If you designed a laboratory experiment to make gold go up, it would look like the last four weeks. The largest Middle East military campaign since the Gulf War. A superpower strike that killed Iran’s Supreme Leader. The Strait of Hormuz closed, pulling roughly 20 million barrels a day of crude off the market. Oil posting its biggest weekly gain on record and Brent touching ~$120. Inflation fear everywhere, consumer confidence wobbling, and equities down 5%+ on the month.
And gold—the asset the entire wealth-management industry sells as the thing you own for exactly this scenario—is down double digits from its war-day peak.
The mechanism matters more than the outcome. This isn’t a “gold is dead” post; it’s a look at what gold actually hedges, which turns out to be something much narrower than what most portfolios assume.
What Happened
The reflex worked for about a day. When the U.S.-Israeli strikes hit on February 28, gold spiked to an intraday high in the ~$5,400 area—a textbook flight-to-safety print. Then the textbook stopped cooperating. Within days it gave back the spike, dropping roughly 4% in a single session in early March. By mid-month it sat around $5,100–5,200, up only ~3% month-over-month while jet fuel surged triple digits. This week it broke below $4,700 and traded down toward $4,400 before bouncing on news that the White House extended its deadline for a deal—an 18% drawdown from the war-day high, during the war it was supposed to hedge.
Flows mirrored the price action. North American funds sold nearly 50 tonnes of gold-backed ETFs into the conflict, CME volumes thinned, and GLD lost roughly 15% over the month while the dollar index gained. The haven asset was distributed, not accumulated, in the middle of a shooting war.
The Mechanism: Gold Is a Real-Rates Asset Wearing a Fear Costume
The framework explaining this is straightforward: gold doesn’t hedge war; gold hedges falling real yields and a weakening dollar. Those conditions often accompany crises—which explains the “war hedge” reputation—but they are not identical, and this crisis pits them in direct opposition.
Follow the chain reaction. Hormuz closes, oil spikes, inflation expectations jump, and central banks turn hawkish. Traders pushed the first Fed cut out to September and started whispering about hikes. Nominal yields rose. The dollar rallied hard, as it typically does in phase one of an energy crisis because the U.S. is energy-rich and the rest of the world needs dollars to pay for $110 crude. The two variables gold actually trades on—real rates and the dollar—both moved against it, while the fear variable moved for it. Fear lost, as it almost always does over any horizon longer than a day.
The 1970s analogy everyone reaches for is misleading on timing. Gold’s great stagflation run came after the initial oil shocks fed through, when real rates went deeply negative because central banks fell behind the curve. In the opening act—supply shock, hawkish repricing, dollar squeeze—the fear premium gets overwhelmed. We are in the opening act.
The Positioning Problem Nobody Mentions
There is a second, more critical driver: gold came into this war up roughly 80% year-over-year. The insurance had already been bought. Two years of central-bank accumulation, de-dollarization narratives, and momentum flows meant that by February 27, an enormous amount of geopolitical anxiety was already priced in above $5,000.
When the actual event arrived, there was no marginal buyer left who wasn’t already long—but there was a large cohort of holders sitting on 80% gains with a fresh reason to fear higher rates. The event didn’t create demand; it created a liquidity moment to monetize the hedge. That is precisely what the 50 tonnes of ETF outflows represent. “Sell the invasion” is a cousin of “sell the news,” happening right when a hedge is overcrowded. A hedge that has already rallied 80% is mostly not a hedge anymore; its convexity has been spent.
What Actually Hedged This
Empirically, the winners in March were clear: energy equities, crude itself, the U.S. dollar, defense primes, and—quietly—long volatility positions established before the spike. The common thread is that they hedge the transmission mechanism (supply shock → inflation → hawkish repricing) rather than the headline (war = fear = gold).
When you buy a hedge, you are implicitly forecasting a transmission channel, not just an event. Get the channel wrong and you can nail the event yet still lose money.
Where It Lands
To be clear, the structural bull case for gold—central-bank buying, fiscal trajectories, and eventual real-rate relief when the inflation impulse forces demand destruction—remains intact. If this conflict drags on and growth cracks while the Fed is pinned by oil, stagflation act two could be very kind to gold, matching the mid-70s playbook.
The narrower takeaway, demonstrated by one of the cleanest laboratory experiments markets ever run, is that gold is a regime hedge, not an event hedge. It insures against negative real rates and dollar debasement over quarters, not a Tuesday when missiles fly. In a supply-shock war that forces central banks to turn hawkish, it can trade actively anti-correlated with the exact risk you bought it to offset.