Central bank gold purchases have reached unprecedented levels over 2022-2025 as reserves diversify away from US Treasuries. We map the structural drivers, identify the highest-quality miners, and assess whether gold's secular bull market has multi-year runway remaining.
Gold has quietly outperformed essentially every major asset class over the past 24 months, with the spot price rising from approximately $1,850/oz in early 2023 to over $3,100/oz by early 2026. This move has occurred against a backdrop of rising real interest rates — traditionally bearish for gold — demonstrating that historical rate-gold correlations have broken down. The underlying driver is not retail ETF demand or speculative positioning; it is sustained, systematic central bank accumulation at a pace unprecedented in modern monetary history.
Global central banks have purchased approximately 1,000+ tonnes of gold annually since 2022, compared to a 200-400 tonne historical average. China's People's Bank of China has been a notable buyer, but purchases are broad-based across emerging market central banks (Turkey, Russia, India, Poland, Hungary, Singapore, Saudi Arabia). The pattern reflects a structural shift: countries that could face US sanctions or frozen asset risk are diversifying reserves away from US Treasuries toward non-confiscatable alternatives. Gold is the primary beneficiary; Bitcoin and other alternatives are secondary.
Our view is constructive: gold is in a structural bull market that likely extends through 2028-2030. We recommend exposure through multiple vehicles: physical gold ETFs (GLD, IAU) for defensive position, high-quality gold miners (Newmont, Agnico Eagle, Barrick) for operational leverage, and selective royalty/streaming companies (Franco-Nevada, Wheaton Precious Metals, Royal Gold) for lower-risk upside. Silver and platinum offer secondary debasement exposure with higher beta. We also discuss the bitcoin-as-digital-gold narrative and identify where the two assets compete versus complement.
Gold's recent performance cannot be explained by traditional frameworks. The standard gold model relates price to real interest rates (negative correlation), dollar strength (negative correlation), and inflation expectations (positive correlation). Over the past three years, real interest rates have risen meaningfully, the dollar has been strong, and inflation expectations have declined — all factors that should depress gold prices. Yet gold has rallied over 60%. Something structural has changed.
The most important structural driver is central bank accumulation. Global central banks collectively held approximately 34,500 tonnes of gold in 2015. By 2025, holdings approached 38,500+ tonnes. The pace of accumulation has accelerated meaningfully since 2022, with annual purchases exceeding 1,000 tonnes for three consecutive years — an unprecedented streak in modern history.
The specific triggers for accelerated accumulation are identifiable. The February 2022 freeze of Russian central bank reserves (approximately $300 billion of assets held at Western custodians) demonstrated that US-dollar reserves can be frozen as a geopolitical tool. Every central bank that might face similar actions reassessed its reserve composition. The rational response has been to accumulate non-confiscatable alternatives: physical gold held in domestic vaults, yuan-denominated assets, bilateral currency swap arrangements. Gold has been the primary beneficiary because it is the oldest, most liquid, and most universally recognized non-confiscatable store of value.
The People's Bank of China (PBOC) has been a particularly important buyer. Officially disclosed PBOC gold holdings grew from approximately 1,850 tonnes in 2015 to over 2,300 tonnes in 2025. However, many analysts estimate actual Chinese gold holdings are meaningfully higher than officially disclosed — the "dark pool" of PBOC gold could exceed another 2,000-3,000 tonnes based on triangulation of Chinese import data, domestic mine production, and disclosure inconsistencies.
China has specific motivations for gold accumulation. First, diversification from US Treasury holdings (approximately $800 billion, declining from peak over $1.3 trillion). Second, support for eventual yuan internationalization — a currency backed by meaningful gold reserves has more credibility in international settlement. Third, hedge against potential US sanctions in any Taiwan conflict scenario. Fourth, domestic savings alternative for Chinese citizens who have lost confidence in property and equities.
Beyond central bank activity, structural fiscal deficits in major Western economies support gold. US federal deficit has been approximately 6-7% of GDP during a strong economic period — historically unprecedented for peacetime non-recessionary conditions. European deficits are lower but still meaningful, with government debt-to-GDP ratios that would require sustained real growth or inflation to stabilize. Japanese fiscal position is the most extreme but Japan has particular structural characteristics that may not apply elsewhere.
Sustained fiscal deficits imply either: (1) continued real growth to service debt, (2) inflation to erode real debt burden, (3) financial repression (forcing holders of government debt to accept below-market yields), or (4) some combination. Gold benefits from scenarios 2 and 3 directly, and from scenario 4 indirectly as it provides a hedge against the financial repression that typically accompanies debt dynamics like current levels.
The broader framework is what some analysts call "the end of financial dominance" — the era since 1971 in which US dollar reserves were the default global asset and the US could run structural current account deficits funded by foreign Treasury purchases. The foreign Treasury demand has weakened meaningfully over the past decade: China has been reducing holdings, Japan has been stable but not growing, and various other holders have diversified. Gold is the primary beneficiary of any sustained shift away from US Treasuries as the default reserve asset, though the transition is gradual rather than sudden.
Gold's supply side is structurally constrained in ways that many commodities are not. Annual mine production has been approximately 3,600 tonnes for over a decade with limited growth despite sustained high prices. This reflects several specific structural factors:
Average mine ore grade has declined meaningfully over the past 30 years as easier-to-mine deposits were exhausted. Current average grades are approximately 1.2 grams per tonne, compared to 2.5+ grams per tonne in the 1990s. Lower grades mean more ore must be processed for the same gold output, requiring more energy, water, and operational cost. All-in sustaining costs (AISC) for major miners have risen from approximately $700/oz in 2016 to over $1,300/oz in 2025.
New gold mines typically require 10-15 years from initial discovery to first production. The development process includes: initial exploration (2-4 years), feasibility studies (2-3 years), environmental permitting (2-5 years), financing (1-2 years), and construction (3-5 years). Even at current elevated prices, the supply response from new mines will take most of the 2020s to materialize. Near-term supply is essentially fixed.
Gold production is concentrated in a limited number of countries. Top producers include China (approximately 9% of global production), Australia (9%), Russia (9%), Canada (6%), US (5%), Peru (4%), South Africa (4%), and various smaller contributors. Several of these jurisdictions have meaningful political and regulatory risk affecting long-term production. South African gold mining, in particular, has been in structural decline for decades as deep mines become uneconomic.
Environmental, social, and governance (ESG) considerations have tightened meaningfully for new gold mines. Projects that would have been approved in the 1990s or 2000s now face additional scrutiny on environmental impact, indigenous community consultation, water usage, and greenhouse gas emissions. Several major projects have been delayed or cancelled due to ESG-related issues. This creates a structural cap on supply growth that reinforces the supply-side bullish case.
For investors seeking leveraged exposure to gold prices, mining equities offer specific advantages and specific risks versus physical gold. Mining companies provide operational leverage — a 10% gold price increase typically translates to 20-40% miner earnings increase given fixed cost structures. They also offer dividend income that physical gold does not. However, miners carry company-specific risks (operational, jurisdictional, management) that physical gold avoids.
Newmont is the largest gold producer globally, with annual production of approximately 5.8 million ounces from operations in North America, South America, Australia, and Africa. The company acquired Newcrest Mining in 2023 for approximately $17 billion, creating a combined portfolio of Tier-1 assets. The scale provides operational diversification and reduces single-mine risk. The 2024-2025 integration has been uneven, with some operational hiccups and management changes, but the underlying asset base is high-quality.
Agnico Eagle is a Canadian-headquartered gold miner with operations in Canada, Finland, Mexico, and Australia. The company is widely regarded as the highest-quality operator in the gold mining sector — disciplined capital allocation, best-in-class safety record, consistent execution, and strong balance sheet. Annual production of approximately 3.4 million ounces.
Barrick Gold is the second-largest gold producer globally with operations in Africa, Nevada (via Nevada Gold Mines joint venture with Newmont), Latin America, and Papua New Guinea. Under CEO Mark Bristow, Barrick has focused on Tier-1 assets and operational discipline. Production of approximately 4.0 million ounces annually. The company also has meaningful copper exposure through the Reko Diq and Lumwana projects.
Beyond the major producers, several mid-tier names offer specific characteristics worth considering. Kinross Gold (KGC) has Russia exposure that has complicated but not destroyed its business. Gold Fields (GFI) has improved execution under recent management. AngloGold Ashanti (AU) has high-quality assets but some jurisdictional challenges. Endeavour Mining (EDV.TO) is a West African specialist with meaningful exposure to developing jurisdictions.
For diversified mid-tier exposure, a basket approach is often preferable to single-name concentration. The VanEck Gold Miners ETF (GDX) provides broad exposure to major and mid-tier producers. The VanEck Junior Gold Miners ETF (GDXJ) provides exposure to smaller, higher-beta producers.
Royalty and streaming companies provide gold exposure with structurally different risk/reward than operating miners. These companies provide upfront capital to mining companies in exchange for either royalty interests (percentage of revenue) or streaming arrangements (right to purchase future production at fixed prices). The model avoids operational risk while maintaining leverage to commodity prices.
Franco-Nevada is the largest and highest-quality royalty/streaming company globally. The portfolio includes over 400 royalty and streaming interests across gold (primary), silver, platinum group metals, and oil & gas. The business model delivers exceptional financial characteristics: gross margins exceeding 85%, free cash flow conversion near 100% of EBITDA, consistent cash returns to shareholders.
The Cobre Panama mine suspension (2023) removed a meaningful revenue contributor but the diversified portfolio absorbed the impact. Franco-Nevada trades at a premium multiple (approximately 30x forward earnings) reflecting the business quality. For investors seeking lower-risk gold exposure with high-quality business characteristics, Franco-Nevada is the benchmark.
Wheaton Precious Metals is the second-largest streaming company, with particular focus on silver streams alongside gold. The company has grown through strategic stream acquisitions and offers diversification across approximately 23 operating mines and multiple development projects. Wheaton has a modestly different risk profile than Franco-Nevada given silver exposure (higher volatility).
Royal Gold is the third major royalty/streaming company, with a portfolio of royalty interests across North and South America, Australia, and Africa. Similar quality business characteristics to Franco-Nevada and Wheaton, smaller in scale. The recent CTP (Cote Gold) royalty and other new contributions are providing growth.
Royalty/streaming companies trade at substantial premiums to operating miners on forward earnings multiples (30x vs 15-22x) but the quality justifies the premium. For investors wanting gold exposure with lower volatility and business quality, we recommend allocating majority of mining exposure to royalty/streaming names rather than operators. A 60/40 split between royalty/streaming and high-quality operators (Agnico, Newmont) provides balanced exposure.
Silver typically offers 1.5-2.0x the beta to gold moves in directional markets. The underlying dynamics are partly similar (store of value, debasement hedge) but partly different. Silver has meaningful industrial demand — approximately 55% of total demand, primarily for electronics, solar panels, and batteries. This industrial exposure provides additional upside during economic expansions but also downside during recessions.
The gold/silver ratio — currently approximately 90x — has historically traded in a range of 60-100x. The ratio tends to compress during precious metals bull markets (silver outperforms) and expand during bear markets. At current 90x ratio in a bull market, silver has catch-up potential. The specific drivers for continued silver strength: industrial demand from solar panel production (which has grown meaningfully despite demand headwinds), supply constraints at current primary silver mines, and generalist fund rotation into precious metals that typically includes silver exposure.
For investors, silver exposure is available through SLV (physical silver ETF) or silver-focused miners (Pan American Silver, Fortuna Silver, Coeur Mining, First Majestic). Pan American Silver (PAAS) is our preferred silver miner — diversified operations across the Americas, meaningful gold byproduct credits, and relatively conservative management.
Platinum group metals (PGMs) have their own specific dynamics. Platinum has been under pressure from declining diesel automotive demand (diesel catalytic converters being the traditional primary application) offset partially by growing green hydrogen demand (platinum is used as catalyst in proton exchange membrane electrolyzers). Palladium, used primarily in gasoline catalytic converters, has been under pressure from electric vehicle substitution of internal combustion engines.
The PGMs are more of a tactical play than a structural one. If green hydrogen deployment accelerates meaningfully, platinum has upside. If EV transition slows, palladium has less downside pressure. But neither has the clean debasement thesis that gold and silver offer. We view PGMs as specialty exposure for investors with specific commodity views rather than as core precious metals positions.
No discussion of gold and debasement would be complete without addressing Bitcoin and its relationship to the precious metals thesis. Bitcoin's "digital gold" narrative has evolved considerably since 2017, with the asset gaining meaningful institutional adoption via spot ETFs (launched January 2024), corporate treasury adoption (MicroStrategy, various smaller companies), and increasing retail familiarity.
The key question for gold investors is whether Bitcoin represents a substitute for gold demand (meaning each dollar into Bitcoin reduces demand for gold) or a complement (meaning Bitcoin expansion of total precious metals awareness supports demand for all stores of value). Empirical evidence from 2020-2025 has been mixed. Some periods show inverse correlation (Bitcoin rallies, gold weakens), others show positive correlation (both rally together during dollar weakness).
Our view is that Bitcoin and gold are partial substitutes and partial complements. Younger investors (Millennials, Gen Z) are increasingly viewing Bitcoin as their "store of value" asset and may allocate less to gold than older cohorts would have. However, Bitcoin's volatility (historically 60-80% annualized) compared to gold's volatility (15-20% annualized) means Bitcoin serves a different portfolio function. Sophisticated investors hold both.
Notably, the dominant source of incremental gold demand over the past 3 years — central banks — has shown essentially zero interest in Bitcoin as a reserve asset. Central banks accumulate gold; they do not accumulate Bitcoin. This is unsurprising — Bitcoin's volatility is incompatible with reserve asset requirements, and the political and regulatory status of Bitcoin varies by jurisdiction in ways that complicate central bank holdings. As long as central bank demand remains the marginal buyer for precious metals, gold's advantage over Bitcoin for that specific demand category is durable.
This does not mean Bitcoin cannot succeed as an investment. Bitcoin has its own bull case and its own addressable market. But the comparison to gold should acknowledge that central bank demand — the single most important marginal gold buyer — does not participate in Bitcoin. Both assets can succeed in their respective niches.
For investors seeking gold/precious metals exposure, we recommend the following allocation framework as a percentage of total portfolio:
Gold and gold miners have meaningful volatility that rewards disciplined rebalancing. Our recommended approach: maintain target allocation percentages within +/-25% bands, trimming winners and adding to laggards when bands are breached. This captures volatility premium while maintaining strategic exposure.
In taxable US accounts, physical gold ETFs (GLD, IAU) are taxed as collectibles at up to 28% rates on long-term gains. Gold miners are taxed as standard equities at standard long-term capital gains rates (up to 20%). For tax-efficient portfolio construction, physical gold exposure works best in IRAs and other tax-advantaged accounts, while miners work well in taxable accounts. For non-US investors, local tax treatment varies — French investors should consider PEA eligibility and wealth tax implications.
The gold/debasement thesis has specific risks that would warrant reassessment:
We view all of these risks as low-probability over 2025-2028 but worth monitoring. The most important single monitoring metric is central bank quarterly gold purchase data, published by the World Gold Council.
In the debasement framework, gold is not the only alternative to fiat currency and sovereign bonds. Investors seeking inflation or debasement hedges have multiple options, each with different risk/reward characteristics. Understanding the relative positioning helps frame appropriate portfolio allocations.
Treasury Inflation-Protected Securities (TIPS) provide direct inflation protection via coupon and principal adjustment based on CPI. TIPS are US Treasury securities, meaning they share US sovereign credit risk and the same vulnerability to sanctions/freezing that central banks are diversifying away from. For US investors without geopolitical concerns, TIPS offer effective inflation protection at yields around 2% real. For investors who share the broader debasement thesis, TIPS are inadequate because they do not address sovereign credit risk.
Real estate has historically been considered an inflation hedge because construction costs (and rental income) tend to track inflation over long periods. The inflation hedging properties are real but vary by property type, geography, and financing structure. Property faces specific risks gold does not: illiquidity, property taxes, maintenance costs, tenant risk, interest rate sensitivity. For diversified portfolios, real estate complements gold rather than substituting for it.
Broader commodity exposure (industrial metals, agricultural, energy) provides inflation correlation but very different cyclical characteristics than gold. Commodities tend to perform well during economic expansions (high demand) and poorly during recessions. Gold tends to perform well during crises (flight to quality) and moderately during expansions. The two categories are not substitutes but complements for investors wanting real asset diversification.
The broader cryptocurrency ecosystem (Ethereum, stablecoins, various altcoins) has less developed debasement narrative than Bitcoin specifically. Ethereum has utility-driven demand (smart contracts, DeFi) that makes it more commodity-like than store-of-value-like. Stablecoins are dollar-denominated and therefore fully exposed to dollar debasement risk. The various altcoins largely lack the network effect and institutional adoption that supports Bitcoin as "digital gold." For investors wanting crypto exposure in a debasement framework, Bitcoin remains the primary vehicle.
Beyond near-term mine output (fixed for several years), the longer-term supply question is what new mines are entering production over 2026-2032 and how this compares to demand growth. Understanding the pipeline helps frame whether structural supply constraints will persist.
Notable gold projects in development include: (1) Agnico Eagle's Detour Lake expansion and Macassa projects in Canada, (2) Barrick's Reko Diq in Pakistan (copper-gold, large), (3) Newmont's Ahafo North in Ghana and various Australia expansions, (4) several juniors with Tier-2 projects approaching production in Mexico, Guyana, and West Africa. Collectively, these projects will add approximately 5-8% to global mine output over 2026-2030 — insufficient to materially change the supply-demand balance but not negligible.
Gold exploration spending peaked in 2012 at approximately $6 billion globally, declined through 2019 to approximately $4 billion, and has been recovering since. Current exploration spend is approximately $5 billion, with the growth concentrated at mid-tier producers (Agnico, Kinross, Alamos). Major producers (Newmont, Barrick) have been more capital-disciplined, which limits discovery potential at their scale. The practical implication: major discoveries that would materially move supply are unlikely to emerge before 2030-2035 given the 10+ year development timeline from discovery.
Several jurisdictions that historically provided new gold production have become more difficult operating environments. Peru has seen mining project opposition and operational disruption. Ecuador has been unstable. Various African jurisdictions (Burkina Faso, Niger, Mali, Congo DRC) have faced political upheaval. Australia and Canada remain friendly but are increasingly saturated for new development. The practical impact is that new gold production increasingly comes from either difficult jurisdictions (with associated political risk) or very expensive stable jurisdictions (with associated capital cost pressure).
The gold debasement thesis is fundamentally tied to the future of the US dollar as global reserve currency. If the dollar's reserve status is genuinely eroding, gold benefits structurally. If the dollar's status is durable, gold benefits only as a modest diversification asset. Understanding this broader question is essential context for the investment thesis.
US dollar share of allocated global foreign exchange reserves has declined from approximately 70% in 2000 to approximately 58% in 2025. The decline has been gradual but consistent. The dollar share has not been lost primarily to any single competing currency — it has been dispersed across euros, yen, pounds, yuan, Canadian dollars, Australian dollars, Swiss francs, and gold. This multi-currency diversification is structurally favorable to gold because no single alternative currency is credible as a full dollar replacement.
Chinese yuan internationalization has been slower than many expected. Despite various initiatives (yuan-denominated oil contracts, Belt and Road lending in yuan, swap arrangements), yuan's share of global reserves remains under 3%. The constraints are structural: Chinese capital controls, limited offshore yuan markets, political concerns about Chinese government reliability, insufficient depth of yuan-denominated financial markets. Meaningful yuan internationalization is probably a 10-20 year project at minimum.
Beyond sovereign currency alternatives, Bitcoin offers a third path — non-sovereign digital asset that cannot be frozen or confiscated by any single government. Bitcoin adoption as a reserve alternative has been limited (El Salvador being the primary example, and a small one at that), but US corporate treasury adoption has been notable (MicroStrategy, Metaplanet, Semler Scientific). If Bitcoin continues to gain institutional and eventual sovereign adoption, it could reduce incremental demand for gold at the margin.
Our view is that the multi-decade structural forces — dollar debasement, sovereign diversification, geopolitical fragmentation, sovereign debt levels — continue to favor gold over 2025-2035. The pace of gold price appreciation may moderate from 2022-2025 rapid gains, but the structural trajectory remains upward. We target approximately $3,700/oz by 2028 (base case) with potential to $4,800/oz (bull case) if dollar reserve concerns accelerate.
Gold price data from LBMA and COMEX. Central bank gold holdings from World Gold Council quarterly Gold Demand Trends reports. Mine production and supply data from Metals Focus and GFMS. Miner operational data from company quarterly MD&A and technical reports.
Price targets derived from Glacium Research scenario analysis. Miner fair value estimates use DCF with gold price scenarios, AISC trajectories, and specific operational assumptions. Royalty/streaming valuations use portfolio sum-of-parts with individual asset DCF and risk-adjusted royalty valuations.
Sources: World Gold Council, Metals Focus, LBMA, company annual reports (Newmont, Agnico Eagle, Barrick, Franco-Nevada, Wheaton Precious Metals, Royal Gold), Bloomberg, Glacium Research primary analysis.