Gold and Commodity ETPs

Why your gold "ETF" isn't an ETF, and why that's fine

Diversification rules define normal funds: UCITS forbids a fund from being one big lump of a single metal, and US '40 Act funds face similar constraints. The industry's solution is elegant:

• Europe: the ETC (exchange-traded commodity). Legally a debt security issued by a special-purpose vehicle, collateralized by physical gold held at a custodian (typically allocated, numbered bars in London vaults), trading on exchange exactly like an ETF. The debt framing sounds alarming and isn't: your claim is secured by the metal itself, bankruptcy-remote from the issuer, with bar lists published. Some ETCs even offer physical delivery.
• US: the grantor trust (SPDR Gold Shares, GLD, the 2004 landmark from Chapter 2, and its cheaper successors). The trust owns the bars; your ETF-like shares own the trust. Same economic substance, different legal costume.

Either way, a physical gold ETP's return is the gold price in your currency, minus a fee of roughly 0.12%–0.40%/yr (the fee is typically paid by the trust slowly selling slivers of metal, so the metal-per-share drips downward by design). Tracking is tight, the product does what the label says, and Chapter 1's "wrapper is a container" lesson gets its cleanest illustration: this container is excellent; whether gold deserves your money is the separate question below.

Broad commodities: the futures machinery and its honest cost

Nobody vaults crude oil, live cattle and natural gas for a retail product. Broad commodity ETPs therefore hold futures contracts and roll them forward as they expire, and this creates the gap that has disappointed a generation of buyers:

• Your return = spot price change + roll yield + collateral interest. The middle term is the trap: when longer-dated futures cost more than near ones (contango, the normal state in many markets, since storage costs money), each roll sells cheap and buys expensive, a persistent drag that can run several percent a year. In the opposite state (backwardation) the roll pays you.
• The historical showcase: periods where oil's spot price rose over a year while oil futures ETPs lost money, the entire gap being roll cost. The product tracked its futures index faithfully; buyers thought they'd bought the headline chart.

Modern "optimized roll" indices soften the drag without abolishing the physics. The chapter's verdict: futures-based commodity ETPs are legitimate professional tools whose return stream is futures-strategy returns, not spot prices, and any purchase made while believing otherwise is a misunderstanding with a fee attached. (Leveraged commodity notes and single-commodity futures products escalate everything here; Part III handles them.)

Does gold deserve a place at all? The fair hearing

Per house rules, both sides, then our view.

The case for: gold has held purchasing power over very long horizons, correlates weakly with stocks and bonds, has repeatedly rallied in inflation shocks and crisis moments (2008's aftermath, 2020, 2022's bond failure year), carries no counterparty or default risk in physical form, and its ETP implementation is cheap and clean.

The case against: it produces nothing, no coupons, no earnings, no compounding engine, so its expected long-run real return is roughly zero-plus, its decade-scale droughts are brutal (1980–2000 lost ~70% real), and its crisis behavior is usually helpful rather than reliably so.

Our view, labeled as such: gold is an optional stabilizer-diversifier, defensible at 0–10% of a portfolio for investors who value its specific insurance properties and will hold it through the boring decades, and never a growth engine or a core. The best predictor of whether gold helps you is not the gold price; it's whether you sized it as insurance or bought it as a bet.

A worked example

Priya allocates €10,000 (5% of her portfolio) to gold as deliberate insurance.

• She picks a physically-backed, allocated ETC at 0.15%/yr, confirms the bar-list and custody disclosures, and, being in Germany, notes the one-year tax clock with quiet satisfaction.
• Her colleague, "diversifying into commodities" after an inflation headline, buys a broad futures-index product without meeting the word contango. Two years later spot commodity indices are up 9%; his product is up 2%; the difference is roll cost working exactly as documented. Nothing malfunctioned except the purchase's premise.
• Priya's gold, meanwhile, does what insurance does: mostly nothing, occasionally everything. Her written policy from Chapter 19 already contains the sentence that keeps it in place through the boring years: insurance is judged on the decade, not the quarter.