Here is the answer most gold ETF comparisons take two thousand words to reach: GLD, IAU, and GLDM all do the same thing. Each is a trust that holds allocated physical gold in professional vaults, and each share represents a fractional claim on that metal. Over any meaningful period, their price charts are nearly identical. The choice between them is not about which one "tracks gold better." It is about fees, liquidity, and what you plan to do with the position.
That choice matters more than usual right now. Gold ETFs took in a record $89 billion of inflows in 2025, and January 2026 set a record for a single month at $18.7 billion, per the World Gold Council. Gold itself printed an all-time high of $5,589 per ounce intraday on January 28, 2026, then corrected roughly 25 percent, trading in the $4,121 to $4,143 range the week of July 6 to 10, per Trading Economics. A wave of new money bought gold near the top and is now learning what a static gold position feels like on the way down. We will get to that in the final section, because it is the part of the gold ETF decision almost nobody writes about.
SPDR Gold Shares (GLD) launched in 2004 and remains the largest gold ETF in the world, with roughly $130 billion in assets, per issuer data, July 2026. Its expense ratio is 0.40 percent, the highest of the three. What you get for that fee is unmatched trading infrastructure: the tightest institutional-size spreads, the deepest order book, and by far the most liquid options market of any gold product. If you trade gold actively, hedge with options, or move institutional size, GLD is the standard. The fee is the cost of admission to that liquidity, and for short holding periods it barely registers.
iShares Gold Trust (IAU) charges 0.25 percent and holds roughly $60 billion plus in assets, per fund data, 2026. It launched in 2005 as the lower-cost alternative to GLD and served that role for over a decade. Its liquidity is excellent for retail purposes, its spreads are tight, and its lower share price (each share represents a smaller fraction of an ounce than GLD) makes position sizing easier in smaller accounts. IAU is a perfectly good fund. Its awkwardness today is purely competitive: it is no longer the cheap one, and it was never the most liquid one.
SPDR Gold MiniShares (GLDM) is GLD's own low-cost sibling, launched in 2018 precisely because cost-conscious buyers were defecting to IAU. At 0.10 percent, it is the cheapest of the three, with roughly $33.5 billion in assets, per issuer data, March 2026. It holds the same kind of vaulted bullion as GLD, from the same sponsor family. Liquidity is more than adequate for buy-and-hold investors, though its options market is thin compared with GLD. If you are allocating a fixed percentage of a portfolio to gold and intend to hold it for years, GLDM is the default answer, and the arithmetic below shows why.
Two other tickers worth knowing as context. abrdn Physical Gold Shares (SGOL) charges 0.17 percent with about $8 billion in assets, per fund data, April 2026, and vaults its metal in Switzerland, which some holders prefer. VanEck Gold Miners (GDX), at 0.51 percent with roughly $23 billion in assets per fund data, 2026, is a different animal entirely: it holds mining stocks, not gold, and behaves like a leveraged, equity-correlated bet on the gold price. It is not a substitute for bullion exposure.
Expense ratios sound abstract until you run them on a real position. Take $100,000 of gold exposure and hold the gold price flat for simplicity:
The GLD vs GLDM gap is about $300 per year, $3,000 per decade, $6,000 over 20 years. And that simple arithmetic understates the true cost, because fees come out of the asset itself: if gold appreciates, you are paying 0.40 percent of an ever-larger balance, and the money skimmed each year never compounds for you. For a multi-decade holder, the drag difference compounds into five figures on a six-figure position. This is why the honest framework is holding period, not brand: the longer you hold, the more the fee dominates, and the stronger the case for GLDM. The shorter and more active your use, the more liquidity dominates, and the stronger the case for GLD.
All three funds are structured as grantor trusts. You own a pro-rata interest in a pile of allocated bullion, audited and held by a custodian. That structure has three consequences investors routinely miss. First, there is no yield: gold pays no dividend or interest, so your entire return is price change minus the fee. Second, you generally cannot redeem shares for physical metal; only large authorized participants can, in baskets. Third, taxes: because the trust holds a collectible, long-term capital gains for US investors are generally taxed at the collectibles rate of up to 28 percent rather than the standard long-term capital gains rate. That is an educational note, not tax advice; the treatment varies by account type and situation, and a tax professional should confirm yours.
Against physical coins and bars, the ETF trade-off is straightforward. Physical gold means no annual expense ratio but real costs elsewhere: dealer premiums of several percent on purchase, storage, insurance, and wide spreads when you sell. ETFs give you exchange liquidity, tiny spreads, and easy portfolio integration, at the price of an annual fee and no metal in hand. For portfolio allocation purposes, the ETF usually wins on cost and convenience. For deep disaster insurance, some investors still want the physical layer. The two solve different problems.
Now the part the fee comparisons skip. Whichever ticker you choose, a gold ETF is a static, always-long position. It owns the same amount of gold at $5,589 as it does at $4,100. It has no concept of trend, no exit rule, no position sizing. When gold fell roughly 25 percent from the January 2026 high, every dollar in GLD, IAU, and GLDM participated in the full drawdown, by design. The funds did exactly what they promise. The question is whether that promise is all you want.
An expense ratio decides tenths of a percent per year. Whether anything manages the position through a 25 percent drawdown decides whole percentages. Investors agonize over the first and ignore the second.
There is a third lane between static ETFs and physical metal: rules-based, systematic gold exposure. Instead of holding a fixed position at all times, a systematic approach applies predefined rules to the same underlying market, scaling position size, stepping aside when trend conditions deteriorate, and re-entering when they improve. The goal is not to predict gold's next move; it is to participate in trends while containing the depth of drawdowns through mechanical exits. That matters most for exactly the investors gold is supposed to serve: the ones holding it as portfolio insurance, who found out in the first half of 2026 that the insurance itself can draw down 25 percent. We cover how institutions structure this in our guide to gold strategies for institutional portfolios. None of this is a recommendation; static and systematic exposure are different tools with different risks, and systematic approaches can underperform in whipsaw markets. But if the drawdown math bothers you, you should know the lane exists.
The decision order, then, looks like this. First, decide whether gold belongs in the portfolio at all, and how much; our piece on how much gold belongs in a portfolio walks through the mainstream 5 to 15 percent guidance. Second, understand what actually moves the metal; see what drives gold prices. Third, choose your vehicle: GLDM-style low-cost ETFs for long static holdings, GLD for active trading and options, physical for disaster insurance, or a systematic strategy if you want the exposure managed by rules rather than left to ride. The best gold ETF is a solved problem. Whether a static ETF is the right form of gold exposure for you is the more interesting question, and it is the one worth sitting with.
It depends on holding period and use case, not brand. For long-term buy-and-hold allocations, the lowest-fee physically backed fund is the rational default, and GLDM at 0.10 percent is the cheapest of the big three, per issuer data, July 2026. For active trading, options strategies, or institutional size, GLD's superior liquidity and options market justify its 0.40 percent fee. IAU at 0.25 percent is a solid middle option. All three hold vaulted physical gold and track the metal almost identically.
They hold the same asset from the same sponsor family, so the choice is fee versus liquidity. GLDM charges 0.10 percent against GLD's 0.40 percent, a difference of about $300 per year on a $100,000 position, which compounds meaningfully over decades. GLD offers deeper liquidity and by far the more liquid options market. Long-term holders generally lean GLDM; traders and options users generally lean GLD.
No. Physically backed gold ETFs like GLD, IAU, and GLDM hold bullion, and gold produces no income. There is no dividend or interest payment. Your total return is the change in the gold price minus the fund's expense ratio, which is deducted from the trust's assets over time.
For US investors holding them in taxable accounts, physically backed gold ETFs are structured as grantor trusts holding a collectible, so long-term capital gains are generally taxed at the collectibles rate of up to 28 percent rather than the standard long-term capital gains rate. Short-term gains are taxed as ordinary income. Treatment differs inside retirement accounts. This is educational only; consult a tax professional for your situation.
GLD, IAU, GLDM, and SGOL are backed by allocated physical bullion held in professional vaults by custodians, with holdings published and audited. Each share represents a fractional interest in that metal. Retail investors generally cannot redeem shares for physical gold, however; only large authorized participants can redeem in baskets. Funds like GDX are different: they hold mining company stocks, not gold itself.