Benchmarking family office returns: what the data can tell you

Most family offices measure themselves against something, but the something varies and the reported results are hard to compare. Citi's 2025 Global Family Office Report found 65% of respondents naming relative investment performance as their main measure of success, and UBS's 2026 report found 68% with a formal financial performance measurement process. The peer data that exist are mostly endowment and private fund indices from Cambridge Associates, which show a median endowment return of 11.7% for fiscal 2025 against 13.4% for a 70/30 index. Family office survey returns are self-reported, collected mid year and expressed as bands, which makes them a weak benchmark for anyone and a useful check on whether an office's targets are realistic.

What benchmarks offices use

No major survey publishes a clean breakdown of benchmark choice, so the evidence is indirect. Citi's 2025 finding that 65% judge success on relative performance implies a reference index or peer group for most offices. The policy benchmark used by endowments, and documented by Cambridge Associates, is a blend of 70% MSCI All Country World Index and 30% Bloomberg Aggregate Bond Index, with private holdings compared to the public market equivalent of the same index. Offices with return targets rather than index benchmarks are also common: J.P. Morgan's 2026 report found about 55% targeting 7% to 10% a year and almost a third targeting above 11%, while Citi's 2026 survey found 41% targeting 7% to 10%. J.P. Morgan notes that its own 10 to 15 year assumptions forecast no asset class above 10%, so the top tercile of targets exceeds what the bank expects any asset class to deliver. The choice of benchmark should sit with the investment committee, whose role is described in the family office investment committee.

Peer data from Cambridge Associates

Cambridge Associates' endowment universe is the closest published proxy for a diversified, alternatives heavy family portfolio. For the fiscal year to June 2025 the median of 322 endowments returned 11.7%, the highest since 2021, with the 5th to 95th percentile spread at 590 basis points, the third lowest in four decades. The three year median of 10.0% lagged the 70/30 benchmark by 300 basis points a year, the largest underperformance in 25 years, and only 8% of endowments beat it. Over ten years the median was 7.6%, rising to 9.3% for endowments with more than 40% in private investments. The firm's private equity benchmarks add asset class detail: US private equity returned 8.7% in calendar 2025 and venture capital 21.1%, with private equity trailing public markets over most horizons under ten years.

Sources: Cambridge Associates, as linked.

Returns reported in family office surveys

Survey returns come in three forms, none directly comparable. Campden Wealth's 2025 Asia-Pacific report gives a point estimate: an average net return of 11% on assets in 2024 for its 76 respondents, with developed market equities earning 18%, gold 18%, private credit 11%, private equity funds 8% and venture capital minus 5%. Expected returns for 2025 were 6% in Asia Pacific and 5% in North America and Europe, and Campden's North American report recorded the North American expectation falling from 11% to 5% with 15% of offices expecting a negative year. Citi gives bands: in its 2025 survey, 59% of offices were up 0% to 10% year to date at June and about 25% were up more than 10%; in its 2026 survey nearly 90% reported positive year to date performance and 26% of Asia Pacific offices were up more than 15%. UBS's 2020 report used a third form, asking whether portfolios had performed in line with or above target benchmarks, which 77% said they had.

The caveats

Four limitations apply to all of these figures. They are self-reported and unaudited. They are collected mid year, so they describe five or six months rather than a full period. They are expressed as bands or averages across offices with very different allocations, so an office cannot locate itself with any precision. And the samples are drawn from bank client bases, which skews towards larger offices that already use external managers; Goldman Sachs's 2025 survey had 67% of respondents above $1bn in net worth. Private holdings add a further problem, since the office's own marks determine its reported return. None of this makes the data useless, though it means a survey average is a reasonableness check rather than a benchmark.

The practical approach is to use a simple public policy index such as the 70/30 blend as the primary benchmark, to compare private holdings with their public market equivalents, and to treat peer surveys as context in the annual report to the family. How that report should present performance is covered in family office reporting to the family, and the valuation backdrop that any forward looking target should acknowledge in what GMO's 7-year forecast says about valuations.