Peak Gold in 2026: Why Record Prices Haven't Produced More Gold

Here is a fact that sits oddly with everything else happening in the gold market: gold prices have roughly doubled in two years, and mine production has barely moved.

In any normal commodity, that would be impossible. High prices call forth supply, supply meets demand, prices settle. Gold does not work that way, and understanding why explains more about the long-term case for owning it than any price forecast.

The numbers

Global mine production reached a record 3,672 tonnes in 2025 — but grew only about 1% year on year, despite average prices above $3,400 an ounce. That is the whole story in one sentence: record output, negligible growth, record incentive.

The World Gold Council's assessment, published in January 2026, is that mined output is likely to peak around 2027 and then gradually plateau rather than fall off a cliff. That is a more measured view than the "peak gold" headlines suggest, and it is worth being precise about, because the difference between a plateau and a collapse matters.

Where it comes from:

Country	Approx. annual production
China	~380 tonnes
Russia	~330 tonnes
Australia	~284 tonnes

No single country dominates the way, say, South Africa once did or the way South Africa still dominates platinum. Gold supply is unusually distributed, which is one reason it has never had a genuine supply shock.

Why supply can't respond to price

Four constraints, all structural, none of which a higher price fixes quickly.

1. Lead times are measured in decades

From first discovery to first pour, a large gold mine routinely takes 10 to 20 years — exploration, resource definition, feasibility studies, permitting, financing, construction. A price signal in 2026 produces metal in the late 2030s. The gold reaching the market this year was committed to when gold traded at a fraction of today's price.

2. Discovery rates are falling

Exploration spending has risen; major new discoveries have not kept pace. The large, high-grade, near-surface deposits were found first, because they were the easiest to find. What remains is deeper, lower grade, more remote, or in jurisdictions where capital is reluctant to go.

3. Grades are declining

Average ore grades have fallen steadily for decades. Lower grade means moving more rock, using more energy and more water, for the same ounce. That raises costs and it raises the environmental and permitting burden simultaneously.

4. Capital discipline

Gold miners spent the 2010s being punished by shareholders for value-destroying expansion at the top of the last cycle. The lesson stuck. Much of the current windfall is going to dividends, buybacks and balance-sheet repair rather than aggressive new development — which is rational for the companies and constraining for supply.

The part that makes gold genuinely unusual

Mine supply adds roughly 1.5% to the above-ground gold stock each year. Almost nothing is consumed. Nearly every ounce ever mined still exists.

All the gold ever mined comes to somewhere around 220,000 tonnes. At $4,349.70 an ounce on our live feed as this was written, that is a total stock worth roughly $31 trillion — and it would form a cube about 22 metres on a side.

The consequence: gold's supply growth rate is both low and extremely stable. It has hovered around 1.5–2% annually for a very long time, through booms and busts. That stability, not scarcity as such, is the actual monetary property. Gold is not valuable because there is little of it — it is valuable because the amount of it changes slowly and predictably, in a way no currency's supply does.

Compare that with silver, which is substantially consumed by industry and where above-ground stocks genuinely deplete. The two metals have opposite supply dynamics, which is one reason they behave differently.

What this does and doesn't mean

Being honest about the limits of this argument matters, because "peak gold" is routinely oversold.

What it does mean: new supply cannot surge to cap a price rally. In a demand shock — central banks buying, investor panic, a currency crisis — the supply side simply cannot respond within the relevant timeframe. That asymmetry is real and it is durable.

What it does not mean:

• Gold is not running out. A plateau at record levels is not scarcity. There are decades of reserves and the WGC explicitly expects a gradual plateau, not a drop.
• Recycling is the shock absorber. High prices pull scrap gold out of drawers and jewellery boxes. Recycled supply is materially price-elastic even when mine supply is not, and it fills part of any gap.
• Above-ground stock dwarfs annual supply. With ~220,000 tonnes in existence and ~3,700 tonnes added yearly, the marginal seller matters far more than the marginal miner. Price is set by what existing holders will part with, not by what comes out of the ground.
• Supply constraints don't set a price floor. Gold fell for years after 2011 while mine supply was equally constrained.

Anyone using peak-gold as a reason gold "must" reach a given number is skipping the third and fourth points.

Miners versus metal

A reasonable question follows: if supply is constrained and prices are high, why not buy the miners instead?

It is a legitimate strategy with a different risk profile. Miners give you operating leverage to the gold price, plus dividends — and also execution risk, jurisdictional risk, cost inflation, labour disputes, and the possibility that a company destroys value even in a bull market for its product. Constrained supply is good for the metal price; it is not automatically good for the companies struggling against those same constraints.

We compare the two in gold mining ETFs vs physical gold. The short version: miners are an equity bet on gold, physical is a position in gold.

What this means for a buyer

The supply picture is a long-horizon argument, not a trading signal. Practically:

1. It supports holding rather than timing. The structural case builds over years, not weeks.
2. Format matters more than the macro. Over a decade, the premium you pay on entry is a cost you control; the gold price is not. Larger bars carry the lowest premium per ounce — a kilo bar spreads fabrication cost across 32.15 ounces.
3. Accreditation protects liquidity. LBMA Good Delivery refiners trade without re-assay anywhere in the world. See the refiner comparison.
4. Divisibility has a price. 1 oz coins cost more per ounce and give you thirty-two exit points instead of one.

We price from live spot at checkout, accept Bitcoin, Ethereum, Monero, USDT, USDC and around thirty other coins, and ship insured in unmarked packaging.

The short version

Mine production hit a record 3,672 tonnes in 2025 and grew about 1%. Output is expected to plateau around 2027. Lead times of 10–20 years, falling discovery rates, declining grades and capital discipline mean the supply side cannot respond to price on any useful timescale.

That does not mean gold is running out, and it does not set a floor under the price. What it means is that gold's supply grows slowly, predictably, and independently of demand — which is precisely the property that made it money in the first place.

General information, not investment advice. Production figures are estimates and are revised.