The Mine Has Years Left. So Why Is It Hard to Finance?

The usual mine valuation counts remaining reserves as value and closure as a discounted cost for later. Evidence from South Africa, Papua New Guinea, Brazil and major mining jurisdictions shows why that divide can be dangerously false: the cost of securing closure can constrain cash long before the mine stops producing.

Anglo American Platinum’s Rustenburg operation was still producing when the company disclosed a R1.817 billion unscheduled closure liability in June 2025.

That is the consequence hiding behind the word “closure”. A working mine, with ore still to extract, had a bill for infrastructure, mining and land rehabilitation, plus the care required after production ends. Post-closure monitoring and maintenance alone accounted for R178.85 million.

The liability was not simply waiting at the bottom of a distant valuation model. Funding required cash held in trust and third-party guarantees. In other words, capital that might otherwise support the operation, refinancing or a transaction had already acquired another job.

That raises an awkward question for anyone buying, lending to or running a mine: what if the asset has reserves, but the money needed to make those reserves valuable is no longer freely available?

Most people looking at a mine see the obvious bargain. Ore remains in the ground. The mine has a stated life. Closure costs appear in the accounts as a provision, generally a future obligation expressed in today’s money. So why not value the asset on the cash it can still generate, subtract the provision and move on?

Because a provision is a forecast. Financial assurance is a demand for something real.

That distinction can decide whether a mine can borrow, refinance or attract a buyer. It is the same basic trap explored in our look at why a strong contract can still leave a company unable to borrow: value on paper is not always cash that a lender can rely on.

The comforting number on the balance sheet

The conventional view is understandable. Closure happens at the end. Rehabilitation means restoring disturbed land and managing the site after mining. A company estimates the cost, books a liability and updates it from time to time. Regulators may require a bond or another form of security. Problem handled.

Except that the three things are not interchangeable.

The provision is an accounting estimate. A rehabilitation bond is one possible promise of payment. Financial assurance is the cash, trust arrangement or third-party guarantee that stands behind the promise if the operator cannot perform.

Imagine you own a house worth £500,000 and expect a £100,000 repair bill in ten years. You may record that concern in your own budget. But if a bank insists today that you put £100,000 into a locked account before it will lend, your financial position has changed immediately. You still own the house. You are simply short of usable cash.

Mines face this version of the problem with much larger numbers and less certainty, especially once water must be treated long after production ends.

Here is the twist: closure is not merely an end-of-life risk brought forward by cautious regulators. It can be a present-day capital-structure constraint. The mine may be producing. Its reserves may have a positive value. But the cash and credit needed to secure its eventual clean-up may be unavailable for everything else.

Rustenburg’s bill was already in the room

Rustenburg matters precisely because it was not a deserted site awaiting a clean-up crew. Anglo American Platinum, the South African mining company, reported the R1.817 billion estimated unscheduled closure exposure while the operation was active.

A third-party guarantee may sound less painful than paying cash. It is not free financial oxygen. The disclosure establishes that Rustenburg’s closure funding involved trust cash and guarantees. The practical financing implication is straightforward: cash placed in trust is restricted, while guarantees are commitments that must be accommodated alongside other funding needs. How any particular lender treats that commitment will depend on its own credit process, the guarantee terms and the borrower’s wider finances.

This is where most people stop looking. They see a closure provision and assume it captures the full obligation. It may be properly prepared for accounting purposes and still fail to answer the questions that matter in a financing or sale: when can security be called for, how much cash becomes restricted, and what other uses of capital become harder as a result?

The R1.817 billion does not, by itself, prove that Rustenburg cannot finance itself. The evidence does not establish that. What it proves is more useful: a large operating mine can carry an unscheduled closure exposure substantial enough to tie up capital well before the last tonne is mined.

The fund that started before closure

In July 2002, the operator of Ok Tedi Mine in Papua New Guinea began contributing to a reserved closure fund every six months.

The mine had not shut. Its cash flow still had work to do. Yet money was being set aside in a jointly held financial assurance fund for a moment that might sit far ahead on the operational calendar.

The Papua New Guinea government accepted US$100 million as financial assurance. Earlier closure costs had been estimated at about US$150 million, then revised to US$91 million. Those changing numbers matter. But the revealing fact is simpler: the money was no longer entirely the operator’s to deploy.

Ok Tedi's closure fund shows how future rehabilitation can begin consuming operating cash decades earlier. Photo: The original uploader was Dr. Blofeld at English Wikipedia. / Wikimedia Commons, Attribution.

Ok Tedi shows the constructive version of closure discipline. A reserved fund offers more credible protection than hoping a future owner, commodity price or operating cash flow will solve the problem. It also constrains liquidity in real time.

Both statements are true. That is the part valuation models too often separate.

Put Rustenburg and Ok Tedi side by side and a pattern appears that neither disclosure says outright. One mine reveals a large closure exposure supported through trust cash and guarantees. The other began pre-funding a jointly held reserve decades before closure. Different structures, same economic fact: closure claims compete with the uses of capital that make a mine operate, expand or change hands.

Water is where the estimate starts to move

Land can be reshaped. Buildings can be removed. Monitoring can be planned. Water is often less obedient.

A 2024 technical article published by Springer found that long-term water treatment can roughly double closure costs, taking estimates from tens of millions to hundreds of millions. The surprise is hidden inside an engineering assumption. A mine can appear to have a manageable closure bill until someone asks a deceptively small question: for how long must the water be treated?

Discounting means reducing a future cost to a smaller value in today’s terms. It is a sensible calculation tool. It is not proof that the future cost, its timing or the security required to cover it has been understood.

For post-closure water treatment, duration matters as much as the nominal estimate. If the estimate excludes a credible long tail of treatment and monitoring, the operator can be underprepared in two ways. The ultimate bill may be larger. And the financial assurance demanded by a government or considered in a financier’s downside analysis may be based on a broader exposure than the accounting line suggests.

This is not an abstract concern. The World Bank’s 2025 mine-closure toolbox describes regimes where the clock has been moved forward.

The rules changed the timing, not the physics

Nevada in the United States, Chile and Quebec in Canada require 100% financial assurance for closure costs on a phased current-period basis, according to the World Bank. In some cases, assurance is required before a mining permit is issued.

The mechanism varies. Cash, bonds, guarantees and other instruments do different jobs. The governing principle does not vary much: do not leave the public carrying the bill if the operator fails.

For an operator, this changes the clock. A distant cost becomes a present requirement. It cannot simply assume that future operating earnings will remain available. It has to show, through hard security, that closure can be paid for.

Some readers will hear “100% assurance” and conclude that it makes mining unfinanceable. That is too neat. It can make a project more credible by reducing the risk of an unfunded legacy for governments and communities. Ok Tedi’s reserved fund illustrates the point.

But stronger assurance is also less liquidity. Any serious valuation has to hold those two ideas together.

The same tension appears beyond mining. A data centre may have a customer contract but remain unbankable until it has power. A mine may have reserves but remain capital-constrained until it has funded closure. In each case, the asset is not just what it can produce. It is what it can produce after its non-negotiable conditions are met. See why the lease alone is not the asset.

Brazil’s missing fund

Barro Alto mine in Brazil offers the uncomfortable opposite of Ok Tedi’s early funding.

A 2024 audit report found no evidence of a trust fund or equivalent long-term security for closure activities, despite a requirement for financial surety covering matters including water treatment and monitoring.

It would be easy to file that under compliance. Missing document, missing fund, regulator follows up. But the financial story is sharper than that. If the required surety is not already in place, enforcement could require an operator to find cash or guarantee capacity quickly. A liability assumed to be self-funded later becomes an immediate liquidity test.

What surprised us was not that a closure fund is helpful. Everyone says that. It is that its absence can create the most acute financing problem of all. A company that has set aside money looks cash-poorer today but may be more resilient. A company that has set aside nothing can look healthier right up to the moment it is asked to secure the obligation.

There is a historical reason governments have become wary. US Senate testimony around 2006 estimated that cleaning up abandoned hardrock mines in the United States could require nearly US$72 billion. The number is not a valuation tool for any operating mine. It is a warning about what happens when assurances are inadequate and the final owner cannot or does not pay.

What the evidence does not prove

The evidence does not say that every closure provision is unreliable, that every mine must lock up the same amount of cash, or that a guarantee always has the same effect as a cash trust.

Jurisdictions differ in timing, enforcement and acceptable forms of assurance. The World Bank records those differences. A well-designed and funded closure arrangement may reduce uncertainty for all parties. The point is narrower, and more important: a valuation cannot treat the provision, the assurance requirement and the mine’s usable liquidity as if they were the same number.

What operators should put on one page

A board does not need a thicker environmental report to see this risk clearly. It needs one working page that joins operations, finance and closure planning.

First, separate the accounting provision from the credible undiscounted closure and post-closure exposure. If water treatment could materially alter the number, show a range rather than one comforting point estimate.

Second, map the timing. When can a regulator demand a top-up? What must be held in trust? What guarantee capacity has been committed? Progressive rehabilitation, meaning restoration carried out while mining continues, should reduce exposure only where the relevant authority actually recognises it.

Third, test usable cash after security. Do not call a mine liquid if its cash is legally restricted, or if available guarantees are already committed.

Fourth, make oversight explicit. A board or board committee should receive regular closure and post-closure reporting covering costs, security instruments, water assumptions and changes in regulatory requirements. The question is not whether management has a closure plan. It is whether the plan still works if ownership changes, cash flow weakens or water treatment lasts longer than expected.

This is basic investment-committee preparation. The deal rarely dies because someone failed to write “closure risk” in a slide. It dies when a decision-maker realises too late that the available cash has already been promised elsewhere.

What experienced capital sees first

An inexperienced investor may ask, “How many years of reserves remain?” An experienced lender is likely to ask a harsher question: “How much cash remains after the unavoidable cost of earning those years of reserves?”

That is the psychology behind mining debt capacity. Debt is repaid from cash that is genuinely free, not cash that exists briefly before it is transferred to a trust or needed to support a guarantee. The precise treatment of closure security varies by lender and jurisdiction, but the issue belongs in any credible assessment of cash available to repay financing.

Smart investors should check four things before accepting a closure provision as reassurance:

• the gap between the recorded provision and the credible full closure exposure;
• the specific form of financial assurance and when it must be increased;
• whether long-term water treatment and monitoring have been explicitly included; and
• whether the obligation remains funded and enforceable if the mine changes hands or the operator weakens.

This does not mean every late-life mine is worth less than zero. It means reserve value alone is not the answer. The mine’s residual value belongs to its owners only after the closure claim has been credibly covered.

GI Network's view: A closure provision tells you what management expects. A residual-liability test tells you whether the mine can still pay, borrow and sell after regulators receive the security they require.

In a mine financing or acquisition, GI Network would build that residual-liability test before capital outreach: reconcile the provision with closure and water-treatment assumptions, identify restricted cash and guarantee calls, stress the remaining cash flow, and prepare the questions an investment committee will ask about assurance and successor exposure.

The Residual-Liability Coverage Test

Use this five-part test before assigning value to a late-life mine.

1. Full bill. What is the credible closure and post-closure cost before discounting, including water treatment and monitoring?

2. Hard security. What cash, trust funding or guarantees must support that bill, and what is already posted?

3. Timing trap. When can the required security rise, relative to the mine’s remaining cash generation?

4. Usable cash. After security, operating needs and existing debt, what liquidity is truly left?

5. Backstop. If ownership changes or the operator falters, who is still on the hook and is that backing credible?

If the answer to those five questions is vague, the provision is not a comfort. It is an invitation to look harder.

A mine does not become safer because its closure cost is placed at the bottom of a valuation model. It becomes safer when someone has established, in money rather than optimism, who can pay the bill.