Family Offices: The Move to Operated Prime

Revision note, 27 September 2026#

The first version of this piece, published on 3 September 2026, read the family-office real-estate weighting from the UBS Global Family Office Report 2025, although the 2026 edition had been out since 28 May 2026. It described the allocation as having barely moved, 10 to 11 per cent. The 2026 edition, built on 307 family offices surveyed between 22 January and 30 March 2026, puts real estate at 11 per cent of strategic allocation in 2025, flat on 2024, and records the 60 per cent of offices planning changes intending to take it to 8 per cent in 2026. The UBS media release calls it reduced exposure to real estate. The 2025 edition had itself recorded a planned return to 10 per cent, which the first version did not report. The two new figures are not like for like. The 11 per cent covers all respondents, the 8 per cent only those planning changes, and it is an intent, not a record of transactions. What remains of the thesis is the compositional argument, which was always triangulated rather than counted and which the new survey does not measure either. It now applies to a line that is flat and planned lower, not to one that edged up from 10 to 11 per cent.

The same review corrects a second reading of UBS. The first version said UBS recorded family offices leaning toward direct holdings within their alternatives. UBS does not split real estate between direct and fund holdings. In private equity, where it does, direct investments fell from 11 to 8 per cent of allocations between 2024 and 2025 while funds went from 10 to 9, and the 2026 plan holds both flat (UBS Global Family Office Report 2026, survey-measured). The preference for direct ownership in this piece is a structural argument, and it is now labelled as one.

The composition moved before the label did#

The family office did not discover real estate in 2024. It changed how it holds it. UBS puts the average strategic allocation to real estate at 11 per cent in 2025, level with 2024 and up from 10 per cent in 2023. The offices planning changes for 2026, 60 per cent of the 307 surveyed, intend to take it down to 8 per cent (UBS Global Family Office Report 2026, survey-measured; the 2026 figure is the stated intent of that subset, not a record of transactions). The line marked real estate is flat, and its stated direction is down. The argument of this piece is that the line is also being re-populated, and the surveys that measure the line do not, for the most part, see the substitution happening inside it.

The population holding that line has itself expanded. Deloitte estimates 8,030 single family offices worldwide in 2024, up from roughly 6,130 in 2019, a 31 per cent rise, with a projection to 10,720 by 2030 (Deloitte, Defining the Family Office Landscape 2024, modelled estimate). Assets under management are put at US$3.1 trillion today, projected to reach US$5.4 trillion by 2030. These are modelled figures, not a census. No registry of family offices exists, and the definition varies by author. The direction is not in dispute. The precision is.

The relevant question for an allocator is not how much sits in real estate. It is what kind. The industry surveys report a single aggregate called real estate. They rarely separate property held passively, a trophy title deed generating little, from property held as an operating business that produces income through the year. That separation is where capital is actually moving, and it is precisely the movement the aggregate conceals.

This piece takes the second reading. It treats the 11 per cent not as a static weighting but as a container whose contents are shifting from dormant to operated. The evidence for the shift is partly measured, partly triangulated, and it is labelled as such throughout. Where a number is an intent rather than a transaction, it is named an intent. Where a count is a model rather than a census, it is named a model.

The count, and its caveat#

Institutionalisation is the word the industry uses, and it is doing real work. Deloitte puts the wealth of families operating a family office at US$5.5 trillion in 2024, up from US$3.3 trillion in 2019, with a projection to US$9.5 trillion by 2030 (modelled estimate). A vehicle that was once a private ledger for a single fortune increasingly runs an investment committee, a mandate, and a policy document. The behaviour that follows is not amateur behaviour. It is closer to that of a small endowment, and endowments hold operating real assets, not only paper claims on them.

The regional distribution matters for property specifically. Deloitte's 2024 count places roughly 3,180 single family offices in North America, 2,290 in Asia-Pacific, 2,020 in Europe, 290 in the Middle East, 190 in South America and 60 in Africa (modelled estimate). The weight sits in jurisdictions where prime real estate is priced against genuine scarcity, and where a family office can plausibly own and run an asset rather than merely subscribe to a fund that does. Concentration of capital and concentration of scarce supply are meeting in the same postcodes.

The caveat travels with every one of these numbers. They are estimates produced by a model, not returns filed with a regulator. The term family office covers a spectrum from a single administrator to a fifty-person institution, and different studies draw the boundary in different places. The surveys that report allocation, UBS chief among them, rest on self-selected samples in the low hundreds. None of this makes the figures useless. It makes them directional at the second decimal and reliable only at the level of trend.

Held to that standard, the trend is unambiguous. More vehicles, more assets, more professional governance, and a rising appetite for real assets that do something rather than merely sit. The interesting claim is not that family offices like property. They always have. It is that the kind of property they are prepared to underwrite has changed, and the survey line has not yet caught up to the change.

Why real estate holds its place#

Real estate earns its allocation for reasons that survive the cycle, not because of a single vintage. It produces income that is contractual rather than discretionary. It carries a tangible claim that can be pledged, insured and passed to the next generation. It moves imperfectly with public equities, which is the whole point of holding an alternative at all. UBS records family offices with 42 per cent of strategic allocation in alternative classes in 2025, of which private equity is the largest single line at 17 per cent and real estate sits at 11 (UBS Global Family Office Report 2026, survey-measured). The alternatives book is not a hedge against equities. It is the portfolio.

The holding horizon is the family office's structural advantage, and property rewards it. A vehicle that measures itself in generations does not need to mark to market every quarter or exit into a soft window. It can underwrite an asset that yields patiently and revalues slowly. That patience is worth little against a passive claim on a fund with a ten-year life and a general partner's clock. It is worth a great deal against an asset the office can hold, improve and operate on its own timetable.

Inflation and currency are the second-order reasons, and they are not decorative. Prime property in a hard-currency jurisdiction is, among other things, a store of value denominated away from the family's domestic risk. For principals whose wealth was created in one currency and whose obligations increasingly sit in another, a prime asset abroad is a currency position with a roof on it. The income it throws off, if it throws off income, compounds that hedge rather than leaving it dormant.

None of these reasons favour the trophy over the operating asset. Several favour the reverse. Income through the cycle, an inflation link that resets, a claim that can be actively managed, all of these are properties of real estate that is worked, not real estate that is merely owned. The structural case for the asset class is, on inspection, a case for the operated end of it.

From holding to operating#

The passive trophy and the operating asset are the same category on a survey line and two different investments in practice. A prime apartment held empty for optionality is a bet on capital value alone, with a carrying cost and no coupon. A prime residence run as a serviced, branded or rental asset is a bet on capital value plus an income stream, with an operator between the principal and the tenant. The first is a call option with negative carry. The second is an allocation with a yield. The 11 per cent line contains both, and the mix inside it is tilting toward the second.

The tilt is legible in what family offices say they want, even where the surveys cannot yet count it. Knight Frank's family-office survey, the Knight Frank 150, found 44 per cent of respondents intending to expand their exposure to commercial property over the following 18 months, and 25 per cent of those with existing residential portfolios considering further purchases (Knight Frank, The Wealth Report 2025, survey-measured intent). Commercial property is, by construction, operating property. It is leased, managed and underwritten on its income. An expansion of appetite for commercial exposure is an expansion of appetite for real estate that works.

The same logic reaches residential once residential is operated. A resort residence with a rental programme, a branded apartment with a hospitality operator, a second home underwritten partly on the nights it lets, each of these is a residential title deed wrapped around an operating business. The family office that wants income and a hard asset in one instrument is drawn to exactly this hybrid. The trophy that pays nothing does not disappear from the portfolio. It stops being the growth edge of it.

This is a shift in what is underwritten, not merely in what is purchased. Buying a trophy asset is a valuation exercise. Buying an operating asset is a valuation exercise plus an operations exercise, and the second dominates the return. The family office moving down this path is not simply spending more on property. It is taking on a different kind of diligence, and, done properly, a different kind of return.

The professionalisation that changes the asset#

The clearest long-run evidence is the trajectory of commercial property inside private portfolios. Knight Frank's Attitudes Survey recorded HNWI allocation to commercial real estate at 2.6 per cent in the first edition of The Wealth Report and at 21 per cent by 2023 (survey-measured). That is not a change in taste. It is a change in capability. Private capital learned to underwrite, hold and operate income-producing real estate at institutional standard, and the allocation followed the capability.

Knight Frank's own reading attributes the shift to professionalisation and to flexibility. Its head of global capital markets describes private investors as able to make quick decisions, draw on diverse capital streams and tolerate risk in ways that let them act through cycles where institutions, sovereign funds and private equity withdrew, notably in London offices when rates rose sharply from 2022 (Knight Frank, The Wealth Report 2026). The observation that matters for this argument is the reason he gives for staying invested through a downturn: even then, the asset was generating income. Operating property pays while it waits.

Flexible firepower is a genuine edge, and it points at operating assets specifically. An institution constrained by mandate and redemption cycle is a poor natural owner of a hotel-backed residence or a value-add operating asset. A family office with a generational horizon and its own decision rights is a better one. The professionalisation Knight Frank documents is precisely what makes the operated asset ownable by private capital. Twenty years ago the family office bought the trophy because operating was beyond its bench. It no longer is.

The direction of travel is therefore self-reinforcing. More capability invites more operating exposure, which builds more capability. The 11 per cent that UBS measures is a snapshot of a portfolio class that is still climbing the operating curve, not one that has settled at the top of it.

Direct over fund#

Operating an asset requires owning it directly, and that is a structural argument, not a survey finding. A fund interest delegates the operation, the fee and the timing to a general partner. Direct ownership keeps all three with the principal. For a vehicle that has built the capability to underwrite operations, the fund wrapper subtracts control and adds cost without adding much the office cannot now do itself. The surveys do not yet show the preference. UBS does not split real estate between direct and fund holdings, and in private equity, where it does, direct investments fell from 11 to 8 per cent of allocations between 2024 and 2025 while funds went from 10 to 9, with both held flat in the 2026 plan (UBS Global Family Office Report 2026, survey-measured).

The fee arithmetic is not incidental. An operating asset held directly captures its net operating income for the family. The same asset held through a fund surrenders a management fee and a carry before the family sees a coupon. Over a generational hold, the compounding difference between gross and net of two layers of fee is not a rounding error. It is a material part of the case for the professionalised family office holding the asset itself.

Direct ownership also aligns the asset with the family's other objectives. A residence the family can use, a hospitality asset that carries the family name, a development the office can shape to its own standard, none of these survive intact inside a commingled fund. The family office that wants the asset to do more than compound, to be used, to be controlled, to be inherited, has to hold it directly. The operating asset is the one where that preference bites hardest.

The cost of the preference is concentration and effort, and it is real. Direct operating ownership is undiversified, illiquid and management-intensive by construction. The office that chooses it is choosing to run a business, not to clip a coupon. That choice is defensible only where the office has, or buys, the operating capability to make it. Where it does not, the fund wrapper it would give up was protecting it from a risk it may not have priced.

The weaknesses, honestly disclosed#

The central weakness of the thesis is that its key term is not directly measured. No major survey reports a clean split between passively held and operated real estate inside the family-office allocation. The shift from held to operated is inferred from adjacent measures, the rise of commercial exposure and the professionalisation trajectory, and it is triangulated rather than counted. The preference for direct holdings is argued here, not measured, and in private equity the UBS figures point the other way. An allocator should treat the direction as well supported and the magnitude as directional only.

The counts themselves are estimates, not returns. Deloitte's 8,030 family offices and US$3.1 trillion in AUM are modelled figures built on a contested definition, with no registry to check them against (modelled estimate). The allocation surveys rest on self-selected samples in the low hundreds, so a one-point move in a reported average is inside the noise. The numbers in this piece are reliable at the level of trend and unreliable at the level of the decimal, and are labelled accordingly.

Intent overstates action, and this argument leans on intent in places. The 44 per cent of family offices intending to expand commercial exposure is a statement of appetite collected at a point in time, not a record of capital deployed (survey-measured intent). Appetite surveys are systematically optimistic. Some of the intended expansion will not happen, some will happen in a different asset, and the survey cannot tell an allocator which.

Finally, the operated asset carries risks the passive one does not. Operating exposure adds management risk, key-person risk, operator-default risk and a deeper illiquidity than a trophy that can at least be sold to another passive buyer. The income that makes the operated asset attractive is the same income that can be interrupted by a bad operator or a soft season. The case for operating over holding is a case for a higher and more skill-dependent return, not for a free one. The UBS survey used here was also collected, between January and March 2026, against an unusually disrupted macro backdrop, which colours any single year's reading.

What this means for the allocator and the principal#

The signal an allocator should take is compositional, not directional on price. The family-office weighting to real estate is not surging. On the latest UBS survey it is flat at 11 per cent, and the offices planning changes intend to cut it to 8. What is moving is the character of the holding, from a passive claim on capital value toward an operating asset that produces income and can be controlled directly. An allocator reading only the headline weighting sees a flat line planned lower. The change this piece argues for sits inside the line, and the headline does not show it.

The principal should underwrite operations as the first-order question, not the second. For an asset held to be operated, the operator, the income model and the management structure determine the return more than the entry price does. The diligence that suits a trophy purchase, comparable values and a view of the market, is necessary but no longer sufficient. The diligence that suits an operating asset is closer to underwriting a business, and a principal who brings only the former to the latter is exposed on the part that dominates the outcome.

The prudent posture is to hold the two readings together. The counts and allocations are trend-reliable and decimal-unreliable; the shift from held to operated is well supported in direction and directional in magnitude; the operated asset offers a higher, income-bearing, more controllable return and demands a heavier, more skill-dependent diligence to earn it. An allocator who accepts all three of these at once is reading the family-office data as it actually reads, rather than as a marketing document would prefer.

Skin in the game disclosure. Victaura, through its parent Greystone B.V. (Netherlands), holds an active operating position in operated prime residential and resort real estate. Readers should assume commentary may be influenced by, or benefit, Greystone's position. This document is classified as marketing material under MiFID II Article 24(3). It is not investment advice.