An ounce isn't an ounce.

A junior drops a maiden resource. Two million ounces. The headline gets screenshotted, the stock moves, the room celebrates.

Almost nobody opens the technical report.

That number is the output of a stack of assumptions — and the assumptions are where you find out what those ounces are actually worth. Here's how I read a resource statement.

First question: which category?

Resources come in three confidence levels. Measured is the highest. Indicated is next. Inferred is the lowest — estimated from limited drilling, with continuity implied rather than confirmed.

That's not a technicality. Inferred ounces can't be converted into reserves, and they can't be used in the economics of a pre-feasibility or feasibility study. The market typically pays far less for them.

2 million inferred ounces and 2 million indicated ounces are not the same asset. The split is the first thing I look for.

The cut-off grade is a dial

Every resource is reported above a cut-off grade. Material below it doesn't count.

That cut-off is a choice. Lower it and tonnes go up, ounces go up, average grade goes down. Same deposit, bigger headline.

This is where the most useful table in the report hides. Many resource statements include a sensitivity table showing the ounces at several different cut-offs.

→ Ounces drop sharply when the cut-off rises a little? Most of them sit in low-grade material.

→ Ounces hold up? There's a genuinely high-grade core.

That one table tells you more about the deposit than the headline ever will.

Find the price in the footnotes

To count as a resource, material needs "reasonable prospects for eventual economic extraction." In practice, that usually means a pit shell or mineable shapes built on an assumed metal price and recovery.

Higher assumed price → bigger shell → more ounces qualify.

The footnotes tell you what price they used. The question isn't whether the deposit works at today's price. It's whether it still works at a conservative long-term one.

Equivalent grades hide the recipe

"3 g/t AuEq" sounds like a gold deposit. It might be mostly silver, zinc, or copper converted into gold terms using assumed prices and recoveries.

Companies have to disclose the individual metal grades behind an equivalent. Go find them.

Then check the recoveries. An equivalent that assumes every metal comes out of the plant in full flatters the grade — and no plant recovers everything.

Where does the study get its ounces?

A PEA can include inferred resources. A pre-feasibility or feasibility study can't. That's why PEAs using inferred carry a mandatory warning: those resources are too speculative geologically to have economics applied the way reserves do.

So when a PEA shows a great IRR, check how much of the mine plan is inferred. Sometimes it's a small slice. Sometimes it's most of it.

And inferred ounces aren't free. Each one waiting to be upgraded is a drill program — and every drill program is a financing. A big inferred number is, quietly, a forecast of future dilution.

Phrases that should slow you down

→ "Historical estimate." Made before the company owned the property. A QP hasn't done the work to classify it as current, and the company can't treat it as current. Context, not a resource.

→ "Exploration target." A conceptual range of tonnes and grade. Not a resource. The disclosure has to say so — read that sentence.

→ "Billions of dollars in the ground." Canadian rules bar companies from disclosing the gross in-situ value of a deposit. If you're seeing that number, it isn't coming from a compliant company disclosure — and it ignores every dollar it costs to get the metal out.

The ounces are the headline. The assumptions are the story.

Before you react to a resource number, open the footnotes and find four things: the category split, the cut-off, the metal price, the recoveries.

Five minutes. They'll tell you more than the press release ever will.

Last week: "Ownership trumps results," on reading financings.