Gold spot prices entered May 2026 at 3,420 dollars per ounce, up roughly 28 percent year-to-date and up 84 percent from the start of 2024. The performance has occurred against a backdrop where the standard predictors of gold prices have given mixed signals. Real interest rates measured by the 10-year TIPS yield are at 1.94 percent, which historically would be a headwind for gold. The dollar index sits at 96.4, which is supportive but not extreme. Inflation expectations are anchored near 2.4 percent on the 10-year breakeven, which is also not the kind of inflation panic that typically drives a gold cycle.
The driver that does match the price action is central bank buying. The World Gold Council reported central bank net purchases of 1,084 metric tons in 2025, the third consecutive year above 1,000 tons. The 2010 to 2021 average was 423 tons per year. The buying has been concentrated in central banks of countries with substantial dollar reserves seeking diversification, including the People's Bank of China, the Reserve Bank of India, the Central Bank of Turkey, and a group of Gulf state central banks. The motivation is dollar diversification rather than inflation hedging, and the demand is largely insensitive to price.
The second driver is geopolitical risk pricing. The Russia-Ukraine conflict, the ongoing Israel-Lebanon-Iran tension, the Taiwan Strait tension, and the broader fragmentation of the post-1945 trade order have produced a sustained bid for what financial markets call safe-haven assets. Gold has been the historical beneficiary of safe-haven demand during periods of geopolitical stress, and the 2022 to 2026 period is the most sustained geopolitical stress regime since the Cold War. The gold price reflects this underlying anxiety rather than any specific event.
The portfolio case for gold in 2026 rests on its low correlation with other asset classes. The 10-year correlation between gold and the S and P 500 is approximately 0.05. The 10-year correlation between gold and the Bloomberg Aggregate Bond Index is approximately negative 0.03. Gold therefore behaves as a true diversifier, meaning it provides return reduction in volatility that is independent of the equity and bond components of a portfolio. The diversification benefit is most valuable in portfolios that already hold significant equity exposure.
The allocation question is where the analysis gets interesting. The classic Harry Browne Permanent Portfolio allocates 25 percent to gold, which most modern asset allocators view as excessive. The Bridgewater All Weather framework allocates approximately 7.5 percent to gold. Ray Dalio has publicly recommended 5 to 10 percent gold allocation in personal portfolios. The major target-date fund providers including Vanguard, Fidelity, and Schwab include effectively zero direct gold exposure in their default allocations. The empirical evidence suggests that 3 to 8 percent gold allocation produces meaningful diversification benefits without dragging long-term returns below diversified equity-heavy portfolios.
The vehicle question matters because the cost differences are real. Physical gold purchased through coin dealers carries a 4 to 7 percent retail markup over spot price plus storage costs. The major gold ETFs include GLD with an expense ratio of 0.40 percent, IAU with 0.25 percent, and the newer SGOL with 0.17 percent. The lowest-cost option for most investors is GLDM at 0.10 percent expense ratio, which is the small-share version of GLD designed for retail investors. For tax-advantaged accounts, the ETF route is straightforward and avoids the storage and authentication issues of physical gold.
The collectible tax treatment is the trap for taxable accounts. Gold ETFs that hold physical gold are taxed as collectibles at a maximum federal rate of 28 percent on long-term capital gains, rather than the standard 15 to 20 percent rate that applies to most equity investments. Investors holding gold in taxable accounts should be aware of this treatment and consider whether they are willing to accept the higher tax rate or prefer to hold gold exposure in a tax-advantaged account where the rate difference does not matter.
The gold mining stock route is a different exposure with different risk characteristics. The major gold miners ETF GDX has returned roughly 47 percent year-to-date in 2026, which is a higher beta version of the gold price exposure. Gold miners have operational risk, capital allocation risk, and geographic risk that bullion does not have. The historical record suggests that miners outperform bullion during the steep rising phase of a gold cycle and underperform during the consolidation phase. Investors who want pure gold exposure should hold bullion ETFs. Investors who want operational beta to the gold price can add a smaller miners position alongside.
The skeptic's argument against gold remains intellectually honest. Gold produces no cash flow. It is not productive capital. The valuation framework for gold is a market clearing price between buyers and sellers without an underlying earnings or coupon stream to anchor it. A 100 percent gold portfolio over the past 30 years has produced returns roughly equivalent to a diversified bond portfolio with significantly higher volatility. The case for gold is purely the diversification case, not a return case in isolation.
The honest answer for most investors in 2026 is that a small gold allocation in the 3 to 6 percent range improves portfolio diversification without significantly affecting long-term returns. The historical pattern is that investors add gold to portfolios after long bull runs and abandon it after long bear runs, which is exactly the wrong sequencing. The investors who benefit most from gold are the ones who established a strategic allocation years before they wanted it and rebalanced consistently across the cycle.
