Gold ETF Risks: Structure, Tracking, and Custody

Tracking and custody risks vary by fund design

A gold ETF is not a single economic exposure. The label can describe a vehicle backed by allocated or otherwise held physical bullion, a fund that obtains exposure through futures, or an equity fund holding gold-mining companies. Each can trade on an exchange and each may be used in a portfolio discussion about gold, but their return drivers, costs and failure modes differ. Analysis fails early when the fund’s stated objective is treated as equivalent to direct exposure to the spot price of gold.

For a physical-gold vehicle, the central question is how a share represents an interest in bullion after expenses. For a futures-based fund, it is how futures contracts, collateral and contract rolling translate into returns. For a mining-equity ETF, company earnings, operating costs, reserves, jurisdictions and equity-market conditions can matter as much as the gold price. Comparing all three against the same gold-price chart can conceal the risks that actually produced a result.

Fund structure determines what is being tracked

A physically backed gold ETF generally seeks to reflect the value of bullion held for the fund, less fees and operating costs. Its benchmark may be a gold price reference, but the prospectus and daily portfolio information determine the mechanics that matter: the form of gold held, custody arrangements, valuation conventions, the fee rate and the treatment of cash. A small, persistent shortfall versus an unadjusted gold reference may be expected when expenses are deducted; it is not automatically evidence of poor management.

A futures-based gold ETF does not own bullion in the same way. It maintains positions in futures contracts and typically holds cash or collateral alongside them. Its performance can diverge from changes in spot gold because a futures price reflects time to delivery, financing conditions and market expectations. When contracts are replaced as they approach expiry, the outcome depends on the prices at which expiring and new contracts are traded. An assessment that ignores the fund’s futures index, roll schedule and collateral policy is measuring the wrong exposure.

Gold-mining ETFs introduce a further layer. Their indexes may be weighted by market capitalisation, production characteristics or other rules, and constituent changes can alter exposures. A miner can underperform gold even when bullion rises if operating costs increase, a mine is disrupted or equity valuations fall. These funds should be evaluated as sector-equity portfolios, not as precise substitutes for bullion.

Creation, redemption and trading liquidity are separate tests

ETF shares trade between investors during the market day, while the fund’s underlying portfolio is valued according to its own market conventions. The creation and redemption process allows authorised participants to assemble or redeem large baskets of shares against the underlying assets or a cash equivalent specified by the fund. That arbitrage mechanism is intended to connect the share price to net asset value, but it does not guarantee that every investor trades exactly at net asset value at every moment.

A reasonable analysis can fail by looking only at average daily share volume. The relevant liquidity also includes the liquidity and trading hours of bullion, futures or underlying mining shares, as well as bid-ask spreads, market-maker capacity and the size of the trade. A share can appear liquid on screen while its quoted spread widens when the underlying market is less accessible or volatile. Conversely, modest secondary-market turnover does not necessarily mean that a fund cannot accommodate activity if creation and redemption remain effective.

Premiums and discounts deserve interpretation rather than a single verdict. Check whether they are recurring or concentrated around periods of market stress, whether the reported net asset value is calculated when the underlying market is open, and whether the fund uses fair-value or other valuation procedures. For a long-term holder, repeatedly paying a premium on purchase or selling at a discount can be as consequential as a small difference in the stated expense ratio.

Fees, distributions and tracking should be read together

The headline management fee is only one source of return drag. Physical funds may incur custody, insurance, trustee, audit and transaction costs within the structure. Futures funds can experience implementation costs associated with rolling contracts, while the income earned on collateral may either offset or add to the gap relative to a benchmark, depending on the methodology. Mining ETFs have portfolio turnover, index-rebalancing effects and the operating economics of their constituents. The appropriate comparison is the fund’s actual return against the return definition of its stated benchmark over comparable periods.

Distribution policy can also confuse comparisons. A bullion-oriented fund may have little or no ordinary income to distribute, whereas a mining-equity ETF can pass through dividends received from portfolio companies after fund expenses. Cash distributions are not extra performance; they reduce the fund’s net asset value by the amount distributed, subject to normal market movements. Investors comparing total return, price return and distribution yield need to use consistent measures.

Tracking analysis should separate expected and unexpected deviation. Expected deviation may arise from fees, sampling, valuation timing, taxes where applicable and the design of a futures index. Unexpected deviation may point to wider dealing costs, unusual premiums or discounts, collateral outcomes, changes in methodology or difficulty implementing the stated exposure. Useful records include the prospectus, benchmark methodology, daily holdings or futures exposure, net asset value history, premium-and-discount data, expense information and distribution history.

Gold-fund flow and holdings data can provide context for investor demand, but flows do not explain a specific ETF’s return by themselves. Nor does a rise in total gold ETF holdings establish that every fund has the same liquidity, custody arrangement or tracking quality. The analysis that holds up is structure-specific: identify what the ETF owns, identify the benchmark it is designed to follow, then test trading execution, costs and return differences against that design.