Single family office or multi family office: the structural decision

Anyone intending to carry wealth into a third generation eventually faces the structural question: establish a single family office of one’s own — or join a multi family office that offers the same services to several families. The question is usually posed as one of cost. It is a question of control. An own office buys confidentiality and control; a shared office buys professionalism and scale effects — and gives up something that never appears in a fee schedule: the exclusivity of the insight.

This article puts both structures side by side: the wealth thresholds below which an own office does not pay off, the robust cost figures for both models, the market data, the hybrid forms in between, and the German supervisory and tax position. Every figure carries its source; where the market knows only conventions, that is stated too.

The threshold for an own family office sits at roughly $100 million of investable assets — and even then, studies put the running cost between about $875,000 and $6.6m a year. Joining a multi family office starts from a tenth of that. What is left on the table is sole control.

The threshold: when an own office pays off

Three sources, three very similar numbers. Deloitte Private cites a common benchmark of a minimum of $100 million in investable assets to justify the start-up and ongoing operating costs of a single family office. Morgan Lewis (2026) places the practical threshold at $100m to $250m where a dedicated own office becomes economically rational. Berenberg (2025) considers founding an SFO generally sensible only from roughly €200 million under management, because only then do scale effects and cost synergies arise. They are benchmarks, not hard lines — but the bandwidth is narrow enough to serve as a decision datum.

The costs thereafter are absolute, not relative: according to the J.P. Morgan Global Family Office Report 2026 (a survey of 333 family offices in 30 countries), an SFO with wealth up to $250m costs around $875,000 a year; at $250m to $500m it is about $1.7m, at $500m to $1bn roughly $3.2m, and above $1bn about $6.6m — 25 to 28 per cent of costs go to external services such as legal, trading and cybersecurity. A Deloitte review of 187 family offices arrives lower: an average of 0.41 per cent of assets under management, just over $1m a year for an office with $250m of active assets. Campden Wealth (2025) puts small family offices below $250m at 50 to 60 basis points and annual budgets of $1.0m to $1.5m.

The widespread rule of thumb of "0.5 to 1.5 per cent of wealth per year" circulates across the industry — it is a convention, not a study value, and the robust study figures partly sit below it. Two qualifications belong alongside: first, per a cost analysis by Westwick, headline benchmarks capture only about 57 per cent of true costs — asset-management fees, bank charges and lifestyle assets are usually left out. Second, the families that carry an own office are not comparing it to a price but to the loss of control in the alternative.

The multi family office: scale effects with a systematic price

A multi family office sells the same services to several families and shares the cost. Fee models are asset-based, flat (retainer) or hybrid; bandwidths evidenced in secondary sources run from 0.2 to 1.25 per cent of AUM (Defiant Capital) and 0.5 to 1.0 per cent plus a retainer (Aleta). The often-quoted range of "0.25 to 1 per cent" is, in that exactness, industry convention rather than study evidence.

Minimums vary accordingly: Defiant Capital cites entry points from $10m to $30m; Morgan Lewis (2026) considers MFOs suitable for families with $25m to $150m; Masttro names $30m to $100m. For families below $10m, most MFOs are simply not the audience. Rödl Partner note that lean, digital MFO offerings are available on the market from €100,000 — a figure that shows what is bought here is not an apparatus but a service.

The systematic price sits in the structure, not the fee. Morgan Lewis (2026) states the advantage of the own office in three words: control, confidentiality, customization — and describes the MFO in reverse: customisation is limited by the platform’s standardised processes, and although reputable multi family offices maintain information barriers, the family’s data, reporting and investment activity flow through common systems and personnel. Whoever has other families as co-users of the same platform does not always control their insight — it depends on the provider’s information architecture.

Discretion as a structural feature

The difference can be traced down to the name. Berenberg (2025) describes that single family offices are not bound to a legal form, are subject to regulation only to a degree, and that the name of the office often reveals nothing about the owning family — invented names are "not a rarity"; some appear as a family holding or under their own firm. The own structure is thus not only a vehicle of administration but also a shell of privacy: it can negotiate, buy and sell without the person behind the mandate becoming visible.

That is precisely where the shared model meets its limit — and why discretion in service providers deserves its own due diligence. How such a review works is described in our piece on vetting luxury service providers; privacy as a good in its own right is the subject of discretion and privacy in premium services.

The market: 8,030 single family offices, and growth that creates structure

Deloitte Private ("Defining the Family Office landscape", 2024, surveying 354 SFOs) counts 8,030 single family offices worldwide — up 31 per cent from 6,130 in 2019. The forecast for 2030 is 10,720 — in Europe up from 2,020 in 2024 to 2,650 in 2030. Assets under management across all family offices are expected to grow from $3.1 trillion (2024) to $5.4 trillion (2030); 68 per cent of today’s SFOs were founded after the turn of the millennium. Germany has no official figure because there is no registration duty — Berenberg (2025) roughly estimates 1,200 SFOs in the DACH region, 600 to 700 of them in Germany.

A growth figure for the MFO sector in the narrower sense is published in none of the primary studies reviewed (Deloitte 2024, J.P. Morgan 2026, KPMG/Agreus 2025, BlackRock 2025, Goldman Sachs 2025, Campden 2025); commercial market researchers’ estimates refer to the family-office services market as a whole, not to multi family offices. Morgan Lewis (2026) records qualitatively that MFOs have professionalised significantly in recent years — with access to deal flow, co-investment opportunities and specialised advisory that smaller own offices struggle to replicate.

Hybrids: the reality in between

Most families do not choose between two extremes but move between stages. Morgan Lewis (2026) describes the virtual family office: a lean internal team of one to three professionals handling coordination, governance and vendor management, while accounting, tax, compliance, legal and investment management are substantially outsourced — best suited to families in the $50m to $200m range and to families in transition after a liquidity event. The common trajectory begins with a multi family office or an outsourced arrangement in the years after a sale, moves into a hybrid model, and can mature into a full own office as assets grow — the way back from SFO to MFO on generational change or a decline in wealth is likewise described.

Two surveys show what the hybrid looks like in practice. Goldman Sachs (Family Office Investment Insights Report, September 2025, 245 decision-makers) puts 70 per cent of investment needs in-house and 30 per cent outsourced. BlackRock (2025 Global Family Office Survey, June 2025, 175 SFOs with over $320bn AUM) measures the capability gaps that are bought in deliberately: reporting 57 per cent, deal sourcing 63 per cent, private-markets analytics 75 per cent. And 38 per cent of respondents in Deloitte’s 2024 survey expect family offices embedded in operating companies to spin out into standalone structures — the structural question is thus also a question of timing, not only of wealth.

The German position: licence-free, not rule-free

Under the BaFin notice "Hinweise zur Erlaubnispflicht gemäß KWG und KAGB von Family Offices" (14 May 2014), managing one’s own family wealth is in principle free of licensing requirements — it is business in one’s own name. The same holds for family wealth administration: where the wealth owner also manages the private wealth of close family members (spouses, life partners, parents, siblings, children, nephews and nieces, uncles and aunts, first cousins) through employees or a controlled company, this too is licence-free family wealth administration — provided the services are not offered to the market. The group privilege of § 2 (1) no. 7 in conjunction with § 2 (6) KWG completes the picture for asset-managing companies.

On the supervisory side, the AIFM Directive sharpens the picture: recital 7 — spelled out by ESMA in its guidelines of 24 May 2013 (ESMA/2013/600) — makes clear that a family-office vehicle investing the private wealth of family members without collecting capital from third parties is not an AIF. A licensing duty under § 20 (1) KAGB or § 32 (1) KWG arises only once an offering to third parties emerges or capital is collected after all. Frequently quoted but often misplaced is § 2 (4) KAGB: it is not a "family-office exemption" but the small capital management company rule — vehicles managing only special AIF and staying below €100m (or €500m without leverage and without five-year lock-ups) get by with registration and a limited catalogue of rules instead of full licensing under § 20 (1) KAGB.

On tax, the term family office is not legally defined in Germany (Rödl Partner); the legal form determines the tax format, and cross-border constellations create their own compliance burden. One relevant lever on the structural question comes from Berenberg (2025): direct holdings of more than 25 per cent in companies can, under certain conditions, qualify as privileged business assets under the German inheritance tax — a point that sits upstream of any fee model. How wealth transition is built structurally is covered in our piece on family office succession planning.

The decision: five questions instead of one formula

The structural question resists a formula, but it sorts into five questions. First, size: below roughly $100m an own office is not justifiable by common benchmark; from about €200m Berenberg calls it sensible. Second, the need for control: anyone who wants to set their own investment philosophy, personnel and technology will chafe against standardisation in a shared model; whoever can rely on platform processes saves substantially. Third, the staffing effort: an own CIO and CFO, succession planning for two to three professionals and any household staff are organisational tasks that, in an MFO, sit elsewhere. Fourth, discretion: whoever wants neither data nor transactions flowing through shared systems pays for exclusivity. Fifth, the family’s own competence: without internal steering, even the own office is an apparatus without a head.

The realistic answer for most families is neither extreme but a stage: virtual, then hybrid, then an own office — and, on generational change, sometimes back again. The counsel of the law firms to review the structure every three to five years or on significant family events is therefore not a formality but the actual control mechanism.

How a family office and a concierge differ as service layers is set out in our comparison of family office vs private concierge.

Frequently asked questions

The quiet way to a decision

The structural question is rarely asked at a freely chosen moment — usually after a sale, an inheritance, a move. Whoever answers it in writing also recognises which services to buy and which to keep — and whoever seeks the discretion of an own office without carrying the apparatus needs, above all, reliable access to vetted providers.

The Allocation1 does not sell wealth structures and charges no commission on any mandate. We structure your requirement and present it exclusively to vetted providers. You may request an invitation.

Access to The Allocation1 is by invitation. We review requests carefully and respond discreetly.