The honest answer first: there is no universal number, and any article that promises one should be read skeptically. What does exist is observed practice, and it clusters. The UBS Global Family Office Report 2025, covering 317 single family offices with an average net worth of $2.7 billion, found hedge funds at roughly 4% of the average portfolio. Systematic sleeves at technology-forward institutions commonly run 2-4% of assets. And asset-manager research on diversifying strategies generally argues for 5-20% before the portfolio-level effect becomes measurable, a figure worth noting, and worth noting that it comes from managers with products to sell.
The number matters less than the logic that produces it. A 3% sleeve arrived at through a clear sizing framework (job definition, drawdown arithmetic, governance capacity) survives committee scrutiny and a bad year. The same 3% arrived at by copying a peer gets cut at the first drawdown, usually at the worst time. What follows is the sizing logic institutions actually use, with the data current as of mid-2026.
Start with the base rates. The UBS Global Family Office Report 2025 put the average single-family-office portfolio at roughly 30% equities, 21% private equity, 4% hedge funds, and 2% gold and precious metals. That 4% hedge fund figure is the broadest available proxy for what sophisticated private capital devotes to absolute-return strategies of all kinds. Systematic sleeves are a subset of it.
The direction of travel matters as much as the level. The 2026 edition of the same report (May 28, 2026; 307 family offices) found a record 60% planning strategic asset-allocation changes over the next twelve months, with capital rotating out of private equity toward hedge funds and liquid diversifiers. Preqin's 2025 survey of roughly 435 institutions found the same tilt, with macro and systematic strategies specifically favored for volatile regimes. Institutional peers are covered in more depth in our review of how pension funds and endowments use systematic strategies.
So the honest benchmark for a first systematic allocation is not 10% or 15%. It is low single digits, deliberately.
The case for the sleeve is diversification of method, not just of asset. Most portfolios are already diversified across asset classes; very few are diversified across decision-making processes. A rules-based strategy takes positions for reasons uncorrelated with a discretionary committee's reasons: it does not read the same headlines, hold the same meetings, or share the same behavioral biases. That is the axis of diversification systematic strategies are built to supply.
The category has scale behind it. The hedge fund industry crossed $5 trillion for the first time in late 2025 and ended the year at a record $5.15 trillion (HFR, January 2026), and within it the HFRI Macro: Systematic Diversified index returned +7.0% in the first quarter of 2026. Flows tell the same story at the liquid end: managed-futures ETFs kept gathering assets straight through a weak 2025 (Simplify's CTA fund added roughly $570 million and crossed $1 billion), behavior Morningstar and ETF Trends commentators described as "a bull market for diversification." Allocators were not buying the recent return; they were buying the method.
Sizing begins with what the sleeve is hired to do. There are three common mandates, and they produce different sizes. A crisis offset (something expected to perform when equities fall) needs to be large enough that its gain in a bad year is visible against the portfolio's loss, which argues for the higher end of any band. A standalone return stream is judged on its own economics and can start smaller. A general diversifier (the most common institutional framing) sits in between: large enough to move portfolio-level statistics, small enough that its worst year is tolerable. Write the job down before discussing the number; committees that skip this step end up cutting the sleeve for doing exactly what it was hired to do.
Systematic returns arrive in lumps, and the recent record makes the point better than any theory. Trend-following strategies were down 9.3% year-to-date at the end of April 2025 (SG Trend Index) and finished that year roughly flat at +2.39%; Morningstar put the typical trend fund at about -2.3% annualized over the three years through August 2025. Then 2026 paid: the SG Trend Index was up 12.3% year-to-date through June 3, 2026, with the broader SG CTA Index up 12.2%, driven largely by energy trends after the Middle East conflict began in late February (CNBC, June 5, 2026).
An allocation is sized correctly when its worst plausible year is a line item the committee can live with, not a reason to call an emergency meeting.
The investors who captured the 2026 recovery were the ones still holding the sleeve after the 2025 whipsaw. That is the real constraint on size: the allocation must be small enough to hold through the years that test conviction, because the payoff years tend to follow them unannounced.
The most robust sizing method starts from the worst case rather than the expected case. Take the strategy's worst plausible drawdown (from live history, stress scenarios, and a margin of skepticism beyond both) and compute what it would cost the total portfolio. The arithmetic is deliberately simple: a 3% sleeve suffering a 20% strategy drawdown costs the portfolio roughly 60 basis points; a 5% sleeve in a 30% drawdown, about 150. Work backward from the number the committee can genuinely tolerate, and the sleeve size falls out. (On how drawdown figures are measured and why they are routinely underestimated, see our primer on maximum drawdown.) This is educational arithmetic, not advice: the tolerance figure is a governance decision worth pressure-testing with your own advisers.
Observed practice sorts into three rough bands:
Two choices affect how quickly an institution moves through those bands.
Start small and scale on verified live results. The scaling trigger should be evidence, not enthusiasm: live performance reconciled against the provider's stated track record, operational conduct through at least one difficult stretch, and clean answers to the questions in our due-diligence framework for trading strategies. A pilot that fails that review should not be scaled. That is the pilot doing its job.
Structure choice affects sizing comfort. An allocator deciding between a commingled fund, a separately managed account, and a licensed strategy run in its own brokerage account is also deciding how much operational risk accompanies the strategy risk, and committees can rationally size up sooner in structures where assets never leave their own custody, because a large category of failure modes is removed. The market has been voting this way: hedge fund SMA assets grew from 3.4% of industry AUM ($66 billion) in 2010 to 7.1% ($315 billion) in 2024, with investor requests for SMA access up roughly 50% in 2024 alone (Dechert, October 2025), and Hedgeweek's 2026 "Separate Ways II" study found 60% of allocators reporting SMA demand up over the past twelve months, with none reporting a decline. The full comparison is in SMA vs. fund vs. licensed strategy.
Revisit on a cadence, not on emotion. The sleeve size belongs in the annual investment-policy review, with defined event triggers: a breach of the modeled worst-case drawdown, a material change at the provider, a shift in the portfolio's risk budget. Between checkpoints, leave it alone: a sleeve resized every quarter on recent performance is no longer systematic.
Algo Alpha structures institutional systematic sleeves of 2-4%, run in your own accounts with zero custody transfer. Details at algoalpha.co/institutional.
There is no universal figure. Family offices average roughly 4% in hedge funds overall (UBS Global Family Office Report 2025), systematic sleeves at technology-forward institutions commonly run 2-4%, and asset-manager literature argues 5-20% for diversifying strategies to have measurable portfolio effect. The defensible number comes from your own drawdown tolerance and governance capacity, decided with your advisers, not from copying a peer.
Below roughly 2%, the sleeve's contribution to portfolio results is statistical noise, but that does not make it pointless. A sub-2% pilot is a legitimate phase whose purpose is operational learning and live verification of the strategy, not portfolio impact. The mistake is judging a pilot on portfolio effect, or leaving it at pilot size indefinitely once it has passed verification.
The UBS Global Family Office Report 2025 (317 single family offices, average net worth $2.7 billion) found hedge funds at about 4% of the average portfolio, versus 30% equities, 21% private equity, and 2% gold. The 2026 edition found a record 60% of family offices planning allocation changes, with capital rotating out of private equity toward hedge funds and liquid diversifiers.
Drawdown. Expected returns are estimates; drawdowns are the events that actually force committees to abandon allocations. Take the strategy's worst plausible drawdown, multiply by the sleeve weight, and confirm the portfolio-level cost is tolerable: a 3% sleeve in a 20% strategy drawdown costs roughly 60 basis points. Sizing from the worst case is what makes the allocation survivable, and surviving the bad years is what earns the good ones.
On evidence, at a defined cadence. Reasonable triggers: live results verified against the provider's stated track record over a meaningful period, at least one drawdown held through with the strategy behaving as documented, and clean operational conduct throughout. Scale at the annual policy review rather than mid-year on enthusiasm, and note that custody-retaining structures such as SMAs or licensed strategies remove a category of operational risk, which is why committees often scale them sooner.