DRC's 96.6 kt cobalt quota — barely half of 2024 actual mine output — is a multi-year structural fact. Indonesia, scaling toward 60+ kt of cobalt-in-MHP this year, is the only at-scale alternative supply geography and the principal beneficiary of the most violent cobalt re-pricing since 2018.
The cobalt market entered its second structural break in three years on 22 February 2025, when the Democratic Republic of Congo (DRC) suspended cobalt exports through ARECOMS Decision 001. What began as an attempt by Kinshasa to defend prices has hardened into a multi-year quota regime — 96.6 kt of permitted exports per year for 2026 and 2027 against an unconstrained 2024 mine output of approximately 220 kt. The result is the most violent re-pricing of cobalt in the post-2018 era: hydroxide payables briefly touched 100% of metal benchmarks in February 2026 (against a 55–75% long-run norm), and SMM's Indonesia MHP cobalt FOB index has more than doubled in nine months.
Against this backdrop, Indonesia has emerged as the only at-scale alternative source. From under 2 kt of cobalt-in-MHP a decade ago, Indonesian output is on track for 53–67 kt this year and 70+ kt by 2027, with twelve operating or under-construction HPAL units totalling close to 1,000 kt/yr of nickel capacity. By 2030, Indonesia is positioned to supply roughly one-fifth of global mined cobalt — a structural transition that will affect every node of the cobalt value chain, from sulphate refining margins in Hebei to cathode procurement contracts in Detroit and Munich.
This report uses SMM's proprietary daily, monthly and forecast datasets to make four arguments: (1) the DRC quota regime is a multi-year, not transient, fact and Indonesia is its primary beneficiary; (2) Indonesian HPAL economics have been transformed by the cobalt byproduct credit — the average new HPAL line is now earning approximately USD 51 million per kt of contained cobalt per year at current MHP FOB prices; (3) Indonesia's 2026 fiscal and royalty package now formally prices cobalt as a sovereign-relevant byproduct, ending an era in which it was effectively rounding error in HPM benchmarks; and (4) downstream demand growth has softened — particularly in China, where LFP captured 81.5% of EV power battery installs in April 2026 — but the supply shock dwarfs the demand softness, leaving the 2026 China cobalt balance in a 5–8 kt deficit on the SMM scenario model.
For Indonesian miners, the implications are commercial: lock in offtake, build local refining, and prepare for the cobalt-passport requirements that will gate EU and IRA-aligned demand from 2027. For Indonesian policymakers, the implications are strategic: cobalt is now a sovereign asset worth a national stockpile policy and a coherent ESG/community framework. For downstream OEMs and cell-makers, the implications are immediate: Indonesia's MHP cobalt is the structural alternative to DRC hydroxide, but it comes with terms — joint ventures, technology transfer, and increasingly, local battery cell investment. The eleven chapters that follow lay out the data, the scenarios, and the actionable conclusions.
The research base for this report combines three SMM proprietary data layers — daily price benchmarks (over 320 trading days of CNY/mt-metal-content cobalt-product pricing from January 2025 through April 2026), monthly forecast models (4 cobalt products × 13 monthly forecast periods × 3 annual periods × low/mid/high price bands), and the SMM cobalt and nickel value-chain scenario simulator (2024–2030 by region) — together with public-domain customs data, regulatory disclosures from the DRC's ARECOMS and Indonesia's ESDM, and the consensus of leading international cobalt research (Cobalt Institute, Fastmarkets, Benchmark Mineral Intelligence, S&P Global Commodity Insights, IEA, and GlobalData).
Per SMM's official cobalt industry chain framework, cobalt moves through five stages (Ore → Concentrate / Intermediate → Refined Cobalt Supply → Downstream Processing → End Use) and resolves at the refined-supply layer into four mainstream products, grouped as cobalt compounds (sulphate / chloride / Co3O4) and metallic cobalt (refined / powder).
2025E end-use shares total 99% due to SMM rounding (28+40+14+1+16). Source: SMM Cobalt Industry Chain Framework, 2025-09 slides 3 + 5.
What this means for Indonesia › Indonesia sits on stage 1 (Ni-Co laterite mining) and stage 2 (MHP intermediate); the policy ambition articulated in Chapter 8 is to climb to stage 3 (sulphate refining and electrolytic cobalt) and selectively into stage 4 (precursor and cell). Every additional stage captured doubles the unit value retained on Indonesian soil. What blocks the climb: domestic sulphuric-acid logistics (HPAL is sulphur-intensive and Indonesia imports the marginal tonne), long-cycle precursor / cell offtake commitments from Korean and Chinese cathode-makers, and IRMA / EU Battery Passport readiness — without sustained progress on all three, the stage 3 – stage 4 ambition stays rhetorical.
From a three-year price slump that bankrupted small Chinese refiners to a 12-month rally that more than doubled the SMM Indonesia MHP index — the cobalt market has compressed an entire cycle into 15 months. This chapter lays out the chronology, the catalysts, and what is genuinely new in the cobalt narrative.
Cobalt's two-decade story is the story of how a single mineral — produced in only a handful of geologies and controlled by a few large miners — interacts with one extremely volatile demand vector: rechargeable batteries. The 2017–2018 bull market took refined cobalt above USD 90,000/t on the back of a perceived shortage; the 2019–2024 bear market took it back below USD 25,000/t as DRC scaled, Indonesian MHP came online, and LFP cathodes squeezed cobalt-bearing chemistries out of the mid-range EV. Entering 2025, SMM tracked refined cobalt at CNY 170,500/t and cobalt sulphate at CNY 130,244/mt metal content (early-January 2025 close) — barely above marginal cash cost for non-integrated Chinese smelters.
The cycle would have been complete had Kinshasa not intervened. On 22 February 2025, ARECOMS — DRC's mineral marketing authority — issued Decision No.001/ARECOMS/2025 suspending all cobalt exports for an initial four months. Within 30 days, refined cobalt was trading at CNY 220,000/t; within 90 days, the export halt had been extended; by October 2025 the halt was replaced with a formal quota framework of 18,125 t for Q4 2025 and 96,600 t/yr for 2026 and 2027.
The chart below collapses the price reset into a single view: five SMM-tracked cobalt products, all moving from a sideways drift through Q1 2025 into a near-vertical rally that began in late February and accelerated as DRC's policy hardened. But the most analytically important feature is the order in which they rose. Per Fastmarkets and SMM's own benchmark series, the year-on-year change to early 2026 reads: cobalt hydroxide CIF China +340% (USD 5.80/lb early 2025 → USD 25.44–25.80/lb Jan–Feb 2026), cobalt sulphate +150–180%, refined cobalt metal +160%, and Co3O4 +140%. This is the SMM-recognised stair-step pattern of a supply-driven re-pricing — intermediate > salts ≈ refined metal > secondary compounds — precisely the opposite of the demand-led 2017 spike where the metal itself led.
SMM has named three reasons why this second rally has been less violent in absolute amplitude than 2017, yet structurally more dislocating: (i) policy suddenness — Kinshasa's 22 Feb 2025 decision arrived without warning, sterilising the usual lead-time-driven inventory build that cushioned 2017's run; (ii) inventory perception and purchasing behaviour — Chinese refiners and OEMs entered 2025 long on stock after 18 months of glut, and the resulting "we have time" reflex slowed the pull-forward bid; (iii) market valuation — the post-2024 cobalt sector traded at multi-year-low EV/EBITDA multiples, anchoring price expectations and dampening the speculative overshoot that amplified 2017. The supply-driven nature of the move is unmistakable: global cobalt demand grew only ~3% in 2025 and the Cobalt Institute's Q4 2025 update put 2026 demand growth at 7% — well below the >15% that drove the 2017 squeeze. The dislocation is in the supply curve: a quota system has effectively removed ~125 kt/yr of DRC capacity, against a global demand pool of ~220 kt.
Price logic matrix — five cobalt products, two drivers
Source: SMM proprietary price assessment + SMM monthly cost-profit dataset + SMM Cobalt Industry Chain Framework, slide 11 (2025-09).
The most important data point is not the price spike but the persistence of the quota framework. The 96.6 kt cap for 2027 means that even if Glencore and CMOC operate flat-out, the DRC ceiling for 2027 will be roughly half of 2024 actual output. With recyclate growth at only ~6 kt/yr through 2030 and no other geology able to add at-scale, Indonesia is the only marginal supplier that can move the global balance — and it is doing so against a backdrop where every additional kilotonne of Indonesian cobalt earns ~USD 50 million at current prices.
The near-term deceleration signal — read carefully. Cobalt-salt prices are no longer in full breakaway mode. SMM's monthly assessment shows the cobalt sulphate (metal-content) index gained only +0.6% m/m in April 2026 versus +1.2% m/m in February 2026; cobalt chloride gained +0.2% m/m; refined cobalt actually retraced −2.8% m/m as social inventory absorbed the marginal sell. The signal is not that the rally is over — hydroxide payables remain near 90%+ and SMM's forecast band points to October 2026 peaks — but that the supply-driven leg of the rally has matured, and the next move is conditional on Indonesian MHP ramp execution, DRC quota strictness through Q3 2026, and downstream LCO procurement cycles. For Indonesian offtake counterparties, this deceleration is the actionable timing cue: the market is in transition, contract terms locked in May–September 2026 will define cash margin through 2027.
Myth one: "the price rally is a speculative spike that will reverse in months." The data argues otherwise. The cobalt hydroxide payable — the ratio of intermediate price to refined metal price — touched 100% in February 2026 against a long-run 55–75% norm. This is not a speculative phenomenon: it is a physical-market signal that Chinese refiners have no alternative but to bid up scarce intermediate to maintain throughput. Spikes reverse when speculators close positions; physical-market dislocations persist until supply or demand structurally rebalances. The DRC quota framework removes the supply side from any near-term rebalancing.
Myth two: "LFP cathode dominance will collapse cobalt demand." The data shows a more nuanced picture. LFP has indeed captured 81.5% of Chinese power battery installs by April 2026, but Chinese power battery installs are only one sub-segment of cobalt demand. Globally, LFP penetration in EV battery installs is ~63% in 2025 with the remaining 37% almost exclusively cobalt-bearing (NMC, NCA). Cobalt's share of each NMC unit has shrunk (NMC811 contains only 10% cobalt versus NMC532 at 20%), but rising premium-EV penetration, consumer electronics renewal cycles, and emerging humanoid-robot demand offset the per-unit decline. Cobalt Institute Q4 2025 modelling shows ~7% global cobalt demand growth in 2026 — slower than 2017–2019 but not negative.
Myth three: "Indonesia will saturate the market and crash prices in 2027–2028." The data suggests this risk is overstated for the first three forecast years. Indonesia's 2026 ramp of 53–67 kt of cobalt-in-MHP barely covers the supply gap left by DRC. Even with the full 2026 capacity additions and a generous 2027 ramp, the global cobalt market remains in deficit or marginal balance through 2027 on the SMM model. The genuine oversupply risk is a 2028+ phenomenon dependent on whether the CMOC stockpile is released — a scenario covered in Chapter 10.
DRC's cobalt export apparatus is now a quota machine. This chapter quantifies the volumes, the producer-level allocations, and the price-payable consequences — and explains why CMOC's stockpiling strategy implies the shock may outlast the formal quota period.
The DRC cobalt policy arc since early 2025 unfolded in four discrete acts. Act I — the surprise total export halt on 22 February 2025 — was framed by the DRC Mines Ministry as a defensive response to a collapsing cobalt price, but in practice was triggered by an internal review showing that hand-dug (artisanal) cobalt had grown to roughly 35% of total exports versus an official EGC franchise. Act II — the June 2025 extension and the parallel formalisation of EGC as exclusive buyer of artisanal output — locked in a multi-stakeholder reform package. Act III — the October 2025 transition from ban to quota — formalised an annual cap of 96,600 t (87,000 t producer "basic" + 9,600 t ARECOMS strategic reserve). Act IV, ongoing through 2026, has been the rollout of producer-by-producer allocations: Glencore's KCC + Mutanda allocation set at 22,800 t for 2026 (carry-in from Q4 2025), CMOC's at approximately 31,200 t against guidance of up to 120 kt of mined cobalt — a 74% reduction in the company's ability to monetise its own production.
SMM tracking of China customs data (indicator a10001579, the official monthly import volume of cobalt hydrometallurgy intermediate products from the DRC) gives an unambiguous quantitative confirmation. Through January 2025 China was importing 50–60 kt/month of DRC cobalt intermediate (physical tons); by August 2025 the figure had collapsed to 4,805 t; by January 2026 to 948 t; and the most recent reading for April 2026 stood at 945 t — a 98% reduction in 12 months. This is the largest, fastest contraction in any major mineral import on record.
Producers have responded in three distinct ways. CMOC declared force majeure on shipments from DRC in March 2025 and has since publicly indicated that excess cobalt is being retained in upstream copper circuits — effectively building DRC-side cobalt inventory that may exceed 50–80 kt by year-end 2026 (no independent confirmation, but consistent with the gap between guidance and quota). Glencore has accepted the quota and is selling to LME and OTC at the inflated prices, with its 2026 quota of 22.8 kt potentially contributing of order USD 1 billion of revenue at H1 2026 cobalt metal levels (indicative SMM analyst calculation). Other smaller producers have been pushed out of the export queue and are reportedly negotiating sub-quota arrangements with Kinshasa.
The CMOC stockpile question matters: if Kinshasa relaxes the quota in 2028 and CMOC releases held inventory, the market could face a 50–80 kt overhang that crashes prices. Conversely, if quotas tighten further (a "pessimistic 87 kt" scenario tracked in Chapter 10), the cobalt market enters a multi-year deficit that pushes hydroxide payables permanently above historic norms.
The clearest signal of how strained the cobalt value chain has become is the cobalt hydroxide payable — the discount that refiners pay relative to the metal benchmark for cobalt contained in intermediates. The long-run norm is 55–75%; in February 2026, hydroxide payables briefly touched 100%, meaning refiners were paying as much for raw hydroxide as they could sell pure cobalt metal for. This is unprecedented and reflects a market in which Chinese refiners have no alternative but to bid up scarce hydroxide just to maintain throughput.
Indonesia hosts the only HPAL build-out in the world that can move the global cobalt balance this decade. Twelve operating or under-construction units — six Chinese-led, three Western- or Indonesian-led, three multi-party — will together command roughly 1,000 kt/yr of nickel capacity by end-2027, with co-produced cobalt rising from ~25 kt in 2024 to a possible 80+ kt by 2030.
The current Indonesian HPAL fleet falls into four cohorts. The foundation cohort — Halmahera Persada Lygend (Obi Island), Huayue Nickel & Cobalt (IMIP), QMB New Energy Materials (IMIP) — was commissioned between 2021 and 2023, totals approximately 230 kt/yr nickel and 25 kt/yr co-produced cobalt, and has now achieved consistent operation above nameplate at HNC and Huafei. The 2024 ramp cohort — PT Huafei Nickel & Cobalt, ESG New Energy Material, Meiming — added another 180 kt/yr nickel capacity. The 2026 commissioning cohort — SLNC (90/10), ENC (72/8), Pomalaa (120/15), Vale–Huayou Sorowako (66/8), Huashan (120/13) — will add roughly 468 kt/yr of nickel and ~54 kt/yr of cobalt over an 18-month window. A 2027–2030 pipeline of further CNGR / Tsingshan / IWIP HPAL lines is in permitting or early construction.
Approximately USD 25–30 billion of capital has been deployed or committed across Indonesian HPAL since 2019, with Chinese majority ownership concentrated in IMIP and IWIP and a more mixed pattern at MBMA (Merdeka), Vale Indonesia, Harita, and Nickel Industries projects. Benchmark Mineral Intelligence estimates that Chinese ownership of Indonesian battery-grade nickel/cobalt processing — currently approximately 80% — will fall to ~54% by the early 2030s as Vale, Harita and other non-Chinese partners expand. The single biggest non-Chinese FID of the past three years was Nickel Industries' Excelsior Nickel Cobalt (ENC) project, which is also the first Indonesian HPAL designed to produce MHP, cobalt sulphate and cathode in a single integrated complex.
The cancellation of the BASF–Eramet Sonic Bay project in June 2024 remains the most consequential negative event in the Indonesian HPAL pipeline. The project would have added 67 kt/yr nickel and 7.5 kt/yr cobalt of capacity through a Western-aligned joint venture explicitly designed to qualify for IRA tax credits and EU Battery Passport. Its loss removed roughly 10% of the planned 2027 cobalt capacity and concentrated the remaining capacity firmly under Chinese-affiliated ownership. Eramet has since pivoted to a New Caledonia / French strategy and BASF has withdrawn from the Sonic Bay HPAL JV and signalled a strategic retreat from Indonesian upstream processing — a strategic retreat that has implications for European auto OEMs seeking IRA-style critical-mineral compliance.
Three numbers anchor the Indonesian capacity story. First, Indonesia's share of global mined cobalt has risen from 6.7% in 2024 to a forecast 14.9–22% in 2026, depending on the source. Second, Indonesia is the only producer geography adding net new capacity at scale — DRC is constrained by quotas, Australia is exiting, the Philippines and New Caledonia are flat. Third, Indonesia's economics now reward cobalt: the byproduct credit at current MHP FOB prices of USD 51,000/mt Co is large enough (~USD 1,000–1,500/t Ni-equivalent credit on a 13:1 Ni:Co ratio) to make a marginal HPAL line cash-positive even at low nickel prices.
The 17 May 2026 Strait of Hormuz tanker-blocking incident lifted spot elemental-sulphur prices from ~USD 145/t to peak USD 240/t within ten trading days — a 65% spike on top of the 2024–2025 baseline. Russia's parallel elemental-sulphur export restrictions (re-confirmed by Rosneft in April 2026) compounded the squeeze. Indonesian HPAL operators import the marginal sulphur tonne; PT Huafei Nickel & Cobalt (Sulawesi, ~140 kt/yr nameplate Ni) partially throttled output through the second half of May 2026, removing an estimated 600–900 t of Co-in-MHP from the spot pool for the month. The feedback into the cobalt payable was immediate: the SMM-implied hydroxide payable, which had eased from the February 100% high back to 86% by mid-April, rebounded to 93–94% over 17–28 May. The lesson for buyers is that the cobalt payable is now sulphur-elastic in a way it was not pre-2025: any sustained Middle East logistics disruption transmits within ten working days into the marginal cobalt unit, regardless of DRC quota dynamics.
Beyond the immediate sulphur episode, Indonesia's capacity ramp is real but not without execution risk. Sulphuric acid supply — the dominant variable cost in HPAL — remains exposed to Hormuz/Russia volatility as documented above. Tailings management remains a separate chronic constraint: Indonesia bans deep-sea tailings disposal (re-confirmed by Coordinating Minister Pandjaitan in 2023), and filtered tailings — the alternative — have failed at IMIP twice (March 2025, February 2026 dam failure). A commissioning slippage of 6–9 months should be assumed for every project not already operating, and ESG-linked offtake terms (EU Battery Passport, IRMA) should be baseline expectations from 2027.
The Indonesian capacity question is not only "how much" but "where to." Of the projected ~60 kt mid-case Indonesian cobalt-in-MHP for 2026, Fastmarkets estimates that the substantial majority — approximately 90% or 53–60 kt — is destined for export to China for refining into sulphate, chloride and Co3O4; the remaining 5–10% represents incremental domestic refining capacity at Huayou and Lygend's joint ventures with CATL and Antam. By 2028, the SMM model projects that incremental Indonesian refining capacity (sulphate, precursor, cell) will rise to 10–15 kt of cobalt, but the majority will still flow to China as MHP. This pattern matters because it determines who captures the margin: the upstream miner gets the MHP FOB price, the refiner gets the sulphate spread, and the precursor / cell-maker captures the integrated value. Indonesia's domestic value capture today is dominated by the upstream margin; the policy goal — articulated by the Ministry of Industry — is to shift the centre of gravity to the midstream over the next five to seven years.
Looking forward to 2030, three scenarios bracket the Indonesian capacity story. In the base case, the operating 2024 fleet plus the 2026–2027 commissioning cohort plus a measured 2028–2030 pipeline brings Indonesian cobalt output to 76–80 kt, equivalent to ~19% of global mined cobalt. In the upside case, an accelerated 2028–2030 build-out (driven by US/EU strategic-mineral commitments, Indonesian domestic battery cell ramp, and continued high cobalt prices) pushes output past 95 kt, or ~23% of global supply. In the downside case, sustained tailings incidents, sulphur supply constraints and a global cobalt price reversion to USD 25,000/mt Co cap Indonesian output at 65 kt and discourage the 2028+ FID wave. The probability-weighted central estimate sits at ~78 kt for 2030, but the upside-downside spread is wide and policy-dependent.
Cobalt's refined supply splits into four mainstream products that move on distinct demand signals. This chapter walks through SMM's monthly supply-demand balance and price forecast for each, from May 2026 through April 2027.
SMM organises the refined cobalt supply layer into two chemical families. Cobalt compounds include cobalt sulphate (Co content factor 0.205 — i.e. 1 physical-tonne sulphate ≈ 0.205 metal-tonne Co), cobalt chloride (factor 0.247) and cobalt tetroxide (Co3O4) (factor 0.728) — the last is a secondary processed compound, sintered from cobalt sulphate or cobalt chloride rather than refined directly from intermediates. Metallic cobalt includes refined (electrolytic) cobalt (≈100% Co content) and cobalt powder. Together these comprise the four mainstream cobalt products that this chapter tracks. The downstream applications differ sharply: sulphate routes to NCM / NCA ternary precursors for EV batteries; chloride routes mainly to Co3O4 for LCO consumer-electronics cathodes; refined cobalt routes to super-alloys, magnetics, tool steel, and aerospace; cobalt powder routes to hard-metal cemented carbides.
Cobalt sulphate is the volume backbone of the cobalt market, accounting for approximately 58,800 t of forecast 2026 demand from ternary precursors alone, plus 11,800 t from Co3O4 (which is itself a sulphate-derived stream in some routes) and 2,500 t from other uses. SMM's monthly model shows the sulphate market in persistent deficit through Q3 2026, with the 2026 annual balance at −7.7 kt and recovering to a +1.5 kt surplus only in 2027 as Indonesian MHP feedstock scales. The 2026 mid-price forecast of CNY 97,219/mt metal content (+74% vs 2025) reflects this tightness; even the low-band forecast (CNY 95,483) sits 74% above the 2025 average.
Cobalt chloride sits between sulphate and Co3O4 in the LCO value chain. SMM's 2026 monthly balance shows 78,159 t/yr of supply (71,452 t of domestic production plus 6,980 t of integrated electrolytic-cobalt-route supply, less 274 t of net exports) against demand of 79,585 t — comprising Co3O4 55,265 t, refined cobalt feedstock 6,980 t, cobalt powder 8,483 t and additives/other 8,857 t. The result is a marginal 2026 deficit of −1,426 t, swinging to a 6,508 t surplus in 2027 as Indonesia-MHP-derived chloride supply ramps. The 2026 mid-price forecast of CNY 116,266/mt metal content (range 114,267 – 118,180) reflects sustained Q1–Q3 strength followed by easing into 2027 (mid 110,450) as supply normalises.
Co3O4 is the smallest of the four mainstream cobalt products by volume but the most cyclical, tied directly to consumer electronics demand for lithium cobalt oxide cathode. Unlike sulphate, chloride or refined metal, Co3O4 is a secondary cobalt compound — it is sintered from cobalt sulphate or cobalt chloride rather than refined directly from intermediates, which is why its 2026 cost-of-conversion sits near 1.5 wan-CNY/physical-tonne (≈2.06 wan-CNY/Co-tonne) on top of the salt-side margin. SMM's 2026 forecast shows Co3O4 production of 91,762 physical tons and a tight supply-demand balance (production roughly equal to the 91,357 t of demand-side absorption after integrated-LCO offsets, while the gross underlying LCO sector demand exceeds 111 kt). Q1 weakness is offset by H2 recovery as consumer-electronics inventory rebuilds. The price forecast reflects this: a 2026 mid of CNY 366,485/mt physical (range 360,430 – 372,539) — 63% above the 2025 average — easing to CNY 348,263 in 2027 as supply normalises.
Refined cobalt — the headline LME-tracked metal — is the smallest market by volume but the most important for price discovery and trade. SMM's 2026 monthly balance shows a sharp contraction in Chinese supply: domestic production of 11,549 t (down 64% from 31,725 t in 2025, reflecting DRC feedstock loss) supplemented by net imports of 3,361 t (China has flipped from net exporter of refined cobalt in 2025 to net importer in 2026) for total 2026 supply of 14,910 t. This collides with domestic demand of 27,637 t — alloys 15,263 t (the dominant end-use), magnetics 2,452 t, other industrial 9,921 t — leaving a structural deficit of −12,727 t. Monthly trade data show refined-cobalt imports of 961 t in March 2026 falling to 826 t in April 2026 then partly recovering to 894 t by December 2026, while exports rise from 413 t to 1,972 t by April 2027 as Chinese refiners arbitrage the international cobalt metal premium. The 2026 annual mid-price forecast of CNY 459,743/t (low 452,035 / high 467,450) — up 69% from the CNY 272,741 average of 2025 — reflects the most extreme dislocation of the four products, with monthly peaks in October 2026 reaching CNY 491,783 before easing to CNY 481,905 by April 2027.
Refined cobalt supply has two layers: the static yield-and-cost matrix that defines who can profitably make what, and the dynamic margin path that has whip-sawed since 2024. SMM's estimation table makes the first transparent; SMM's daily cost-profit dataset makes the second auditable.
Every refined-cobalt product can be traced back to a feedstock × process pair with a known yield and an industry-typical unit conversion cost. SMM's cobalt product estimation table (an internal but widely circulated SMM industry-chain reference) tabulates the six commercially relevant routes. Two structural facts emerge from the table that are critical to interpreting Indonesia's cost position:
Note: The estimation table's "Co3O4" rows are denominated in wan-CNY per physical tonne (the "wan" prefix is abbreviated in the original SMM table); at 72.8% Co content this converts to ≈2.06 wan-CNY/Co-tonne. Source: SMM Cobalt Product Estimation Table (2025-09 industry chain reference).
Insight 1 — MHP and DRC intermediate are at near-parity for cobalt sulphate. The MHP → sulphate route costs 2.8–3.0 wan-CNY/Co-t against 2.7–3.0 for the DRC intermediate route — a gap of <5%. For a precursor buyer this is the empirical basis for why Indonesian MHP is a structural substitute for DRC hydroxide on the dominant volume product. For an Indonesian HPAL operator it is the empirical basis for prioritising sulphate-bound offtake over electrolytic-cobalt-bound offtake, where MHP carries a modest cost handicap.
Insight 2 — Co3O4 is a downstream of the salts, not of the intermediate. The estimation table makes this unambiguous: there is no intermediate → Co3O4 row. Every Co3O4 tonne is sintered from cobalt sulphate (99% yield) or cobalt chloride (the dominant route in practice). This puts Co3O4 producers structurally downstream of the sulphate / chloride margin, with a 1.5 wan-CNY/physical-tonne conversion add-on that compresses when feedstock prices outrun finished Co3O4 selling prices — exactly the dynamic of Q1–Q2 2026. For Indonesian operators the policy implication is clear: do not prioritise domestic Co3O4 capacity. The integrated salt-then-Co3O4 economics structurally favour staying upstream of the Co3O4 step and exporting sulphate or chloride to Chinese sinterers — who already operate at scale and absorb the cyclical LCO-demand risk. Indonesian downstream investment should target sulphate and precursor (NMC/NCA), not Co3O4 and LCO.
Insight 3 — Re-electrolysis from salts is dramatically cheaper than primary electrolysis from intermediate. Sulphate / chloride → refined cobalt at 1.0–1.3 wan-CNY/Co-t versus intermediate → refined at 3.5–3.8: a ~3× cost gap that explains why during periods of high refined-cobalt price (such as H1 2026), Chinese smelters with surplus salt inventory will pivot to re-electrolysis rather than waste payable-rich intermediate on a process where they earn less margin.
The 2026 HPM revision (Chapter 8) explicitly prices cobalt at +USD 2.3/wmt on HPAL ore and +USD 3.2/wmt on NPI ore. Combined with three exogenous variables — sulphuric acid cost (driven by the Hormuz Strait incident of May 2026), captive-coal electricity tariffs, and the ad-valorem royalty layer of PP 18-19/2025 — the HPM mechanism transmits these costs into the MHP cash cost stack. The downstream effect: every 30% step-up in sulphur cost or every USD 5/MWh increase in CFPP tariff raises the MHP → sulphate route cost by approximately 0.15–0.20 wan-CNY/Co-t, eroding part of the <5% advantage versus DRC intermediate. The byproduct credit (USD 51 million/kt Co at May 2026 MHP FOB) currently absorbs this — but the loop is real, and any sustained sulphur or power-cost shock will narrow Indonesia's cost edge over time.
The SMM cost-and-profit dataset tracks daily cash production cost and profit for cobalt sulphate refining from purchased cobalt intermediate. The data tell a story that runs opposite to the popular "DRC shock means refiners win" narrative: from January 2025 to April 2026, non-integrated sulphate refining has been persistently loss-making, with daily profit per metal tonne ranging from CNY −700 to CNY −6,000 throughout the period. Through January–February 2025 the losses were a modest CNY −1,000 to −1,200, typical of pre-shock low-price equilibrium. By July 2025 the loss widened to CNY −2,300 as intermediate cost spiked faster than sulphate selling prices; by April 2026 the daily loss reached CNY −3,557 per metal tonne. The intermediate price moved up 2–3× over 12 months; sulphate selling prices moved up by less because precursor buyers refused to absorb the full pass-through. The structural lesson is that integrated Chinese refiners with offtake or equity stakes in Indonesian MHP capture an internal transfer-price advantage that effectively absorbs the loss, while non-integrated refiners — who must buy intermediate at spot — face an existential margin squeeze. Several non-integrated Chinese sulphate refiners have either idled capacity or shifted to toll-processing arrangements over Q1 2026.
The most strategically important profit narrative in the 2024–2026 cycle is the electrolytic (refined) cobalt reversal — and SMM has named it precisely: a transition from high-profit warehousing to deep loss and passive destocking. In the pre-ban period (2024 through early 2025) Chinese electrolytic-cobalt producers ran a high-profit regime supported by three structural cushions: (i) accumulated high-margin inventory from the prior high-price era; (ii) bonded-warehouse arbitrage between SHFE-deliverable and Indonesian/DRC-sourced material; and (iii) financial-product speculation (including cobalt-linked investment funds and physical-backed structured notes) that absorbed metal off the spot market and slowed the natural price decline. Together these mechanisms postponed the cash-cost capitulation that would normally have closed the merchant capacity by mid-2024.
The post-ban regime reversed all three. The 22 February 2025 export halt and the subsequent quota framework forced refiners to source from a shrunken hydroxide pool at payable structures touching 100%; the cost-line lifted by ~CNY 234,000/t over 16 months while refined cobalt selling prices on the SHFE did not fully follow. By April 2026 the SMM daily cost-profit dataset shows electrolytic cobalt at −69,325 CNY/t for the intermediate route and −63,795 CNY/t for the sulphate route — both routes in losses larger than the historical price range, with the two cost lines converging near a common loss floor. Crucially, despite these losses, social-warehouse inventory has continued to flow out: it has fallen from 225 days in May 2025 to 63 days by April 2026, a 72% draw-down (Chapter 6). This is the textbook signature of passive destocking — physical inventory exiting because end-users are taking it, not because producers are voluntarily releasing it, even as refiners themselves run cash-cost negative. The economic pattern is unmistakable: integrated producers (Huayou via Indonesian HPAL equity, Jinchuan via partial DRC integration) ride out the cycle on transfer-pricing advantages; merchant refiners are being structurally pushed toward consolidation, idling, or exit.
Chloride refining has shown smaller absolute swings than the sulphate and refined-cobalt routes but is now also moving into losses as the cost line tracks intermediate prices upward. SMM data show daily chloride profit moving from approximately flat (CNY ±500/t) in 2025 to mildly negative through Q1 2026, with non-integrated chloride refiners in cash-cost territory by April 2026. Co3O4 refining shows a more mixed picture, with both intermediate-fed and chloride-fed routes oscillating around break-even in 2026; the integrated route (via intermediate) maintains a small positive margin while the chloride-fed merchant route shows compressed margins — reflecting Co3O4's tighter consumer-electronics demand and slower price pass-through. The cost-curve charts in this chapter make the divergence visible: cost lines have moved up materially across all four products, but selling prices have not kept pace for non-integrated routes.
The dominant strategic conclusion from the cost data is that integration matters more in 2026 than in any prior year of the cobalt market. A refiner with equity in an Indonesian HPAL line captures the cobalt byproduct credit and pays an effective intermediate transfer price below market — generating CNY 5,000–8,000/t of incremental margin over a non-integrated peer. This explains the strategic intensity around Indonesian HPAL offtake contracts in late 2025 / early 2026: every major Chinese refiner not already integrated has been actively negotiating multi-year MHP supply agreements with Indonesian producers.
Cobalt inventories are the early-warning system for the next price move. SMM's segmented inventory data shows where stocks have built and where they have run down, and which products carry the most reflexive price risk.
The SMM inventory-days dataset tracks three buckets for cobalt sulphate: non-integrated upstream refiners, integrated producers (refiners with their own precursor capacity), and non-integrated downstream (precursor plants buying sulphate on the merchant market). The data through April 2026 show a dramatic compression: integrated producers' inventory days have collapsed from 58.9 days in May 2025 to 20.7 days by April 2026 — a 65% reduction reflecting the urgent need to convert in-process cobalt as quickly as possible while prices rise. Non-integrated upstream inventory has stayed stable at ~3.7 days; non-integrated downstream has fallen from 7.1 to 2.6 days as precursor plants run hand-to-mouth.
Chloride inventories show a different pattern: upstream refiners' inventory days have actually risen (from 18.6 to 37.5 days through 2026) as production has stayed ahead of LCO-driven demand, while downstream Co3O4 plants' chloride days have stayed roughly flat at 23 days. Co3O4 itself shows the most uneven inventory pattern of any cobalt product — downstream cathode plants briefly held 78 days of Co3O4 inventory in February 2026 before drawing down to 54 by April; upstream Co3O4 refiners' inventories have built from 28 days to 36 days as LCO demand has been slower than feared.
Refined cobalt's three-bucket inventory structure (upstream smelter / social / downstream materials) reveals a market in transition. Social inventory — the visible warehouse stock that LME and SHFE participants monitor — has fallen from 225 days in May 2025 to 63 days by April 2026, a 72% draw-down that has been the principal source of physical cobalt supply through the DRC shock. Upstream smelter inventory has been more variable but is trending lower; downstream materials plants are running on minimal stock (~10 days). The implication: any further DRC tightening, or any new geopolitical event, will hit a refined cobalt market with limited inventory buffer and high price elasticity.
Aggregating the four product-level inventory series produces an important second-order signal. The integrated layer (vertically integrated Chinese refiners with both smelting and precursor capacity) has destocked aggressively — sulphate integrated days at 20.7 in April 2026 vs 58.9 a year earlier — because these players see the price ramp as an opportunity to monetise stored cobalt at peak. The merchant layer (non-integrated upstream and downstream) is in the opposite regime: inventory is being held thin out of necessity rather than choice, as feedstock procurement has become operationally difficult. This divergence is the strongest argument for the SMM neutral scenario's "marginal balance" call for 2027: if H2 2026 Indonesian MHP arrives on schedule, the integrated layer will be inclined to rebuild some inventory, while the merchant layer will be forced to maintain hand-to-mouth running. The combined effect is a 2027 destocking-then-restocking transition that helps the system cope with the structural deficit without a runaway price spike. The risk: if Indonesian capacity slips by 6+ months (a real possibility given the sulphur and tailings constraints in Chapter 3), neither layer has the inventory cushion to absorb the shock — and 2027 H1 could see a second leg of price strength even sharper than 2026.
Cobalt demand growth is no longer dictated solely by NEV sales. LFP has captured most of the Chinese power battery market, premium long-range EVs preserve a high-Ni cobalt-bearing share, and humanoid robots have entered the demand model.
By April 2026, lithium iron phosphate (LFP) accounted for 81.5% of Chinese power battery installations according to the China Automotive Battery Innovation Alliance (CABIA), with ternary chemistries holding 18.5%. The shift has been driven by three factors widely reported in the industry: vertically integrated LFP supply chains enabling sub-CNY 0.4/Wh cell costs; new LFP-class platforms achieving energy densities approaching the lower end of the NMC range; and the proliferation of mid-range NEV models in the CNY 70,000–150,000 price band that prioritise cost over range. Globally, LFP's share of EV battery installs rose from 51% in 2024 to ~63% in 2025 and is forecast at 63–65% in 2026.
The LFP advance has been more than offset by three factors that have kept cobalt demand growing at ~5–7% per year. First, premium long-range NEVs (the Mercedes / Porsche / Tesla S-tier and equivalent Chinese super-luxury models) almost exclusively use high-nickel NMC811 or NCA cathodes that retain 8–10% cobalt content. As global EV penetration extends to premium segments, this floor is structural. Second, consumer electronics (lithium cobalt oxide / LCO) demand has shown a renaissance in 2025–2026 as AI-PC and smartphone refresh cycles drive ~10% YoY shipment growth; LCO accounts for roughly 16% of global cobalt demand and is essentially insensitive to LFP substitution. Third, the global cobalt market has added approximately 4 kt/yr of "new" demand from energy storage (residential and grid), aerospace alloys (driven by global engine OEM order book recovery) and tool steel — small in percentage terms but stabilising on the down-side.
The most discussed new cobalt demand vector is humanoid robotics. Tesla's Optimus V2 uses a 2.3 kWh high-nickel battery pack per Nickel Institute (April 2026); other US and Chinese humanoid developers have similarly standardised on NCM811 or NCA cathodes for the energy-density and longevity profile required. TrendForce estimates 75 GWh of humanoid robot battery demand by 2035 in its bull case; the bear case (36Kr) is ~5 GWh. Even the bull case translates to only ~3 kt of cobalt demand per year — small in market terms but symbolically important as a demand category that LFP cannot serve.
The fastest-growing global battery segment is stationary energy storage (BESS), which is essentially 100% LFP and therefore cobalt-free. SMM tracks 700+ GWh of forecast 2026 BESS deployment globally — a 65% YoY increase. BESS demand growth is therefore neutral for cobalt; if anything, it absorbs lithium supply that would otherwise weigh on LFP cell prices.
Indonesia's domestic EV and cell-making sector is small but structurally important. HLI Green Power (Hyundai-LG joint venture, Karawang) has 10 GWh of NCMA cell capacity operational since 2024 with expansion to 30 GWh on the drawing board. CATL's Karawang complex — a USD 5.9 billion integrated mining-precursor-cell-recycling joint venture with Antam, IBC, Brunp and Lygend — broke ground in June 2025 with Phase 1 6.9 GWh due Q4 2026 and full operations targeted for 2028. Indonesia's 2030 target is 140 GWh of domestic cell capacity. For Indonesian downstream players, the implication is clear: there is a real domestic offtake channel emerging, but it lags the upstream HPAL ramp by 18–36 months.
SMM organises cobalt end-use into five primary classes, which together cover the global demand pool. The 2025E shares (per the SMM Cobalt Industry Chain Framework, slide 5) are: Power (EV traction batteries) 28%, up from 20% in 2019 — EV penetration is the principal accelerant despite the LFP wave; 3C consumer electronics (largely LCO) 40%, essentially flat versus 2019's 40%, as AI-driven smartphone and PC refresh cycles offset LFP-substitution losses at the low end; Alloy & electroplating 14%, down from 18% as super-alloy growth is outpaced by battery demand; Catalyst 1%, down from 2% — a small specialty stream; and Other (magnetics, aerospace, tool steel, pigments, animal feed) 16%, down from 21%. The 2025E total sums to 99% reflecting SMM rounding (2019 totals 101% for the same reason). In tonnage terms, the SMM bottom-up estimate puts 2026 global cobalt demand at approximately 222 kt against the Cobalt Institute's 219.6 kt; cross-validation against three external models (Fastmarkets 218 kt, Benchmark 224 kt, S&P Global 220 kt) places SMM within the consensus band.
Growth trajectories — the SMM CAGR decomposition (slide 7). SMM's five-class CAGR decomposition over the 2018–2025E period shows EV cobalt demand +12% CAGR (the dominant accelerant), 3C +12% CAGR (AI-driven), alloy +5%, magnetics +3%, other +3% — for an overall +8% global cobalt demand CAGR. The 2025–2030E projection moderates sharply: EV slows to −1% CAGR (LFP substitution dominates incremental units), 3C to +1% (LCO growth decelerates), alloy to +4% (aerospace/defence resilience), magnetics to +4% (renewable infrastructure), other to +3%. Overall global cobalt demand CAGR +2% 2025–2030E — meaningfully slower than 2018–2025E but unmistakably positive, and the slowdown is concentrated in the EV segment that LFP can substitute, not in the segments that LFP cannot serve.
The 2026 Indonesian policy package is the most significant change in nickel/cobalt fiscal architecture since the 2020 ore export ban. For the first time, cobalt is explicitly priced inside the HPM benchmark and royalty structure.
Indonesia's nickel downstreaming doctrine — anchored by the January 2020 ban on raw nickel ore exports — has been the single most consequential mining policy of the 2020s, attracting more than USD 25 billion of HPAL and RKEF investment and transforming Indonesia from a 5% global cobalt supplier into a 15% supplier within five years. The doctrine's logical extension to cobalt was always implicit but operationally absent until the 2026 ESDM 144/2026 package. The new HPM (Harga Patokan Mineral) reference price now treats cobalt as an independent commodity inside the ore reference price — HPAL ore cost +USD 2.3/wmt and NPI ore cost +USD 3.2/wmt — formally pricing the cobalt component of every ton of nickel-cobalt ore for the first time.
Minister Bahlil Lahadalia confirmed in December 2025 that the 2026 RKAB (national mining quota) for nickel ore will be cut to ~250–270 Mwmt versus 379 Mwmt in 2025 — a 30–34% reduction designed to address persistent oversupply and the resulting low realised nickel price. The RKAB cut has a direct cobalt implication: with HPAL feedstock supply constrained, the cobalt byproduct produced by each HPAL line per unit of ore becomes proportionally more valuable.
The RKAB framework was originally designed for nickel supply discipline; cobalt was an unpriced rounding error. The 2026 cycle is the first in which the cobalt byproduct credit is large enough to meaningfully shift the calculus. Each tonne of HPAL ore contains ~0.10–0.15% cobalt; at 2026 MHP FOB of USD 51,000/mt Co, every 1 Mwmt (1 million wet metric tonnes) of RKAB allocation embeds roughly USD 50–75 million of cobalt value. The 110 Mwmt RKAB cut from 2025 to 2026 therefore locks in an approximately 110–170 kt potential cobalt-production ceiling — a quota effect that runs in parallel with the DRC quota and reinforces global tightness. SMM modelling suggests the RKAB cut alone adds ~USD 1,000/t to realised cobalt prices through 2026 H2. For Indonesian policymakers the actionable insight is that the RKAB volume lever is now a dual-commodity lever: every adjustment moves both nickel realised price and global cobalt balance, and the fiscal implications should be evaluated against both revenue streams jointly, not in nickel terms alone. Specific proposal: ESDM should publish a quarterly "RKAB cobalt shadow-price" disclosing the USD 50–75 million of cobalt value embedded per Mwmt of RKAB allocation, so the volume-tightening lever is priced in two currencies simultaneously.
Government Regulation PP No.18 and 19/2025 (effective April 2025) restructured Indonesian nickel and cobalt royalty into a tiered ad-valorem framework with rates rising to 19% at high price levels. Combined with the additional USD 1.0–1.5/wmt royalty per nickel-percent on ore introduced in 2025, the effective state-take per tonne of nickel-equivalent HPAL output has risen by ~30% versus the pre-2024 regime. For cobalt specifically, the royalty applies to the contained cobalt portion of MHP at the new HPM reference — meaning the state captures more of the cobalt windfall than under prior regimes.
Indonesian government officials have begun framing the country's nickel/cobalt market position as analogous to OPEC's role in oil ("ONEC" — Organisation of Nickel-Exporting Countries). The framing has practical content: the RKAB cut + HPM revision + royalty restructure create a coordinated supply-and-pricing toolkit that Jakarta can use to support nickel and cobalt prices when oversupply threatens — and that, equally, gives downstream consumers (Korean and Japanese cell-makers, European OEMs) a clearer set of policy variables to model. Whether the framing extends to a formal producers' alliance is unclear, but the toolkit itself is real and consequential.
The February 2026 US–Indonesia Agreement on Reciprocal Trade, signed by President Prabowo and President Trump, includes a USD 33 billion package with explicit critical-mineral language: Indonesia confirms its commitment to refuse raw mineral exports and the US confirms market access for processed Indonesian critical minerals. The agreement also includes implicit Chinese-ownership ceilings that have not been formally specified but are widely understood in Jakarta to apply to projects seeking IRA-style critical-mineral tax credits. The agreement is consistent with the downstreaming doctrine but creates real tension with the 80%-Chinese-led nature of current HPAL capacity. Resolving this tension — likely through additional Western-led FIDs and a measured de-concentration of Chinese ownership — will be a defining policy challenge through 2027.
Cobalt's ESG narrative has historically focused on artisanal mining in the DRC. In the Indonesian era, the ESG frontier has shifted: filtered tailings, coal-powered processing, indigenous land rights, and the EU Battery Passport.
The DRC cobalt ESG narrative is well-known to international buyers: artisanal mining of approximately 15–25% of supply, child labour reports, and high-conflict cobalt routes. Indonesian cobalt presents a structurally different ESG profile. There is no artisanal cobalt mining in Indonesia — all cobalt is produced as a byproduct of industrial HPAL nickel processing. There is no conflict-mineral concern — Indonesia is a stable democracy with strong central-government control over mining concessions. However, three new ESG concerns dominate the Indonesian cobalt narrative: tailings management, coal-powered processing emissions, and indigenous land rights at Halmahera and Sulawesi sites.
HPAL produces approximately 1.4–1.6 tonnes of sulphuric-acid-bearing tailings per tonne of contained nickel. Indonesia bans deep-sea tailings disposal (DSTP), confirmed by Coordinating Minister Pandjaitan in 2023 and reinforced by Ministerial Regulation in 2024. The only permitted alternative is filtered tailings storage in engineered dams — a technology that has now failed twice at the IMIP complex (March 2025 and February 2026 dam failures). The Earthworks-WALHI report "Filtered Tailings in Indonesia: The Catastrophic Failure of a Disruptive Technology" (2025) documents these incidents and argues that the seismic, high-rainfall, dense-vegetation environment of Sulawesi and Halmahera makes filtered-tailings dams structurally inadequate. The implication for buyers is that any cobalt offtake contract from Indonesian HPAL after 2026 must include tailings-management audit terms aligned with IRMA / Initiative for Responsible Mining Assurance standards.
Indonesian HPAL is currently power-supplied largely by coal-fired captive power plants (CFPP) at IMIP and IWIP. According to IEEFA's 2024 analysis, this gives Indonesian HPAL-derived nickel a carbon intensity of approximately 13–15 tCO2 per tonne of nickel-in-MHP — lower than RKEF/NPI (28–68 tCO2/t Ni) but materially higher than Western HPAL operations powered by gas or renewables. For cobalt, the implied co-product allocation at the 13:1 Ni:Co ratio gives approximately 170–195 tCO2 per tonne of cobalt — a number that will be visible in the EU Battery Passport from February 2027. Harita's 300 MW solar project at Obi Island and similar planned installations at IMIP and IWIP are the principal mitigation; full decarbonisation is realistically a 2030+ project.
The Hongana Manyawa indigenous group on Halmahera island has been the focus of international attention following the Survival International / AEER advocacy campaigns of 2024–2025. The Sonic Bay cancellation in June 2024 was partly attributed to these concerns. Subsequent Indonesian government engagement has produced a more structured framework for indigenous consultation on new HPAL FIDs, but the issue remains a binding constraint on the pace of permit issuance for Halmahera-based projects.
Two ESG-linked regulatory frameworks will shape the Indonesian cobalt market from 2026 onward. The EU Battery Passport (Regulation 2023/1542) enters mandatory phase from February 2027, requiring cell-level traceability including cobalt provenance, GHG intensity, and recycled content. IRMA (Initiative for Responsible Mining Assurance) certification has been achieved by Vale and Harita and is in the pipeline for HNC, QMB and Lygend; CATL Indonesia is not yet IRMA-aligned. By 2027, IRMA-equivalent reporting will be table stakes for any cobalt offtake contract destined for the EU or for Korean OEMs subject to Korea's K-Battery / US IRA disclosure rules.
SMM's monthly forecast model now extends through April 2027 with low / mid / high price bands by product. Three scenarios — neutral, pessimistic and bullish — bracket the plausible trajectory.
The neutral scenario assumes DRC quotas remain at the 96.6 kt cap with full 80% redirection to China, Indonesia's 2026 capacity ramp delivers 59.8 kt of cobalt-in-MHP, and global cobalt demand grows 7% per the Cobalt Institute Q4 2025 update. Under this scenario, SMM's monthly model shows refined cobalt averaging CNY 459,743/t in 2026 (low 452,035 / high 467,450), with H2 2026 monthly peaks near CNY 491,783 in October before easing to CNY 481,905 by April 2027 as Indonesian supply normalises. Cobalt sulphate averages CNY 97,219/mt metal content (low 95,483 / high 98,955) with peaks in October 2026 and gentle Q1 2027 retracement. The 2026 product-level balances under this scenario: sulphate −7.7 kt, chloride −1.4 kt, refined −12.7 kt; 2027 recovers to sulphate +1.5 kt, chloride +6.5 kt, refined +3.5 kt — a meaningful structural shift driven by the 2026 H2 Indonesian capacity wave.
The pessimistic scenario assumes the DRC adjusts the 2026 producer basic quota down to 87 kt (eliminating the strategic reserve from the export pool) and reduces redirection to China to 70%. This more aggressive variant of the pessimistic scenario (deeper than the structured-model −8.3 kt China-balance pessimistic case cited in the Executive Summary) is the upper bound of plausible 2026 deficit. The SMM scenario model shows the 2026 China cobalt balance widening to a 13–15 kt deficit, refined cobalt averaging CNY 500,000+/t through Q2 2026, and hydroxide payables sustained at 90%+ for 12+ months. Under this scenario, Indonesian MHP FOB prices could test USD 60,000/mt Co and remain there into 2027.
The bullish scenario is an SMM analyst overlay rather than a separate quantified output in the structured forecast model (which contains only neutral and pessimistic cases). It assumes Kinshasa reverses the quota framework in 2027 or 2028 — driven by fiscal pressure, IMF program requirements, or a change of mining minister. Under this analyst-derived scenario, the CMOC stockpile (estimated at 50–80 kt, derived from the gap between CMOC's standalone mining guidance of up to 120 kt and its 2026 quota of ~31.2 kt over 2.5 years of restricted exports) returns to market over 18 months, pushing the 2027 cobalt balance into 25–45 kt surplus and crashing refined cobalt below CNY 250,000/t by Q4 2027. For Indonesia, the bullish scenario is the principal downside risk: a price reversion to USD 25–30,000/mt Co MHP FOB would compress byproduct economics back toward 2023–2024 marginal-cash-cost levels and slow the post-2027 capacity expansion. The probability weight of 15% reflects the political and fiscal incentives in Kinshasa to maintain the price-supporting quota framework, rather than a quantified model output.
For offtake buyers, the prudent strategy in 2026 is to lock multi-year MHP supply at the current SMM Indonesia FOB benchmark with annual price-formula adjustments tied to a 50/50 weighting of Fastmarkets hydroxide and SMM refined cobalt — capturing some of the neutral-scenario margin while protecting against the pessimistic-scenario upside and not over-paying in the bullish-scenario downside. For sellers, the prudent strategy is to maintain optionality on physical delivery, with a higher proportion of spot sales than has been typical to participate in the H1 2026 price strength.
The SMM monthly forecast for the four cobalt products provides the most granular view available of the 2026–2027 price path. Cobalt sulphate: monthly average price climbs from CNY 95,364/mt metal content in April 2026 to a peak of CNY 98,674 in October 2026 before easing to CNY 96,594 by April 2027 — a moderate 3.5% rise then 2% retracement that reflects the expected H2 2026 Indonesian MHP supply ramp. Cobalt chloride: monthly price climbs from CNY 115,531 in April 2026 to a peak of CNY 117,200 in October 2026 before stabilising at CNY 114,000–115,000 through Q1–Q2 2027. Co3O4: monthly price moves from CNY 364,857 in April 2026 to CNY 368,569 in October 2026 before easing to CNY 358,942 in February 2027 — reflecting a smaller demand-supply imbalance in the consumer-electronics value chain. Refined cobalt: the most volatile of the four, monthly price climbs from CNY 415,881 in April 2026 to a peak of CNY 491,783 in October 2026 (driven by inventory exhaustion and import constraint) before retracing to CNY 481,905 in April 2027. Each of these forecasts assumes the neutral scenario; the pessimistic scenario adds approximately 12–15% upside to each price band, and the bullish (analyst-derived) scenario subtracts approximately 30–40% from each.
Beyond the 12-month forecast horizon, the critical question is whether 2028 marks a structural inflection. Three forces converge in 2028: (1) the full ramp of the 2026–2027 Indonesian commissioning cohort adds approximately 25 kt of cobalt-in-MHP; (2) recyclate flows from EU and Korean black-mass facilities exceed 30 kt for the first time; (3) the DRC quota framework is up for renewal — Kinshasa may extend it, tighten it, or relax it depending on the fiscal and political context. If the quota framework persists, 2028 is broadly balanced; if it relaxes and the CMOC stockpile flows back, 2028 is a substantial surplus year and cobalt prices return to the 2023–2024 marginal-cash-cost equilibrium. This binary outcome is the principal long-term planning challenge for both Indonesian producers and downstream offtake counterparties.
This chapter translates the data and forecasts into specific action items for four stakeholder groups: Indonesian miners, downstream battery players, Indonesian policymakers, and ESG / civil society participants.
Lock in cobalt byproduct value. At today's MHP FOB of USD 51,332/mt Co, a 120 kt/yr Ni HPAL with 13 kt/yr contained cobalt produces ~USD 670 million of annual cobalt sales — more than the typical EBITDA of a comparable RKEF NPI line. Lock 3–5 year offtake contracts before Chinese inventories normalise and before the bullish scenario crystallises.
Prepare for the 2026 HPM cobalt royalty. Build the additional USD 0.22+/wmt royalty per cobalt component into 2026–2027 mine plans, with sensitivity around the ad-valorem rate at high price levels.
Accelerate downstream investment. Pomalaa, SLNC and ENC commissioning in 2026–2027 will add ~25 kt/yr of cobalt capacity. Use the proceeds to finance domestic refining (precursor, sulphate, cathode) rather than only exporting MHP — Indonesia's downstream value chain captures 2–3× the upstream margin.
IRMA / EU Battery Passport readiness. Begin third-party audit cycles in H2 2026 to be ready for February 2027 EU compliance and 2028 Korean K-Battery requirements.
Indonesia is the only at-scale alternative to DRC. The 96.6 kt DRC quota framework is a multi-year structural fact. Long-term offtake from Indonesian HPAL must be re-priced upward — the implicit USD 25–30,000/mt Co assumption in pre-2025 LCFA-based contracts is no longer commercially feasible.
Cobalt-bearing chemistries still matter. Despite LFP's 81.5% share of Chinese power battery installs (April 2026), high-nickel NMC and NCA retain 25%+ of global EV battery installs by capacity and ~37% by value. Premium long-range EVs, consumer electronics, and humanoid robotics preserve the cobalt demand floor.
Indonesian battery cell capacity reaches ~20 GWh by end-2026. For OEMs assembling in Indonesia (Wuling, Vinfast, Toyota PHEV), this matters for IRA-style local-content qualification and ASEAN trade preferences.
Diversify the procurement risk profile. Consider a 60% MHP / 30% sulphate / 10% recycled mix by 2028 to mitigate any single-source disruption.
Establish a national cobalt strategic reserve. Analogous to the DRC's 10% ARECOMS reserve, an Indonesian strategic stockpile of 5–10 kt of contained cobalt would give Jakarta a counter-cyclical lever and a fiscal hedge for the next price downturn — at current prices, a 5 kt reserve costs USD 250 million but provides ~USD 1 billion of optionality value.
Continue the diversification of HPAL ownership. The Sonic Bay cancellation in 2024 underscored the cost of single-source dependence. Active support for Western and Korean FIDs — through investment incentives, regulatory predictability, and a workable indigenous-rights framework — would meaningfully de-risk the long-term cobalt supply story.
Lead on tailings safety. The IMIP filtered-tailings dam failures (March 2025, February 2026) are a binding constraint on permit approvals. A national tailings safety standard — co-developed with the international mining-standards community (e.g. ICMM and IRMA) — would unlock the next wave of EU- and IRA-compliant investment.
Position for the EU Battery Passport. The 2027 EU compliance deadline is an opportunity for Indonesia to lead. A national cobalt traceability platform — co-developed with SMM and global ESG verifiers — could become a Southeast Asian standard.
Filtered tailings management is the central technical challenge. The Sulawesi / Halmahera geophysical environment requires either a step-change in tailings dam engineering, a hybrid approach combining filtered tailings with engineered backfill, or relocation of new HPAL FIDs to more geotechnically suitable sites.
Indigenous land rights at Halmahera require a permanent framework. The Hongana Manyawa case has set a precedent that must be operationalised as a standard FPIC (Free, Prior and Informed Consent) protocol for all new HPAL FIDs.
Coal-powered processing is the largest single decarbonisation lever. Indonesia's HPAL fleet is currently approximately 90% coal-powered. A coordinated transition to gas, solar, and hydropower could halve the carbon intensity of Indonesian cobalt within 5–7 years.
Recycling capacity is the long-term ESG win. Building 30+ kt/yr of black-mass processing capacity in Indonesia by 2030 would close the cobalt loop, reduce dependence on virgin extraction, and align with EU Battery Regulation recycled-content thresholds.
Twelve operating, under-construction, or cancelled Indonesian HPAL units as tracked by SMM. Nickel and cobalt capacities are nameplate at full ramp. Status as of May 2026.
| Project | Owner | Location | Status | Ni cap (kt/yr) | Co cap (kt/yr) | Online |
|---|---|---|---|---|---|---|
| Halmahera Persada Lygend (HPL/ONC) | Lygend (Ningbo Lygend) / Harita | Obi Island, North Maluku | Operating Operating + expansion | 120 | 13 | 2021 |
| Huayue Nickel & Cobalt (HNC) | Huayou 57% / Tsingshan 21% / Nickel Industries 10% | IMIP, Morowali, C. Sulawesi | Operating Operating >135% nameplate | 60 | 7 | 2021 |
| Huafei Nickel & Cobalt | Huayou 51% / Eve Battery 17% / Glaucous 30% | IMIP, Morowali | Operating Operating (Q1 2024) | 120 | 14 | 2024 |
| QMB New Energy Materials | GEM 36% / Tsingshan / CATL / IMIP | IMIP, Morowali | Operating Operating | 50 | 5 | 2022 |
| ESG New Energy Material | MBMA (Merdeka) / GEM JV | SCM / IMIP | Operating Operating (2024) | 30 | 3.3 | 2024 |
| Meiming New Energy Material | MBMA / GEM | IMIP / SCM | Operating Operating (2025) | 30 | 3 | 2025 |
| Sulawesi Nickel Cobalt (SLNC) | MBMA + partners | IMIP | Under construction Under construction | 90 | 10 | 2026 |
| Excelsior Nickel Cobalt (ENC) | Nickel Industries 55% / Tsingshan | IMIP, Morowali | Under construction Under construction (FID 2023-10) | 72 | 8 | 2026-2027 |
| Kolaka Nickel Indonesia (Pomalaa) | Huayou 73.2% / Vale Indonesia 18.3% / Ford 8.5% | Pomalaa, SE Sulawesi | Mech. completion Aug 2026 Mech. completion Aug 2026 | 120 | 15 | 2026 H2 |
| Vale-Huayou Sorowako/Bahodopi HPAL | Vale Indonesia + Huayou | Sorowako / Bahodopi, S. Sulawesi | Under construction Under construction | 66 | 8 | 2026-2027 |
| Huashan Nickel Cobalt | Huayou 68% / Glaucous 32% | IWIP, Halmahera | Under construction Under construction | 120 | 13 | 2026-2027 |
| Sonic Bay (CANCELLED 2024-06-24) | Eramet 51% / BASF 49% | IWIP, Halmahera | Cancelled CANCELLED | 67 | 7.5 | n/a |
Monthly average prices for the five SMM-tracked cobalt products and selected lithium/nickel benchmarks. All prices CNY/mt unless noted.
| Product | Unit | Dec 2025 | Jan 2026 | Feb 2026 | Mar 2026 | Apr 2026 | MoM % |
|---|---|---|---|---|---|---|---|
| Cobalt Intermediate (CIF China) | USD/lb | 20 | 23 | 23 | 27 | 28 | +4.7% |
| Cobalt Sulphate (≥20.5%) | CNY/mt physical | 92,800 | 92,750 | 93,250 | 93,350 | 93,500 | +0.2% |
| Cobalt Sulphate (metal content) | CNY/mt Co | 452,000 | 451,000 | 453,500 | 458,000 | 460,976 | +0.6% |
| Cobalt Chloride (≥24.2%) | CNY/mt physical | 110,900 | 112,050 | 112,750 | 113,500 | 113,750 | +0.2% |
| Refined Cobalt (≥99.8%) | CNY/t | 455,000 | 451,000 | 437,500 | 428,000 | 416,000 | −2.8% |
| Co3O4 (≥72.8%) | CNY/mt physical | 500,000 | 500,000 | 485,000 | 372,000 | 365,000 | −1.9% |
| Indonesia MHP FOB Co | USD/mt Co | 47,058 | 49,116 | 49,601 | 51,232 | 50,715 | −1.0% |
| Nickel sulphate (battery grade) | CNY/mt | 27,500 | 27,800 | 27,900 | 28,200 | 28,350 | +0.5% |
| Lithium carbonate (battery grade) | CNY/mt | 82,500 | 81,200 | 80,100 | 78,500 | 76,800 | −2.2% |
SMM operates three parallel cobalt pricing systems and two flagship indices. This appendix consolidates the formulas, weighting rules, and constants that recur across the report.
MB benchmark route: MB cobalt-intermediate price = MB standard-grade electrolytic cobalt price × payable coefficient.
SMM reverse-implied route: SMM cobalt-intermediate = (SMM cobalt-sulphate price / 0.205 − tolling fee) / VAT / 2204.6 / FX × yield.
Constants: 2204.6 = lb/tonne; 0.205 = cobalt content of cobalt sulphate. The MB and SMM routes are jointly used (dual benchmarking).
LME route: MHP = LME nickel × Ni payable + MB cobalt × Co payable.
SMM reverse-implied route (dual-factor):
MHPCo = (SMM cobalt sulphate / 0.205 − tolling) / FX / VAT × yield × Co content
MHPNi = (SMM nickel sulphate / 0.222 − tolling) / FX / VAT × yield × Ni content
MHP = MHPCo + MHPNi
Constants: 0.205 = Co content of cobalt sulphate; 0.222 = Ni content of nickel sulphate.
Ternary scrap (NCM):
Price = SMM cobalt sulphate / 0.205 × Co coefficient × Co content
+ SMM nickel sulphate / 0.222 × Ni coefficient × Ni content
+ SMM industrial lithium carbonate / 0.187 × Li coefficient × Li content
LCO scrap: identical formula, Li term excluded.
Constants: 0.205 / 0.222 / 0.187 (Co / Ni / Li content factors).
Enterprise-weight architecture (SMM cobalt sulphate index, slide 29): Sellers (upstream — cobalt feedstock + cobalt sulphate suppliers) carry 50% of total weight; Buyers (downstream — precursor + Co3O4 + cathode plants) carry 50% of total weight. For any participating enterprise A, its weight W1% = Y1% (sell-side share) + Q1% (buy-side share). The index is then I = Σ Pn × Wn. A penalty-down-weighting and periodic re-evaluation mechanism applies if any enterprise systematically deviates.
The cobalt-intermediate tolling-fee index is a composite index incorporating both the participating-enterprise price quotes and three fundamental factors:
SMM cobalt-intermediate tolling index = (P1×W1 + P2×W2 + … + Pn×Wn) + Wa×Pa + Wb×Pb + Wc×Pc
where:
Price factor (X%) — buyer X% (cobalt salts + metallic cobalt), seller X% (cobalt ore)
Fundamental factors (1−X%) —
(Wa, Pa) inventory factor
(Wb, Pb) supply-demand factor
(Wc, Pc) profit factor
Source: SMM Cobalt Industry Chain Framework, 2025-09 (slides 26–32). Constants and weighting rules are SMM proprietary; reproduced here under the data-pro permitted-disclosure provisions.
Indonesia (ESDM × Antam × KEMENPERIN) could co-develop an SMM-methodology-aligned Indonesia MHP Cobalt FOB Index that mirrors the four-step process documented above: (i) daily quotation collection from operating Indonesian HPAL producers and their Chinese / Korean / European MHP buyers; (ii) weight rebalancing with the same 25% maximum single-enterprise cap; (iii) seller (Indonesian HPAL) 50% / buyer (precursor & cell-maker) 50% architecture; (iv) AB-post cross-checked daily publication. An Indonesia-anchored cobalt index would become the local HPM cobalt component and royalty reference — analogous to LME nickel's role in Indonesian nickel — and would migrate price discovery from the SMM-implied / MB-derived hybrid that currently dominates to a sovereign-aligned benchmark. The methodology infrastructure already exists at SMM; the policy step is publication branding and regulatory anchoring.
| MHP | Mixed Hydroxide Precipitate — the cobalt and nickel intermediate produced by HPAL processing of nickel laterite ore. Typically 30–40% Ni and 3–5% Co by weight. |
| HPAL | High Pressure Acid Leaching — the dominant hydrometallurgical route for processing low-grade nickel laterite ore into battery-grade nickel and cobalt intermediates. |
| NPI / RKEF | Nickel Pig Iron, produced via Rotary Kiln-Electric Furnace pyrometallurgical processing. NPI is the dominant feed for Indonesian stainless steel and a competitor to HPAL for ore. |
| RKAB | Indonesia's annual national mining quota / work plan and budget. The 2026 RKAB cut to ~250–270 Mwmt represents a 30%+ reduction vs 2025. |
| HPM | Harga Patokan Mineral — Indonesian government reference price for ore-based royalty calculations. The 2026 HPM revision explicitly prices cobalt for the first time. |
| ARECOMS | Autorité de Régulation et de Contrôle des Marchés des Substances Minérales Stratégiques — DRC's mineral marketing authority that administers the cobalt export quota. |
| EGC | Entreprise Générale du Cobalt — DRC's state-affiliated franchise that has exclusive rights to buy and export artisanal cobalt. |
| LCO | Lithium Cobalt Oxide — the high-energy-density cathode chemistry used in consumer electronics (smartphones, laptops, tablets); ~16% of global cobalt demand. |
| NMC / NCM | Nickel Cobalt Manganese — the dominant family of EV cathode chemistries. NMC811 = 80% Ni, 10% Co, 10% Mn — the high-nickel premium-EV standard. |
| NCA | Nickel Cobalt Aluminum — the cathode chemistry used by Tesla / Panasonic in 21700 cells; ~5% cobalt. |
| LFP | Lithium Iron Phosphate — cobalt-free cathode chemistry that dominates Chinese EV power battery installs (81.5% in April 2026) and global energy storage. |
| IRMA | Initiative for Responsible Mining Assurance — the leading independent ESG certification standard for mining operations; widely adopted by major Western OEMs for cobalt and nickel supply chain compliance. |
| FOB | Free On Board — the Incoterm under which the seller delivers goods at the port of shipment with risk transferring at that point. SMM's Indonesia MHP cobalt index is reported FOB Indonesia. |
| IRA | US Inflation Reduction Act (2022) — provides EV tax credits contingent on critical-mineral and battery-component sourcing from US-aligned countries. |
| EU CRMA | European Union Critical Raw Materials Act (2024) — establishes the EU's strategic project framework for critical minerals including cobalt. |
| FPIC | Free, Prior and Informed Consent — the international standard for indigenous community consultation on extractive projects. |
| DSTP | Deep Sea Tailings Placement — disposal of mining tailings into deep ocean, banned in Indonesia since 2023. |
| BESS | Battery Energy Storage System — stationary grid-scale or residential battery installations, predominantly LFP and therefore cobalt-free. |