# What the office model cannot see

- **Source:** EdgeProp Singapore
- **Published:** 2026-08-25T00:19:58.000Z
- **Author:** Jaelle Ang
- **Original:** https://www.edgeprop.sg/property-news/what-office-model-cannot-see
- **Topics:** Commercial & Industrial, Investment & Capital Markets, Regional Markets

## Featured rationale

The shift may favour actively operated offices with accountable managers and occupier-use metrics, potentially influencing asset repricing, ownership suitability and investment decisions across the segment.

## AI summary

Singapore office assets remain predominantly lease-managed despite higher interest rates, slower cap-rate adjustments and regulatory deadlines making passive ownership and deferred capital expenditure costlier.

## Original article

Two buildings, same road, same vintage, same land cost.

In the first, one person’s professional standing rests on how the building did last month. She knows the rate, the margin, the three competitors she is measured against and exactly where she sits among them. If the arrival is miserable on Tuesday morning, it is her problem by Tuesday afternoon.

In the second, there is a fund manager, an asset manager, a property manager and a leasing agent. All four are good at their jobs. None of them would say the arrival is theirs, and each would be right.

The first building is a hotel. The second is an office. That nobody finds this strange is the most interesting thing about it.

Read also: Frasers Property's walk, run, cycle campaign raises over $214,000 for seniors and persons with disabilities

Singapore runs its hotels as businesses. It runs its malls as businesses. Its offices are still run as leases — and the spread that made that acceptable has gone.

What a general manager actually does

I spent several years building hotels in Bangkok before I ever ran anything, which is a useful order to do it in. You learn early that the building is the easy part.

A general manager walks her building before the guests are up. Not an inspection — a reading. What the lobby smells like at seven. Whether the doorman is talking to anybody. How long the queue at the desk has been standing, and whether last night’s complaint has quietly become a review.

By the time she sits down she has formed a view of the day that no report will confirm for another month, and she is usually right, which is either experience or witchcraft; and after a while, you stop asking.

Then she looks at the numbers, and they are the right numbers. Not how full the building is — anyone can fill a building, by ruining the rate — but what it earns per available room, and how that sits against three named competitors on the same street.

If she is behind them she will know by how much. She will also be asked, which is the part that concentrates the mind.

Read also: Prime Orchard malls' retail rents inch up 0.3% in second quarter; limited new supply ahead

None of this is heroic. It is the ordinary conduct of somebody whose reputation is attached to an address.

The strange thing is not that hotels are run this way. It is that we looked at the tower next door, with its lobby and its lifts and its several thousand people, and concluded it needed nothing of the sort.

What hospitality gave up, and what it got back

Hotels were once owned and run by the same people.

Then the industry pulled the two apart — the real estate here, the operating business there — and discovered several things it had not been looking for.

Once the operator no longer owned the building, the operator had to be paid for something other than owning it. So the fee became a base on revenue plus an incentive that pays only after the owner has taken his return.

The operator eats last. One mechanical fact, doing more work than any quantity of alignment language in a management agreement.

Paying on performance requires performance to be measurable, and measurable against something.

So hospitality standardised its accounts until two hotels on two continents could be read line against line, then built the benchmarking to say whether a result was good or merely positive.

Read also: Therme, Marriott to explore wellbeing initiatives in travel and hospitality

Commercial property mandates rarely contain anything of the sort. We report. We do not compare.

And once you are measuring performance rather than occupancy, you start measuring the guest rather than the room.

Beneath all of it sits one person. The general manager is not a facilities role. She runs a business that happens to have a building around it.

Who is accountable for the lobby?

Ask that of a Singapore office tower.

Not the Reit manager, whose fee turns on assets under management and whose horizon is a distribution.

Not the property manager, engaged against a scope that specifies cleaning frequencies and response times, and who would be exceeding both his mandate and his margin by proposing the ground floor be reconceived.

Not the leasing agent, paid on transactions, measured on filling the space in front of him rather than on what the building becomes in year seven.

Not the asset manager, who may see the whole thing perfectly clearly from three removes.

Follow the question far enough and it does not get answered. It dissolves.

This is not a shortage of talent. Singapore’s property managers are as good as any in Asia and the institutional platforms here are genuinely sophisticated. Everyone is doing their job.

The trouble is that the sum of four well-executed mandates is not a business. It is a building in good repair.

In defence of the lease

At which point the landlords deserve a hearing, and a better one than my side of this argument usually gives them.

The lease is not laziness. It is a machine for turning something unpredictable into something predictable, and predictability is precisely what most owners of Singapore offices are contractually obliged to produce.

A Reit distributing every quarter cannot absorb the moods of an operating business. Its unitholders did not buy volatility. They bought an income stream with a covenant behind it, and a manager who quietly swapped the one for the other would be explaining himself at the next results briefing, at length.

There is a structural dimension too. The rules governing property funds here limit how much of a Reit’s income may come from operating rather than renting. So a Reit manager who woke up tomorrow persuaded by this article would find the vehicle itself standing in his way.

That is not an excuse. It is a real constraint, and one the industry is oddly shy about naming — perhaps because it sounds less like strategy than like paperwork.

And the custodial model was right for a long time. For two decades it produced perfectly good outcomes at low cost, and paying for operating intensity would have meant paying for something the market was handing out free.

Prudence is not a character flaw. It becomes one only when the conditions that rewarded it quietly expire and nobody sends a notice.

So the honest version of my argument is narrower than the one I started with, and better for it.

The constraint is real — but it constrains the vehicle, not the asset class. Private capital has no such limit. Neither does a family owner, a developer holding for the long term, or a structure built for the purpose.

Which turns the question inside out. If a building now needs running, and your structure cannot run it, the conclusion is not that the building should be left alone to think about what it has done.

It is that you are the wrong owner for it.

Some assets belong in a passive vehicle. A steadily increasing number do not, and the sorting is already underway — quietly, deal by deal, in a market that has not yet agreed to call it that.

Custodial is not the same as active

Any large landlord will object, reasonably: we do active asset management, we have the teams.

They do, and much of it is skilled. But most of what this market calls active management is more accurately custodial. Renewals. Capex on a cycle. Service charge discipline. New finishes when the old ones date.

Necessary, defensive, and all of it in service of maintaining a revenue model rather than changing one.

Active, in the sense worth arguing about, changes what the building is for — who occupies it, how they use it, what it can charge, and the part underwriters skip: who will be allowed to buy it from you at the end.

We took a tired heritage shophouse that the market had priced as retail-and-storage and rebuilt what it was for, from the conservation works to the tenant mix to how the ground floor met the street. Net operating income grew 233%. None of that was available in the entry price.

The reverse holds too. A building we opened at 5% occupancy stayed above 95% for the following decade. The occupancy is not the point. A decade of retention is not a leasing outcome. It is a thousand small operating decisions, none of which anybody wrote down.

Why this has stopped being free

For most of the last cycle none of it mattered, because the spread did the work. You bought well, signed long, held costs, and let falling cap rates and cheap debt supply the return.

Then rates moved and stayed moved. Cap rates adjusted more slowly than sentiment.

And a regulatory clock started running on the physical fabric of the stock itself, which turns deferred capex from prudence into a liability with a date on it.

Doing nothing was always the cheapest option. It is now the most expensive.

Look across the street, not overseas

The convention at this point is to point abroad. I would rather point across the street.

Singapore does not need to import this model. It runs two asset classes on it already.

Our hotels have general managers. Our malls have centre managers, curated tenant mixes and turnover rent — the landlord’s income moving with the tenant’s trade.

That makes a mall owner structurally curious about footfall, about adjacency, about whether the third floor is quietly dying. Mall owners here think about their tenants’ businesses because they are paid to.

Office is the only one of the three where the landlord’s income is deliberately insulated from whether the occupier is doing well.

We call this the defensiveness of the lease. It is also why the office landlord is the last to know.

The sectors sort themselves along a single line: whether the owner’s income moves with the occupier’s success.

Hospitality never had the choice — a hotel is so obviously a business that pretending otherwise would require effort.

Retail learnt the hard way, because e-commerce made not learning fatal, and turnover rent is the scar tissue.

Industrial and logistics dodge the question rather than answer it: a single tenant on a long lease does its own operating, and the landlord is passive by design rather than by temperament, which is perfectly respectable when the occupier’s business is the only reason the building exists.

Office is the odd one out. Multiple occupiers, shared ground, a common experience that somebody has to author — every characteristic of a managed asset, and none of the machinery.

Mixed-use makes the point most cruelly of all. In a single scheme, under one owner, the mall downstairs is run by a person with a name and a target. The tower above it is run by a document.

Three uncomfortable requirements

None of this needs a new team, or a rebrand of the existing one.

It needs three things, each awkward in its own way.

Someone accountable for the whole building. Named, with authority over the ground floor and the tenant mix, not merely the budget variance. This is the cheapest of the three and the most resisted, because it makes visible an accountability that currently enjoys being nobody’s.

A fee that eats last. Pay a manager on assets under management and he will optimise for assets under management. There is nothing cynical in the observation. It is only what incentives do.

And a measure of the occupier rather than the tenancy. Occupancy counts leases, not people. There are buildings in this city that are full on paper and dark by four.

The gap between leased and used is the leading indicator of the next repricing, and anyone who has stood in a lobby at half past eight knows it. Almost nobody underwrites it.

The question nobody asks the office

Return to the two buildings on the same road.

Next month, somebody will ask the first building’s general manager how it did. She will have an answer, measured against named competitors, with consequences attached to it.

If the answer disappoints twice running, she will be having a rather different sort of conversation by the third.

Nobody will ask the second building anything at all. Its owner will receive a rent roll and a service charge reconciliation.

Both will be accurate. Neither describes whether the building is any good — because nothing produced in the ordinary running of an office tower was ever designed to answer that, and nobody in the chain has been asked to try.

That is the whole of it. Not ambition, not talent, not capital. Simply the difference between a building somebody has to explain and a building nobody does.

We have known how to do this in Singapore for decades. We filed it under hospitality, and under retail, and never thought to ask why the tower on the same street was exempt.

## Chinese translation

> Translation model: openai_codex_cli

### 办公楼模式看不见的东西

新加坡把酒店作为企业来运营，把商场作为企业来运营。它的办公楼仍然按租约来运营，而曾经支撑这一模式的利差已经消失。

两栋楼，同一条路，同一年代，同样的土地成本。

第一栋楼里，一个人的专业声誉取决于这栋楼上个月表现如何。她知道房价、利润率、用来衡量她的三家竞争对手，以及自己在它们之中的准确位置。如果周二早晨抵达体验糟糕，那么到周二下午，这就是她的问题。

第二栋楼里，有一位基金经理、一位资产经理、一位物业经理和一位租赁代理。四个人都很擅长自己的工作。没有人会说抵达体验归自己负责，而每个人这样说都是对的。

第一栋楼是酒店。第二栋楼是办公楼。没有人觉得这很奇怪，正是其中最有意思的地方。

相关阅读：Frasers Property 的步行、跑步、骑行活动为年长者和残障人士筹得逾新币214,000元

新加坡把酒店作为企业来运营。它把商场作为企业来运营。它的办公楼仍然按租约来运营，而曾经让这种做法可以接受的利差已经消失。

总经理到底做什么

在真正经营任何东西之前，我曾在 Bangkok 建酒店多年，按这个顺序来做很有用。你很早就会明白，建筑本身是容易的部分。

一位总经理会在客人起床前巡视她的建筑。不是检查，而是读懂。早上七点大堂闻起来是什么气味。门童是否在和任何人交谈。前台队伍已经排了多久，以及昨晚的投诉是否已经悄然变成一条评论。

等她坐下来时，她已经对这一天形成了一个看法，而任何报告都要再过一个月才能证实；她通常是对的，这要么是经验，要么是巫术；过了一段时间，你就不再追问。

然后她看数字，而且这些数字是正确的数字。不是楼有多满，任何人都可以通过毁掉房价来填满一栋楼，而是每间可售客房收入是多少，以及它相对于同一条街上三家具名竞争对手的位置。

如果她落后于它们，她会知道差了多少。她也会被问到，这才是让人集中注意力的部分。

相关阅读：Prime Orchard malls 第二季度零售租金小幅上涨0.3%；未来新增供应有限

这些都不是什么英雄之举。这只是一个声誉与某个地址绑定的人所做的日常管理。

奇怪的不是酒店这样运营。奇怪的是，我们看着隔壁那栋有大堂、有电梯、有数千人的办公楼，却得出结论认为它不需要这类东西。

酒店业放弃了什么，又拿回了什么

酒店曾经由同一批人拥有并运营。

后来，行业把两者拆开，一边是不动产，另一边是运营业务，并发现了一些它本来并未寻找的东西。

一旦运营方不再拥有建筑，运营方就必须因为拥有之外的事情获得报酬。因此，费用变成了基于收入的基本费用，再加上一项只有在业主取得回报之后才支付的激励费用。

运营方最后吃饭。一个机械性的事实，比管理协议里任何数量的利益一致措辞都更有作用。

按业绩付费要求业绩必须可衡量，并且必须能与某些东西比较。

因此，酒店业把账目标准化到两大洲的两家酒店都能逐行对读，然后建立基准比较，用来说明某个结果是好，还是仅仅为正。

相关阅读：Therme、Marriott 将探索旅游与酒店业的健康福祉计划

商业地产授权很少包含这类东西。我们汇报。我们不比较。

而一旦你衡量的是业绩而非入住率，你就开始衡量客人，而不是房间。

所有这一切之下，都坐着一个人。总经理不是一个设施管理角色。她经营的是一家恰好被一栋建筑包围着的企业。

谁对大堂负责？

拿这个问题去问一栋新加坡办公楼。

不是 REIT 经理，因为他的费用取决于管理资产规模，而他的时间视野是一笔分派。

不是物业经理，因为他受聘于一份规定清洁频率和响应时间的工作范围；如果他提议重新构想首层空间，那既超出了他的授权，也超出了他的利润率。

不是租赁代理，因为他的报酬来自交易，衡量标准是填满眼前的空间，而不是这栋楼到第七年会变成什么。

也不是资产经理，尽管他也许隔着三层距离把整件事看得非常清楚。

沿着这个问题追问得足够远，它不会得到回答。它会消散。

这不是人才短缺。新加坡的物业经理不逊于亚洲任何地方，这里的机构平台也确实成熟。每个人都在做自己的工作。

问题在于，四项执行良好的授权加总起来，并不是一家企业。它只是一栋维护良好的建筑。

为租约辩护

说到这里，业主们值得被听一听，而且应该得到比我这一方通常给予他们的更好的陈述机会。

租约不是懒惰。它是一台把不可预测变成可预测的机器，而可预测性正是大多数新加坡办公楼业主在合同上有义务提供的东西。

一个每季度分派的 REIT 无法承受运营业务的情绪波动。它的单位持有人买的不是波动性。他们买的是背后有契约支撑的收入流；如果经理人悄悄把前者换成后者，他就得在下一次业绩简报会上长篇解释自己。

这里也有结构性层面。新加坡监管房地产基金的规则限制了 REIT 收入中可来自运营而非租金的比例。因此，如果一位 REIT 经理明天醒来被本文说服，他会发现载体本身挡在他的面前。

这不是借口。这是真实约束，也是行业奇怪地不愿点明的约束，也许因为它听起来不像战略，更像文书工作。

而托管式模式在很长时间里都是对的。二十年来，它以低成本产生了完全不错的结果；为运营强度付费，本来意味着为市场免费提供的东西付费。

审慎不是性格缺陷。只有当曾经奖励它的条件悄然到期而没有人发出通知时，它才会变成缺陷。

所以，我论点的诚实版本比我一开始提出的更窄，也因此更好。

约束是真实的，但它约束的是载体，而不是资产类别。私人资本没有这样的限制。家族业主、长期持有的开发商，或为此目的而设计的结构也没有。

这让问题反转过来。如果一栋楼现在需要被运营，而你的结构无法运营它，结论并不是这栋楼应该被放着不管，让它反思自己做了什么。

结论是，你不是它合适的业主。

有些资产适合放在被动载体里。越来越多的资产并不适合，而这种分类已经在进行中，悄悄地、一笔交易接一笔交易地，在一个尚未同意这样称呼它的市场里发生。

托管不等于主动

任何大型业主都会合理地反驳：我们做主动资产管理，我们有团队。

他们确实如此，而且其中许多工作很专业。但这个市场称为主动管理的大部分事情，更准确地说是托管。续租。按周期进行资本开支。控制服务费。旧装修过时时换上新装修。

这些都是必要的、防御性的，而且全都服务于维持一种收入模式，而不是改变一种收入模式。

值得争论的主动，指的是改变这栋楼的用途：谁使用它，他们如何使用它，它能收取什么租金，以及承销人会跳过的部分：最终谁将被允许从你手中买下它。

我们接手过一栋老旧的保育店屋，市场把它按零售加储物空间定价；我们从保育工程到租户组合，再到首层如何与街道衔接，重建了它的用途。净营运收入增长了233%。这些都没有体现在进入价格里。

反过来也成立。我们开业时入住率为5%的一栋楼，在随后的十年里一直保持在95%以上。入住率不是重点。十年的留存不是租赁结果。那是一千个细小运营决定的结果，而这些决定没有任何人写下来。

为什么这已不再免费

在上一个周期的大部分时间里，这一切都无关紧要，因为利差完成了工作。你买得好，签长约，控制成本，然后让下降的资本化率和廉价债务提供回报。

后来利率上升，并持续在高位。资本化率调整得比情绪更慢。

与此同时，监管时钟开始对存量物业的实体结构本身倒计时，这把递延资本开支从审慎变成了一项有到期日的负债。

什么都不做一直是最便宜的选择。现在它是最昂贵的。

看街对面，不要看海外

按惯例，此时应当指向海外。我宁愿指向街对面。

新加坡不需要进口这种模式。它已经用这种模式运营两个资产类别。

我们的酒店有总经理。我们的商场有中心经理、经策划的租户组合和营业额租金，即业主收入随租户交易额而变动。

这使商场业主在结构上会关心人流、关心相邻关系、关心三楼是否正在悄然衰退。这里的商场业主会思考租户的生意，因为他们是按此获得报酬的。

在三者之中，办公楼是唯一一个业主收入被有意与占用方经营好坏隔离开来的资产类别。

我们把这称为租约的防御性。这也是为什么办公楼业主总是最后一个知道的人。

这些板块沿着一条单一界线自我分化：业主收入是否随占用方的成功而变动。

酒店业从来没有选择，因为酒店显然是一门生意，假装不是反而需要费力。

零售业是吃了苦头才学会的，因为电子商务让不学习变得致命，而营业额租金就是留下的疤痕组织。

工业和物流不是回答这个问题，而是绕开它：单一租户签长期租约并自行运营，业主在设计上而不是性格上被动；当占用方的业务是这栋楼存在的唯一理由时，这完全值得尊重。

办公楼是异类。多个占用方、共享首层、共同体验需要有人来创作，这些都是受管理资产的特征，却没有任何相应机制。

综合用途项目最残酷地说明了这一点。在一个单一项目里、同一个业主之下，楼下商场由一个有姓名、有目标的人运营。楼上的办公楼由一份文件运营。

三个令人不适的要求

这一切并不需要一支新团队，也不需要给现有团队换个品牌。

它需要三样东西，每一样都以自己的方式令人尴尬。

有人对整栋楼负责。具名，并且拥有对首层和租户组合的权限，而不仅仅是预算差异。这是三者中最便宜的一项，也是最受抵制的一项，因为它让一种目前乐于无人承担的问责变得可见。

一种最后吃饭的费用机制。按管理资产规模给经理人付费，他就会为管理资产规模而优化。这个观察没有任何犬儒之处。激励机制就是这样运作的。

以及衡量占用者而非租户关系。入住率数的是租约，不是人。这个城市里有些建筑在纸面上满租，到了四点却一片昏暗。

已出租与已使用之间的差距，是下一轮重新定价的领先指标；任何早上八点半站在大堂里的人都知道这一点。几乎没有人把它写进承销。

没有人问办公楼的问题

回到同一条路上的两栋楼。

下个月，会有人问第一栋楼的总经理表现如何。她会有一个答案，以具名竞争对手为基准，并附带后果。

如果答案连续两次令人失望，到了第三次，她将进行一场相当不同的谈话。

没有人会问第二栋楼任何问题。它的业主会收到一份租金清单和一份服务费核算。

两者都会准确。两者都无法描述这栋楼是否足够好，因为办公楼日常运营中产生的任何东西，从来都不是为了回答这个问题而设计的，而链条中的任何人也从未被要求尝试。

这就是全部。不是雄心，不是人才，不是资本。只是必须有人解释的建筑与无人需要解释的建筑之间的区别。

几十年来，我们在新加坡一直知道该怎么做。我们把它归档在酒店业之下，归档在零售业之下，却从未想到要问，为什么同一条街上的办公楼可以例外。