This is investment research, not personal financial advice.
The rally was about profit, but the harder question is cash conversion
AMP Limited rose about 5.3% to A$2.295 in late-morning trade after its 1H26 release put three numbers in front of the market: underlying NPAT up 33% to A$174 million, Platforms net cashflows up 33% to A$3.1 billion, and a further A$150 million on-market buyback (ASX 2026; AMP 2026a). The move added roughly A$260 million of equity value on the ASX header data used for this article.
That is not a wild repricing for a company that just printed half-year underlying profit equal to 61% of FY2025 underlying NPAT. It is also not a clean compounding story yet. AMP is still a wealth-platform, superannuation, New Zealand, partnerships and bank-runoff group whose reported result blends operating progress with capital release. The commissioning question is narrow: did the market correctly reward a higher-quality wealth result, or did it capitalise capital returns that still depend on a shrinking bank and lower legacy noise?
The observational answer sits between those two readings. The rally looks broadly proportionate if 1H26 marks a new flow base for North and Superannuation & Investments. It looks generous if the A$150 million buyback is treated as recurring operating cash rather than surplus capital from a portfolio still being simplified. The result changed the evidence, but it did not finish the proof.
What the result actually changed
The trigger document was AMP's 6 August 2026 result release. Management said underlying NPAT increased 33% to A$174 million, statutory NPAT increased 57% to A$154 million, AUM reached A$167.6 billion, and Platforms net cashflows rose 33% to A$3.1 billion for the half (AMP 2026a). Superannuation & Investments delivered its first positive half-year net cashflow result since 2017, while China partnerships contributed A$56 million, more than double the prior comparable half.
The causal chain is straightforward. AMP's equity value has spent years being judged less on a single earnings multiple and more on whether the post-simplification group could stop leaking flows, release trapped capital, and turn underlying profit into cash that belongs to shareholders. The 1H26 release hit each part of that chain. It showed flow momentum in the wealth engine, a better statutory conversion after legacy matters, and a concrete capital return through an interim dividend plus buyback.
But the composition matters. AMP generated A$236 million of surplus capital in 1H26 and returned A$201 million through dividends and buybacks, then announced the further A$150 million buyback (AMP 2026a). The bank capital-release strategy produced an A$89 million surplus capital position at half-end. That is useful, but it is not identical to self-funding growth from the platform business. A market reaction that simply capitalises A$174 million of half-year underlying NPAT and adds the buyback risks double-counting the same capital base.
The better frame is a repricing of quality. Yesterday's AMP was still marked by legacy settlements, advice exits and a bank question. Today's AMP has more evidence that North and retirement can carry the story. The stock's 5% move implies the market is raising confidence in that evidence, not declaring the turnaround complete.
The business now has two engines and one release valve
AMP today is not the old vertically integrated financial-services conglomerate. The continuing group is led by Platforms, Superannuation & Investments, New Zealand Wealth Management, Partnerships & Group, and AMP Bank. Platforms, centred on North, earns administration and investment-platform revenue from advisers and clients. Superannuation & Investments earns AUM-based revenue from member balances and retirement products. New Zealand is a smaller wealth-management business. Partnerships includes China Life Pension Company and other partnership or sponsor exposures. AMP Bank is increasingly a capital-management variable rather than the long-term strategic centre of the group.
That mix creates two operating engines and one release valve. The operating engines are platform flows and retirement member economics. They compound when adviser relationships, product functionality and investment balances increase AUM without requiring equivalent increases in controllable costs. The release valve is bank and surplus capital management. It can support per-share value through dividends and buybacks, but it does not widen the moat unless the remaining business can grow after that capital is returned.
The 1H26 data pack helps separate the two. AMP reported total revenue of A$671 million, controllable costs of A$315 million, EBIT of A$206 million, underlying NPAT of A$174 million and statutory NPAT of A$154 million for the half (AMP 2026b). Cost-to-income improved to 60.5% from 63.0% in 1H25, and underlying ROE rose to 9.8% from 7.4%. Those are operating-quality markers.
The flow figures are the more important evidence. Platforms net cashflows of A$3.1 billion show adviser-market relevance. Superannuation & Investments turning positive for the first half since 2017 matters because it attacks the old bear case directly: that AMP's brand and product set could not retain members without price or capital support. One positive half does not prove durability, but it changes the burden of evidence.
The history still shows a repaired company, not a smooth compounder
AMP's recent financial history is messy because the company has been simplified, businesses have been sold, legacy costs have moved below the line, and capital has been returned. The table below uses reported underlying NPAT as the cash-earnings proxy, statutory NPAT for accounting conversion, and reported underlying ROE as a practical return metric. ROIC and incremental ROIC labels in the frontmatter are author mappings to AMP's disclosed underlying ROE and changes in underlying NPAT over the simplified equity base, not source-reported industrial ROIC.
| Period | Revenue / operating income (A$m) | Statutory NPAT (A$m) | Underlying NPAT / owner-earnings proxy (A$m) | EPS | Return metric | Balance-sheet note |
|---|---|---|---|---|---|---|
| FY2021 | 3,197 | -252 | 356 | -7.6c statutory | 8.4% underlying ROE | post-royal-commission repair still dominant |
| FY2022 | 2,397 | 387 | 184 | 12.0c statutory | 4.6% underlying ROE | portfolio simplification and AMP Capital sale |
| FY2023 | 1,718 | 265 | 196 | 9.3c statutory | 5.0% underlying ROE | buybacks reduced shares on issue |
| FY2024 | 1,784 | 150 | 236 | 5.7c statutory | 6.4% underlying ROE | advice exit and capital returns continued |
| 1H2026 | 671 | 154 | 174 | 6.1c statutory | 9.8% underlying ROE | A$236m surplus capital generated in the half |
The table should not be read as a clean industrial compounding sequence. FY2021 and FY2022 include businesses and costs that no longer sit inside the same economic perimeter. FY2024 is a better base for the current group, and 1H2026 is the freshest evidence. On that base, the result is strong: half-year underlying NPAT of A$174 million compares with FY2024 underlying NPAT of A$236 million and FY2025 underlying NPAT of A$285 million reported in the 1H26 data pack (AMP 2026b; AMP 2024).
The owner-earnings bridge starts with the A$174 million underlying NPAT. Statutory NPAT was A$154 million after A$20 million of items below underlying NPAT. Surplus capital generation was A$236 million, but that figure is not the same as operating free cash flow. It includes capital release and balance-sheet management. For valuation, the cleaner mid-cycle base is not A$236 million doubled. It is annualised underlying NPAT adjusted down for the possibility that China partnerships and unusually strong flows moderate, then checked against statutory conversion.
That gives a working owner-earnings range of A$320 million to A$380 million for a base case. The lower end assumes 1H26 contains some catch-up and market tailwind. The upper end gives AMP credit for a more durable North and S&I flow reset. The difference is the valuation argument.
Moat evidence improved, with two obvious holes
The moat case is no longer just cost cutting. North appears to have adviser relevance, retirement demand is visible, and member cashflows have shifted in the right direction. Those are better signs than a one-off statutory beat because wealth platforms scale when flows bring revenue without matching cost growth. Cost-to-income moving down by 2.5 percentage points also supports operating leverage (AMP 2026b).
The first hole is margin. Platforms can grow AUM and still disappoint if administration margins compress or if adviser acquisition requires product spend that offsets scale. AMP's own result points to functionality and the North Interactive Wealth Portal as part of the adviser appeal (AMP 2026a). That may help retention, but it also means the platform must keep investing. A wealth platform without ongoing product investment becomes a melting ice cube in a competitive adviser market.
The second hole is capital quality. The further A$150 million buyback is a strong per-share signal, but a buyback funded by release of excess capital is finite. A buyback funded by recurring operating cash is more valuable. AMP's 1H26 release still ties capital management partly to the bank release strategy. Until the bank runoff is further advanced, the market has to split capital returns into recurring and non-recurring buckets.
Peer context also matters. HUB24 and Netwealth have trained the market to reward clean platform flow growth with premium ratings when organic flows, margins and technology spend line up. AMP is not that clean. It has platform growth, but it also has a bank, partnerships, legacy history and statutory noise. That is why the same flow improvement should not mechanically receive the same multiple as a pure platform peer (HUB24 2025).
Valuation: the post-result price sits near the middle, not the edge
AMP's A$5.25 billion equity value at A$2.295 per share means the market is paying about 15.5 times a A$340 million owner-earnings base, or 13.8 times a A$380 million base. On a simple earnings approach, that is no longer distressed, but it is not full platform-peer pricing either. It is a mixed-business multiple for a company proving that its wealth flows can offset the declining strategic weight of the bank.
The severe downside case uses A$250 million to A$285 million of sustainable owner earnings and a 13 times or lower multiple after allowing for renewed outflows, elevated below-the-line costs, and slower bank capital release. That produces A$1.45 to A$1.70 per share after allowing for the current share count.
The bear case uses A$300 million to A$325 million of owner earnings and an 13 to 14 times multiple. The assumption is not failure. It is a weak but stable AMP where the half-year result fades to a lower run-rate and surplus capital is useful but not repeatable. That gives A$1.75 to A$2.05 per share.
The base case uses A$340 million to A$380 million of owner earnings, a 14 to 15 times multiple, and modest credit for the A$150 million buyback. It produces A$2.15 to A$2.55 per share. This range straddles the post-result price, which is why the late-morning rally looks proportionate rather than stretched on the evidence available today.
The bull case requires the most discipline in wording. It is not simply 1H26 doubled. It assumes positive S&I flows persist, Platforms remains above A$3 billion per half of net flows, China partnerships remain resilient, and statutory conversion stays close to underlying profit. A 16 to 17 times multiple on A$390 million to A$430 million of owner earnings supports A$2.75 to A$3.20 per share.
A reverse valuation says the market is now pricing roughly A$340 million to A$360 million of sustainable owner earnings at a mid-teens multiple. That is close to annualised 1H26 after a haircut, not a heroic assumption. The sensitivity is concentrated in two variables: net cashflows and statutory conversion. A one percentage-point change in the multiple is worth about 15 to 17 cents per share on the base earnings range. A A$40 million change in sustainable owner earnings is worth about 24 to 27 cents per share at the same multiple.
What would settle the argument
The crux is flow durability. The FY2026 result and the 1H2027 update need to show whether Platforms net cashflows stay near the 1H26 level and whether Superannuation & Investments can stay positive. If S&I slips back into material outflow, the first positive half since 2017 will read as a milestone rather than a reset. If it stays positive while North keeps gaining adviser flows, the moat evidence strengthens.
The second crux is capital release. AMP Bank generated a surplus capital position of A$89 million at half-end, and capital release remains part of the group's capital-management strategy (AMP 2026a). The market will need to see whether releases arrive without dragging down earnings quality or leaving costs behind. That resolves through the FY2026 balance sheet, the buyback completion updates, and any further bank strategy disclosure.
The third crux is statutory conversion. In 1H26, A$174 million of underlying NPAT became A$154 million of statutory NPAT. That is a far better conversion than the worst legacy years. If below-the-line items stay under control, the market can place more weight on underlying NPAT. If they re-expand, the rally's quality premise weakens.
Source notes
Confidence is partial rather than full. The triggering announcement, data pack, presentation and annual reports from FY2021 to FY2024 were fetched and read. The legal identity was cross-checked against the ASX company page and ASX header API, but the repo identity helper could not auto-resolve the page. The main missing item is the completed FY2026 annual report, because today's release is a half-year result. ROIC-style frontmatter fields are author mappings from AMP's disclosed underlying ROE and should be read as return-on-equity proxies, not source-reported industrial ROIC.
References
- ASX 2026: ASX company page and header data for AMP Limited, used for the late-morning price, equity value, shares-on-issue reconciliation and legal identity.
- AMP 2026a: AMP's 6 August 2026 1H26 result release, the triggering announcement for the price move, profit growth, AUM, cashflows, buyback and surplus-capital facts.
- AMP 2026b: AMP's 1H26 data pack, used for revenue, costs, EBIT, underlying NPAT, statutory NPAT, EPS, ROE and business-unit detail.
- AMP 2026c: AMP's 1H26 results presentation, used for management's explanation of the flow and capital-management story.
- AMP 2024, AMP 2023, AMP 2022 and AMP 2021: AMP annual reports used for the multi-year financial-history table and capital-allocation context.
- RBA 2026: cash-rate context for bank economics and wealth-market discount-rate conditions.
- HUB24 2025: platform peer context for flow-quality and market-rating comparison.
- APRA 2026: banking-system regulatory context for AMP Bank capital and deposit-market framing.
The main evidence gap is the absence of a completed FY2026 annual report. The 1H26 row is an interim row, and the ROIC-style metric used here is a services-company mapping from AMP's disclosed underlying ROE rather than a source-reported industrial ROIC calculation. Verification is therefore partial, even though the triggering announcement, data pack and four annual reports were retrieved and read for this run.