INCO: The Quality Trap That Might Not Be a Trap

Action Plan: Entry, Exit, & Broken-Thesis Signals

Before the long argument, here's the whole plan in one screen: where I enter, where I exit, and, most important, what tells me the thesis is wrong. Discipline over hype. Updated 10 Aug 2026, the price has already moved from Rp 4,560 to Rp 5,500 since this thesis was written; Entry 1 has effectively passed.

Investment Summary: Three Legs & How INCO Makes Money

INCO is, at its core, the only nickel producer on the IDX that the Western battery supply chain can actually touch. Three business legs feed one balance sheet.

The base business is simple: a tier-1 laterite mine + smelter complex in Sorowako, South Sulawesi, running on its own 365MW hydro dams, producing ~70kt/year of nickel matte. Every ton is sold in USD to Vale Canada and Sumitomo Metal Mining under long-term take-or-pay offtakes, priced off LME. Since July 2025 a second stream switched on, direct ore sales from the new Bahodopi and Pomalaa mines, scaling fast: 1.2% of revenue (Q1'25) → 10.3% (FY2025) → 22.8% (Q1'26) → 26.2% (H1 2026, actual, Catatan 24), and 29.2% in Q2 2026 alone, so the diversification trend keeps accelerating.

Here's the distillation: volume is the newer swing factor too, alongside price and quota. Matte output was guided at 67,645t for 2026 (official), but H1 2026 actual sales (27,659t) track ~18% below that run-rate. What moves the P&L is (a) realized ASP, LME × 82% payability since July 2025, confirmed again at US$14,764/t in Q2 2026, and (b) how much ore the government lets the new mines ship, still only 30% of what INCO asked for as of the 29 Jul interim filing (the revision that was supposed to land by 31 Jul did not).

And the market's framing has shifted. Back in July, on trailing earnings INCO looked absurd, P/E ~40x against ANTM at ~8x and NCKL at ~9x. Now, with H1 2026 fully booked (net profit US$104.4M, +313% YoY) and the price up to Rp 5,500, trailing P/E has come down to a more defensible ~24x, while on Macquarie's 2026 estimate (US$337M net profit) the stock is roughly 10x. The debate is no longer "is the multiple too high", H1 already answered that the earnings trajectory is real (net profit Q1 US$43.6M, Q2 US$60.8M, +39% QoQ). The live question now is whether volume (matte tracking below guidance, ore still quota-capped) catches up to the margin recovery already banked.

Investment Thesis: Three Legs, Why It's Cheap

Three legs, one conclusion: at Rp 4,560, on the earnings that matter, this may quietly be the cheapest "expensive" stock in Indonesian nickel. Let's dissect each leg.

Q1 2026 looked like normalization (net margin 17.3%, back to 2021-22 levels). H1 2026, now fully reported, proved it was understating the story: net margin 19.2%, and on gross margin (28.0%) and operating margin (24.7%) INCO is already above the FY2023 cycle peak (28.2%/24.5%), not just "back to normal." Q2 2026 alone: revenue US$290.4M (+15% QoQ), net profit US$60.8M (+39% QoQ), EBITDA margin 44.7%.

The ~3-point gap between H1 2026 net margin (19.2%) and the FY2023 peak (22.3%) is fully explained by two charges that didn't exist in 2023: the 10% IUPK Net Profit Sharing (US$11.6M in H1) and finance costs that only started appearing as the SLL facility activated (US$6.6M). Strip those out and operationally INCO is running ahead of its best-ever cycle year, a materially stronger read than the original "back to normal" framing.

Three mechanical drivers, all structural: payability step-up (82% of LME since Jul-2025, confirmed again at Q2's US$14,764/t ASP), the ore business materializing (0 → 26.2% of revenue in five quarters, at confirmed actual cash costs), and cost position held (US$9,339/t FY25 four-year low, still bottom-quartile globally).

The same country that created the 2023-2025 glut is now engineering the floor beneath it. The 2026 RKAB quota came in ~250-270Mt against smelter demand of ~340-360Mt, a 90-100Mt paper deficit, and ESDM has publicly refused a blanket increase. Weda Bay alone was cut 71% (42Mt → 12Mt, per Meridian Research); INCO's own aggregate RKAB read from the same source shows a comparably deep cut.

LME rallied from ~US$15,000 (2025 avg) further than this thesis originally captured: a spike to ~US$19,350-19,600/t around May-June 2026 (Goldman/Meridian both flag this peak), before correcting back to the US$16,700-17,150 range it's consolidated in through early August (US$16,970/t, 7 Aug, TradingEconomics). Goldman's 2026 forecast progression, US$14,800 → US$17,200 (Feb) → US$18,500 (May), has proven directionally right on the upside spike, though the market gave most of that back.

For INCO specifically, the regulation cuts both ways: the same quota regime supporting LME is still choking INCO's own ore ramp at 30% of requested volume (Bahodopi 2.31Mt vs 8.8Mt asked, −74%; Pomalaa 5.8Mt vs 18.06Mt, −68%). The 31 July revision window has now closed, and the LK Interim 30 Jun 2026 (Catatan 39h) confirms the quota is still 30%, with no approved revision and no new deadline. Management says it plans to file a revision request, but as of the interim filing (29 Jul), there's no certainty of approval.

INCO's structural edge isn't the US$7.2B capex number, it's who owns it. Every major Indonesian HPAL project is majority-Chinese, locking output out of US IRA (FEOC rules) and increasingly EU CRMA premium chains. INCO, Vale Canada 33.88% + MIND ID 34% + Sumitomo 11.48%, hydro-powered, TRIFR down to 0.12, is effectively the only IDX-listed nickel name with a clean FEOC/CRMA story. The Pomalaa JV even has Ford in it.

Now the cost of the story, stated plainly: the J-curve is real and now three years deep. FCF was +US$135M in 2023 → −US$125M (2024) → −US$251M (2025) → −US$216.6M in H1 2026 alone (already worse than all of FY2024). Cash fell from US$376M (end-2025) to US$110.7M (30 Jun 2026). Capex H1 2026 actual: US$255.0M (annualized ~US$510M, still under the US$568M FY guidance, consistent with a H2-heavy pattern).

Company Profile: Who INCO Really Is

Not just numbers, who INCO is and why its ownership structure is the moat.

Founded 1968 as PT International Nickel Indonesia, IPO 1990, renamed Vale Indonesia in 2012. The 2024 divestment reshaped control: MIND ID went 0% (2021) → 20% (2022) → 34% (Jul 2024, via rights issue at Rp3,050), making it controlling shareholder inside the Danantara/BUMN ecosystem, with Vale Canada down to 33.88% and Sumitomo at 11.48%. Share count rose 6.1% to 10.54B, mind the EPS base across 2024.

From Ore to Offtake (and to the HPAL Plants)

How exactly does INCO turn rock into cash? Two chains, two swing factors, one balance sheet.

Macro, Industry & Competitive Landscape

Top-down, nickel in mid-2026 is a policy-managed market. Demand is soft, supply is abundant, yet price holds ~US$16,000-17,000 because one government controls two-thirds of mine supply and has decided to defend the floor.

5.1 PESTLE, Indonesian Nickel, July 2026

5.2 Porter's Five Forces

5.3 Competitive Landscape, The Big 4 + The Shadow Giant

The relative-value truth, stated honestly: for pure risk-adjusted nickel exposure today, NCKL (9x, proven HPAL) and ANTM (7.8x, gold buffer) are better value. The case for paying INCO's premium rests entirely on three things the cheap names can't offer: FEOC/CRMA provenance, the tier-1 cost + hydro asset, and the 2026-27 earnings inflection.

Fundamental Analysis: Sensitivity, Value, Decomposition

Now let's prove the numbers work, and, more important, cut through the headline that misleads.

6.1 Macro→Micro Sensitivity, What Actually Moves the P&L

6.2 Value Creation, ROIC vs WACC, the Honest Ledger

The uncomfortable core: INCO currently destroys value on paper. ROE 2.74% (FY25) vs WACC ~10.4%. DuPont shows a triple hit in 2024-25, margin collapsed, asset turnover fell, leverage rose, all at once. The rehabilitation case: margin is already recovering (proven), asset turnover is the 2027 story (HPAL revenue meeting a bloated asset base), leverage now works for equity (SLL at ~6% vs ~11.6% cost of equity).

6.3 Revenue Decomposition (Price × Volume × Mix)

Matte = the locked block. Price = LME × payability; the July-2025 step-up to 82% is worth ~7-9% of ASP permanently. Ore = the swing block, quota-capped, HPM-priced. The market still models INCO as "matte × LME", the ore block, a quota decision away from doubling, is where estimate revisions come from.

6.4 COGS Bottom-Up

Two lines are inflecting deliberately: contractors (scaling with ore wmt) and royalty (progressive regime + ore ramp), the cost of the growth thesis, not slippage.

6.5 Ratio Outlook, Where Each Ratio Goes, and Why

The theme: 2026 fixes the income statement, 2027 fixes balance-sheet productivity, 2028 fixes cash flow.

Valuation: Scenarios, Methods, Sensitivity

Updated 10 Aug 2026 with H1 2026 actuals and re-anchored to the current price (Rp 5,500) and fresh market data (LME ~US$17,000, USD/IDR ~18,000). The read has changed: DCF-only methods still say roughly fair-to-slightly-rich at the current price, but SOTP (crediting HPAL) still points meaningfully higher. The valuation debate is now less about "will earnings recover" (settled) and more about volume/RKAB delivery and how much HPAL value to credit.

7.1 Three Scenarios, Basis & Sensitivity (Updated)

Cross-check: H1 2026 actual net income (US$104.4M, annualized run-rate ≈US$209M) already sits above the updated Bear case and close to Base, achieved with Furnace 3 back online, but without any RKAB revision. The Base case here already prices in continued RKAB constraint through 2026, it is not dependent on the quota being resolved.

7.2 Multi-Method Valuation & Football Field (Updated, Base Case)

The read is different from July: back then every intrinsic method sat below the price and only the growth pipeline (SOTP) justified paying up. Now, after the H1 beat re-rated the price but also re-anchored the core-business cash flows to a materially net-cash balance sheet, DCF and Comps read roughly fair-to-slightly-cheap, and SOTP still implies meaningful additional upside if the HPAL option pays off. The debate has moved from "does INCO deserve any premium at all" to "how much HPAL value should be credited."

Investment Risk

INCO's risk profile is unusual, the top risks are dated, binary, and mostly regulatory rather than operational. That makes them mappable, and it makes position sizing the main defense.

8.1 Risk Register

8.2 Risk Mapping to Thesis Legs (Updated)

Leg #1 (matte) is threatened by R2 (price) and now also R4 (the volume shortfall that's already showing up in H1 actuals). Leg #2 (ore) is made of policy, R1 is its direct mirror image, and it's the risk that's already been realized rather than resolved. Leg #3 (HPAL) carries R1b + R3 together, feedstock competition and execution timing. R6 is the anti-thesis tax: it grows precisely when the thesis works.

8.3 Scenario Heat Maps, How the Risk Board Shifts

Entry, Exit, & What Makes This Thesis Broken

Discipline over hype. Where I enter, where I exit, and, most important, what tells me I'm wrong.

How to Entry, scale in, don't go all-in (Updated)

The original entry levels (below Rp 4,200 / Rp 3,600) were never revisited after the H1 beat, the price moved straight through them to Rp 5,500. Entry 1 (revised): below Rp 5,000, a pullback into the DCF/Comps blended band (§7.2); start 30-40% of intended position. Entry 2 (revised): below Rp 4,350, closer to the Comps-only read (the most conservative intrinsic method); add the remaining 60-70%. If the RKAB revision is finally approved and the price gaps up first, the thesis is confirmed but the entry is worse, re-run the sensitivity before paying up.

How to Exit, scale out in tiers (Updated)

Exit 1: Rp 5,500, already here; this was the original re-rating target. Consider trimming 20-30% given the RKAB catalyst has not resolved. Exit 2: ~Rp 6,600 (Base-case blended fair value, updated model), requires visible HPAL commissioning and/or an RKAB revision. Sell to core. Exit 3: ~Rp 8,300+ (Bull scenario), only holdable if LME sustains >US$18,000 and the RKAB quota is revised up and Pomalaa HPAL ramps cleanly from its August start.

What makes our thesis broken? (Status update)

Closing (Updated 10 Aug 2026)

Three weeks on, the picture has sharpened rather than changed direction. The earnings-recovery half of this thesis is now proven, not projected, H1 2026 margins sit above the prior cycle peak. But the market has already paid for a good chunk of that (price up ~21% to Rp 5,500), and the regulatory half of the thesis, the RKAB revision, has not resolved, sitting exactly where it was in July. INCO is no longer a stock where the debate is "is the earnings recovery real", it's now a stock where the debate is "does the government revise the quota, does Pomalaa's August start-up hold, and does matte volume catch up to guidance." Worth continuing to watch closely; less asymmetric than it was at Rp 4,560, but not resolved either way.