Copper Crosses Half the Company, and the Dividend Crosses Free Cash Flow: Initiating BHP at Hold

Results vs. Consensus

H1 FY26 Scorecard

BHP reports twice a year, and it publishes a quarterly operational review roughly four weeks ahead of each financial result. Volumes, realised prices and FY26 volume and unit-cost guidance were therefore already in the market on 20 January. What was genuinely new on 17 February was the margin outcome, the dividend, confirmation of the Jansen Stage 1 cost reset, and the Antamina silver stream. The scorecard below reflects that: the Street had published depth on two lines, underlying attributable profit and the interim dividend, and those are the only two against which a beat or miss is claimed.

Two independent estimates exist for each of the two forecast lines and both are shown rather than averaged, because the spread is the information: the market's central case for the dividend sat well below what the board declared, while its central case for profit was close to right.

Year-Over-Year Comparison (H1 FY26 vs. H1 FY25)

BHP defines Underlying EBITDA margin as Underlying EBITDA excluding third-party product EBITDA divided by revenue excluding third-party product revenue. The reported 58.4% therefore does not equal US$15,462M divided by US$27,902M, and the margin figures throughout this note are BHP's reported figures on BHP's definition.

Sequential Comparison (H1 FY26 vs. H2 FY25)

BHP does not publish a sequential half-on-half bridge. The comparison below derives the H2 FY25 column by subtracting the reported H1 FY25 figures from the reported FY25 (year to 30 June 2025) figures, both of which appear as columns in the same income statement and cash flow statement. It is the more demanding read of the half, because it removes the flattering effect of comparing against a weak base.

H2 FY25 column derived as FY25 (year ended 30 June 2025) less H1 FY25 (six months ended 31 December 2024) for each line, using the two comparative columns disclosed alongside H1 FY26 in the same statements.

The sequential table contains the single most useful fact in the result: attributable profit rose 22.5% half-on-half while net operating cash flow fell 9.7%. Earnings and cash moved in opposite directions across the same six months. That gap is working capital, and BHP names the cause directly.

Quality of Beat

Revenue. The US$2,726M of incremental revenue is a price story with a small volume assist. Group copper production was flat at 984.1 kt against 987 kt, while the average realised copper price rose 32% to US$5.28/lb. Iron ore volumes rose 2% to 133.8 Mt and the WAIO realised price rose 4% to US$84.71/wmt. Coal went backwards on price in both products, with steelmaking coal realisations down 9% to US$188.58/t and export energy coal down 23% to US$95.76/t. There is nothing inorganic in the number: no acquisition closed in the half, and the two portfolio transactions announced are cash-for-future-metal arrangements that had not completed.

Margins. The 730bp of margin expansion is where the operational case actually lives, and it is better than the headline suggests. Escondida cut unit costs 16% to US$1.12/lb and Copper South Australia cut 53% to US$0.74/lb, both helped by by-product credits from gold, silver and uranium that generated US$2.1bn of revenue across the copper portfolio, up 46%. WAIO's C1 cost of US$17.66/t rose only 0.9% against Australian inflation of 2.5% applied in BHP's own waterfall, while record material mined was up 9%. Set against that, WAIO's total unit cost including third-party royalties and inventory movements rose 7% to US$19.41/t, which sits in the upper half of the US$18.25 to US$19.75/t FY26 guidance range. Cost control was real; it was not uniform.

EPS and cash. Basic EPS rose 27.6% against a 22.0% rise in underlying EPS, and the gap is exceptional items: a US$562M after-tax charge in H1 FY26 versus US$666M in H1 FY25, so the statutory line flattered by comparison. Below the operating line, net finance costs rose 54.3% to US$705M on a larger debt stack, and the adjusted effective tax rate was essentially flat at 36.6%, or 43.0% including royalties. Free cash flow of US$2,906M grew 9.7%, less than half the rate of EBITDA, because US$2.3bn went into working capital.

Read the copper column on its own. Net price contributed US$2,971M and the segment gained US$2,941M. Every dollar of copper's EBITDA growth, and then some, came from price. Volumes cost the segment US$281M as Escondida's concentrator feed grade fell from 1.03% to 0.93% and Spence's planned cathode grade fell from 0.77% to 0.60%, and controllable cost savings of US$323M were largely handed back through exchange rates and inflation. The operational excellence BHP describes is genuine, and in this half it funded roughly nothing of the earnings growth.

Segment Performance

Group Underlying EBITDA by Segment

BHP's own stated contribution percentages of 51% copper, 48% iron ore and 1% coal are calculated on the sum of the three operating segments of US$15,673M, excluding the negative Group and unallocated line. The column above is calculated on Group Underlying EBITDA so that it reconciles to the income statement. On BHP's basis the prior-year split was copper 39%, iron ore 56%, coal 4%.

Asset-Level Detail

Asset-level EBITDA figures are as presented by BHP in its asset summaries and are stated on differing bases across equity-accounted and consolidated assets, so they are not additive to the segment totals above. The BMA and NSWEC percentage changes are declines. Unit costs are at realised exchange rates of AUD/USD 0.66 and USD/CLP 947 in both periods.

Copper

This is the half copper became the majority of BHP. US$7,952M of Underlying EBITDA at a reported 66% margin, up from 54%, on production that did not grow. Segment Underlying ROCE went from 13% to 22%. The composition matters: by-products contributed US$2.1bn of revenue, up 46%, from 259 koz of gold, 8.9 Moz of silver and 1.5 kt of uranium, and by-product credits are what drove Copper South Australia's 53% unit-cost reduction and roughly three-quarters of Escondida's 16%. BHP is running a copper business whose unit economics are increasingly levered to gold and silver prices.

"In copper, we generated a record $8 billion of EBITDA in the half, over half the group total at a margin of 66%. Our copper assets each produce significant amount of byproducts. We are not only the world's largest copper producer, but also a global top 20 gold producer and the world's third largest uranium producer, valuable positions at a time of near record prices of these commodities."
— Vandita Pant, Chief Financial Officer

Escondida is doing the heavy lifting and is doing it against a declining grade. Concentrator feed grade fell to 0.93% from 1.03%, and production still held flat at 646.1 kt on record concentrator throughput and improved recoveries. Unit costs fell to US$1.12/lb, below the bottom of the US$1.20 to US$1.50/lb FY26 guidance range. Full SaL leaching contributed 28 kt and is expected to yield roughly 410 kt of cathode over a decade. Copper South Australia produced 147.7 kt with record refined gold output of 113 koz, its second consecutive half above US$1bn of EBITDA. Spence was the weak asset, down 10% on planned lower cathode grade and port swells, with unit costs up 2%.

Assessment: The copper business is performing, but the earnings are price-driven and the volume trajectory for the rest of FY26 is not. Group FY26 guidance of 1,900 to 2,000 kt against 984.1 kt delivered implies an H2 of 916 to 1,016 kt, flat to slightly down at the midpoint, with Escondida guided to 554 to 629 kt in H2 against 646.1 kt in H1. The volume growth investors are underwriting sits in FY27 and beyond, not in the current year.

Iron Ore

WAIO produced a record first half at 129.8 Mt on a BHP-share basis, 146.6 Mt on a 100% basis, an implied run-rate of 300 Mtpa adjusting for the Car Dumper 3 renewal outage. The realised price of US$84.71/wmt rose 4%. Underlying EBITDA of US$7,499M rose 5% and segment ROCE was 43%, down a point. The cost position is the durable asset here: C1 of US$17.66/t against US$17.50/t a year earlier, an increase of under 1% while mining 9% more material in a 2.5% inflation environment.

"At Western Australia Iron Ore, we've increased our lead as the world's lowest cost major producer. In fact, we've reduced costs in real terms post-COVID, the only major Pilbara producer to do so. This is a critical advantage as competition in this market intensifies."
— Mike Henry, Chief Executive Officer

The phrase to note is "as competition in this market intensifies." BHP's own market commentary expects seaborne demand to plateau at current levels while supply increases, "alongside the new supply from Guinea." The company puts cost support at US$80 to US$100/dmt on a 62% Fe CFR basis, formed by roughly 180 Mt of higher-cost tonnes. Against a CY25 average of US$102/dmt, that is a market trading close to its own floor with new low-cost supply arriving.

Assessment: Iron ore is the funding asset, not the growth asset, and management is treating it correctly: incremental low-capital-intensity projects to hold more than 305 Mtpa from Q4 FY28 at under US$17.50/t, sustaining capex guided to roughly US$6.50/t, and a tenure-boundary agreement with Rio Tinto to unlock up to 200 Mt at Yandi. The FY26 guide implies an H2 of 121 to 132 Mt against 129.8 Mt delivered, so the year is a volume plateau. The risk is not execution, it is that a plateauing volume base meets a plateauing price with new Guinean supply on the other side of the trade.

Coal

Coal is now 1.5% of group EBITDA and shrinking on both sides. BMA's Underlying EBITDA fell 34% to US$0.3bn on a 9% decline in realised steelmaking coal prices, despite volumes up 2% and the highest first-half stripping volumes in five years. NSWEC fell 75% to US$0.04bn on a 23% decline in export energy coal realisations. Management is managing it as a wasting asset: Saraji South into care and maintenance in Q2 FY26, roughly 750 roles removed across Queensland, and NSWEC on a path to cease mining by the end of FY30.

Assessment: The Queensland royalty regime is the binding constraint, and BHP says so plainly: BMA is acting "against a backdrop of the material impact of the Queensland Government's coal royalties on business returns". At 1.5% of EBITDA, coal has stopped being a swing factor for the equity. The medium-term plan to lift BMA to 21.5 to 22.5 Mt at under US$110/t is the option worth watching, but it is a FY28-and-beyond story that requires the metallurgical coal price to cooperate.

Potash

Potash consumed US$1.0bn of capital in the half and generated US$(166)M of EBITDA, which is what a pre-production asset does. The market backdrop improved materially: prices averaged roughly US$340/t FOB Vancouver in H2 CY25, about 30% above the prior year, and the CY26 China contract settled at US$348/t, earlier than usual. CY25 demand hit an all-time high of roughly 75 Mt. Jansen Stage 1 is 75% complete for first production in mid-CY27 followed by a two-year ramp; Stage 2 is 14% complete for FY31.

Assessment: The commodity thesis for potash is holding up better than the project economics. At US$8.4bn for Stage 1 against roughly US$1bn per stage of expected steady-state EBITDA, the payback arithmetic on Stage 1 alone runs beyond eight years before Stage 2 capital is counted, and Stage 2's number is not yet published. This is the part of the growth pipeline where we want more information before crediting it in a valuation.

Key Topics & Management Commentary

Overall Management Tone: Assured and structural, with the emphasis placed on twenty-five-year track records and 2035 production pathways rather than on the current half. Management led with strategy and portfolio position and handled the two genuinely awkward items, the Jansen cost reset and a dividend larger than free cash flow, as single lines inside a longer growth narrative rather than as topics in their own right. The presentation was pre-recorded and carried no analyst question segment, so there was no mechanism by which either could be pressed.

1. Copper Becomes the Majority of BHP

The single structural fact of the half is that copper contributed 51% of Underlying EBITDA on BHP's stated basis, up from 39%, on the back of a 59% increase in segment EBITDA. Management framed this as the payoff from a decade of deliberate portfolio action rather than as a price windfall, and the four-year production record supports the framing even if this half's earnings do not.

"This half marks a milestone for BHP with Copper contributing the largest share of our overall earnings, at 51% of Underlying EBITDA. BHP is the world's largest copper producer and with strong performance at Escondida, and solid contributions from our other operations in Chile and South Australia, we have increased FY26 group copper guidance to 1.9 – 2.0 Mt."
— Mike Henry, Chief Executive Officer

Henry put the same point differently on the call, attributing the shift to specific operational decisions rather than to the copper price: more reliable operations at Olympic Dam, grade and sequencing discipline at Escondida, and the OZ Minerals acquisition. On his numbers, copper's earnings share is "up 30 percentage points over the past 3 years".

Assessment: The mix shift is real and it is the most valuable thing about this company, because it converts BHP from an iron-ore proxy with a copper option into a genuine diversified producer with the sector's best copper position. The caution is that a 12-percentage-point jump in earnings share in a single half is mostly the copper price moving faster than the iron ore price, and the same arithmetic runs in reverse. Judge the mix shift on the volume path, which is credible, not on this half's earnings share.

2. Escondida Guidance Raised Twice, and the More-Than-500 kt Claim

Escondida's FY26 guidance rose to 1,200 to 1,275 kt from 1,150 to 1,250 kt, and FY27 guidance was set at 1,000 to 1,100 kt against a prior medium-term range of 900 to 1,000 ktpa. Medium-term guidance for FY28 to FY31 stays at 900 to 1,000 ktpa at grades below 0.80%. Management aggregated these into a claim about cumulative incremental volume.

"Including the 400,000 tonnes of incremental production over 2027 to 2031 that we announced last year, the recent increase in guidance means we now expect to deliver over 500,000 more tonnes over the next 5 years compared to what we announced at the Chile site visit in 2024. At today's prices and margins, that would be an additional $5 billion of EBITDA over that period."
— Mike Henry, Chief Executive Officer

The mechanism is unglamorous and therefore credible: low-capital-intensity productivity work across the Laguna Seca concentrators, and a life extension for Los Colorados beyond FY29 followed by demolition to reach the high-grade PL2 zone earlier. The Escondida New Concentrator, worth 220 to 260 ktpa, remains a permit application in H2 FY26, an FID in CY27 or CY28, and first production in CY31 or CY32.

Assessment: This is the highest-quality disclosure in the result. Brownfield tonnes at an operating asset with a stated capital intensity of US$15,000 to US$21,000 per tonne of copper equivalent and a 13% to 16% IRR are worth more than any greenfield ounce in the portfolio. Note the framing though: the US$5bn of incremental EBITDA is quoted "at today's prices and margins", and today's price was a record. The volumes are the commitment; the dollar figure is a spot-price illustration.

3. Jansen Stage 1 Reset to US$8.4bn, Stage 2 Under Review

In July 2025 BHP gave a preliminary updated Jansen Stage 1 estimate of US$7.0bn to US$7.4bn. In January 2026 the detailed review landed at US$8.4bn including contingencies, roughly US$1bn above where analysts had carried it. First production remains mid-CY27, which is the original schedule, followed by a two-year ramp. Stage 2 expenditure is "under review" with an update promised in Q4 FY26.

"In January, we completed a detailed review of the cost and schedule estimates for Stage 1. First production remains on track for mid-2027, but we updated our cost estimate to $8.4 billion."
— Mike Henry, Chief Executive Officer

The written release adds that BHP "has implemented a response plan to address cost and schedule risks for JS1 which has improved productivity, strengthened project management and enhanced oversight of execution contracts", and that Stage 2 will "implement the project execution improvements identified in the detailed review of JS1." Local conditions are part of the story: BHP notes industrial construction costs in Saskatoon rising more than 12% over two years.

Assessment: Holding the schedule while resetting the cost is the better of the two ways to miss, and the honesty of the disclosure is to management's credit. But this is the only major project BHP currently has under construction, the estimate has now moved twice inside seven months, and the second stage of the same project is being advanced in parallel with its cost estimate withheld until Q4 FY26. The growth pipeline is the reason to own this stock at a premium; capital-estimate volatility on the one live project is the reason not to pay that premium yet.

4. The Antamina Silver Stream: US$4.3bn Upfront

Announced on the day of the result, BHP sold a silver stream over its 33.75% Antamina share to Wheaton Precious Metals for US$4.3bn upfront plus 20% of spot silver on each delivery. BHP delivers the equivalent of 33.75% of Antamina silver at a 90% fixed payable rate, stepping down to 22.5% after 100 Moz. Effective date 1 April 2026, completion on or around the same date, subject only to customary closing conditions with no regulatory approvals required. Copper, zinc and lead exposure is untouched.

"On completion, we will receive $4.3 billion in cash, an amount just shy of broker estimates of our share of Antamina's entire value. This follows our agreement in December in relation to our share of WAIO inland power consumption."
— Vandita Pant, Chief Financial Officer

She made the same comparison in the written release, describing the upfront consideration as comparing "favourably with the consensus estimates of our entire share of Antamina." BHP also states the transaction "is not expected to increase BHP's reported debt levels."

Assessment: This is genuinely good execution. Silver was a non-core by-product, silver prices are historically strong, and BHP has converted a stream of future by-product revenue into cash without giving up an ounce of copper or a share of the asset. The accounting claim deserves scrutiny over time, because a US$4.3bn upfront payment against future physical delivery is economically a financing whatever its balance-sheet classification. Timing is the point that matters for this half: the cash arrives around 1 April 2026, and the dividend it helps fund was declared in February.

5. Up to US$10bn of "Unlockable" Capital

Together the Antamina stream and the December WAIO inland power agreement with Global Infrastructure Partners, worth US$2bn for a 25-year tariff linked to power use, unlock more than US$6bn. Management repeatedly framed a larger number behind it.

"Across the group, we see potential for up to $10 billion in capital that could be unlocked and reinvested into higher returning opportunities and/or increase shareholder returns."
— Mike Henry, Chief Executive Officer

Pant characterised the pair of agreements as "examples of BHP's razor-sharp approach to capital portfolio and asset management", emphasising that the WAIO arrangement leaves ownership untouched: "BHP will retain full operational and strategic control of WAIO."

Assessment: A US$10bn program is roughly 5% of market capitalisation and, at a capex run-rate of US$11bn a year, more than ten months of the entire investment budget. That is material and it is the most interesting new disclosure of the half after the copper mix. Two things are unspecified: which assets make up the remaining US$4bn, and whether the proceeds go to the balance sheet, to growth, or to returns. Investors are being asked to credit the number before the composition or the destination is known.

6. The Dividend: 60% Payout at the Top of the Price Cycle

The board declared US$0.73 per share, US$3.7bn in aggregate, at a 60% payout ratio against the 50% minimum in the Capital Allocation Framework and the 50% actually paid a year earlier. Payment date 26 March 2026, NYSE ex-dividend 6 March. Management tied the decision explicitly to the framework's semi-annual assessment.

"Every 6 months, we assess shareholder returns through our capital allocation framework to ensure dividends reflect performance. Based on our strong results, confidence in our outlook and cash flows, we have determined a half year dividend of $3.7 billion, a payout ratio of 60%."
— Vandita Pant, Chief Financial Officer

Against that, free cash flow for the half was US$2,906M, and cash dividends actually paid during the half were US$3.1bn to BHP shareholders plus US$1.0bn to non-controlling interests. Net debt rose US$1.8bn to US$14,686M, and gearing moved from 19.8% at 30 June 2025 to 20.9%.

Assessment: The declared dividend equals 127% of half-year free cash flow. The gap is real but it is bridged, twice over, by the US$6.3bn of monetisations landing in H2, and net debt at US$14.7bn sits comfortably inside the unchanged US$10bn to US$20bn target range with A1 and A ratings intact. This is not a stretched balance sheet. It is a payout decision taken at the top of a copper price cycle and funded partly by non-recurring asset monetisations, which is a different proposition from a payout funded by operations, and the yield should be underwritten accordingly.

7. Working Capital and the Provisionally Priced Receivable

Net operating cash flow of US$9,372M rose 13% year on year but fell 9.7% against the derived H2 FY25 half, and BHP attributes the drag directly.

"Net operating cash flow increased 13% predominantly due to higher realised copper and iron ore prices. This included an offsetting increase in working capital of ~US$2.3 bn, primarily as a result of larger receivables driven by higher copper prices."
— from the half-year results announcement

The balance-sheet detail behind that sentence is worth spelling out. Provisionally priced trade receivables, carried at fair value through profit or loss, rose from US$2,581M at 30 June 2025 to US$3,989M at 31 December 2025, an increase of US$1,408M. Provisionally priced trade payables rose from US$493M to US$626M. This is also why the average realised copper price of US$5.28/lb sits above the US$4.72/lb average LME price BHP cites for the period: provisionally priced cargoes are remeasured to the period-end forward curve, and copper finished near US$5.67/lb.

Assessment: This is the most important thing in the result that management did not lead with. Roughly US$4.0bn of the balance sheet is copper sold but not yet finally priced, marked at a copper price near its record. If copper is materially lower when those cargoes settle, the mark reverses through the income statement, and the realised-price line that produced 96% of the EBITDA growth partially unwinds. It is not an accounting concern, it is a beta measurement: the earnings are more levered to the copper price than the production profile alone implies.

8. The Downside Stress Test Management Volunteered

The CFO put two figures on the record about cash generation under different price paths, one of the few genuinely forward-looking quantifications in the presentation.

"At spot prices, we expect to generate around $60 billion in attributable free cash flow over the next 5 years. That's cash flow after funding our investment in growth. Even in an extreme and prolonged low price environment, one in which prices fell 20% to 40% below current levels and stayed there for 5 years, we would generate around $10 billion in attributable free cash flow over that period."
— Vandita Pant, Chief Financial Officer

Assessment: The spread is the disclosure. A 20% to 40% price decline takes five-year free cash flow from roughly US$60bn to roughly US$10bn, a reduction of about 83% for a price move of 20% to 40%. That is the operating leverage of a fixed-cost extractive business stated plainly, and management deserves credit for volunteering it. It also quantifies exactly why we are not paying a premium multiple here: at US$10bn of five-year free cash flow against roughly US$189bn of market capitalisation, the downside case does not support the current dividend, let alone the growth program.

9. Copper South Australia: The Second Growth Engine

Copper SA has delivered more than 310 ktpa consistently since the Carrapateena and Prominent Hill integration, and every asset in it now has a growth program running. Prominent Hill's PHOX shaft project is on track for H2 FY27 at roughly US$0.9bn and extends mine life to at least 2040. Carrapateena's decline to the block cave base is 95% complete, lifting throughput to 12 Mtpa from FY29. Olympic Dam's Southern Mining Area Decline completes in FY28 and unlocks up to 2.5 Mtpa of additional vertical capacity. The Smelter and Refinery Expansion has a potential FID in CY27.

"If it were a stand-alone business, today, Copper South Australia would be a global top 15 copper producing asset, the fifth largest gold producer on the ASX and produce around 5% of the world's uranium."
— Mike Henry, Chief Executive Officer

Assessment: The stated path runs to more than 500 ktpa in phase one and up to 650 ktpa in phase two toward the late 2030s, at capital intensities of US$16,000 to US$21,000 per tonne of copper equivalent. This is 100%-owned, in a stable jurisdiction, with by-product optionality across gold and uranium that is currently doing real work on unit costs. It is the most underappreciated asset in the portfolio and the reason a Hold here is a waiting call rather than a negative one. A promised growth update "later this year" is the next catalyst.

10. Vicuña and the Greenfield Optionality

The Lundin joint venture published an updated technical assessment on the day of the result: Josemaria and Filo with a combined 47 Mt of contained copper, 97 Moz of gold and 1.8 Boz of silver, first-quartile cash costs, and peak production potential above 500 ktpa of copper, 0.8 Mozpa of gold and 20 Mozpa of silver. Recent drilling added another 9 Mt of contained copper, which Henry described as "equivalent to another 1.5 Josemarias." Vicuña will spend roughly US$800M on a 100% basis in CY26, has applied to Argentina's RIGI regime for 40 years of fiscal stability, and could take a Stage 1 FID as early as the end of CY26.

Assessment: Genuine world-class optionality, staged sensibly, and shared with a partner. It is also Argentina, a decade away from Stage 3, and pre-FID. We would carry Vicuña at option value rather than in a base case, and would want the RIGI outcome and the Stage 1 FID before revisiting. The same discipline applies to the 45% Resolution interest in Arizona, which remains a permitting story.

11. The Commodity Outlook Management Is Underwriting

BHP's published view is roughly 3.0% global growth in CY26, a China that met its "around 5%" CY25 target with the 15th Five-Year Plan expected to lift household demand, and India above 7%. On copper, the company sees demand growing from roughly 34 Mt today to more than 50 Mt by CY50 and a market that "still requires 10 Mt of additional, as-yet-uncommitted new supply, to be able to balance by CY35", with data-centre copper demand potentially growing sixfold to nearly 3 Mtpa. On iron ore it expects Chinese steel to plateau around 1 Bt into the late 2020s with seaborne demand flat and supply rising.

"Given the strong demand outlook, combined with the impact of disruptions at our competitor's mines, grade declines and the slow mine development pipeline, we anticipate a continued tight copper market over the next few years."
— from the half-year results announcement

Assessment: The copper view is consensus and we largely share it. The more useful disclosure is the asymmetry between the two big commodities: copper tight for years, iron ore plateauing on demand with supply rising including from Guinea. That asymmetry is the entire strategic rationale for the portfolio BHP is building, and it argues for owning this company at some price. It does not by itself argue for owning it at this price.

12. A Structurally Higher Cost Environment

The cost commentary was more candid than the margin outcome required. BHP flags Australian inflation above the RBA's 2% to 3% target band, Chilean regulatory change adding to labour costs "in the coming years", Saskatoon industrial construction costs up more than 12% over two years, and sulphuric acid prices "elevated." The waterfall shows US$270M of inflation and US$287M of adverse exchange rates together consuming more than the US$522M of controllable cash cost savings.

"Overall, these dynamics point to a structurally higher cost environment compared to pre-COVID norms. This reinforces the importance of productivity and cost discipline, while highlighting the competitive advantage of low-cost, diversified producers in a market where price support levels have shifted upward."
— from the half-year results announcement

Assessment: Note the second half of that sentence: BHP is arguing that structurally higher costs raise the price floor, which is convenient for a low-cost producer and is probably right. It is also the tell for why the FY26 unit-cost guidance ranges were left unchanged rather than lowered despite Escondida printing below its range. Management is not banking a cost win it expects inflation to take back.

Guidance & Outlook

Implied H2 ramp. Subtracting the reported half from the FY26 ranges gives an H2 that is flat to slightly lower on volume across the board. Group copper of 916 to 1,016 kt against 984.1 kt delivered, with Escondida at 554 to 629 kt against 646.1 kt and Copper SA at 162 to 192 kt against 147.7 kt. Group iron ore of 124 to 135 Mt against 133.8 Mt. NSWEC energy coal of 6.9 to 7.9 Mt on the upper-half steer, against 8.1 Mt. Capital and exploration expenditure of roughly US$5.7bn against US$5.3bn, a 9% step up. The second half is a price story with a mix rotation inside copper, not a volume ramp.

Guidance style. BHP guides conservatively on volume and then raises, which is what happened at Escondida twice inside a month. It guides unit costs in wide ranges and then steers to an end of the range rather than resetting the range, which is what happened at Escondida (bottom end), BMA (upper half) and NSWEC (upper half). The one place the pattern broke is Jansen, where a preliminary estimate given in July 2025 was superseded by a materially higher definitive estimate in January 2026.

What is not guided. There is no FY26 revenue, EBITDA or EPS guidance, which is normal for the sector. There is also no stated intention on the payout ratio for the FY26 final dividend, no allocation plan for the US$6.3bn of incoming monetisation proceeds, and no Jansen Stage 2 capital number until Q4 FY26.

What They're NOT Saying

With no analyst question segment on this call, the omissions carry more weight than usual. Nothing below was asked and dodged; all of it simply went unaddressed in a presentation management controlled end to end.

1. That the dividend exceeded free cash flow. The presentation described "growth in cash returns", a 60% payout ratio and "confidence in our outlook and cash flows". It did not put the US$3.7bn declared dividend next to the US$2,906M of free cash flow that generated it, or next to the US$1.8bn increase in net debt over the same six months.
2. Where the monetisation proceeds go. US$6.3bn completes in H2 FY26 with up to US$10bn identified, and the stated destination is "higher returning and more value-accretive uses, including growth and shareholder returns." That formulation covers every possible answer. Whether this money retires debt, funds Escondida and Copper SA, or returns to shareholders is the most consequential open capital-allocation question at the company, and it was left open.
3. Which assets make up the remaining US$4bn. Two transactions are disclosed and a US$10bn total is asserted. The gap is unspecified, and the two completed examples, a by-product stream and an infrastructure tariff, do not obviously generalise to another US$4bn without touching something closer to the core.
4. The Jansen Stage 2 number. Stage 2 is 14% complete and being advanced "in parallel", with expenditure "under review" and an update deferred to Q4 FY26. Stage 1 has just moved from a US$7.0bn to US$7.4bn preliminary range to US$8.4bn definitive. Building a second stage while withholding its capital estimate for another two quarters is a choice, and the silence points one way.
5. The provisionally priced receivable. US$4.0bn of receivables marked at period-end copper prices, up US$1.4bn in six months, appears in a financial-instruments note and nowhere in the presentation. Management discussed the working-capital build as a cash-flow timing item and did not connect it to the realised-price line that produced the earnings.
6. Why the FY26 unit-cost ranges were not lowered. Escondida printed US$1.12/lb on both realised and guidance exchange rates, below a US$1.20 to US$1.50/lb range that was steered to "the bottom end" rather than re-guided. Either the H2 cost base at BHP's largest copper asset is expected to deteriorate sharply, or the range is being kept deliberately wide. Neither was explained. The same question does not arise at Copper South Australia, whose US$0.74/lb realised print becomes US$1.34/lb at the AUD/USD 0.65 rate the guidance is set on, comfortably inside its range.
7. The China Mineral Resources Group. BHP's iron ore commentary covers Chinese steel output, port inventories, cost support and Guinean supply. It does not mention the centralised buying entity that has been renegotiating seaborne iron ore contract terms, which was on the pre-result watch list and bears directly on the realised-price line of the segment that still produces 48% of group EBITDA.
8. What happens to Escondida after FY27. FY26 and FY27 guidance were raised; FY28 to FY31 medium-term guidance stayed at 900 to 1,000 ktpa with grades below 0.80%. The near-term raises therefore pull volume forward into a period of falling grade rather than lifting the medium-term plateau, and the presentation framed the raises without that context.
9. Samarco beyond the settlement schedule. The provision fell to US$5.3bn from US$5.8bn, with FY26 cash of roughly US$2.2bn and FY27 of roughly US$0.6bn. The release refers to "updates from the UK group action" as one input to the provision movement without further detail, and the presentation did not mention Samarco at all.
10. Anything about capital returns beyond the ordinary dividend. With more than US$110bn returned since 2016, a balance sheet at 20.9% gearing inside a US$10bn to US$20bn net debt target, and US$6.3bn of monetisation cash arriving, the buyback question is live. It was not raised or ruled out.

Market Reaction

• Pre-print setup (ADR): BHP closed at US$73.38 on 13 February 2026, entering the print up 21.6% year to date against the S&P 500 down 0.1%, up 42.3% over twelve months, and up 13.1% over the prior thirty days. The 52-week closing range was US$40.22 to US$75.13, placing the stock 2.3% below its 52-week closing high. Sell-side positioning as at 10 February was 4 buy, 8 hold and 1 sell.
• Timing: BHP released to the ASX before the open on Tuesday 17 February AEDT, which is Monday evening US Eastern. The NYSE was closed on Monday 16 February for Presidents' Day, so the ADR's single Tuesday session absorbed two Australian sessions.
• Reaction session (ADR, 17 February): opened US$72.75, a 0.9% gap down; traded a range of US$71.79 to US$74.35, which is 2.2% below to 1.3% above the pre-print close; closed US$74.29, up 1.2% or US$0.91. Volume of 4.6M shares against a 4.5M thirty-day average, a multiple of 1.0x.
• Reaction session (local line, 17 February): the ordinary shares closed at A$50.36 on 16 February, opened at A$53.00, traded to an all-time high of A$54.20, and closed at A$52.74, up 4.7% on the day and up 3.1% from the 13 February close of A$51.13.
• Peers (17 February closes): Rio Tinto down 1.2%, Freeport-McMoRan down 2.8%, Vale down 4.5%, S&P 500 up 0.1%. BHP was the only large diversified or copper name higher on the session.

The relative move is the signal, not the absolute one. A 1.2% gain on 1.0x volume looks like the market shrugging. Set against a peer group that fell between 1.2% and 4.5% on the same session, it is a 2.4 to 5.7 point relative gain, and it happened while the S&P 500 was flat. Something knocked the mining complex down on 17 February and BHP alone was exempt. The result did work.

The ADR understated the reaction, for a mechanical reason. The Australian line closed 1.5% lower on 16 February, the last session before the print, while the ADR was shut for the US holiday and could not participate. When the ADR reopened on 17 February it was marking against a Friday close that had already run ahead of the Sydney price, which is why it gapped down 0.9% and traded 2.2% below the pre-print close intraday before recovering to close at the high end of its range. The A$52.74 close on the local line, up 4.7%, is the cleaner read of what holders of the actual result thought of it. Currency was not a factor: AUD/USD moved from 0.7069 on 13 February to 0.7085 on 17 February.

The give-back matters too. The ordinary shares traded to an all-time high of A$54.20 and closed at A$52.74, surrendering roughly a third of the intraday gain, then fell a further 0.9% on 18 February. That is a market that liked the dividend, respected the copper mix, and declined to re-rate the multiple on it. Our read of the fade is that the buyers who chased the headline dividend met sellers who had already seen the volumes on 20 January and had done the free-cash-flow arithmetic.

Street Perspective

Debate: Is the copper mix shift a re-rating event or a price artefact?

Bull view: Copper crossing half of group EBITDA is the moment BHP stops being valued as an iron-ore proxy. The Street applies a structurally higher multiple to copper earnings than to iron ore earnings, and a company whose earnings mix has moved 12 percentage points toward copper in a single year, with a credible 40% volume growth path to FY35, deserves a mix-weighted re-rating that has not yet happened.

Bear view: Earnings share is an output of relative prices, not of portfolio construction. Copper EBITDA rose 59% on flat volumes. Run the same portfolio at a copper price 25% lower and iron ore flat, and copper's share falls back below 45%. Paying a copper multiple for a mix ratio produced by a copper price near its record is the classic cycle mistake.

Our take: The bears have this half and the bulls have the decade. The volume path is real and mostly brownfield, and by FY30 the mix argument will be structural regardless of price. But right now the mix ratio is doing work in the narrative that the volumes have not yet done in the plant, and the correct response is to underwrite the volume path rather than the earnings share. That is a reason to own BHP at a fair price, not to pay up for it today.

Debate: Is the 60% payout sustainable or a cycle-peak gesture?

Bull view: BHP has returned more than US$110bn since 2016, the balance sheet is A1-rated at 20.9% gearing inside an unchanged target range, and US$6.3bn of monetisation cash lands in H2. Management has proved for a decade that the framework produces the right payout for the conditions, and 60% at the top end of a strong half is exactly what the framework is designed to do.

Bear view: A payout equal to 127% of free cash flow, funded on arrival by asset sales rather than operations, is a distribution decision dressed as a policy outcome. If copper mean-reverts, the payout ratio goes back to 50% on a smaller base, and the yield investors just bought halves. The framework's 50% minimum is the number to plan around; the extra 10 points is discretionary and was granted at the best possible moment.

Our take: The bear case describes the mechanism correctly and overstates the risk. The balance sheet can carry this comfortably and the monetisations are contracted, not hoped for. But nobody should capitalise a 3.6% trailing yield as a forward yield. Model the 50% minimum, treat anything above it as a copper-price dividend, and the stock looks fairly valued rather than cheap.

Debate: Does the growth pipeline justify a premium to the diversified peer group?

Bull view: No other diversified miner has this combination: the largest copper position in the world, a 3% to 4% copper-equivalent CAGR to FY35 that is predominantly brownfield, the lowest-cost major iron ore business funding it, and a new potash business that is differentiated from every other commodity in the portfolio. Scarcity of genuine copper growth means BHP should trade above Rio Tinto and Vale on any forward measure.

Bear view: The pipeline is mostly permits, studies and FIDs. The Escondida New Concentrator is a CY27 or CY28 FID for CY31 or CY32 production. Vicuña is pre-FID in Argentina. Resolution is a permitting question. Copper SA's growth update is promised but not delivered. The one project actually under construction, Jansen, has had its cost estimate raised twice in seven months with the second stage's number withheld. That is a pipeline with a poor recent track record of capital estimation.

Our take: The bear case is the more useful discipline here. A growth pipeline earns a premium when the market has evidence that capital estimates hold, and the one live data point runs the other way. Escondida's brownfield increments are the exception and are the highest-confidence part of the plan. We would pay a premium for Escondida and Copper SA, carry Vicuña and Resolution as options, and require the Q4 FY26 Jansen Stage 2 number before crediting potash beyond Stage 1.

Debate: Is iron ore a stable annuity or a slowly deflating one?

Bull view: WAIO is the lowest-cost major producer globally, has cut costs in real terms post-COVID, delivers roughly US$10 per tonne more free cash flow than its nearest major competitor, and has a low-capital path to more than 305 Mtpa. BHP's own cost-support estimate of US$80 to US$100/dmt puts a floor under the price. This is a decades-long annuity that funds the copper build.

Bear view: BHP's own outlook says seaborne demand plateaus while supply increases including from Guinea, and Chinese steel plateaus around 1 Bt. An annuity where volume is flat, price drifts toward a cost floor, and a new low-cost entrant is arriving is a deflating annuity. Segment ROCE already slipped from 44% to 43% in a half where realised prices rose 4%.

Our take: Both are right on different horizons. Through FY28 the annuity holds and the cost lead widens. Beyond that, Simandou-scale supply meeting plateaued demand compresses the whole cost curve, and the correct assumption is a lower mid-cycle iron ore price than the last five years delivered. That is precisely why the copper pivot matters, and precisely why paying a full price today for the pivot's early innings is premature.

Model & Valuation Framework

This is an initiation, so the table below sets the drivers we intend to carry rather than revising an existing model. Aardvark Labs does not currently publish a financial model for BHP.

Where the stock trades. Trailing twelve-month basic EPS, taking the derived H2 FY25 figure of 90.7 US cents plus H1 FY26's 111.1 US cents, is 201.8 US cents per ordinary share, or US$4.04 per ADS given that each ADS represents two ordinary shares. Against the 17 February close of US$74.29 that is roughly 18.4 times trailing statutory earnings. On an approximate underlying basis, using FY25 underlying attributable profit of about US$10.2bn less H1 FY25's US$5,082M to derive the H2 FY25 half, trailing underlying EPS is close to US$4.46 per ADS and the multiple is roughly 16.7 times. Market capitalisation is approximately US$189bn on the 5,077M weighted average share count, and enterprise value approximately US$203bn including net debt, or roughly 6.6 times annualised first-half Underlying EBITDA.

Yield. Declared dividends over the trailing twelve months are the FY25 final of US$0.60 plus this interim of US$0.73, or US$1.33 per ordinary share and US$2.66 per ADS, a trailing yield of 3.6% at the reaction-day close. On a 50% payout assumption applied to a lower copper price, we would plan around something closer to 3%.

Valuation impact. Nothing in this result changes our view of intrinsic value enough to justify chasing the print. The earnings beat was small, the dividend beat was a payout decision, and the volumes were known four weeks earlier. What did change is the quality of the FY27 copper picture, via the Escondida raise, and the balance-sheet flexibility, via the monetisations. Both accrue to FY27 rather than FY26. We would become buyers on a pullback toward the mid-US$60s per ADS, or on evidence that H2 unit costs held at H1 levels while the monetisations closed and the payout returned inside free cash flow.

Thesis Scorecard: Establishing Coverage

This is our first published note on BHP, so the scorecard below establishes the pillars rather than grading a standing thesis. Each is stated with the signpost that would move it, and each will be carried forward and graded at the FY26 result.

Overall: A high-quality operator executing a strategically correct pivot, priced as though the pivot is already complete. The bull pillars are all on track and none is in doubt. The bear points are all emerging rather than contained, and all three concern the same underlying question: how much of this half was the business and how much was the copper price.

Action: Hold. Own the asset base at a fair price, decline to pay up for a half in which price supplied 96% of the earnings growth. We would upgrade to Outperform on any of three developments: a pullback toward the mid-US$60s per ADS, an FY26 result showing H2 unit costs held with the payout back inside free cash flow, or an explicit allocation of the monetisation proceeds to the Escondida and Copper South Australia growth programs.

Bottom Line

BHP's December half was a good result delivered by a very good company at a very good moment for its most important commodity, and those three things are not the same thing. Copper crossing half of group earnings is a genuine milestone, and the operating detail underneath it, record Escondida throughput against a falling grade, a 53% unit-cost reduction at Copper South Australia, WAIO holding C1 costs flat while mining 9% more material, is the work of an organisation that runs its assets better than its peers run theirs.

But the company's own waterfall says 96% of the earnings growth came from price, and its own balance sheet says US$4.0bn of receivables are marked at a copper price that finished the period near a record. The dividend that generated the headline was 127% of the half's free cash flow, bridged by two asset monetisations that had not yet closed. And the one major project actually under construction had its capital estimate raised for the second time in seven months while the second stage's number was deferred another two quarters.

The market read this correctly on the day. A 1.2% ADR gain against a peer group down between 1.2% and 4.5% says the result did real work; an all-time-high print on the local line that gave back a third of its gain by the close says nobody re-rated the multiple on it. We agree with both halves of that verdict. Initiating at Hold, with a clear path to Outperform and no desire to chase the print to get there.