Indonesia holds the world's fourth largest gold reserves, and the next task is measuring what the country actually produces

On 12 August 2026, MIND ID convened MINDialogue in Jakarta under the theme "Komoditas untuk Negeri: Strategi Penguatan Cadangan Mineral untuk Stabilitas dan Kedaulatan Ekonomi RI", commodities for the nation, a strategy for strengthening mineral reserves in service of Indonesia's economic stability and sovereignty. The panel brought together the state mining holding, the Ministry of Energy and Mineral Resources, and the investment ministry. What emerged was less a dispute than a shared diagnosis, and it centred on gold.

The specific proposal was procedural. Tri Winarno, Director General of Minerals and Coal at the Ministry of Energy and Mineral Resources, said the ministry is working to bring miners currently operating without licences into the permit system through the Izin Pertambangan Rakyat, the people's mining permit known as IPR. Maroef Sjamsoeddin, Chief Executive of MIND ID, set out the same direction from the industry side, describing a transformation of artisanal mining into IPR status supported by regulatory harmonisation and technical guidance. Todotua Pasaribu, Deputy Minister of Investment and Downstream Processing and Deputy Head of BKPM, framed the objective in terms of added value, arguing that the country's natural resource endowment has to generate more of it domestically.

That is a governance agenda, and it is a constructive one. It is also, read closely, a measurement agenda. The figures presented in Jakarta describe a national gold system whose totals do not currently reconcile, and reconciling them is the work that makes every downstream policy choice possible.

The numbers that do not yet close

Start with the endowment. The United States Geological Survey's Mineral Commodity Summaries 2026 places Indonesia fourth in the world for gold reserves, at 3,600 tonnes as of 2025, behind Australia at 13,000 tonnes, Russia at 12,000 tonnes, and South Africa at 5,000 tonnes. That is a substantial national asset by any accounting.

Now the flows. Speakers at MINDialogue put annual domestic gold demand at 190 to 200 tonnes. Formal supply, the gold that moves through licensed channels and can be traced, was given as 53.5 to 86.2 tonnes a year. Small scale and artisanal mining was estimated to produce 50 to 120 tonnes annually, a band whose upper end approaches and may exceed the entire formal supply.

Set those against independent production data and the picture becomes more interesting still. The USGS estimates Indonesian gold mine production at 90 tonnes in 2025, down from 94 tonnes in 2024. Irwandy Arif, chairman of the Indonesia Mining Institute, has put 2024 output from all gold companies operating in Indonesia at approximately 160 tonnes. Those two figures describe the same country in the same year and differ by roughly seventy tonnes. Irwandy has also noted that of 112 gold mining licences across Indonesia, 80, or 71 percent, have documented resource data.

Indonesia is not short of gold. It is short of an agreed number for how much gold it produces, and that single missing number sits underneath the stockpile proposal, the reserve asset proposal, and the formalisation proposal alike.

None of these figures is wrong in the sense of being carelessly produced. They are measuring different things through different instruments: reported company output, modelled national production, formal channel throughput, and estimated informal activity. The discrepancy is a description of the current state of the evidence base, not a failure of any one institution. But a range of 50 to 120 tonnes is a factor of 2.4 between its ends, and policy cannot be costed on a factor of 2.4.

Why gold is harder to measure than nickel or tin

Maroef offered the geological reason directly. Tin is concentrated in Bangka Belitung and the Riau Islands. Nickel is concentrated in Sulawesi and Maluku. Gold is not concentrated anywhere. As he put it, the deposits run from Aceh to Papua, across essentially the whole archipelago, and can be worked in northern Kalimantan or in Java with modest equipment.

That dispersion changes the governance problem in kind, not just in degree. Nickel and tin governance can lean on a small number of large, fixed, observable sites. Gold governance cannot. It has to work across thousands of small, mobile, geographically scattered operations, many of them the economic base of the communities around them. Instrumenting that is a different exercise, and it is the exercise that GIS and remote sensing are built for. Satellite and aerial observation can establish where activity is occurring, over what area, and how that area has changed between periods, at a national scale and at a repeatable cadence that field inspection alone cannot match.

The traceability point raised at the forum runs in parallel. MIND ID described the absence of an integrated traceability system as the reason artisanal gold struggles to enter formal supply chains or meet international refining standards such as those of the London Bullion Market Association. The proposed architecture has three tiers: formalisation into IPR upstream, licensed aggregator institutions to handle ore in the midstream, and a transparent, tracked formal system downstream.

The valuation case sitting inside the formalisation case

Formalising a mine is not only about issuing a document. An IPR carries obligations, and among them are the environmental ones: reclamation planning, reclamation guarantee deposits, and compliance with the standards that permit-holders answer for to the Ministry of Environment and Forestry. Bringing tens of tonnes of annual production into the permit system therefore brings a corresponding volume of land, water, and ecosystem interaction into the accountable perimeter for the first time.

That perimeter needs a baseline. Reclamation guarantee amounts are set against disturbed area. Environmental economic loss under Permen LH No. 7 Tahun 2014 is calculated against measurable change in ecological condition. Neither calculation can be performed retrospectively on an operation whose starting condition was never recorded. The moment of formalisation is the one moment when that baseline can be captured cleanly, and it is available only once per site.

There is also a straightforward fiscal argument. Gold that moves outside formal channels generates no royalty and no recorded export value. Bank Indonesia reported foreign exchange reserves of US$145.3 billion at the end of July 2026, against US$145.6 billion at the end of June, equivalent to 5.5 months of imports and comfortably above the international adequacy benchmark of around three months. Maroef's argument that domestically produced gold could serve as a strategic asset supporting those reserves depends entirely on that gold being counted, refined to standard, and held within a system that can verify it. His related proposal for a national mineral stockpile, modelled on how Malaysia manages crude palm oil supply and pricing, rests on the same precondition.

What to watch over the coming months

Three things will indicate whether the Jakarta discussion converts into implementation. The first is regulatory harmonisation on IPR, which Maroef named explicitly and which matters because reclamation obligations for small scale mining are currently addressed across more than one instrument. The second is whether the aggregator institutions in the midstream tier are given a defined legal form, since a traceability system without an accountable intermediary is difficult to audit. The third is whether a single national production figure emerges that the ministry, the state holding, and independent bodies can all work from.

Sirkularium's view

For government and public institutions, the constructive reading of MINDialogue is that Indonesia is choosing to formalise rather than simply to police, and formalisation is the option that produces data. Every permit issued under IPR is a site that can be mapped, baselined, monitored, and eventually valued.

Sirkularium's position is that the valuation work should be built into the formalisation process from the start rather than added later. That means pairing each IPR issued with a baseline drawn from GIS and remote sensing, verified against ground observation, and recorded using the same methodology the country already applies when it prices environmental loss. Done at the point of entry, that baseline costs comparatively little and makes reclamation guarantees defensible, ecosystem services valuation possible, and natural capital accounting for the gold sector achievable within a few years.

The alternative is to formalise first and measure afterwards, at which point the counterfactual is gone and every subsequent figure becomes contestable. Indonesia has the reserves, and it now has a clear policy direction. Commissioning rigorous, independently verified economic valuation as ongoing practice, rather than as a response to disputes after they arise, is what turns a fourth place ranking in the ground into a measured contribution above it.