How we value resource companies

Why the P/E ratio does not work here

Most exploration and development companies have no revenue and no earnings. A price-to-earnings ratio simply cannot be calculated. What they do have is an asset in the ground, documented in a technical report — and a long, expensive road to production.

So what gets valued is the project, not the operating business.

The decision that governs everything else

Before the individual building blocks comes a question most write-ups skip: where in the calculation does the risk sit?

There are two workable answers. They arrive at the same place, and they are mutually exclusive.

Convention A — low discount rate, multiple below 1.0

The project's present value is discounted at 5% real. The risk adjustment does not sit in the rate; it sits in the P/NAV multiple, which is therefore well below 1.0. This is how the gold industry has worked for decades. Ernst & Young set it out verbatim in a valuation filed with the US Securities and Exchange Commission in 2019:

"it is common in the gold mining industry to consider a 5% real discount rate in conjunction with a net asset value ('NAV') multiple; where a NAV multiple of less than 1.0 captures incremental risk relative to the 5% real discount rate and a NAV multiple of greater than 1.0 implicitly ascribes value to upside…"

That the 5% is deliberately low is confirmed elsewhere in the same document: EY put the weighted average cost of capital for that same asset at 11.5%. The gap is not an arithmetic slip. It is the method.

Convention B — risk built into the discount rate, multiple near 1.0

The rate carries the risk: 8% to 12% real, after tax, depending on jurisdiction, permitting status and stage. The multiple then sits close to 1.0, because the haircut has already been taken.

What does not work is mixing them

A present value discounted at 10% and then multiplied by 0.3× counts the same risk twice — once in the rate, once in the factor. The result looks conservative. It is merely double-counted, and it leaves almost no project standing.

That mixture is exactly what the previous version of this page contained. It has been corrected.

Our choice: convention A

We discount at 5% real, after tax, and carry the risk in the multiple. Two reasons:

1. It is the documented practice of the sector, set out in a formal valuation filed with a securities regulator — not industry folklore.
2. It is the only one that keeps our numbers comparable with the multiples published in analyst research. A P/NAV built on convention B cannot be placed alongside those reports.

Where a case calls for a departure — pronounced jurisdiction or permitting risk, for instance — we say so, and we do not additionally cut the multiple for the same circumstance.

Which leads to the sentence this page really exists for:

A P/NAV quoted without its discount rate is not a number. It is a gesture.

0.6× on a NAV discounted at 5% and 0.6× on a NAV discounted at 10% describe two entirely different views of the same project. Set them side by side and you are comparing nothing. So in every analysis we state the discount rate, the price assumption and the study stage. Without that trio we publish no P/NAV.

In detail: P/NAV.

The three approaches we combine

1. NAV — net asset value via a project DCF

The basis is the technical study under NI 43-101 (Canada) or JORC (Australia). From it we take the production profile, operating costs, capital requirement and mine life, and discount the resulting cash flows. A report filed under those standards is a disclosure document, not a seal of quality: it certifies how the figures were produced, not that the project works.

Discount rate: 5% real, after tax, per convention A. The multiple carries the risk.

Commodity price: long-term consensus assumption, not spot. A DCF built on an all-time high is not a valuation, it is a bet.

Cost base: the operating costs in the study, checked against reported all-in sustaining costs — a metric whose contents vary with the issuer's interpretation and which is therefore rarely comparable at face value.

2. P/NAV — the discount the market demands

Junior names almost never trade at full NAV. That is not a market error: financing risk, dilution, permitting and years stand in between.

Stage-based ranges circulate widely. They appear in presentations and forums; no standard prescribes them and no bank publishes them as a rule. They are convention — useful as a reminder of the ordering, useless as an expected value:

Observed values support mainly the lower half of that table. For the valuation cited above, EY surveyed the multiples analysts applied in their research to eight gold companies — mostly producers and near-producers. The result:

The median therefore sat where the convention places an explorer — for companies that are already mining or close to it. A table that puts producers at 0.8× to 1.2× across the board describes a strong market, not a general rule.

In every analysis we state which band we apply and why, and we treat that band as what it is.

3. EV per resource unit — for comparison only

EV/oz Au, EV/t Cu, EV/t LCE. Useful for placing a company against its peer group. Never as a standalone measure — the metric ignores grade, recovery, depth, infrastructure and jurisdiction, which is to say almost everything that makes the difference.

Resource categories: not all of it counts equally

The categories do not describe how much metal is in the ground. They describe how confident one is that it is there. So they enter the NAV on a sliding scale rather than being added up:

This ladder is not a standard; nothing prescribes it. There is no standard for the size of the haircut at all: analyst research commonly applies anywhere between a 30% and a 60% discount to inferred material. We follow the weighting EY applied in the valuation cited above — inferred resources at 30% of their contribution, that is, a 70% haircut. It sits at the conservative end of normal practice, and it is our choice, not the only defensible one.

Which of the two treatments we use for inferred material is stated explicitly in each analysis, together with the cut-off grade and the assumed price. It is the easiest lever with which to flatter a NAV. That is why we disclose it.

What the categories mean one by one — and why measured does not automatically become proven: Inferred, indicated, measured.

Dilution: the most common error

An explorer without cash flow funds every further drill season through equity raises. Shares outstanding can multiply several times over before production.

So we calculate NAV per share after expected dilution to production — not on today's share count. That is the difference between a valuation and a brochure.

It also means accounting for outstanding warrants and options, and for the cash runway in months at the current burn rate. The runway decides whether a company negotiates its next financing from strength or from need — and therefore on what terms.

Study stages and their accuracy

Two sets of numbers stand against each other here, and the difference is the heart of the matter.

What the studies say about themselves

In the reports themselves, accuracy is usually given as a symmetrical band. A PFS puts its cost estimate at "±25%", a feasibility study at "±15%".

These bands are industry convention, not a standard. Neither NI 43-101 nor the CIM Definition Standards prescribe them. They appear in the documents because they have become customary.

What the cost standard actually says

The reference the industry appeals to is the AACE 47R-11 cost estimate classification for mining and mineral processing. There, the bands are not symmetrical:

The axis against which AACE plots these bands is titled Growth from Estimated Costs. The standard does not expect the error to be evenly distributed in both directions. It expects it on the high side.

Read "±50%" for an early-stage study and an 80% cost increase looks like an outlier. Under AACE it falls inside the band. The previous version of this page understated the bad side by half — which is precisely the side that reaches the shareholder.

AACE assigns the feasibility study to class 3 and the pre-feasibility study to class 4. On the PEA the standard stays deliberately open: it is "usually … Class 5, but not always". We do not resolve that ambiguity. "PEA" is a term from disclosure law, not from cost engineering; two studies carrying the same label can represent two different levels of planning maturity.

What practice shows

Bertisen and Davis examined 63 mining projects in the Engineering Economist in 2008, comparing actual capital costs with the bankable feasibility study in each case. The result: a mean overrun of +14% — and close to half the projects fell outside ±15%, that is, outside what a feasibility study claims for itself.

The reason is mundane. At an early stage the missing items are the ones nobody knows about yet: infrastructure, dewatering, reclamation, interest during construction. A line item nobody estimated can only move one way.

We therefore carry the accuracy of the cost base asymmetrically, following the logic of AACE 47R-11, rather than as the "±" printed in the report. A PEA cannot declare reserves and is expressly preliminary; we weight PEA-derived results accordingly. More on this: PEA, PFS, FS.

What this method cannot do

• It values the known project. Exploration success beyond the current resource is not included — upside from new discoveries stays out.
• Technical studies are snapshots. Capital costs have risen sharply in recent years; older studies often understate the capital requirement considerably.
• Permitting and jurisdiction risk can only be approximated, not cleanly quantified. Under convention A it lands in the multiple, and the multiple is a judgement, not a measurement.
• The NAV itself is not fixed. Two houses can reach present values a factor of two apart on the same project if one uses spot and the other long-term consensus. The P/NAV inherits that spread in full.
• In a takeover, an acquirer often pays for strategic reasons no DCF captures.

What we do not do

We do not model spot prices at all-time highs. We do not count inferred resources in full. We do not combine a risk premium in the rate with a risk discount in the multiple. We do not ignore dilution. And we do not take a figure from a corporate presentation without checking it against the technical report.

What changed in version 2

Date of this version: 11 August 2026. The following statements in version 1 were incorrect and have been corrected against the primary sources.

1. Discount rate and multiple. Version 1 required 8–10% real and P/NAV bands of 0.2× to 0.7×. That counts the same risk twice. Version 2 sets out the two conventions separately and states that we work under convention A: 5% real, risk in the multiple. The basis is the 2019 EY valuation filed with the SEC, which describes the pairing explicitly.
2. P/NAV bands. They are now labelled as convention and set against observed values — mean 0.54×, median 0.40×, range 0.30× to 1.08×. Version 1 presented them as established figures.
3. Inferred resources. Instead of a blanket "50% discount, or excluded", the ladder is 100% / 60% / 30%, stated explicitly as our choice within a customary range of 30% to 60% discount.
4. Study stages. The symmetrical bands (±50% / ±25% / ±15%) now appear where they belong: as what the reports say about themselves. Alongside them stand the asymmetrical AACE 47R-11 classes and the empirical finding of Bertisen and Davis.

One clarification that belongs here: no valuation was published under version 1. The site operates in "start without an analyst" mode — without fair values and without price targets. No published assessment rests on the old bands.

Sources

• Cost Estimate Classification System — As Applied in Engineering, Procurement, and Construction for the Mining and Mineral Processing Industries, AACE International Recommended Practice No. 47R-11 — web.aacei.org
• CIM Definition Standards for Mineral Resources and Mineral Reserves, Canadian Institute of Mining, Metallurgy and Petroleum, 10 May 2014 — bcsc.bc.ca
• Independent Valuation Report, Ernst & Young LLP, filed with the U.S. Securities and Exchange Commission, 2019 — discount rate, WACC, surveyed NAV multiples and the weighting of resource categories — sec.gov
• Bertisen, J. / Davis, G. A.: Bias and Error in Mine Project Capital Cost Estimation, The Engineering Economist 53(2), 2008 — tandfonline.com

Version 2 · 11 August 2026