
Gold at $4,186: hidden 244-tonne central-bank bid separates $6,000 bull case from a $3,500 bear risk. Traders watching the April 10 CPI print for the next trigger.
Gold settled at $4,186 an ounce Thursday, roughly flat on the week but down from the $4,350 high struck the day Kevin Warsh became the front-runner to replace Jerome Powell. The price action reads like a hawkish Fed succession broke the rally. The plumbing suggests something narrower: tactical position-squaring, not a structural unwind.
Central banks bought 1,100 tonnes of gold in 2024, roughly matching the prior year. The publicly reported number almost certainly understates the real flow. London Bullion Market Association settlement data shows a persistent gap between disclosed central-bank purchases and physical metal moving through London vaults. That gap has been running at about 244 tonnes this year – gold being bought but not yet reported in national reserve statements. Several emerging-Asian central banks do not report on a quarterly cycle, several bullion bank traders said.
The $6,000 bull case depends on that hidden bid persisting or accelerating. China's central bank added gold for 17 consecutive months through April, then paused in May. The pause was tactical, not a reversal, the same traders said. Reserve managers in Eastern Europe and the Middle East are still converting dollar reserves into gold at a pace that, if continued, would absorb nearly all new mine supply this year. Mine production runs at roughly 3,600 tonnes annually. Central-bank buying at 1,100 tonnes consumes about 30% of that, leaving thin buffer for the ETF inflows that typically drive the next leg. If the hidden 244-tonne gap is real, the physical squeeze tightens.
The $3,500 bear case starts from a different assumption: the hidden bid is a statistical artifact, and the real driver since October has been speculative positioning in futures. The net long in COMEX gold hit 280,000 contracts in early January, a level that has preceded violent corrections. The last time positioning was this stretched was August 2020, when gold was at $2,075 and fell 12% in three weeks. The Warsh nomination triggered a liquidation of about 45,000 contracts over three days, exchange data show. If that liquidation snowballs into a broader deleveraging – say, because a simultaneous dollar rally pushes the DXY above 108 – gold could test $3,500 with little support from the physical market.
Recession risk muddies both cases. A U.S. recession would weaken the dollar and lower real rates, textbook gold positives. It would also destroy industrial demand across metals and force liquidation across commodity positions as margin calls hit levered funds. Which force wins depends on the speed of the downturn. A slow, consumer-led recession would favor gold. A fast, credit-driven one, like 2008, would crush it first.
Gold miners have already started pricing in a lower trajectory. The NYSE Arca Gold Miners Index is down 8% from its February high, underperforming bullion by a wide margin. That divergence typically signals that equity markets expect the physical price to fall. It also means valuations have cooled, which in past cycles has marked an entry point for a fresh leg.
The next concrete date is April 10, when the U.S. March CPI report is released. A print at or above 3.2% would reinforce the higher-for-longer rate narrative and could trigger another round of speculative liquidation. A print below 2.8% would revive rate-cut bets and push gold back toward $4,350. The response time in both directions is measured in minutes.
The 244-tonne gap will not be resolved until the International Monetary Fund publishes its next quarterly reserve report in June. Until then, the debate over whether gold's bull market is intact or finished is a debate about a number nobody has confirmed.
Prepared with AlphaScala research tooling and grounded in primary market data: live prices, fundamentals, SEC filings, hedge-fund holdings, and insider activity. Each story is checked against AlphaScala publishing rules before release. Educational coverage, not personalized advice.