Mining Stock Valuation Methods: P/NAV, EV/EBITDA, EV per Ounce and When Each Breaks

Summary box

• A mine is a depleting asset with a defined end. Perpetuity-based multiples built for going concerns systematically misprice miners.
• P/NAV is the sector's primary valuation anchor: market capitalisation divided by discounted after-tax net asset value.
• EV/EBITDA works for diversified producers and fails for single-asset miners with short remaining lives.
• EV per resource ounce or pound is a screening ratio, not a valuation. It ignores grade, depth, recovery, jurisdiction and capital intensity.
• No single metric works across the lifecycle. Explorer, developer and producer are three different valuation problems.

Why mining equities need their own toolkit

Three properties distinguish a mining company from a typical industrial:

Depletion. Every tonne mined is a tonne that will never be mined again. Reserves are inventory that is consumed, and the company must replace them through exploration or acquisition simply to stand still. A miner with ten years of reserves and no replacement pipeline is a bond with commodity optionality, not a growth business.

Finite, modellable life. Unusually for equities, the asset has an engineered production schedule with an end date. That makes a discounted cash flow model genuinely tractable — and makes terminal-value assumptions largely unnecessary.

Price-taking. Revenue is a function of a commodity price the company does not control. Operating leverage to that price is often extreme.

Together these mean the sector's anchor metric is a DCF-derived net asset value, and everything else is a shorthand for it.

P/NAV: the anchor

Net asset value is the sum of the discounted after-tax free cash flows of each operating and development asset, plus balance-sheet items — cash, debt, investments, corporate G&A capitalised as a negative, and any exploration or royalty value ascribed separately.

P/NAV is market capitalisation divided by that NAV, usually quoted as a multiple: 0.6x, 1.1x, and so on.

What it is good for: it is the only widely used metric that captures grade, cost, mine life, capital intensity, tax regime and timing in one number. It is the metric mining M&A is actually negotiated on.

Where it breaks:

• It is model output, not observed fact. NAV depends on the analyst's price deck, discount rate and production assumptions. Two competent analysts can differ by 40% on the same asset.
• The discount rate carries enormous weight. Conventionally around 5% for gold and 8% for base metals, but that convention encodes a view about jurisdiction and commodity risk that is rarely re-examined per asset.
• It rewards long mine lives mechanically. At a 5% discount rate, cash flows in year 25 still carry meaningful weight. At 10% they are close to worthless. Long-life projects are therefore disproportionately sensitive to a parameter chosen by convention.
• Exploration and optionality are handled inconsistently. Some analysts ascribe an in-situ value per ounce to resources outside the mine plan; others assign zero. The difference can be most of the market cap for an explorer.

Practical use: compare P/NAV across a peer set built on a single consistent price deck and discount rate. Cross-broker P/NAV comparisons are close to meaningless.

EV/EBITDA: useful for producers, dangerous for short-life miners

Enterprise value divided by EBITDA is the generalist's multiple, and it does travel to mining — with a caveat that matters more here than almost anywhere else.

EV/EBITDA implicitly assumes the earnings stream continues. For a diversified major with decades of reserves across many assets, that is a defensible approximation. For a single-asset producer with six years of reserve life, an 4x EV/EBITDA multiple is not "cheap" — it may be exactly right, because the earnings stop.

Adjustments worth making:

• Reserve life adjustment. Compare EV/EBITDA against remaining reserve life. A scatter of the peer set on those two axes explains a large share of apparent multiple dispersion.
• Sustaining capex. EBITDA ignores it. In mining, sustaining capital is a permanent, non-discretionary cost of remaining in business. EV/(EBITDA − sustaining capex) is a more honest denominator.
• Closure liability. Reclamation and closure obligations are real, senior, and frequently understated in enterprise value. Add them to net debt.
• Attributable versus consolidated. Joint ventures and minority interests need consistent treatment on both sides of the ratio.

EV per ounce and EV per pound: a screen, not a valuation

Enterprise value divided by contained metal in resources or reserves gives a headline like "$38 per resource ounce."

This is genuinely useful for one thing: rapidly ranking a large universe to find outliers worth investigating. It is not a valuation, for at least six reasons:

1. Grade is ignored. An ounce in 0.4 g/t ore and an ounce in 6 g/t ore cost radically different amounts to extract.
2. Depth and method are ignored. Open-pittable and 1,200 metres underground are different businesses.
3. Recovery is ignored. Contained metal is not recoverable metal, and refractory ore may recover 60% where free-milling ore recovers 94%.
4. Category is ignored. Unless you split it, Inferred ounces count the same as Proven reserves. They should not.
5. Capital intensity is ignored. Two identical resources, one needing $200 million to build and one needing $2 billion, are not equally valuable.
6. Jurisdiction is ignored. Tax, royalty, permitting timeline and expropriation risk do not appear anywhere in the ratio.

If you use it, use it in a tightly defined cohort: same commodity, same broad grade band, same mining method, same jurisdiction tier, same resource category. Then it becomes a reasonable relative screen.

For base metals the equivalent is EV per pound of contained copper, zinc, nickel or lead, with the same caveats. Polymetallic deposits require conversion to a common metal-equivalent basis, and metal-equivalent calculations are themselves price-assumption-dependent — check the prices used before comparing.

Metrics by lifecycle stage

Royalty and streaming companies deserve a note: they persistently trade above 1.0x P/NAV, often well above, because they carry no operating cost inflation, no sustaining capital and no closure liability, while retaining full exposure to metal price and to resource growth on the underlying properties. Comparing a royalty company's P/NAV to a producer's without accounting for that structural difference produces a false conclusion every time.

Metrics that matter alongside the multiple

All-in sustaining cost and position on the cost curve. Margin resilience through the cycle is a function of where you sit on the industry cost curve, not of absolute cost.

Reserve life index. Reserves divided by annual production. Below eight years, reserve replacement becomes the dominant question.

Reserve replacement ratio. Reserves added versus reserves depleted, over three to five years. Sustained sub-1.0x means the company is liquidating.

Free cash flow yield. After sustaining capital and tax. The number that survives contact with reality.

Net debt to EBITDA. In a price-taking, high-operating-leverage sector, leverage above roughly 2.0x becomes structurally dangerous at the bottom of the cycle.

Dilution history. For developers and juniors, share count growth is the most reliable predictor of per-share value destruction. A project that doubles in NPV while the share count triples has made shareholders poorer.

Single-asset concentration. One mine, one jurisdiction, one permit, one processing plant. Concentration is the dominant risk factor for most junior producers, and no multiple captures it.

A workable sequence

1. Establish the stage: explorer, developer, single-asset producer, diversified producer, royalty.
2. Pick the anchor metric for that stage.
3. Build the peer set on shared characteristics — commodity, stage, jurisdiction tier, mining method.
4. Normalise the inputs: one price deck, one discount rate, one tax basis, one currency.
5. Cross-check with a second metric on a different logic. If P/NAV and EV/EBITDA disagree sharply, the disagreement is the finding.
6. Overlay the qualitative gates: reserve life, balance sheet, dilution history, jurisdiction, single-asset risk.
7. Ask what has to be true for the valuation to be right, and what observable event would prove it wrong.

Step 7 is the one that separates analysis from narrative.

How Mining Terminal handles valuation inputs

Mining Terminal holds the raw inputs — resource and reserve tonnes and grades by category, project economics by study stage, production and cost metrics, capital structure and dilution history, ownership, and transaction comparables — extracted from source filings and tagged with their effective date and reporting basis.

Transaction values and financial metrics are carried in the currency the filing disclosed. Aggregates spanning multiple currencies are labelled as mixed rather than converted at a rate the source never used, because a silently converted total is a fabricated number wearing a decimal point.

To build a normalised peer set or compare project economics on a consistent basis, get in touch or ask Nara.

Related reading

• Mineral Reserves vs Mineral Resources: What the Categories Actually Mean
• PEA vs PFS vs Feasibility Study
• AISC Explained: What All-In Sustaining Cost Includes, Excludes, and Hides

This article is educational and is not investment advice. Mining Terminal is a data platform, not a broker, dealer or investment adviser.