Amex Exploration’s Self-Funding Claim: Does the Arithmetic Hold?

Junior mining companies almost never pay for their own mine builds. The convention is dilution: issue shares, take on project debt, sign away future ounces through a royalty or streaming deal, and hand a slice of the upside to whoever writes the cheque. Amex Exploration is claiming it can skip all of that and fund CAD 193.9 million in Phase One capital at its Perron Gold Project without issuing a single dilutive share, using a 40,000-tonne bulk sample and pre-commercial production revenues that run ahead of the official production declaration.

This is worth scrutinising precisely because it is already in motion. The Phase One Feasibility Study is complete, underground development is underway under mining contractor CMAC-Thyssen, and the decline has reached roughly 100 metres of depth. This is not a theoretical financing model waiting for a market window. The arithmetic behind Amex Exploration self-funding its build either holds or it does not.

What follows here is deliberately practical. This piece walks through the numbers behind management’s self-funding thesis, identifies the structural decisions that make dual-phase cost allocation possible in the first place, and flags the specific conditions under which the model breaks down. After reading, you will know whether the thesis is arithmetically coherent, not just aspirationally appealing.

How the bulk sample and pre-commercial revenues are supposed to cover CAD 194 million

Start with the target the self-funding thesis has to reach.

The Phase One number to beat: CAD 193.9 million in initial capital. Everything below has to add up to this or the company needs outside money.

The thesis assembles that figure from three separate revenue and cost-allocation components, and following them in sequence is the only way to see where the assumptions carry the most weight.

The first component is a cost reclassification. The full bulk sample programme is expected to cost around CAD 60 million, but roughly CAD 40 million of that qualifies directly as Phase One capital expenditure. The reason is structural: the infrastructure built for the bulk sample is engineered to Phase One specifications, so a single spend counts against both programmes. That CAD 40 million is capital Amex would have had to spend anyway.

Bulk sample programmes serve a dual function that conventional exploration drilling cannot replicate: they generate grade and metallurgical data at meaningful tonnage scale while simultaneously producing saleable material, and the ratio between those two outputs is what determines whether a bulk sample pays for itself or requires external funding support.

The second component is bulk sample revenue. The programme is expected to recover between 20,000 and 28,000 ounces, with 25,000 ounces used as the planning figure. At a planning gold price of around USD 4,000 per ounce and a 1.35 CAD/USD exchange rate, that produces roughly CAD 135 million in gross proceeds before operating costs, treatment, refining, and royalties. That last clause matters, and we will return to it.

The third component is pre-commercial Phase One production. The Feasibility Study models CAD 68.1 million in pre-production revenue as a direct capital offset, generated from toll-milled ore during the pre-production window, based on an earlier planning assumption of USD 3,500 per ounce.

Add the three together and the picture becomes clear.

On paper, the offsets clear the target. That headroom is what makes the thesis more than marketing language at current gold prices. The open question is how much of it survives contact with reality: the bulk sample revenue line is gross, not net, and the recovered-grade assumption underneath the 25,000 ounces is the single most consequential number in the entire chain.

Why the infrastructure decisions are what make this arithmetic possible

The more useful question is not whether Amex can afford this. It is why this cost structure is available to the company at all, when it is not available to most of its peers.

The answer is that the dual-phase allocation only works because Amex designed the bulk sample infrastructure to Phase One specifications from the outset. The same physical spend serves both programmes simultaneously rather than sequentially, which is what lets CAD 40 million of bulk sample cost be reclassified as Phase One capital without anyone stretching the definition.

The power grid decision is the clearest illustration. Rather than run the site on diesel generation, Amex invested CAD 7.7 million in a grid connection locking in hydroelectric power at CAD 0.055 per kilowatt-hour. That single spend satisfies both bulk sample and Phase One power requirements. There is no future upgrade to budget for.

The same logic runs through the rest of the build:

• Power supply: The CAD 7.7 million grid connection is sized for Phase One demand, not just bulk sample throughput, so the hydroelectric rate is locked in for both.
• Water treatment: The plant under construction is built to Phase One capacity, avoiding a later expansion.
• Underground development: Portals, ramps, headings, and stopes are built to Phase One permit specifications, eliminating subsequent upgrades. Development is being delivered by CMAC-Thyssen, chosen for its Abitibi regional experience, and has reached roughly 100 metres of depth.
• Toll milling: Outsourcing processing removes the largest single capital item of all.

The signal here is worth sitting with. This is not a financing story constructed after the fact to explain a funding gap. The cost allocation is grounded in physical infrastructure decisions made before production started, which is a meaningfully different indication of planning discipline than investors usually get from a junior development story.

The Lamaque toll-milling arrangement and what the LOI gap means

The toll-milling decision deserves separate treatment because it dwarfs the other items in scale.

A toll-milling arrangement means Amex ships its ore to a third-party plant, pays a per-tonne processing fee, and receives gold doré in return, without building any processing infrastructure of its own. Amex has signed a Letter of Intent with Eldorado Gold for processing at the Lamaque facility in Val-d’Or, and has already submitted a regulatory project notice (an avis de projet in Quebec) designating Lamaque as the destination. Phase One is modelled as a contract mining and toll-milling operation running 1,100 tonnes per day.

By outsourcing processing, Amex avoids building a mill and a tailings management facility. Those two items alone would dwarf the current CAD 193.9 million capital requirement, which is precisely why the self-funding model is viable without them.

The caveat is that this remains an LOI, not a binding agreement. That distinction carries real execution risk, addressed in full below. The partial cushion is that at least four to five alternative mills in the Abitibi region actively seek feed material, so Lamaque falling through would not leave Amex without a processing route.

What makes this model viable for Amex when it fails for most juniors

Most advanced junior developers reach for the same three financing tools: equity issuances, project debt, and royalty or streaming packages. Each works, and each extracts a price. Equity dilutes existing shareholders. Debt requires bankable reserves and adds fixed obligations. Streaming deals hand away future ounces at a discount. All three transfer option value away from current holders, which is exactly the outcome Amex is trying to avoid.

Conventional junior mining financing strategies each extract a measurable price from existing shareholders: equity dilutes, debt adds fixed obligations, and streaming deals surrender future ounces at a discount, which is precisely why Amex’s cost-allocation approach is structurally unusual rather than merely aspirational.

The self-funding alternative is coherent here for reasons specific to the Champagne Zone, and those reasons do not travel well to other projects.

What 12 g/t means: The Champagne Zone hosts a defined resource of roughly 74,750 tonnes averaging just over 12 grams of gold per tonne. In the Abitibi, where many operating mines process ore in the low single digits, that is an exceptionally high-grade feed, and grade is what makes selling a small tonnage worth CAD 135 million in the first place.

Four conditions have to hold together for the model to work, and they rarely coincide:

• Grade: At just over 12 g/t, a small tonnage generates outsized revenue.
• Metallurgy: The ore is free gold in quartz with no deleterious elements, making it clean, saleable feed that mills want.
• Toll-milling availability: Multiple Abitibi facilities can process the ore, so no mill needs building.
• Gold price environment: Sustained elevated prices widen the arithmetic buffer.

There is also a regulatory reason most juniors never attempt this. NI 43-101, the Canadian standard governing how mineral project disclosures are presented, limits how pre-commercial and bulk-sample economics can be communicated to investors. Building an equity narrative around pre-production revenue is difficult when the rules constrain how that revenue can be framed, so most companies default to the conventional financing path.

The gold price environment then does the rest of the work. The Feasibility Study base case uses US$3,500 per ounce, while September 2026 spot sits near elevated levels. That gap turns a tight funding model into one with genuine headroom. The Feasibility Study also reports a post-tax IRR of 114.6% and a payback of roughly 0.5 years from commercial production start.

If any residual gap remains, management has stated a preference for forward gold sales over equity, illustratively pre-selling around 10,000 ounces in 2027 at USD 4,000-plus per ounce for roughly USD 50 million, with delivery due around 2029. The read for you is straightforward: the grade and metallurgy are the preconditions for everything else, so any deterioration in recovered grade is the variable most likely to unwind the structure.

Gold offtake agreements and forward sales occupy a similar structural position in project financing: both convert future production into present-day capital certainty without the dilution of equity, though they differ materially in how counterparty credit risk and delivery obligations are structured.

Where the thesis could break down: four execution risks worth pricing in

The arithmetic is coherent under current assumptions. That is not the same as saying it is safe. Four specific mechanisms could invalidate it, and ordering them by how far they sit outside management’s control is more useful than a generic list of caveats.

Execution risk in underground mining compounds across contractor performance, ground conditions, permitting sequencing, and processing availability, and the interaction between these variables is precisely what makes the Lamaque LOI the most time-sensitive milestone in Amex’s funding chain.

1. Milling agreement dependency (least controllable). The Lamaque arrangement is still an LOI, not a binding contract, as of late September 2026. Any delay, pricing revision, or counterparty change would directly threaten the funding timeline, because pre-commercial revenue depends on ore actually being processed. The four to five alternative Abitibi mills provide partial protection, but a switch would still cost time.
2. Gold price sensitivity (partly controllable via forward sales). The headroom in the model is a function of elevated prices. At the US$3,500 base case, the buffer compresses sharply and a funding gap becomes plausible.
3. Permitting sequencing (partly controllable). The bulk sample is fully permitted, but commercial production requires further authorisations. Commercial production is defined as sustained output of 660 tonnes per day for three months, and any delay in securing the relevant permits could strand Amex in the pre-commercial phase without a clean ramp-up path.
4. Grade variability (most within management’s operational control). High-grade vein systems are prone to the nugget effect, where gold is distributed unevenly and a 40,000-tonne sample may not capture the full structural complexity. Pretium’s Brucejack bulk sample is the cautionary precedent: early grade optimism in a high-grade vein system was later revised downward. If recovered grade comes in below the roughly 12 g/t average, the entire revenue chain weakens.

The milling agreement is the risk that sits most squarely outside Amex’s hands, which is why converting the LOI into a binding contract is the single most time-sensitive milestone to track before committing capital.

Gold price sensitivity and the point at which external capital becomes necessary

The price question deserves a scenario view rather than a single number.

At elevated spot prices, the offsets clear the CAD 193.9 million target with meaningful headroom. At USD 3,000, the model tightens considerably and the margin for error on grade and processing costs narrows. At the US$3,500 feasibility base case, the residual gap would likely require forward sales or a small debt facility to bridge.

That is where the timing of any pre-selling decision becomes material to you. Locking in forward sales at today’s elevated prices could secure the economics even if spot falls later, so the question is not only whether Amex hedges but when.

Whether the self-funding thesis holds up

Under current gold prices, the arithmetic works. The CAD 40 million in applicable capex, the bulk sample revenue, and the CAD 68.1 million in pre-commercial Phase One revenue combine to roughly CAD 230 million against a CAD 193.9 million target, leaving genuine headroom rather than a knife-edge balance. The self-funding thesis is arithmetically coherent, not merely aspirational.

Two variables decide whether it stays that way. The most consequential internal one is recovered grade from the Champagne Zone, because the roughly 12 g/t average is what makes a small tonnage worth so much. The most consequential external one is the conversion of the Lamaque LOI into a binding milling agreement, because pre-commercial revenue cannot flow without a confirmed processing route.

For an investor, the monitoring framework reduces to three trackable conditions:

• Bulk sample recovered grades track the roughly 12 g/t average.
• The Lamaque LOI converts into a binding, definitive milling agreement.
• Gold prices remain materially above the feasibility base case.

None of this removes the ordinary reality of the project. Phase One is still a five-year, 1,100-tonne-per-day contract mining and toll-milling operation, and commercial production still means sustaining 660 tonnes per day for three months. Underground execution risk is inherent regardless of how the capital is raised.

The useful analytical question is not binary. It is which of the four execution risks materialises first, and at what point the remaining upside justifies the residual uncertainty. The conditions that make this model rare (grade, metallurgy, infrastructure, and price) are real, but they have to hold simultaneously, and that is what merits genuine scrutiny rather than reflexive scepticism.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results. Financial projections are subject to market conditions and various risk factors, and forward-looking statements are speculative and subject to change based on market developments and company performance.