Photo by Vitaly Gariev on Unsplash
The short version
- A gold sleeve of roughly 5–10% has support in the long-horizon diversification literature; silver does not have the same case and behaves more like a cyclical industrial holding than a hedge.
- Fees matter more than they look: at GLD's 0.40% versus GLDM's 0.10%, the spread compounds to a meaningful drag on a multi-decade gold sleeve, even though both vehicles track the same bullion.
- The often-missed second-order effect is US tax treatment — physical-metals ETFs are taxed as collectibles (up to 28% on long-term gains), not at long-term capital-gains rates, which materially changes the after-tax return profile in a taxable account.
The interesting question for a long-horizon investor in 2026 is not whether gold will rally next quarter. It is whether a small, durable precious-metals sleeve still earns its seat in a multi-decade portfolio when the 10-year Treasury yields 4.47% in real terms close to historical fair value, and how the choice between vehicles — GLD, GLDM, IAU on the gold side, SLV and SIVR on the silver side — quietly changes the after-fee, after-tax outcome.
This piece keeps the framing structural rather than tactical. It looks at what the vehicles actually own, what they cost, how the US tax code treats them, and what sizing the academic diversification literature supports.
What a precious-metals ETF actually is
The largest US-listed precious-metals ETFs are organized as grantor trusts holding physical bullion in vaulted custody. SPDR Gold Shares (GLD) and iShares Gold Trust (IAU) hold London Good Delivery gold bars. SPDR Gold MiniShares (GLDM) does the same in a lower-cost wrapper. iShares Silver Trust (SLV) and abrdn Physical Silver Shares (SIVR) hold physical silver. The trust structure means each share represents a fractional, declining claim on a fixed bar inventory — the per-share metal content drops slightly each year as the expense ratio is paid in metal.
None of these vehicles pay a dividend. None hold miners. Tracking versus spot is mechanical and tight — the differences worth analyzing are fees, liquidity, and tax treatment, not security selection.
The vehicle landscape
| Ticker | Name | Underlying | Expense ratio | Inception | Issuer fact sheet |
|---|---|---|---|---|---|
| GLD | SPDR Gold Shares | Physical gold | 0.40% | 2004 | ssga.com |
| GLDM | SPDR Gold MiniShares | Physical gold | 0.10% | 2018 | ssga.com |
| IAU | iShares Gold Trust | Physical gold | 0.25% | 2005 | ishares.com |
| SLV | iShares Silver Trust | Physical silver | 0.50% | 2006 | ishares.com |
| SIVR | abrdn Physical Silver Shares | Physical silver | 0.30% | 2009 | abrdn.com |
Expense ratios above are taken from each issuer's current fact sheet. Specific point-in-time AUM and realized-return statistics are omitted from this edition because the source price feed used to compute them did not return in time for publication; the article focuses on the structural and cost framework, which moves on a much slower clock than spot price.
Gold and silver are not the same trade
The temptation is to bracket gold and silver together as “the metals.” The data — and the underlying economics — argue otherwise. Gold demand is dominated by central bank reserves, jewelry, and investment flows; industrial use is a rounding error in the global gold balance. Silver is roughly half industrial: solar cells, electrical contacts, EV components. That structural difference shows up in observable behavior — silver's realized volatility runs materially higher than gold's, and silver drawdowns tend to be deeper and longer.
For a long-horizon investor, the implication is that gold and silver play different roles. Gold sits in the portfolio as a low-correlation, central-bank-demanded reserve asset; its job is to behave differently from equities during specific macro regimes (currency stress, real-rate compression, geopolitical tail events). Silver behaves much more like a high-beta industrial commodity with a precious-metal overlay. Bundling them together under a single “precious metals” weight obscures this. The longer treatment of this asymmetry sits in a previous head-to-head on GLD vs SLV; the silver-specific case is examined in an analysis of SLV for long-term investors.
Cost compounds: the gold-ETF fee landscape
All three large gold vehicles hold the same underlying bullion. The differences are fees, liquidity, and tax-lot mechanics. GLD charges 0.40%, IAU 0.25%, GLDM 0.10% — per the current issuer fact sheets. On a 30-year holding period, the 30 basis points between GLD and GLDM compound into roughly 9% of cumulative value drag relative to a zero-fee benchmark, before considering bid-ask spread differences and the slow per-share metal decline that the expense ratio causes.
For an investor planning to hold a gold sleeve for decades, the cost case for GLDM or IAU over GLD is straightforward. The counter-case for GLD is institutional: GLD has the deepest options market, the tightest spreads at large size, and the longest track record. None of those matter to a buy-and-hold individual investor running a few percent of a retirement portfolio in gold. They matter to a hedge fund putting on a sized position with options overlays.
The fee gap between GLD and GLDM looks like a rounding error on a quote screen. Compounded across a multi-decade sleeve, it is the largest controllable variable in the gold allocation decision.
The collectibles tax problem most retail summaries skip
The structural feature that catches most US investors unprepared: the IRS treats physically-backed precious-metals ETFs as collectibles. Long-term gains are taxed at up to 28% under current rules, not at the 15%/20% long-term capital-gains rates that apply to equity ETFs. Short-term gains are taxed at ordinary income.
This is not a small footnote. In a taxable account, a gold position that returns the same pre-tax CAGR as an equity position can deliver a meaningfully lower after-tax CAGR over a multi-decade horizon, depending on the holder's marginal bracket. The practical implications: precious-metals ETF positions are good candidates for tax-advantaged accounts (IRA, 401(k), HSA depending on availability), and the cost case for the lowest-fee vehicle in the sleeve is reinforced because the after-tax fee drag is fully captured at the taxable distribution layer.
Note that this treatment is jurisdiction-specific. Investors outside the US should check local tax handling of physically-backed metal trusts before assuming the cost arithmetic in their account.
Sizing the sleeve: what the diversification literature supports
Modern portfolio research does not establish gold as a primary return engine. The realized real return on gold over very long windows is approximately zero — gold preserves purchasing power on average; it does not compound it. The case for owning it rests on its correlation profile, not its expected return. In specific macro regimes (high inflation surprise, real-yield compression, severe equity drawdown coinciding with currency stress), gold has historically diversified equity risk in ways that high-quality bonds did not — most visibly in 2022, when long-duration Treasuries failed as a diversifier while gold did not.
The conventional academic and practitioner range for a gold allocation in a long-horizon balanced portfolio is roughly 5–10%, sized as a strategic, behavior-tested position rather than a tactical bet. Silver, when included at all, is typically a 0–3% satellite tilt — and that tilt is more an industrial-commodity expression than a hedge. With the 10-year Treasury at 4.47% (FRED, asof 2026-05-14) and CPI year-over-year at 3.95% (FRED, asof 2026-04), real yields are positive, which historically reduces gold's tactical appeal at the margin — but the strategic case for a small sleeve does not hinge on the spot real-yield level. It hinges on regime diversification.
Sizing the sleeve is downstream of the broader portfolio question, which is treated in the rebalancing literature; a discussion of what rebalancing discipline actually adds covers the band methodology that governs when to add to or trim a metals sleeve. Macro framing for the same sleeve sits in a previous piece on gold, the dollar, and risk assets.
Scoreboard: best-in-class by criterion
| Criterion | Winner | Reason |
|---|---|---|
| Lowest cost (gold) | GLDM (0.10%) | 30 bp lower than GLD on identical underlying |
| Lowest cost (silver) | SIVR (0.30%) | 20 bp lower than SLV; meaningful over decades |
| Deepest liquidity | GLD / SLV | Tightest spreads at institutional size; matters less to long-term retail holders |
| Suitability for long-term core | Gold (small sleeve) | Silver's industrial cyclicality argues against core inclusion |
FAQ
How much gold should a long-term portfolio hold?
The defensible range from diversification research is roughly 5–10% as a strategic sleeve. Anything larger is a directional view on gold itself, not a diversification decision.
Is GLDM as good as GLD for a buy-and-hold investor?
For a multi-decade hold, the 30 bp fee gap favors GLDM. GLD's advantages — deeper options liquidity, tighter institutional spreads — apply to active and large-size users, not long-horizon retail holders.
Are precious-metals ETFs taxed differently from equity ETFs?
In the US, yes. Physically-backed precious-metals ETFs organized as grantor trusts are treated as collectibles for tax purposes — long-term gains up to 28%, short-term at ordinary rates. Holding inside a tax-advantaged account avoids this.
Does silver belong in a long-term core portfolio?
Silver behaves much more like an industrial commodity than a hedge — higher volatility, deeper drawdowns, stronger cyclicality with manufacturing demand. If it is held at all, it is typically a small satellite position rather than core.
Does gold still hedge inflation when real yields are positive?
The relationship is regime-dependent. Gold's strongest historical performance has come during real-yield compression rather than headline inflation alone. With the 10-year at 4.47% and CPI at 3.95%, real yields are modestly positive — a headwind, not a structural disqualifier. Strategic gold sizing does not require betting on the current rate regime.
What this analysis can and can't tell you
The argument here rests on structural facts — fee schedules, tax treatment, and the broad shape of historical correlation behavior. It does not include realized return data for the current article cycle (the price-data source did not return in time), so any conclusion about which vehicle has outperformed over a recent window is deferred. The collectibles tax treatment described is US-specific; investors in other jurisdictions face different rules. The diversification case for gold is built on long-window historical data covering only a few full macro regimes, so single-regime risk applies — past correlation behavior is not a contract.
Scenarios where each piece fits
An investor with a 100% equity portfolio considering a first defensive sleeve: a gold position of 5–7% via GLDM or IAU, ideally inside a tax-advantaged account, is a reasonable starting point — it adds a low-correlation asset without committing to long-duration bonds at the current real-yield level.
An investor already running a meaningful long-duration bond allocation: gold is closer to a complement than a substitute. Sizing the gold sleeve smaller (3–5%) and treating bonds and gold as two different regime hedges is more defensible than treating gold as a bond replacement.
An investor interested specifically in the solar/EV thematic: silver via SIVR is a way to express that view, but it is an industrial-commodity bet, not a hedge — sizing should reflect that, typically as a 1–3% satellite rather than a core holding.
Editor's read
If forced to pick a single structure, the editor leans toward a small (roughly 5–8%) gold sleeve held in a low-cost vehicle — GLDM or IAU — and a zero or minimal silver allocation. The strategic case for gold rests on regime diversification, where 30 basis points of recurring fees and a 13-point tax-rate differential matter more than they look on a single statement. The case for silver as a long-term core position has not been clear to the editor; as a thematic satellite expressing a view on industrial demand, it can be justified, but sizing should reflect that it behaves like a commodity, not a hedge.
The editor holds a small gold sleeve via a low-cost physical-gold ETF; does not currently hold silver.
Methodology
Expense ratios, inception dates, and fund structure were taken from each issuer's current fact sheet (State Street Global Advisors for GLD and GLDM, iShares for IAU and SLV, abrdn for SIVR), accessed 2026-05-18. Macro reference values are from FRED: 10-year Treasury constant maturity (asof 2026-05-14), CPI year-over-year (asof 2026-04-01). Realized return and volatility computations from yfinance were not available for this article cycle and are omitted; the framework and cost arithmetic do not depend on them. Tax treatment description reflects current US Internal Revenue Code handling of grantor-trust physical-metal ETFs and is not jurisdiction-portable.
This article is for educational purposes and does not constitute personalized financial advice. See full Disclaimer.