Why gold fell during a war that should have lifted it

Why gold fell during a major war looks like a contradiction. Gold is supposed to benefit when geopolitical risk rises, yet by June 26 the benchmark used in a midyear drawdown study was about 25% below its January record. The safe-haven case did not disappear. A faster chain of events took control of the price: the conflict restricted oil supply, energy costs lifted inflation fears, markets pushed expected US interest rates higher, and the dollar strengthened. Those moves raised the cost of holding a non-yielding asset. Gold then faced a second problem. The trade was crowded after a long rally, so a normal macro reversal turned into a much larger liquidation.

The war pushed yields higher instead of sending money into gold

A conventional geopolitical shock helps gold when investors become more worried about growth, financial stability, or the safety of other assets. They move toward cash, government bonds, and gold. Falling bond yields can reinforce the move: they shrink the income investors give up by holding gold instead of bonds.

The Iran war produced a different sequence. Shipping disruptions through the Strait of Hormuz restricted energy supply. The US Energy Information Administration reported that Brent crude futures rose from $61 a barrel at the start of 2026 to $118 at the end of the first quarter, the largest inflation-adjusted quarterly increase in its data since 1988.

That changed the question investors were asking. The early question was whether the conflict would create enough fear to increase safe-haven demand. The later question was whether expensive energy would keep inflation high enough to prevent rate cuts or force tighter policy.

The Federal Reserve's July Monetary Policy Report confirms much of the chain. Energy prices surged after the conflict began. Two-year nominal Treasury yields rose about 60 basis points from the start of the year, and the largest yield increases came at shorter maturities as markets priced a higher path for the federal funds rate. The trade-weighted dollar also appreciated.

For gold, that combination is difficult. An April market review also found a conflict-led rise in US real yields. Higher real yields improve the return available on inflation-protected dollar assets. A stronger dollar makes the dollar gold price more expensive for many non-US buyers. The result was a conflict that lifted inflation risk while also lifting the opportunity cost of gold. Our guide to interest rates and gold explains why the expected path of real yields matters more than a single policy decision.

Higher yields triggered the turn, and positioning made it violent

A June 24 research note called the repricing of interest-rate expectations the primary driver of the decline. Higher Treasury yields supported the dollar, while weaker ETF demand removed a source of buying that had helped drive the earlier rally.

Rates alone do not explain the speed or depth of the move. Gold entered the year after a powerful rally, with large exposures held by ETF investors, retail buyers, futures traders, and systematic funds. A crowded position can remain profitable while new buyers keep arriving. It becomes fragile when the macro reason for owning it changes and several holders try to reduce risk at once.

The April market review found that global gold ETFs lost about 80 tonnes between the start of March and March 24. It also identified estimated selling by commodity trading advisers after gold broke its 50- and 55-day moving averages, and said broader cross-asset deleveraging likely spilled into gold. Investors facing losses or margin calls elsewhere may have sold liquid gold positions to raise cash and reduce portfolio risk.

That distinction resolves the apparent disagreement in the research. Rate repricing explains why the market turned. Positioning and liquidity explain why the fall became larger than the change in yields or the dollar would normally imply. The rate shock was the trigger; the crowded trade was the amplifier.

May brought a related shift. A global fund-flow review found that higher rates and dollar strength had raised gold's opportunity cost, while investors rotated toward technology and other risk assets. Gold was therefore hit from both sides: some holders were forced to sell, and some discretionary investors simply found other assets more attractive.

Central-bank demand was too slow to stop fast selling

Central banks and official institutions were net buyers in the first quarter. A Q1 2026 demand report put net official-sector buying at about 244 tonnes, despite a pickup in selling by some central banks during the quarter. The net additions help explain why the longer-term gold thesis survived even as the market corrected.

But structural support is not the same as short-term price control. Central banks usually buy through patient programs executed over months or years. ETF redemptions, futures selling, and margin-driven liquidation can hit the market within hours. The daily price is set by the marginal buyer and seller at that moment, not by the largest category of holders over a full year.

This also clarifies what people mean when they call central-bank demand a floor. It can absorb supply and limit a decline without preventing one. If fast investment flows turn negative, the price can fall until slower physical and official demand becomes large enough to meet them. How central-bank buying moves the gold market covers that difference between a structural bid and a price spike.

Gold did not stop being a safe haven

Gold is a safe-haven asset, but it is not a contractual hedge that must rise whenever war begins. Its price rises when new demand exceeds the metal being sold. In a liquidity shock, gold's strengths can work against its price for a while: it trades around the clock, has deep liquidity, and can be sold quickly when investors need cash.

The kind of crisis matters too. A growth scare that lowers real yields can support gold. A supply shock that lifts inflation, yields, and the dollar can hurt it first. If inflation later damages confidence in monetary policy or currencies, gold may benefit, but that is a different part of the cycle. Gold's record as an inflation hedge is strongest over long periods and inconsistent over short ones.

The size of the 2026 correction is severe but not outside gold's history. A study of drawdowns since 1971 counted eight previous declines of more than 20% from a record high. Their average depth was 36% and their median was 29%. That history does not identify a bottom or predict a recovery. It shows that a large correction can occur without disproving gold's longer-term monetary role.

The evidence points to rates as the trigger

The strongest synthesis of the cited evidence is a chain, not a single headline. War disrupted oil supply. Higher energy prices lifted inflation concerns. Markets priced a higher path for US rates, pushing yields and the dollar upward. That change broke the logic of a crowded trade, and ETF outflows, systematic selling, and likely cross-asset deleveraging made the decline much larger. Net central-bank buying remained a slower source of support. Watch real yields, the dollar, and investment flows alongside the live XAU/USD price; the war headline alone cannot explain the move.