Most commercial property has stopped behaving like a bond. The owners built to replace the passive model are patient, involved and rarer than the market needs.
Prime office space in Britain is being taken up almost as fast as it is listed, around 83 per cent of take-up in London and the South East and 57 per cent regionally going to Grade A, while secondary stock sits empty and regional values sit some 40 per cent below their 2022 peak. This is more than a flight to quality in the usual cyclical sense. The market is repricing two different things at once, a passive asset it no longer seems to want and an operated one for which it is paying a premium to the owners getting the shift right.
For a generation, a good commercial building was a predictable, bankable cashflow with capital upside attached. It was, in effect, a bond with a roof and one whose coupon and capital value tended to rise together.
An owner, whether an institution, a REIT, a fund or a private landlord, bought it, let it to a strong-covenant tenant on a long full repairing and insuring (FRI) lease with upward-only rent, passed the running costs down and did the most profitable thing available, which was doing very little thereafter. The income was contractual, the tenant was captive, the management was transactional and the coupon arrived regularly. Pension books were built on this, and rightly so.
The same owners now survive by doing the opposite. Those letting bare floors have lost the tenants; those selling finished, serviced, managed space are charging a premium for it. Great Portland’s and British Land’s managed products command double-digit premiums to conventional rent, and managed-office supply in central London has multiplied several times over in five years. Value has moved from the lease to the operation of the building itself.
The collapse of terms
The popular story that COVID emptied the office is substantially wrong, and asset managers know it. What made a building behave like a bond was never the bricks. It was the lease, the 15-to-25-year term, the upward-only five-yearly review and the FRI structure that pushed every material cost down to the occupier, producing income so protected that the academic literature described it, without irony, as a bond with an equity kicker. It was rational for as long as the conditions held, and they held for decades.
Those conditions went (and tenant expectations were the first to turn) well before 2020. Occupiers had been pressing for shorter terms, breaks, flexibility and fit-out contribution for years; COVID accelerated a demand that long pre-dated it. A major landlord’s campuses now run on a weighted average term under six years. Managed and fitted products and all-in occupancy charges have moved the moment that sets the rent away from the five-yearly review and towards renewal, refit and re-letting, which comes round far more often and far less predictably. An upward-only clause, still standard, protects nothing on a lease that breaks before the review falls due. The machinery that manufactured duration no longer produces it, so the income can no longer be priced as if it did.
That said, the bond has survived at the very top. Long-income, index-linked property, the supermarket on a thirty-year RPI lease or the sale-and-leaseback to a government covenant, still trades as a bond because it still behaves like one, and investors who want fixed income still buy it.
The trouble is how small that slice has become and how much of the market’s capital it absorbs. Prime stock is perhaps seven per cent of supply and took a quarter of all lettings in the first three months of this year.
Lenders have followed it almost exclusively, competing to finance and refinance the same trophy assets in a margin war that has driven prime spreads down while debt funds take share, which leaves the broad middle of the market, the sub-100,000 and sub-50,000 square foot buildings most owners actually hold, financed in fragments and at a materially higher price. Beneath that sits a wall. Around £33 billion of UK commercial property loans, roughly a fifth of the £174 billion outstanding, mature this year into a far more hostile rate environment, with interest cover compressed to levels last seen before the previous crisis and perhaps one loan in six carrying no covenant that would let a lender intervene early. Set that against secondary commercial rents falling by as much as 19 per cent year-on-year and the exposure is plain. The owners who refinance comfortably will be the ones who have already turned their buildings into products. A good many of the others were never set up to operate or curate what they built or bought, whether by size or by design, and they are about to find that the refinancing conversation has become an operating conversation.
For years the lease held ownership and consequence together. Now that it has gone, the consequence has come home to the owner, which is where it belonged all along.
The building became a product
There is a macro reason this is becoming an issue now. For 15 years, near-zero gilt yields flattered every property valuation and made passive holding look like skill. As yields normalised the discount rate rose, secondary values fell to where they now sit and the only reliable way left to defend value became improving and operating the asset harder. The repricing did more than punish weak buildings. It ended the era in which owning was enough and running was optional. A building run as a product, with pre-letting investment and an operation (not just “administration”) function carries costs the coupon model never contemplated, and they recur. Office fit-out runs at roughly £213 to £316 per square foot, and lifting a building from EPC D to B costs from around £113 per square foot and well beyond once amenity is added. At the sophisticated owners these are standing portfolio programmes, British Land’s multi-year energy spend, Grosvenor’s roughly £90 million London retrofit, the price of staying competitive. The evidence that operating pays is now in the returns.
Retail, written off a decade ago, leads UK commercial property returns at 8.4 per cent in the year to May, and the centres drawing capital are the ones that have evolved beyond a single retail function and are being bought on the quality of that evolution.
Then there is the part that never reaches the capex line. Tenants now buy a relationship, service, flexibility, shared fit-out risk, performance data and a reason to stay that owes nothing to a break penalty. Curating that is an operating discipline, and most owners hold neither the discipline nor the people in-house. The building now needs running, and running requires holistic skills and a different approach.
Who is built to hold it and who is not yet
So the real question beneath the hybrid-work headlines is one about owners. If a building is now an operating business, who is built to own one and hold it for the decade it takes to compound?
Consider what the asset now rewards. It rewards capital that can hold through a full cycle without a fund deadline forcing a sale. It rewards an owner who can decide in weeks and carry a position others would find too idiosyncratic. And it rewards a particular posture towards control, the owner who keeps the few decisions that set the value of the place, hands the daily running to a specialist operator and retains both the right and the information to judge that operator’s performance. Pension funds were built for a different asset. The capital built for this one is patient private capital, and at its most developed, the family office.
The fit deserves a precise statement. Family offices are not, based on the evidence, driving this shift; the visible innovation still comes from the REITs and specialist operators. What the family office has is the right shape, direct investment, a preference for control and patience measured in generations rather than investment fund cycles. In real estate it behaves as capital seeking selective control, direct or co-invested equity with governance rights and reporting and a firm disinclination to run the boiler. The fully developed version is the great estates such as Grosvenor, Cadogan and Howard de Walden, the type of family capital that never stopped operating its property. The important caveat is that these are large operating estates with deep teams (and pockets), where six in 10 family offices run on fewer than half a dozen staff. The profile that fits perfectly today is narrow. It is essentially the family that already runs a platform. Which is where the opportunity sits. The mid-sized single-family office and the multi-family office that still hold property passively, through funds and REITs, are the capital that could fill the gap the institutions have vacated. They will fill it through the route the sector already uses for everything complex, the club deal and the operating-partner joint venture, and they will fill it without becoming facilities managers. One family leads and carries the diligence; trusted co-investors follow; a specialist operator supplies the running and curating capability. The family brings patient capital and keeps selective control; the operator runs the place.
The scarce seat in that structure belongs to whoever can read both the capital stack and the trading floor, holding the sponsor’s value-creation plan and the covenant headroom in one hand and the tenant mix and the footfall in the other. Private equity has a name for that seat, the “operating partner”, and has been hiring for it in other sectors for a decade. Property has barely begun.
The commonest special situation in the sector now is a completed place that has opened, disappointed on yield and discovered that no one in the structure can read both the P&L and the floor well enough to fix it. To hold such an asset sustainably takes three things, an operating partner the owner genuinely trusts, the data to supervise that partner at arm’s length and a holding structure built for the operating decade. This owner is the private-sector cousin of the accountable operator this series has argued the public realm lacks. The same role, priced as an investment.
That arm’s-length supervision has only recently been made practical and democratised. Property was always data-poor and fragmented, and even where granular geo-demographic data existed it came at a hefty cost. The change that matters has little to do with the fashion for AI running buildings, which a serious owner can smell as a gimmick. It lies in the falling cost of gathering, analysing and making sense of operating data, so that active ownership is now tractable for owners who could never previously have justified the machinery. It lets patient capital supervise an operator without building a platform. It enables the model. The people in it still have to be found.
What would let the capital in
If the right capital is identifiable, the fair question is why so little has arrived. The signals are muddy. The tax treatment of building works still turns on the repair-versus-improvement line, relieving a repair now but capitalising an improvement slowly, so the cash for a retrofit goes out today while the relief dribbles back over years, penalising exactly the active investment the market now demands. The regulation compounds the problem. Since April 2023 no commercial building rated F or G may be let at all, so EPC E has been the working floor across the market. In June 2026 the government confirmed that privately let commercial buildings over 1,000 square metres will need to reach EPC B from 2031, dropping the interim step it had previously proposed (an EPC C requirement by 2027), leaving smaller stock at the E floor, conditioning the whole thing on cost-effectiveness and deferring the secondary legislation that would make it law. It is a direction of travel an investor cannot yet underwrite. Underneath sits a government low on fiscal room and on market confidence, which makes the certainty capital needs harder to supply precisely when it is most useful. None of this is a call for a scheme. It is a question worth leaving with a policymaker. What would make the case clear enough for this capital to commit? Faster relief for the repositioning and decarbonisation the state says it wants. A firm energy standard in law, because a case cannot be built on a trajectory. And for genuinely difficult assets, the town centres and mixed-use schemes this series has followed, a structure that pairs public mandate with patient private capital and a named operating partner, of the kind Homes England already runs through its regeneration vehicles. Another grant, or a sale to whoever will take it, solves little or nothing.
The passive coupon has gone. The asset now needs an owner who stays, runs it and holds it through the decade that decides whether it was worth owning.
Those owners exist and are more identifiable than the hand-wringing suggests, though there are too few of them. Fewer still are the true operators and curators, as distinct from the administrators staffing most asset managers, who can run a commercial mix well enough to deserve the capital; the demand this shift creates would outrun their number and their capability tomorrow. Whether the rest arrive turns on three things being converging at the same moment. The rules, the relief and the operators. Two of those sit in the government’s gift. The third is the industry’s to build, and it has barely begun.
Related Insights in the series: