What Drives the Gold Price? Real Rates, the Dollar, and Fear

Gold earns no interest. It pays no dividend, carries no coupon, and does nothing but sit in a vault costing money to store. By the logic that prices most financial assets, it should be the dullest thing in the market. And yet, in 2025, the gold price set 53 separate all-time highs, and by early 2026 an ounce had briefly traded near $5,600 (World Gold Council, Gold Demand Trends: Full Year 2025; market data, January 2026). Something is clearly moving it — the question is what.

The honest answer is that no single number explains the gold price, and anyone who claims a clean formula is selling something. But three forces do most of the work, and once you understand how each one pushes and pulls, the moves stop looking mysterious. They are the real return on cash, the value of the US dollar, and the price the world is willing to pay for safety. This article takes each in turn — and then looks at why, in recent years, the textbook relationships have started to bend.

The Textbook Answer — and Why It’s Worth Questioning

Ask a trained analyst what moves gold and you will usually hear the same short list: real interest rates, the dollar, and risk sentiment. It is a good list. For most of the last two decades it did a remarkably good job of explaining where gold went and why.

But a framework is only useful if you know why each piece belongs on it — and if you notice when it stops fitting the facts. In our companion pillar, Gold and Precious Metals: Why They Still Matter, we introduced these three drivers in a paragraph each. Here we open the engine and look at the mechanism inside each one, because the mechanism is what tells you when the rule will hold and when it will break.

Real Interest Rates: Gold’s Center of Gravity

Start with the driver that sits underneath the other two.

Gold’s defining feature is that it pays you nothing to hold it. A dollar in a bank account earns interest; a Treasury bond pays a coupon; even a boring money-market fund throws off a yield. Gold just sits there. That means every ounce of gold you hold carries an opportunity cost — the return you gave up by not holding something that pays. Economists call this the “carry” trade-off, and it is the closest thing the gold market has to a law of gravity.

The number that matters is not the headline interest rate but the real interest rate — the rate after subtracting inflation. If a one-year Treasury pays 4% and inflation is running at 4%, your real return is zero: your money grows in dollars but buys no more than before. In that world, gold’s lack of a yield costs you nothing, because cash isn’t really paying you either. But if that same bond pays 4% while inflation is only 1%, cash now delivers a genuine 3% real return — and holding gold instead means giving up something real. The higher the real return on safe assets, the more expensive it is to sit in gold.

The market’s cleanest gauge of the real interest rate is the yield on Treasury Inflation-Protected Securities, or TIPS — US government bonds whose value adjusts with inflation, so their quoted yield is a real yield stripped of inflation expectations. For roughly two decades, gold and the real yield on TIPS moved like two ends of a seesaw. From 2003 to 2022, their rolling correlation averaged around −0.7 (goldsilver.com; RBC Wealth Management) — a strong inverse link, meaning gold tended to rise when real yields fell, and slip when they climbed. When central banks cut rates toward zero after 2008, and again in 2020, real yields went deeply negative and gold ran to successive records. The seesaw worked.

This is why real rates are the center of gravity: they set the baseline cost of choosing gold over cash. Hold that idea, because it is also the rule that recently broke — and understanding the break requires first understanding the rule.

The Dollar: Gold’s Other Side of the Coin

The second force is more mechanical, and it starts with a simple fact of plumbing. Gold is quoted in US dollars per troy ounce on every major market in the world. That single convention ties the metal’s price to the value of the dollar itself.

Picture a buyer in Europe or India who wants an ounce of gold. They don’t think in dollars — they think in euros or rupees. To buy gold, they must effectively convert their own currency into dollars first. So when the dollar strengthens against other currencies, that same ounce suddenly costs more in euros or rupees, even if nothing about gold has changed. Higher local prices cool demand from the largest non-US buyers, and that softer demand tends to pull the dollar price of gold down. When the dollar weakens, the reverse happens: gold gets cheaper abroad, demand firms, and the dollar price tends to rise. Gold, in this sense, is partly a bet against the dollar — when the dollar loses value, it takes more dollars to buy the same real thing, and an ounce of gold is a very real thing.

There is a second channel too, and it loops back to real rates. A strong dollar and high US interest rates often travel together, because money flows toward the currency that pays the best real return. So a firm dollar frequently coincides with exactly the high-real-yield conditions that raise gold’s opportunity cost. The two forces reinforce each other, which is why the historical gold–dollar correlation has been even tighter than the gold–rates one, running around −0.8 (CME Group; goldsilver.com).

The relationship is a tendency, not a wire. Some days gold and the dollar rise together — usually when fear is the dominant force and investors are buying both as havens at once. But as a background current, the rule holds: gold and the dollar sit on opposite sides of the same coin. For a fuller picture of what pushes the dollar around in the first place, see What Moves Currency Prices? Rates, Risk, and Macro.

Fear: The Asset With No Counterparty

The third force is the hardest to measure and the easiest to feel. Gold is where money goes when trust is the thing being repriced.

To see why, it helps to notice what gold isn’t. A bank deposit is a promise from a bank. A bond is a promise from a government or a company. A banknote is a liability of a central bank. Every one of these is somebody else’s IOU, and every IOU carries the risk that the other side fails to pay — what markets call counterparty risk. Gold is different: a bar of gold is nobody’s promise. It has no issuer who can default, no government that can inflate it away by printing more, and no board that can dilute it. It is a bearer asset — value that belongs to whoever holds it, backed by nothing but itself.

That is exactly the property investors want when they stop trusting the promises. In the 2008 banking crisis, when the question was whether financial institutions themselves would survive, gold drew a flood of buyers. It does the same during inflation scares, when the concern is that paper money is quietly losing value, and during geopolitical shocks, when the worry is that assets tied to a particular country might be frozen or seized. In each case the appeal is identical: gold is the asset you own outright, answerable to no one. We traced why metal earned this role across five thousand years in A Brief History of Money: From Cowrie Shells to Crypto and What is Money? — the short version is that gold’s independence from any issuer is not a modern discovery but its oldest job.

Fear demand is spiky and unpredictable — it arrives with crises and fades with them — which is why it explains gold’s sharpest moves but not its long trends. For the slow, structural story, we have to look at who has been buying gold not in a panic, but on purpose, year after year.

The Regime Change: When Central Banks Became the Marginal Buyer

Here is the puzzle that breaks the textbook. From 2022 through 2025, US real interest rates were high and the dollar was firm — the exact combination that is supposed to weigh gold down. Under the old rules, gold should have drifted lower. Instead it did the opposite, climbing to record after record. Something outside the classic framework was clearly at work.

The measurements show it plainly. The gold–real-yield relationship that had averaged around −0.7 for two decades essentially dissolved: through 2022 and 2023, the correlation between gold and TIPS yields fell to the low single digits (RBC Wealth Management). The seesaw simply stopped seesawing.

What replaced it was a change in the marginal buyer — the buyer whose demand sets the price at the edge. For years, the swing buyer of gold was the investment crowd, moving in and out of gold-backed exchange-traded funds as real yields rose and fell. That is the channel through which the real-rate rule worked. But from 2022 onward, a different buyer stepped to the front: central banks. Official institutions bought more than 1,000 tonnes of gold in each of 2022, 2023, and 2024, and another 863 tonnes in 2025 — still far above the 2010–2021 average of roughly 470 tonnes a year, even after cooling from the peak (World Gold Council, Gold Demand Trends: Full Year 2025). Poland alone added 102 tonnes in 2025, lifting its reserves to 550 tonnes.

Central banks do not buy gold to earn a yield or to time real rates — they buy it as a neutral reserve asset that no other government controls. Two motives dominate.

The first is diversification away from the dollar, and China is the clearest case. The People’s Bank of China has added gold in every month of a buying streak running twenty-one consecutive months to July 2026, lifting official holdings to roughly 2,366 tonnes — the longest such run since 2015. Even after that accumulation, gold accounts for less than 10% of China’s total reserves, which is the usual explanation for why the buying has not stopped: relative to the size of its dollar holdings, China’s gold position is still small (World Gold Council; PBoC monthly reserve data, July 2026).

The second motive is sanctions insurance, and it has a specific origin. In February 2022, following Russia’s invasion of Ukraine, the G7 and allied states immobilized roughly $300 billion of Russian central-bank reserves — foreign currency and securities held inside Western financial institutions. For every reserve manager watching, the implication was uncomfortable and hard to unsee: reserves held as claims on another country’s banking system can be switched off by that country. Gold sitting in your own vaults cannot. It is nobody’s liability, and spending it requires nobody’s permission. This is the counterparty argument from the previous section, applied at national scale — and it is buying driven by strategy rather than by the opportunity-cost calculation, which is precisely why it overrode the usual rules.

The most recent data puts that argument under a live test — and largely supports it. Through the first half of 2026, the gold price came off its peak: after a record quarterly average of $4,873 an ounce in the first quarter, the second-quarter average fell about 8% to $4,506 (World Gold Council, Gold Demand Trends: Q2 2026, published 30 July 2026). Gold-backed ETFs did exactly what the opportunity-cost model says they should when prices soften and rate expectations firm: they went into outflow, shedding 45 tonnes over the quarter. The price-sensitive buyer sold.

Central banks did the opposite. They bought 289 tonnes in the same quarter — 62% more than a year earlier, and the strongest second quarter in the WGC’s series — while the price fell the whole way through. That is about as close to a natural experiment as this argument is likely to get: when the two kinds of buyer face the same conditions, they move in opposite directions, because they are not solving the same problem.

Two honest qualifications belong here. The first quarter of 2026 was weak for official buying — just 57 tonnes, after a downward revision to the WGC’s own data — so the half-year total of 345 tonnes is actually the softest first half since 2022. And the WGC’s own outlook expects central banks to have another strong year, but a lower one than 2025. Sovereign demand behaves like a floor, not an escalator: it is lumpy quarter to quarter, and the claim worth making is about who sets the price at the margin, not about a number that only goes up.

The useful way to hold this is not that the old framework is wrong, but that it is incomplete. A widely used summary among analysts is that real yields still set the pace and magnitude of gold’s swings, while sovereign demand now sets the floor beneath them. The seesaw is still bolted to the ground — but a new, heavy weight has been placed on one end, and it changes how the whole thing tilts.

What About Supply?

Notice what has barely come up: how much gold is mined. That is not an oversight. Supply is one of the least important short-term drivers of the gold price, and understanding why reveals something essential about what gold is.

Almost all the gold ever mined still exists. Because gold does not corrode or get consumed, roughly every ounce pulled from the ground across all of history is still sitting in a vault, a piece of jewelry, or a coin somewhere. That accumulated above-ground stock is enormous, and each year’s new mine production adds only about 2% to it. In 2025, mine output reached a record 3,672 tonnes — and even that record was a rounding adjustment to the total stock, not a flood (World Gold Council, Gold Demand Trends: Full Year 2025).

This is what makes gold behave like a monetary asset rather than an ordinary commodity. The price of wheat or oil is set at the margin by this year’s harvest or this year’s output, because those things get eaten and burned. Gold’s price is set by shifts in demand for the entire existing stock — by whether the holders of the world’s gold want to hold more or less of it. Mine supply is a slow, nearly fixed variable; demand is the fast one. Even recycling, the other source of supply, barely responds: despite a 67% rise in the dollar gold price during 2025, the amount of old gold sold back for scrap rose only 3%, because much of the world’s gold is held for reasons that a higher price does not shake loose.

Putting It Together

None of these forces acts alone, and that is the whole point. Real rates set the baseline cost of choosing gold over cash. The dollar amplifies or offsets that cost and adds its own mechanical pull. Fear can override both for weeks at a time. And underneath it all, a structural buyer — central banks accumulating a neutral reserve — has changed how much the first two matter.

The practical lesson is not a prediction but a way of reading the tape. When gold moves and the reason isn’t obvious, the questions to ask are the same four every time: what are real yields doing, which way is the dollar going, is there a fear event in the background, and is there a large strategic buyer in the market? Usually one or two of them explains the move. When none of them does, that is itself information — it may mean the market’s understanding of gold is shifting again, as it plainly did after 2022.

A necessary word of caution: understanding the drivers is not the same as forecasting the price. Banks and research houses publish gold price targets, and those targets vary widely — which is itself worth knowing, because a wide spread among well-resourced forecasters is a fair indication of how little consensus exists. This article names none of them and endorses none of them. Nothing here is a suggestion to buy, sell, or hold gold or anything else. Markets carry risk, the forces above interact in ways no model captures cleanly, and gold can fall as hard as it rises. This is a framework for understanding, not a signal.

The Bottom Line

Gold’s price looks irrational only if you expect it to behave like a stock or a bond. It isn’t one. It is the world’s oldest monetary asset — value with no issuer — and its price is really the market’s running answer to a single question: how much do you trust everything else?

When real yields are high, the dollar is strong, and confidence is calm, the answer is “quite a lot,” and gold has little to offer. When real returns on cash vanish, the dollar sags, fear spikes, or the institutions that manage the world’s reserves quietly decide they want more of the one asset nobody can freeze or print — the answer tilts the other way, and gold moves. The three classic drivers tell you how it responds. The recent regime change is a reminder that the market can rewrite its own rules, and that the deepest driver of the gold price has always been the same thing that made it money in the first place: it is what people reach for when they stop trusting the promises.

The Currency Stack provides educational and research content only. Nothing here is financial, investment, or trading advice, or a recommendation to buy or sell any asset. Markets carry risk; do your own research and consider professional advice before acting.

Further reading: This article deepens our pillar guide Gold and Precious Metals: Why They Still Matter, which covers why gold endures and how its market is built. For the same rate-and-risk forces seen from the currency side, read What Moves Currency Prices? Rates, Risk, and Macro. For how metal earned its role as the asset with no counterparty, see A Brief History of Money: From Cowrie Shells to Crypto and What is Money? Terms in bold are defined in our glossary. Sources: World Gold Council, “Gold Demand Trends: Q4 and Full Year 2025” for the 2025 demand, supply, central-bank, and price-average figures, and “Gold Demand Trends: Q2 2026” (published 30 July 2026) for the first-half 2026 figures and quarterly price averages; correlation history from RBC Wealth Management, CME Group, and goldsilver.com; PBoC monthly reserve data for Chinese official holdings; Brookings, the Congressional Research Service, and European Parliament research for the scale of immobilized Russian reserves. Figures verified 18 August 2026 and move daily.