compliance assessment or investigation. It could also be narrower – focused specifically on the Triple Zero failures and the circumstances surrounding the customer deaths. The honest answer is that we don’t know. What we do know is that TPG has disclosed active investigations and simultaneously assessed contingent liabilities as immaterial. Time will tell which assessment was correct.
The annual report’s contingent liabilities note states “no matters expected to result in a material effect on the financial position.” This may prove to be the correct assessment. But it sits alongside active investigations into incidents where people died, which at minimum suggests a high threshold for what management considers material.
Governance: The Bonus That Tells the Story
In a year of two customer deaths from Triple Zero failures, active ACMA investigations, a multi-year compliance uplift program, and $7 million in pre-tax profit, the Board assessed that all risk gateways had been passed and awarded CEO Iñaki Berroeta a $250,000 discretionary bonus for “outstanding leadership in a period of significant transformation.”
The Board also approved an equity grant to the CEO in December 2025 – three months ago – while the ACMA investigations were active and the compliance remediation was underway.
These are small numbers in absolute terms against a $7.3 billion market capitalisation. But they send a governance signal:
the Board is rewarding executive performance against a financial and operational backdrop that includes fatal service failures, active regulatory scrutiny, flat premium growth, and a pre-tax profit that rounds to zero.
The TIO complaints metric was added to the FY26 executive STI scorecard – an implicit acknowledgment that complaint performance needed executive incentive alignment – but the FY25 bonus and equity grant were approved without that alignment in place.
For investors focused on governance quality and alignment of executive incentives with shareholder outcomes, these data points warrant attention.
The Roger Montgomery Test
If you applied a quality-focused value investing framework to TPG – the kind of analysis Roger Montgomery at Montgomery Investment Management would run – the business fails on almost every dimension.
Return on capital: ROIC of 5.42% is below cost of capital. On any reasonable estimate of WACC, the business is still destroying value on incremental invested capital. Montgomery’s foundational principle is that intrinsic value is driven by returns above cost of capital. TPG has never achieved this since the merger. The handset receivables securitisation mechanically improves ROIC by reducing the denominator – adding roughly 70 basis points in FY26, getting to just over 6%. But that’s balance sheet monetisation improving the ratio, not operational improvement generating better returns on actual business assets.
Competitive moat: Narrow and not widening. The network is shared (MOCN with Optus), spectrum is leased from government at increasing cost, no single brand commands premium positioning, and pricing power is demonstrably weak – postpaid is flat despite price increases. In a three-player oligopoly, TPG is the weakest participant: the lowest ARPU, the thinnest earnings, the weakest brand, and the most precarious balance sheet heading into spectrum renewal. The number one player (Telstra) has brand and network leadership. The number two (Optus) has a deep-pocketed parent in Singtel. The number three has neither.
Capital allocation: The $3 billion capital return may have been premature given the $2 billion spectrum liability on the horizon. The dividend exceeds earnings by 6x and is funded from the cash flow gap between D&A and CAPEX – sustainable only if the asset base doesn’t need replacing at the rate it’s being depreciated. That’s exactly the tension with CAPEX restraint.
Earnings quality: Poor. Statutory NPAT relies on a non-recurring tax benefit. OFCF relies on a one-off securitisation. Management uses four different profit definitions (statutory, pro forma, guidance basis, underlying NPATA) – when a company needs four definitions of profit to tell its story.
The simplest explanation is usually that none individually tell a good story.
Owner earnings: Using the Buffett-influenced framework that Montgomery favours – net profit plus D&A minus sustainable maintenance CAPEX – you get roughly $600–700 million. Against a market capitalisation of roughly $7.3 billion, that’s an owner earnings yield of 8–10%. Not terrible in isolation, but not compelling for a business earning below its cost of capital, with a $2 billion spectrum liability approaching, a depleting tax shield, and no premium revenue momentum.
The verdict: A business earning low returns on capital at a “cheap” price is often more expensive than a business earning high returns on capital at a “fair” price. Telstra at a higher multiple but with genuine pricing power, premium brand positioning, ROIC above WACC, and infrastructure assets generating recurring income is the quality pick. TPG at a lower multiple but with none of those characteristics is, in Montgomery’s framework, a value trap.
The Provisions Question
Other provisions jumped from $2 million to $115 million in FY25, with $118 million “adjusted during the year.” Management attributes this entirely to Vocus Transaction separation obligations.
That’s plausible. Complex carve-outs generate tail costs – IT system cloning, transitional service agreements, infrastructure unwinding, employee provisions. At 2.5% of a $4.7 billion deal value, the quantum is within normal range for a major separation.
But it’s a single line item with no breakdown. And the contingent liabilities note says “no matters expected to result in a material effect on the financial position” – despite active ACMA investigations into incidents where customers died.
The opacity is the issue, not the number. A single line item with no breakdown, uncertain timing, and uncertain amount creates a $115 million bucket that external analysts cannot independently verify against its stated purpose. When the most serious regulatory matters facing the business don’t appear in the contingent liabilities framework, it raises questions about what else might be absent.
The $115 million provision will unwind into actual cash outflows over FY26 and beyond. That’s future cash flow drag already expensed but not yet paid – on top of spectrum costs, on top of a dividend that exceeds earnings, on top of ongoing MOCN costs, on top of a declining fixed business.
A Valuation Anchor
What is TPG worth on adjusted numbers?
Taking the normalised FCF to equity of $396 million as the FY25 base, growing to ~$430 million in FY26, and assuming the $600 million FY27 target is deferred to FY28 once spectrum costs are absorbed, a DCF at a 9% cost of equity and 2% terminal growth produces a range of roughly $3.10 to $3.80 per share depending on spectrum cost assumptions and payment phasing.
On an owner earnings basis – $600-700 million capitalised at a 9–10% required return – the range is $3.20 to $3.80 per share.
On an EPS basis, using the broker median of ~8 cents for FY26 and a 15x multiple (appropriate for a telco earning below its cost of capital with limited growth), you get roughly $1.20 per share – though this understates value because EPS doesn’t capture the cash generation from the D&A/CAPEX gap.
The most useful framing is probably the cash flow approach: TPG is worth $3.10 to $3.80 on normalised cash flows, with the current price of ~$4.02 sitting at or slightly above the top of that range. The stock is priced for the bull case – continued EBITDA growth, successful spectrum payment phasing, no dividend disruption, and no adverse regulatory outcomes. There is limited margin of safety for anything going wrong.
Morgan Stanley’s $3.50 Underweight target sits in the middle of this range. UBS’s $3.95 Neutral target sits at the top. The stock appears fully valued on a risk-adjusted basis, with the skew of potential outcomes tilted to the downside by the spectrum obligation.
The Silence on the Call
Forty-four minutes of analyst Q&A.
Eight brokerages represented: Barrenjoey, UBS, Evans & Partners, JP Morgan, Jarden, Macquarie, Morgans, Jefferies, CLSA.
Not one question about Triple Zero. Not one about ACMA. Not one about compliance. Not one about governance.
An annual report that discloses two customer deaths, active regulatory investigations, a multi-year compliance uplift program, and the introduction of TIO complaints as an executive remuneration metric for FY26 – and the professional sell-side let management walk through the entire call without a single governance question.
That silence is not the absence of a problem.
It is the absence of accountability.
When the Music Stops
TPG’s FY25 result was not a collapse.
The playbook has been:
- Sell the EGW assets ✓
- Return capital to shareholders ✓
- Pay down debt ✓
- Present the result as “transformational” ✓
- Blend reporting categories in a way that makes postpaid stagnation harder to isolate ✓
- Validate a $600 million FY27 FCF estimate on the earnings call ✓
And now:
- Spectrum bill of $2 billion arriving FY27-30 ✗
- Negative FCF in FY27 per UBS ✗
- ROIC still below cost of capital ✗
- Postpaid growth: zero ✗
- NBN declining ✗
- Tax shield depleting ✗
- Dividend exceeding earnings by 6x ✗
- EPS cut by 73% at one covering broker ✗
- CEO granted $250k bonus and equity while ACMA investigates – see Post #67 for full governance analysis ✗
The slides were polished.
The economics are precarious.
And the $2 billion spectrum bill doesn’t care how transformational the LinkedIn post was.
📩 Right of Reply:
TPG Telecom Limited, Vodafone Australia, and any executives or representatives referenced in this article are invited to provide clarification, correction, or additional context. Verified responses will be published in full and in context.
⚖️ Disclaimer:
This article represents independent commentary and analysis based on publicly released financial results, investor call transcripts, ACMA consultation documents, published broker research, and market data. All views expressed are opinions, not statements of proven fact. References to spectrum pricing reflect ACMA’s preferred position as published in December 2025, which remains subject to further consultation and finalisation. References to the TIO’s Systemics function and potential ACMA investigation scope are presented as analytical commentary on plausible possibilities, not assertions of fact. This does not constitute legal, financial, or investment advice. Readers should seek independent professional advice before making any decisions. All entities and individuals referenced retain the presumption of lawful conduct unless determined otherwise by a competent authority.
Leave a Reply