How does the central bank purchase and store gold?

Global central bank gold reserves reached roughly 36,000 tonnes in 2025 with a market value exceeding $4 trillion, and gold now accounts for about 20% of global foreign-exchange reserves (versus the euro's 16% and the dollar's 46%). CICC analysts report that central banks are increasingly acquiring gold via discreet London OTC ownership transfers and opaque domestic-currency procurements, while also repatriating holdings (e.g., Germany repatriated 674t in 2013–17 and India moved >65% of its reserves to domestic storage in 2025), creating tracking 'black boxes.' These shifts concentrate custody around the NY Fed, Bank of England and BIS, imply sustained and underreported central-bank demand, and are broadly supportive for gold prices while complicating reserve and FX analytics.

Analysis

Market structure: Central banks (reserve managers) and domestic refiners/miners are the clear winners—they gain pricing leverage and optionality by bypassing LBMA channels and avoiding FX outflows. LBMA-clearing banks, custodians in New York/London and ETF issuers face reduced fee capture and potential market-share loss as an estimated 36,000t (~$4tn) becomes more domestically held; expect LBMA OTC volumes to shrink by a material margin (>10–20%) over 12–24 months. Mining equities with strong jurisdictional control and low cost-inflation exposure gain asymmetric upside to gold moves.

Risk assessment: Tail risks include sudden central-bank sales (liquidity shock), geopolitical seizures of shipped metal, or regulatory bans on domestic procurement channels; each can move spot >5–10% intraday. Near-term (days–weeks) price volatility will rise on repatriation headlines; medium-term (3–12 months) continued stealth purchases could tighten physical premiums and increase futures–spot basis volatility; long-term (1–3 years) structural de-dollarization could reprice FX reserves and reduce euro demand by several percentage points. Hidden dependency: concentrated custody (NY/BoE/BIS) creates single‑point operational risk and insurance-cost asymmetry that will widen physical–paper arbitrage.

Trade implications: Expect higher gold implied vols, steeper miners’ beta and occasional dislocations between GLD/IAU flows and actual central‑bank accumulation; bond markets may see safe‑asset rebalancing (slightly higher long yields if gold replaces sovereign debt in reserves) and downward pressure on EUR funding demand. Options markets should price fatter tails—use 3–9 month call skew to capture asymmetric upside while selling near-dated vol if liquidity premiums spike. Commodities with fiscal-linked producers (gold, uranium) will trade on geopolitical custody signals rather than pure supply/demand.

Contrarian angles: The market underestimates that central banks avoid ETFs—so ETF inflows are not a reliable proxy for reserve demand; current consensus that gold ETFs = central-bank buying is likely overstated. Repatriation reduces available lendable inventory, potentially making paper markets more fragile and creating opportunities for negative-carry physical trades; historical parallel: post‑Bretton Woods reserve reallocations (1970s) produced multi-year volatility before a new equilibrium. Unintended consequence: tighter physical market could produce episodic price spikes even if macro growth weakens, favouring outright physical exposure and insurance through miners rather than short-duration paper trades.

Key Decisions for Investors

• Establish a 2–3% tactical long in physical/ETF gold (preferred IAU for fees) over 6–12 months; scale in 25% tranches, add more if spot breaks above $2,300/oz or if IMF/WGC quarterly data shows >50t of unexplained central-bank accumulation.
• Allocate 1–2% to high-quality producers: buy Newmont (NEM) 1% and Barrick (GOLD) 1% as operational-leverage plays; set stop-loss at -20% or hedge with 6-month out-of-the-money covered calls to monetize time decay if gold stalls.
• Buy a 3–6 month GLD call spread (e.g., buy 6-month ATM calls, sell a higher strike ~10–15% out) sized to 0.5–1% NAV to capture upside while limiting premium; simultaneously sell short-dated gold volatility (1–2 months) only after intraday spikes to harvest elevated IV.
• Short EURUSD size 0.5–1% notionally (or buy EUR put 3‑month) as a tactical play on reduced euro reserve share; add if ECB/Eurostat balance-sheet signals >50bp of further sovereign-bond issuance or if WGC reports additional central-bank shifts away from euro in next 90 days.