How much more gold could central banks buy?

BofA estimates central banks below a 30% gold-reserve allocation would need to buy about 20,333 tonnes of bullion—more than two decades of purchases at recent rates—supporting a continued bullish gold outlook. Global central banks bought a record 1,092 tonnes in 2024, while China alone is estimated to have a 5,628-tonne shortfall versus BofA's target allocation. BofA sees ongoing reserve diversification, accelerated since Russia's 2022 invasion of Ukraine, as a durable source of gold demand.

Analysis

The investable implication is a persistent, price-insensitive official-sector bid that can reduce bullion's sensitivity to real-rate headwinds. Unlike ETF flows, reserve managers are not benchmarked to quarterly performance and tend to transact opportunistically on pullbacks; this should raise the floor under GLD and gold-miner cash flows over the next 6-18 months, while also making physical-market tightness more consequential than speculative futures positioning suggests.

The cleanest equity beneficiaries are senior, unhedged producers with long reserve lives and low geopolitical risk: NEM, AEM and GOLD. Their operating leverage means a sustained $100/oz gold-price increase can produce a materially larger percentage increase in FCF, but miners remain exposed to diesel, labor and local-currency cost inflation; soaring energy prices can absorb part of the bullion upside. Royalty companies FNV and WPM offer lower cost-inflation exposure and should outperform if gold rises gradually rather than spikes.

Consensus may overstate the mechanical significance of reserve-allocation targets: central banks optimize for liquidity, sanctions risk and domestic politics, not a static portfolio frontier. Large prospective buyers can also favor domestic production, bilateral arrangements, or slower accumulation to avoid moving the market. Near-term gold downside is therefore meaningful if higher oil feeds inflation expectations, pushes U.S. real yields and the dollar higher, and triggers liquidation in rate-sensitive commodity positioning; the next 1-3 months hinge on real yields and official purchase data rather than long-run allocation arithmetic.

BAC has only indirect exposure: a stronger gold complex may support commodities-market activity, but a higher-for-longer rate repricing and risk-asset volatility are more important to its valuation. APP and SMCI are article-adjacent promotional references rather than beneficiaries of the underlying macro mechanism; no fundamental read-through should be assigned.

Key Decisions for Investors

• Initiate a 6-12 month long FNV or WPM versus short GDX pair: royalties retain gold upside with substantially less diesel, labor and capex inflation than producers. Target 10-15% relative return; exit if gold falls below its 200-day moving average while U.S. 10-year real yields rise above recent highs.
• For direct bullion exposure, accumulate GLD on pullbacks rather than chase breakouts over the next 1-3 months. Use a 5-7% downside risk budget, with the thesis invalidated by two consecutive quarters of materially weaker disclosed official-sector demand combined with sustained dollar and real-yield strength.
• Maintain a watchlist long NEM/AEM only after confirming all-in sustaining cost guidance is stable despite energy inflation. The missing data is each producer's fuel-cost sensitivity and hedge profile; absent stable cost guidance, favor FNV/WPM over miners.
• Do not position in APP, SMCI or BAC from this catalyst. For BAC specifically, reassess only if rate volatility translates into demonstrably higher trading revenue without a deterioration in credit provisions; oil-driven inflation is otherwise a net macro-risk signal for bank multiples.