Gold: Central Bank Floor Meets a 17% Crash
Key Takeaways
- Central banks bought 863 tonnes in 2025 and are projected to buy 755 tonnes in 2026 — the structural buying programme continued right through the 17% crash.
- The crash was driven by leveraged speculator liquidations after the Fed killed rate-cut expectations, not by any change in sovereign demand fundamentals.
- The structural floor sits at $4,200-$4,500 (near the 200-day moving average), not $5,000 — central bank buying supports prices over months, not against day-to-day speculative selling.
- Wall Street maintains bullish targets: J.P. Morgan $6,300, UBS $6,200, Deutsche Bank $6,000 — viewing the crash as a positioning flush, not a fundamental reversal.
Updated April 5: Central banks are still buying gold. The price crashed anyway. Gold has fallen 17% from its $5,627 all-time high to $4,680, blowing through the sovereign-demand floor that this article argued was structurally supporting prices above $5,000. The thesis was half right: central bank buying is real, persistent, and accelerating in breadth. But it operates on monthly tonnage timescales while leveraged speculators liquidate in hours.
The 863 tonnes central banks purchased in 2025 and the 755 tonnes projected for 2026 didn't prevent the crash because they were never designed to. Sovereign reserve managers don't defend price levels. They accumulate physical metal on dips over multi-year programmes. The buyers who set gold's price at the margin — leveraged futures traders, momentum funds, options market makers — all headed for the exit simultaneously when the Fed killed rate-cut expectations and the dollar surged to 120.9.
The central bank floor exists. It just sits at $4,200-$4,500, not $5,000.
The Sovereign Buying Machine Kept Running
Central bank gold demand hasn't wavered during the crash. The structural drivers are unchanged: the 2022 freeze of Russian reserves catalysed a permanent reallocation away from dollar-denominated assets. China has extended its buying streak to 15 consecutive months. Poland added 102 tonnes in 2025 alone. Uzbekistan lifted reserves to 399 tonnes — 86% of total reserves, up from 57% in 2020.
The buyer base continues to broaden. Malaysia purchased gold for the first time since 2018. South Korea announced plans to add gold ETFs to reserves for the first time since 2013. The World Gold Council's 2025 survey found 95% of central banks expected global gold reserves to increase — near-unanimity that reflects a strategic consensus rather than tactical positioning.
The 2026 projected total of 755 tonnes looks like deceleration but is partly an arithmetic illusion — at $4,680 per ounce, 755 tonnes represents a similar dollar commitment to 1,000 tonnes at $3,500. Central banks think in reserve percentages, not absolute tonnage.
Why Sovereign Buying Didn't Prevent the Crash
Central banks bought gold throughout February and March. The price crashed throughout February and March. Understanding why requires distinguishing between who sets the floor and who sets the marginal price.
Sovereign buyers are price-insensitive and accumulate steadily — 5 tonnes in January, with monthly averages around 27 tonnes. They don't compete with leveraged futures traders who can move hundreds of tonnes of notional exposure in a single session. When CFTC data showed near-record speculative longs and the macro narrative shifted (rate cuts evaporated, dollar strengthened), the resulting margin-call cascade overwhelmed months of central bank accumulation in days.
Think of it this way: central banks add a few tonnes per month to physical vaults. Speculators can liquidate the equivalent of hundreds of tonnes in paper gold futures in a single week. The structural bid sets a floor over time; the speculative unwind sets the price at the margin.
The crash was a positioning event, not a fundamental one. Central banks didn't sell. ETF outflows were modest. The forced liquidation of leveraged longs drove the entire move — which is why the structural thesis remains intact even as the price thesis required a 17% adjustment.
Macro Backdrop: The Fed Changed Everything
Gold was rallying in defiance of positive real yields and a firm dollar through Q1. That defiance had limits.
The Fed funds rate remains at 3.64%, but the expected path changed dramatically. Markets entered 2026 pricing 2-3 rate cuts; the Fed's latest projections show one at best. The Iran conflict's oil shock pushed inflation expectations higher, not lower, removing the rate-cut catalyst that gold bulls had been counting on.
The 10-year Treasury yield at 4.31% offers genuine real returns against CPI running at roughly 2.2% annualised. The trade-weighted dollar at 120.9 is near cycle highs. Both metrics represent headwinds that central bank buying can't offset at the margin.
The traditional model that prices gold as a function of real yields and dollar strength has reasserted itself after a two-year holiday. Central bank demand can push gold above model-implied fair value, but it can't permanently decouple the metal from macro gravity.
Wall Street Still Sees Higher Prices
Every major bank that published a gold target above $5,000 has maintained it through the crash. J.P. Morgan stands behind $6,300 year-end. UBS targets $6,200. Deutsche Bank calls for $6,000. Wells Fargo's range is $6,100-$6,300.
Their argument: the crash was a positioning flush, not a fundamental regime change. US debt-to-GDP continues to rise. Central bank diversification away from Treasuries is a multi-decade trend. The de-dollarisation impulse that drove sovereign gold buying was catalysed by geopolitics (Russia sanctions), not by gold's price — so gold's price falling doesn't reverse the catalyst.
Morgan Stanley at $4,400 represents the bear case — still only 6% below current levels. The spread between the most bullish and most bearish Wall Street forecasts ($6,300 vs $4,400) is $1,900, reflecting genuine uncertainty about how quickly structural demand can offset cyclical headwinds.
The consensus is clear: the crash was an opportunity, not a verdict. Whether that consensus proves right depends entirely on the Fed's next move.
Where the Real Floor Sits
The 200-day moving average at $4,302 represents the structural floor — the level where central bank physical demand, long-term ETF holders, and value-oriented allocators converge. Gold has tested and held this zone during every significant correction in the current bull market.
Above that, the $4,500 area has attracted dip-buying interest as the initial panic subsided. The current $4,680 level sits in the middle of what's likely to be a wide consolidation range of $4,300-$4,900 until the macro picture clarifies.
For investors, the positioning implications are straightforward. The 5-10% portfolio allocation to gold makes more sense at $4,680 than it did at $5,007 when this article was first published. Dollar-cost averaging into gold ETFs or physical positions over the next quarter allows you to accumulate near the structural floor without trying to call the exact bottom.
Central banks are rewriting the price floor. They're just writing it at $4,300, not $5,000.
Conclusion
The central bank thesis survives the crash test — with a lower price tag. Sovereign gold demand of 755+ tonnes annually creates a genuine structural bid that didn't exist before 2022. The buyer base is broadening, the motivation (reserve diversification away from dollars) is strengthening, and no central bank that started buying has reversed course.
But structural demand sets a floor, not a ceiling. The ceiling is determined by macro variables — real yields, dollar strength, Fed policy expectations — and all three currently argue against gold. The metal will likely consolidate in the $4,300-$4,900 range until either the Fed pivots dovish or the Iran conflict de-escalates enough to normalise oil prices and relieve inflation pressure. Patient accumulators, like the central banks themselves, have the right approach: buy the dip, think in years, and ignore the noise.
Frequently Asked Questions
Sources & References
www.jpmorgan.com
fred.stlouisfed.org
fred.stlouisfed.org
Disclaimer: This content is for informational purposes only and does not constitute financial advice. Consult qualified professionals before making investment decisions.