Gold Spot vs SPDR Gold ETF (GLD)

SPDR Gold Shares (GLD) is a physically-backed gold ETF launched November 18, 2004, the first US-listed gold ETF and still the largest by assets. The fund holds London Good Delivery gold bars vaulted at HSBC in London. As of early 2026 the fund holds approximately 900 tonnes of physical gold (down from its peak of ~1,350 tonnes in 2012 as AUM has fluctuated with gold's relative attractiveness versus other assets). AUM approximately $130 billion at April 2026 gold prices.

Each GLD share initially represented 1/10 ounce of gold, but the expense ratio erodes this over time. As of early 2026, each share represents approximately 0.0917 ounce of gold (the fund sells small amounts of gold to pay the 0.40 percent annual expense). This structural drag is the primary difference between holding GLD and holding physical gold directly.

GLD tracks the London PM Gold Fix, the benchmark gold price set twice daily in London by a panel of banks. In practice the tracking is close but not perfect. Daily tracking error (the difference between GLD's total return and the spot gold return) typically ranges from -0.02 percent to +0.02 percent, driven mostly by timing differences between the London PM Fix and the market close.

Annual tracking error is more systematic: GLD underperforms gold by approximately 0.40 percent annually (the expense ratio) plus minor secondary effects from cash drag and small premium/discount fluctuations in the shares. Over 20 years, the cumulative drag is approximately 8 percent compound: $10,000 invested in gold would be worth approximately $26,500 after a 2.65x spot move; the same $10,000 in GLD would be worth approximately $24,300 after fees. For long-term investors this drag is meaningful; for tactical positions it is negligible.

GLD was the original gold ETF but is no longer the cheapest. iShares Gold Trust (IAU), launched 2005, has expense ratio 0.25 percent, 15 basis points lower than GLD. SPDR Gold MiniShares (GLDM), launched 2018 by the same SPDR family, has expense ratio 0.10 percent, 30 basis points lower than GLD. Both alternatives use similar physically-backed structures (vaulted gold bars in London for IAU, various vaults for GLDM) and deliver nearly identical gold-tracking performance at lower costs.

GLD's advantage over IAU and GLDM is liquidity: GLD trades approximately $1-2 billion in notional daily, roughly 10x IAU and 30x GLDM. For large institutional positions and options-market makers, GLD's deeper liquidity is worth the higher expense ratio. For retail buy-and-hold investors, IAU or GLDM provides better long-term economics with sufficient daily liquidity for position sizes below $10-50 million.

GLD shares trade at a premium or discount to the net asset value (NAV) of the underlying gold. In liquid normal markets, the premium/discount is typically within plus or minus 0.05 percent of NAV, because authorized participants create or redeem shares to arbitrage any meaningful gap. The creation/redemption process requires delivering or receiving 100,000 GLD shares (approximately $43 million at current prices) in exchange for approximately 9,170 ounces of physical gold.

In stress markets the spread can widen. March 2020 COVID stress saw GLD briefly trade at a 1-2 percent discount to NAV as authorized participants struggled with gold delivery logistics amid refinery and vault closures. These episodes typically resolve within days to weeks as logistics normalize. For retail investors, the takeaway is that GLD is a reliable gold proxy in normal conditions but may fail to track gold perfectly during acute stress.

Owning physical gold (coins, bars, bullion) eliminates expense-ratio drag but introduces other costs: vault storage fees (0.5-1 percent annually for secure storage), insurance, dealer markup on purchase (2-5 percent typically, higher for smaller denominations), and illiquidity for rapid selling. Total cost of ownership for physical gold is typically 1.0-2.5 percent annually including markup amortization, higher than GLD's 0.40 percent.

The case for physical gold: counterparty-risk elimination (no ETF sponsor, no custodian, no market structure dependency), tangible ownership, privacy (to varying degrees depending on purchase documentation), and protection against specific tail risks (financial system collapse, government intervention in ETF structures). For most investors with portfolios above $100K, physical gold storage fees mean GLD or IAU are more cost-effective. For large allocations above $1 million, physical gold in allocated storage becomes competitive on a fee basis. For "just in case" holdings of modest size, a small allocation of physical coins plus a larger GLD position is a common hybrid.

GLD is taxed as a collectible at the investor's ordinary income rate up to 28 percent maximum for long-term capital gains (versus the 20 percent maximum for ordinary long-term capital gains on equities). This is because the IRS treats precious metal ETFs holding physical metal as collectibles under IRC Section 408(m). IAU and GLDM have the same tax treatment.

This is a meaningful disadvantage versus gold mining stocks or gold-related futures contracts, which receive standard capital gains treatment. For tax-sensitive investors in high brackets, the effective return difference between GLD and gold futures can exceed 100 basis points annually over long holding periods. IRA-eligible gold holdings (in self-directed IRAs) can avoid this treatment but introduce additional custodian fees and complexity. Pre-tax investment accounts should also watch for the collectibles treatment when rebalancing or selling gold positions.

GLD daily holdings data is published every market day showing the exact tonnage of gold held. Changes in holdings (creations minus redemptions) are a clean retail/institutional sentiment signal: rising holdings indicate gold demand exceeding supply from authorized participants, falling holdings indicate redemptions exceeding creations.

GLD holdings peaked at approximately 1,353 tonnes in December 2012 during the post-financial-crisis gold rally. Holdings fell to approximately 600 tonnes by 2015-2016 as gold's appeal waned during the dollar rally. Post-2019 they ranged 1,000-1,300 tonnes with correlation to gold price but some divergences. Current April 2026 holdings near 900 tonnes are relatively low given gold's price surge, indicating retail/institutional US investors have not fully participated in the gold rally that has been driven primarily by central bank buying outside the US. This is a bullish contrarian signal: if US ETF demand recovers, it would add incremental price support on top of central bank demand.

Beyond GLD, IAU, and GLDM, the gold ETF universe includes leveraged products (UGL at 2x, DUST and GDXD as inverse gold miners ETFs), gold mining ETFs (GDX, GDXJ, SGDM), and silver/precious metal basket ETFs. For gold exposure specifically, the choice is between physical-backed ETFs (GLD, IAU, GLDM) and futures-backed alternatives.

There is no popular US-listed futures-backed gold ETF because the contango roll cost has historically been punitive (futures-backed ETFs typically lose 2-4 percent annually versus spot through negative roll yield). Physical-backed ETFs are the dominant retail and institutional gold exposure vehicle. For active traders seeking leverage or short exposure, UGL (2x leveraged gold) provides daily resetting leverage but with compounding decay that makes multi-month holding destructive. Pure gold exposure via GLD, IAU, or GLDM remains the best choice for most investors.

Gold mining ETFs like GDX (VanEck Gold Miners ETF) and GDXJ (VanEck Junior Gold Miners) offer amplified exposure to gold price movements. Gold miners typically have 1.5-2x beta to the gold price because their cost structures are relatively fixed (labor, energy, royalties) while revenue moves with gold. A 10 percent rise in gold typically produces a 15-20 percent rise in gold miners.

The trade-off is operational risk that pure gold does not have: mine-specific issues (strikes, geological setbacks, safety incidents), management execution, acquisition/divestiture activity, and jurisdiction risk (especially for GDXJ which includes more geographically diverse small producers). For views on gold's direction, pure gold ETFs (GLD, IAU, GLDM) are the cleaner expression. For bullish gold views with tolerance for operational noise and desire for upside leverage, GDX provides amplified exposure. GDXJ adds more volatility and smaller-company risk for potentially higher beta.

The primary signal for GLD tracking relative to physical gold is the authorized participant (AP) activity. If APs are actively creating and redeeming shares, GLD will track gold tightly. During stress episodes (like March 2020 COVID), logistics can disrupt AP activity and create temporary tracking gaps. Monitor GLD daily holdings data from the State Street fund page for any sudden decline that would indicate redemption stress.

Secondary signals: gold ETF flows across GLD, IAU, and GLDM combined (total US gold ETF holdings moved from ~2,800 tonnes in 2020 peak to ~2,300 tonnes in 2024 to current ~2,400 tonnes, suggesting room for ETF demand to re-accelerate); gold price levels relative to central bank fair-value estimates (Goldman Sachs 2026 forecast approximately $4,500-5,000 target); and US investor sentiment surveys on gold allocation (typically have 2-5 percent allocation, could rise to 10 percent in fear regimes). The 2026 Iran-Hormuz geopolitical premium on gold prices has not yet translated into substantial US retail ETF inflows, which creates upside potential if US sentiment catches up to the global central-bank-driven rally.