For most of the past two decades, gold was a fringe line item in institutional portfolios: a small, grudging hedge that investment committees rarely discussed. That era is over. Central banks purchased more than 1,000 tonnes of gold in each of 2022, 2023 and 2024 and added a further 863.3 tonnes in 2025, down 21% year over year, yet still roughly 1.8 times the 2010-2021 average of 473 tonnes, per the World Gold Council’s February 2026 data.
Price followed the official-sector bid. Gold rose roughly 50-65% in 2025, setting more than 50 all-time highs, and peaked at $5,589 per ounce on January 28, 2026. It has since retraced to roughly $4,000-$4,400 through June 2026, a meaningful drawdown, but still up around 21% year over year. (For the structural drivers behind the move, see our primer on what drives gold prices.)
Institutional allocations have not kept pace. Investors hold roughly 2.8% of assets under management in gold by J.P. Morgan’s estimate (up from around 2%, but below the 4-5% that common strategic recommendations suggest), and many Western pension funds sit under 1%. That gap is where the current allocation debate lives, and increasingly it is not just about how much gold, but about how to hold it: passively, or systematically.
The strongest evidence is behavioral. In the World Gold Council’s 2025 central-bank survey, 95% of respondents said they expect global official gold reserves to rise, and a record 43% said they plan to increase their own holdings. This is not a tactical trade; it is a structural reallocation of reserve assets, driven by sanctions risk, sovereign fiscal concerns, and diversification away from any single reserve currency.
Gold has surpassed US Treasuries’ share of central-bank reserves for the first time since 1996.
Private institutional flows followed. Gold ETFs took in a record $26 billion in the third quarter of 2025 and $89 billion for the full year. In China, regulators approved the top ten insurers to hold up to 1% of assets in physical gold. And on the sell side, J.P. Morgan Research has published a $6,000-per-ounce target for end-2026: an analyst forecast, not a fact, to be treated as one input among many, but a marker of how far mainstream institutional expectations have shifted.
If reserve managers, the most conservative capital on earth, are running toward gold while many pensions sit under 1%, the institutional question becomes uncomfortable in the other direction: what is the case for holding less gold than a central bank?
Conceding the allocation case does not settle the implementation question, because passive gold has three structural weaknesses.
It pays nothing. Gold has no coupon, no dividend, and no earnings stream. Every dollar of return must come from price appreciation: a pure bet on the price path, with a real carry cost once custody, storage or ETF fees are netted against a positive short-term rate.
It participates fully in its own drawdowns. A passive holder captured all of gold’s run to $5,589 in January 2026, and then all of the retracement to the $4,000s over the following months. Nothing in a bar or an ETF reduces exposure when the trend breaks, and gold has historically endured drawdowns deep and long enough to test any investment committee’s patience.
It is more volatile than its reputation suggests. Gold’s annualized volatility has typically run around 10-18% (comparable to the S&P 500’s roughly 14.3%), with spikes that decay with a half-life of about 1.6 months, per World Gold Council and State Street data. An asset that moves like equities but yields nothing is what critics mean by the “inert hedge”: it hedges the portfolio in a crisis, and does little else.
Systematic gold strategies exist to address exactly those weaknesses: gold exposure becomes rules-based and conditional rather than constant. Three families of approach dominate.
A trend overlay holds gold when the price is in a defined uptrend (measured by moving averages, breakout channels, or momentum signals) and reduces or exits exposure when the trend breaks. The logic is the same as in managed futures and CTA programs: capture the persistent moves, sidestep the deepest drawdowns, accept some whipsaw as the cost of the insurance.
The broader trend-following category illustrates both sides of that bargain honestly. The SG CTA Index was up 12.2% and the SG Trend Index up 12.3% year-to-date through June 3, 2026, a strong stretch, though attribution matters: those gains were driven largely by energy trends after the Middle East conflict began February 28, 2026, not by gold. The preceding year showed the patience cost in full: the SG Trend Index was down 9.3% through end-April 2025 in a whipsawing tape before finishing the year up 2.39%, and Morningstar data show the typical trend fund lost about 2.3% annualized over the three years through August 2025. Trend following pays for its crisis convexity with stretches of frustrating chop. Allocators should underwrite that trade-off explicitly.
A volatility-targeted gold strategy scales position size inversely to volatility: when gold’s volatility rises, exposure shrinks; when it falls, exposure expands toward the target. Because volatility clusters (and, in gold, spikes around exactly the macro events that drive the worst price gaps), this mechanically de-risks the position when conditions are most treacherous, producing a smoother return stream per unit of exposure for institutions managing to a total-portfolio risk budget.
The third family caps what any single position can lose. Rules-based stop levels, per-position risk limits, and exposure caps turn open-ended gold exposure into a series of defined-risk positions, and allow the strategy to hold no position at all when its conditions are not met. Passive gold is always fully invested; a defined-risk strategy treats flat as a legitimate state. That changes the allocator’s question from “what is my gold worth today?” to “what is the most this sleeve can lose before the rules intervene?” That is a question with an actual answer.
Framed properly, the choice comes down to four operational dimensions.
In practice, most institutions are not debating a 15% gold book; they are debating a sleeve. UBS’s Global Family Office Report 2025 found family offices holding about 2% of the average portfolio in gold and precious metals, and the 2026 edition (May 28, 2026) found a record 60% planning strategic asset-allocation changes over the next twelve months, an unusually large door for new sleeves to walk through.
A common institutional framing is a 2-4% systematic sleeve: large enough that the diversification and crisis-hedge properties register at the portfolio level, small enough that strategy-level drawdowns cannot impair the mandate. Whether that sleeve is pure gold, gold-plus-trend, or a broader systematic allocation with gold as its flagship exposure depends on what the rest of the portfolio already holds. We cover the sizing framework in more depth in how much to allocate to systematic strategies.
Once the sleeve is sized, provider selection is a due-diligence exercise. At minimum, an allocator should demand clear answers to three questions:
We publish a full framework for this process in our guide to trading strategy due diligence.
Gold has completed its migration from fringe hedge to core reserve asset; the implementation is now migrating too, from static holdings to rules-based exposure. The institutions best positioned for the next decade will treat gold not as a relic or a trade, but as a sleeve to be engineered.
Algo Alpha’s institutional program includes systematic gold strategies licensed to run in your own accounts: full transparency, zero custody transfer.
There is no single answer, but the reference points are consistent: institutions currently hold roughly 2.8% of AUM in gold (J.P. Morgan estimate), common strategic recommendations sit at 4-5%, and family offices average about 2% (UBS, 2025). A 2-4% sleeve is a common institutional framing: large enough to matter in a crisis, small enough that gold-specific drawdowns cannot impair the broader mandate.
Gold’s annualized volatility of roughly 10-18% is comparable to the S&P 500’s (~14.3%), so it is equity-like rather than bond-like risk. The institutional response is sizing and structure: hold it as a small sleeve, and consider systematic implementations (volatility targeting, trend overlays, defined-risk rules) that reduce exposure when volatility spikes rather than riding it passively.
A rules-based approach to gold exposure. Instead of holding a constant position, the strategy follows predefined rules (trend and momentum signals, volatility targets, per-position risk limits) that determine when to hold gold, how much to hold, and when to stand aside entirely. The goal is to retain gold’s diversification and crisis-hedge properties while managing the drawdowns and volatility that come with a static holding.
It is a trade-off across four dimensions. Passive gold is cheaper and operationally simple, but pays no income and participates fully in every drawdown. Systematic exposure costs more and requires provider due diligence, but is designed to truncate drawdowns and manage volatility, and a licensed strategy can run in the institution’s own account, preserving custody. Many allocators combine the two: a passive core with a systematic overlay.
Diversification away from any single reserve currency, insulation from sanctions risk, and concern over major-sovereign fiscal trajectories. The scale is historic: over 1,000 tonnes of net purchases in each of 2022-2024 and 863 tonnes in 2025 (World Gold Council, Feb 2026), and in the WGC’s 2025 survey a record 43% of central banks said they plan to increase their own gold holdings, with 95% expecting global official reserves to rise.