Trimegah Bangun Persada (Harita Nickel)

1. 01Mine one island, two ore types

   Obi Island yields both saprolite (higher grade) and limonite (lower grade). Each feeds a different processing route, so a single mining operation supplies two product chains.

   EconomicsThe limonite is the point. In a matte-only or NPI-only model that low-grade material is waste; here it is HPAL feed. Monetising the whole ore body rather than half of it is the structural reason NCKL earns what it earns.

2. 02Two processing routes in parallel

   Saprolite goes through RKEF smelters into ferronickel and NPI for stainless steel. Limonite goes through HPAL into MHP, then on to battery-grade nickel and cobalt sulfate.

   EconomicsGross margin fell from 54.73% (FY2021) to 32.69% (FY2025) as volume scaled roughly 3.6 times, so the added tonnes came in at lower unit margin than the first tonnes. The trade was accepted deliberately: absolute gross profit still rose from Rp 4.50 T to Rp 9.69 T.

3. 03Refine downstream on the same island

   MHP is refined on site into nickel sulfate and cobalt sulfate rather than shipped out as intermediate, keeping the downstream margin inside the group.

   EconomicsIntegration on one island removes a logistics leg at every stage. It is also why NCKL reports in rupiah while INCO, MBMA and MDKA report in US dollars: much of the chain settles domestically. That reporting choice has a direct valuation consequence, set out under returns below.

4. 04Carry the ore, then collect

   Inventory dominates the working-capital cycle: 107.52 inventory days in FY2025 against 23.87 receivable days and 33.09 payable days.

   EconomicsThe cash cycle of 98.30 days is an ore-stockpile cycle, not a credit cycle. Customers pay quickly; the capital is tied up in material sitting in the chain. That is the working-capital cost of running two processing routes off one mine.

Cost structureCost of goods sold absorbed 45.27% of revenue in FY2021 and 67.31% in FY2025, the mirror image of the gross-margin path from 54.73% to 32.69%. Operating margin fell less far, 43.61% to 28.63%, meaning overheads grew slower than revenue and gave back part of the gross-margin loss. Compare the shape with INCO: NCKL lost margin because it added lower-grade volume by choice, while INCO lost margin because its cost base would not move when the price did. Falling margin is not one diagnosis.

Cash cycleCash conversion cycle: 121.68 days (FY2021), a spike to 221.41 days (FY2022) when inventory days reached 304.60 during the HPAL ramp, then 104.05, 97.96 and 98.30 days. The FY2022 spike is the build showing up in working capital: material accumulated in a chain that was not yet running at rate. Since FY2023 the cycle has been stable near 98 to 104 days, which is the steady-state cost of the integrated model.

Unit economicsRevenue per rupiah of gross fixed investment tells the ramp story better than margin does: asset turnover ran 0.39 times (FY2021), fell to 0.28 times (FY2022) while capacity was being built but not yet sold from, then settled at 0.53, 0.52 and 0.48 times. The asset base began paying only from FY2023.

Good, with one year that must be set aside. FY2022 net income of Rp 5.92 T exceeded EBIT of Rp 3.86 T by Rp 2.06 T, and net margin (61.82%) came in above even gross margin (48.81%). The engine’s tax-burden ratio printed 1.579 that year, above 1.0, which is impossible for a genuine tax factor and is exactly the condition that triggers the ROIC clamp described in the ratio engine. The item is not identified in the seeded statements. Every other year is clean: in FY2025 net income of Rp 8.95 T sits below EBIT of Rp 8.48 T only by the ordinary financing and tax wedge, and operating margin, EBITDA margin and net margin move together. Read the FY2022 ROE of 49.21% as inflated by an unidentified below-the-line item and the surrounding range of 20.93% to 34.90% as the honest one.

Cash conversionClean and improving. FY2025 operating cash flow of Rp 8.60 T against net income of Rp 8.95 T is 0.961 times, close to one-for-one, and the ratio against EBIT rose from 80.98% in FY2024 to 101.39% in FY2025. Unlike INCO, the conversion here is not a depreciation artefact: depreciation of Rp 1.89 T is only 0.22 times EBIT of Rp 8.48 T, so the cash is coming from trading profit. Free cash flow has been positive three years running and reached Rp 7.66 T in FY2025.

Mixed is the accurate label because the window contains a completed transition rather than one behaviour. NCKL was a builder in FY2021 and FY2022, spending Rp 6.88 T and Rp 6.41 T of capital expenditure against operating cash flow of Rp 1.83 T and Rp 3.52 T, and running free cash flow of negative Rp 5.05 T and negative Rp 2.89 T. From FY2023 it has been a harvester: free cash flow of positive Rp 2.30 T, Rp 4.40 T and Rp 7.66 T, a dividend yield of 4.76%, and debt to equity down from 0.69 to 0.20. Across the full five years capital expenditure absorbed 75.70% of operating cash flow and cumulative free cash flow was positive Rp 6.43 T, equal to 6.54% of cumulative revenue. Set against the other three nickel-complex names tracked here, that 6.54% compares with INCO at 0.80%, MBMA at negative 22.17% and MDKA at negative 19.48% (MDKA consolidates MBMA, so those two overlap rather than stand apart): NCKL is the only one of the four to have converted its build into cumulative free cash worth more than a rounding error.

DeploymentCapital expenditure fell 86.35% across the window, from Rp 6.88 T (FY2021) to Rp 0.94 T (FY2025), and FY2025 spending was just 3.17% of revenue and 0.50 times depreciation of Rp 1.89 T. The released cash went to two places: the balance sheet, where debt to equity fell from 0.69 to 0.20 and the current ratio rose from 0.55 to 2.13, and shareholders, at a 4.76% dividend yield. Spending below depreciation deserves scepticism rather than applause, and it is recorded as a red flag below.

Returns trendThis is the finding that separates NCKL from its peers, and it survives the currency problem rather than depending on it. ROIC ran 26.00%, 16.56%, 18.37%, 17.28% and 16.92% against an engine WACC of 13.62% and a discount band of 12% / 14% / 16%. NCKL cleared its cost of capital in all five years, by 12.38, 2.94, 4.75, 3.66 and 3.30 percentage points. The comparison must be made this way and not on raw ROIC, because NCKL reports in rupiah and is discounted on a rupiah band, while INCO, MBMA and MDKA report in US dollars and are discounted at 11.08%, 10.95% and 12.14%. NCKL therefore carries the HIGHEST hurdle of the four and is still the only one clearing it: on spread over its own hurdle, FY2025 reads NCKL +3.30 points, INCO negative 9.70 points, MBMA negative 9.82 points, and MDKA not measurable. A raw ROIC comparison across four companies reporting in two currencies would misrank them. One caution about the peer set itself: MDKA consolidates MBMA in full, and MBMA’s own revenue equals 75.71% of MDKA’s consolidated revenue in FY2025, so those two are not independent observations and the same nickel platform is counted twice in any list naming both.