// DEEP-CONTEXT INTELLIGENCE · SGAI · Singapore · 2026

The Singapore property agency, decoded.

A record 36,816 licensed agents now chase a shrinking pool of deals — while the number of agencies has fallen below 1,000 for the first time. It looks like a glamorous, high-income business. The reality underneath: a genuine statutory licence guards the door, but it guards the person, not the firm — so the value walks out every night. Here is the picture the recruitment talks skip: how to actually get licensed, what agents really earn, why the agency keeps almost nothing, and where an independent genuinely wins.

36,816

licensed agents (1 Jan 2026 — a record)

997

agencies left (first time below 1,000)

~90%

of agency revenue is paid back out as commission

~45%

RES exam pass rate (unofficial — it's hard)

The verdict, in one number

The MOAT Score: is a property agency worth building?

Before the details, our one-number read. The MOAT Score grades a sector's economic quality on the value-investing lens of Graham, Buffett and Munger — four pillars (Margin, Operating moat, Appetite, Treadmill), each out of 25. Real estate agency is the rare SME sector with a real licence, which lifts it above the no-moat trades — but the licence belongs to the agent, not the agency, which caps it well below the sticky licensed sectors.

M — Margin: 10/25O — Operating moat: 11/25A — Appetite: 16/25T — Treadmill: 12/25DMOAT
The MOAT Score
49/100

Winner-takes-most

Winner-takes-most — hard.

The MOAT Score · M · O · A · T

How the score is built

The MOAT Score sums four pillars — each scored 0–25 — from the value-investing lens of Graham, Buffett and Munger. No black box: here is the working.

  • Margin

    10/25

    Does the average operator actually keep money — real net margin and return on the capital tied up?

    This sector: the agency keeps only ~10% of revenue (89.7%SourcePropNex AR2025, FY2025 of revenue is paid out as commission) and nets single digits; the agent keeps a high split but carries every cost and has no salary floor.

    Buffett 1979 — “a high earnings rate on equity capital… without undue leverage”; 1986 owner earnings.

  • Operating moat

    11/25

    Pricing power and a durable competitive advantage — can a typical operator raise prices and have customers shrug?

    This sector: a genuine statutory CEA / RES licence — the rare SME barrier — but it attaches to the person; agents and whole teams switch firms freely (100+ left one agency in weeks in Feb 2026SourceMothership / Business Times, 2026).

    Buffett, FCIC 2010 — pricing power is “the single most important decision”; 1991 franchise; 2007 moat.

  • Appetite

    16/25

    Demand durability — steady, recession-resistant repeat demand vs fragile, discretionary or faddish.

    This sector: structural transaction demand (~90% home ownership, the BTO→resale→upgrade cycle), but violently cyclical — private volumes swung from 33,557SourceURA, 2021 to 19,044SourceURA, 2023 as cooling measures bit.

    Graham, Security Analysis Ch.2 — inherent stability “derives from the character of the business”.

  • Treadmillinverted · less is better

    12/25

    Capital intensity and structural drag — rent, churn, fashion, discounting. Scored inverted: less treadmill, more points.

    This sector: asset-light (no fit-out, no inventory) is the real advantage — dragged by a rising portal-advertising bill (PropertyGuru's revenue per agent rose +22%SourcePropertyGuru, FY2023), one-off re-acquisition on every deal, and zero income through a cooling-measure trough.

    Buffett 2007 — “the worst sort of business… requires significant capital… Think airlines.”

Total · Winner-takes-most — hard

M + O + A + T, out of 100

49/100

The MOAT Score is a transparent SGAI judgement on a sector’s economic quality through a value-investing lens — not a verdict on any individual business, and not a comment on an owner-operated livelihood (a sector can score low on capital returns yet work as a job).

The read: the licence guards the door, but the value walks out every night

Singapore's estate-agency industry is built on a paradox. On one hand it has the thing café, salon and renovation founders would kill for: a real statutory barrier. You cannot legally sell property here without passing the CEA's RES examination — a notoriously hard two-paper exam — and registering through a licensed agency. That keeps the door narrow.

On the other hand, the barrier protects the wrong party. The licence, the client relationships and the personal brand all belong to the agent, who can walk to a competitor any time. So the agency has almost no hold on its most valuable asset. We watched this play out in February 2026, when a high-profile agency was hollowed out in weeks — over 100 of its ~157 agents left after a leadership scandal, top teams simply relocating en masse to PropNex and ERA. Back in 2022, more than 700 agents left ERA in a single exodus. The firms are recruiting machines precisely because retention is impossible — this is the salon's “the stylist walks out the door” problem, scaled to a S$1-billion industry.

And the money tells the same story. PropNex booked S$1.12 billion of revenue in 2025 — an enormous number — but roughly 90% of it was paid straight back out to its agents as commission. The agency keeps a ~10% gross slice, and after running the brand, the systems and the leadership overrides, nets ~6%. For the agent, the upside is real but the floor is zero: 100% commission, no base salary, no employer CPF, and a market that the government can cool overnight. It is not a property business. It is a licensed, commission, cyclical sales business — and that is a very different thing to be in.

The comparison

Where the commission dollar actually lands

A S$1.6M condo at a 2% seller fee generates S$32,000 of commission. Watch it get split — by the co-broke convention, then by the agent's split with the agency — until the firm keeps almost nothing.

Source: SGAI worked example on SG commission conventions, 2026 (commissions are not regulated; rates negotiable)

The map: ~52,000 sales a year, two listed giants — and the “market size” number is junk

The industry rides on roughly 52,000 sale transactions a year (~10,800 new launches + ~14,600 private resale + ~26,200 HDB resale) plus well over 100,000 rentals (URA / HDBSourceURA + HDB transaction data, via research houses, 2025). Ignore the aggregator reports that put the “Singapore real estate market” at US$35–66BShaky figure — treat with cautionmarket-size aggregators (IMARC, Mordor, Technavio)These measure the WHOLE property market (development/commercial/value), not agency commission revenue — and they contradict each other by ~US$20B. Useless as an agency-industry size. — that is the value of all property, not agency revenue, and the figures disagree by tens of billions. The only defensible top-line is the two listed firms' combined ~S$1.8B revenue — of which ~90% is commission passed through to agents.

36,816

registered agents

1 Jan 2026 (CEA) — a record high

997

licensed agencies

down from 1,118 in 2023 (CEA)

~88%

of agents sit in the top 5 firms

heavy concentration (CEA register)

~S$1.8B

combined listed-firm revenue

PropNex + APAC, FY2025 — ~90% pass-through

The trend

The numbers you hear are usually a year out of date

Agencies keep consolidating while agents keep multiplying. A profession under real compression shrinks — Singapore's is growing, which tells you commissions are holding up, not collapsing.

2023: 1,11820232024: 1,09020242025: 1,04620252026: 9972026997

2026agencies fall below 1,000 for the first time (CEA, 1 Jan 2026)

Source: CEA Industry Statistics, registered agencies as at 1 January each year (updated 16 Jan 2026).

The players: a recruiting league table, not a profit league table

The big agencies compete on headcount, because headcount is the business: more agents → more commission flow → a bigger thin slice. PropNex is decisively #1. But notice the margins — even the giants net single digits, and the one who nets more (PropNex) does so on a swing in new-launch commissions, where the developer pays and the agency's real moat (scale + developer panels) actually lives.

The comparison

The two listed giants, and what they actually keep

  • PropNex (SGX: OYY)

    The largest — ~38% of all SG agents

    SG agents
    ~14,200
    % paid to agents
    89.7%
    Net margin
    6.3%
    FY2025 signal
    Record S$1.12B rev; net S$70.4M; new-launch +134%
  • ERA / APAC Realty (SGX: CLN)

    The #2, master-franchisor across APAC

    SG agents
    ~8,400
    % paid to agents
    89.6%
    Net margin
    3.0%
    FY2025 signal
    Rev S$675.6M; net S$20.6M; thinner fixed-cost base
  • Huttons Asia

    The #3 (private)

    SG agents
    ~5,760
    % paid to agents
    Net margin
    FY2025 signal
    No public accounts — grew by poaching in 2022
  • Ohmyhome (Nasdaq: OMH)

    The fixed-fee disruptor

    SG agents
    % paid to agents
    Net margin
    FY2025 signal
    Persistently loss-making; pivoted AWAY from flat-fee broking
yes partial noSource: PropNex AR2025 + APAC Realty FY2025 SGX filings; CEA register; Ohmyhome SEC 20-F. Net margins computed from audited P&L.

The cautionary tale: Ohmyhome

The Nasdaq-listed flat-fee disruptor promised to break the ~2% model with a S$2,888 full-service HDB sale. Instead, its brokerage revenue shrank from 56% to ~21% of the business as it pivoted into property management, losses more than doubled, and it ran a 1-for-10 reverse split to keep its listing. The lesson for anyone betting on fee disruption in Singapore: the flat-fee model didn't generate profit, and the incumbent commission has barely moved. Tech raised the agent's value-add here; it did not commoditise it.

The customer: a high-trust, high-stakes, once-every-few-years decision

A home is the biggest transaction most Singaporeans ever make, wrapped in eligibility rules, ABSD, CPF and financing — so most still want a professional to carry the pricing, paperwork and negotiation, and above all to not get cheated. Satisfaction is high and rising: the CEA's 2024 survey found 92%SourceCEA Public Perception Survey, 2024 of consumers satisfied with their agent (up from 77% in 2021), and 70%SourceCEA survey, 2024 check the Public Register before engaging one. The catch for the agent: each client transacts once every several years, so there is no repeat revenue — every deal is a fresh customer, acquired again from scratch.

HDB resale (the volume engine)

The largest, most price-conscious, most research-heavy segment — and the one most exposed to DIY: HDB’s free Resale Flat Listing portal now lets owners sell agent-less after the 5-year MOP.

Private resale & upgraders

Higher tickets, condo and landed; the HDB-to-private upgrade is the bread-and-butter "asset progression" pitch — and the most policy-sensitive (15-month wait-out, ABSD on the second property).

Rental (high-frequency, low-ticket)

Well over 100,000 leases a year at ~half-a-month fees — fast cash flow, thin per-deal, increasingly each party pays their own agent under the 2024 co-broke best-practice guide.

Foreign & luxury buyers

Crushed by the 60% foreigner ABSD since April 2023 — foreign share of prime-district buying roughly halved. The surviving niche (GCB, ultra-prime) is relationship-gated and off-market.

What aspiring agents (and their clients) Google at 1am

The whole funnel sits in a handful of high-intent queries — and most existing answers are recruiter pitches or out-of-date numbers. These are exactly the questions this report answers straight:

  • “how to become a property agent in Singapore”
  • “RES exam pass rate / is the RES exam hard”
  • “how much do property agents earn in Singapore”
  • “property agent commission Singapore HDB”
  • “PropNex vs ERA which is better”
  • “is it worth becoming a property agent 2026”
  • “can I sell my HDB without an agent”
  • “RES course cost / how long to get licensed”

The economics: two very different P&Ls under one roof

There are two businesses here. The agency runs a high-revenue, razor-thin-margin model — ~90% of revenue is commission it must pay out, leaving a ~10% gross slice and a single-digit net. The agent runs the opposite shape: keeps 70–90% of their own commission, but carries every cost and has no salary floor. Below is the agency's dollar; then model the agent's take-home for yourself.

The margin breakdown

Where the agency's revenue dollar goes

A listed Singapore agency, modelled per S$100 of revenue. Commission to agents is the overwhelming line — the agency lives on the thin remainder.

Revenue (per S$100)100%
Commission paid to agents (cost of services)Staff, premises, systems, marketing, overridesNet margin (the agency keeps)
Revenue (per S$100)100%
  • Commission paid to agents (cost of services)
    90%
    10% left

    ~89.7% at PropNex, ~89.6% at APAC/ERA · PropNex + APAC FY2025 filings

  • Staff, premises, systems, marketing, overrides
    4%
    6% left

    brand, training, technology, leadership overrides · SGAI calc on filings

Net margin (the agency keeps)

What the owner actually keeps

6%

Verdict: A ~6% net for the market leader, ~3% for #2 — the agency is a thin-margin clipping operation on enormous turnover, not a high-margin business. The year-to-year swing comes almost entirely from the higher-margin new-launch (developer-paid) mix.

Illustrative model on PropNex (net ~6.3%) and APAC Realty (net ~3.0%) audited FY2025 results. The ~6% net line is the residual after the ~4% operating cost shown. Not financial advice.

The value-investing verdict · Graham · Buffett · Munger

Real estate agency

No

Would a value investor own the average operator here?

A value investor would not want the average agency or the average agent — a real licence guards the door, but the firm has no hold on the agent, and the agent has no salary floor in a market the state can cool overnight.

Pricing powerConstrained price-taker
Price-takerPrice-maker

Commissions are unregulated but anchored hard by convention (HDB ~2%/1%, private ~2%); negotiable downward, and fee disruptors press on the bottom — but they have stuck for a decade.

Buffett, FCIC 2010 — pricing power is “the single most important decision”

Moat Stable
Licence / IP

A genuine CEA/RES statutory licence — but it attaches to the agent, not the firm. The firm’s only durable moat is developer/new-launch relationships (scale-gated); agent loyalty is near-zero.

Buffett 2007 — an enduring moat protects returns on capital

Return on capitalMediocre

Asset-light, so little capital is tied up — but the agency’s thin net and the agent’s no-floor income make returns mediocre and volatile.

Buffett 1979 — a high earnings rate on capital, unleveraged

Capital intensity / treadmillLight

No fit-out, no inventory, no kitchen — the structural advantage over café/clinic/studio. You fail cheaply, and you start cheaply.

Buffett 2007 — the worst business needs much capital, earns little

Demand durabilityDiscretionary

People always buy, sell and rent — but transaction volumes swing violently with cooling measures and rates (33,557 in 2021 → 19,044 in 2023), and the agent has no income through the trough.

Graham, Security Analysis Ch.2 — inherent stability is qualitative

Cyclicality / policy riskHigh

ABSD, TDSR, LTV, SSD and the 15-month wait-out can halve volumes overnight — the defining risk.

URA volume swings 2017–2025; MAS/IRAS cooling measures

If not the average — what a winner needs

A specialised niche (luxury/GCB, landed, rental, relocation), a personal brand the firm can’t take, the highest split you can negotiate (or your own low-overhead agency), and developer-panel access for the new-launch margin.

Assessment uses the value-investing lens on SG agency + agent unit economics (2023–2026). A lens on economic quality, not a verdict on an owner-operated livelihood or a successful agent’s career.

Now the agent's side — model a year's closings and see what you actually keep after your split and your portal bill:

Resale (HDB / private) — your side

6
S$700,000
1.5%

New launch (developer pays) + rental

2
S$1,500,000
3.5%
8
S$3,500

Your split + your costs

80%
S$1,500
S$800

Your take-home / year (pre-tax)

S$117,400

S$9,783/mo · from 16 deals

Gross commission

S$182,000

Your costs / year

S$28,200

The firm's clip on you: S$36,400 (20% of your gross). That thin slice — not your commission — is why the listed agencies net only single digits, and why a rival can out-bid for you with a better split.

Verdict: A strong year. This is the top tail the awards celebrate — but the dispersion is brutal: most agents never reach it, and the figure swings hard with the market cycle.

Illustrative, not tax advice. Commissions are not regulated in Singapore — rates are negotiable; the figures here are conventions (HDB seller ~2% / buyer ~1%, private resale ~1–2%, new launch developer-paid ~2–5%, rental ~½–1 month). Take-home is pre-tax: an agent is a self-employed commission agent (personal income tax, compulsory MediSave, no employer CPF). Co-broke typically splits a side ~50/50, so model your share. Add GST (9%) only if your agency is GST-registered.

How to actually get licensed (the real barrier, sequenced)

Unlike a café, the bottleneck here isn't premises — it's the exam. The CEA (Council for Estate Agencies) runs a genuine statutory gate, and the RES exam is where most aspirants stall.

Becoming a salesperson (RES), in order

  1. Meet the bar — at least 4 GCE O-Level passes (or the WPLN equivalent), and be 21+ to register.
  2. RES course (~S$700–900) at a CEA-approved provider, ≥75% attendance; the completion certificate is valid 2 years.
  3. Pass the RES exam — two papers, 2.5 hrs each, 60/100 to pass each; held 3× a year; fee S$512.30 (incl. GST). This is the real gate.
  4. Register through a licensed agency — you cannot operate solo, and you can only be registered with one agency at a time (S$60 + ~S$280/yr).

Running your own agency (the KEO path)

To hold an estate-agent (EA) licence you need a qualifying Key Executive Officer: a pass in the REA exam (within 2 years), plus 3 years' experience and 30 transactions over the last 3 years (or 3 years managing an EA). The agency must be ACRA-registered, carries an application fee of S$120 and an annual licence fee scaling from S$330, and must hold professional indemnity insurance. Notably, there is no minimum paid-up-capital requirement — the “fit and proper” test and insurance substitute for a capital floor, which is exactly why this is an asset-light business to start. From 2026, every agent's mandatory CPD includes an AML/CFT course.

Where an agent or a boutique actually wins

Because the asset is the person, the winning moves all build something the firm can't take from you — a niche, a brand, a higher split. Volume-chasing in the open market is the trap; differentiation is the way out.

Own a niche

Luxury/GCB, landed, rental, en-bloc, expat relocation — relationship-gated segments where one off-market deal dwarfs a year of HDB sides, and where AVMs and DIY portals can’t compete.

Build a personal brand

The strongest independent moat in SG residential — content, video home-tours, an audience that follows YOU. Stacked Homes and PropertyLimBrothers proved it; just remember it’s fragile to your own conduct.

Negotiate the split / go independent

Every point of split you keep is pure margin. A top producer at a low-overhead boutique — or as KEO of their own small EA — keeps far more than the franchise’s ~10% clip lets them.

Get on the developer panel

New-launch is the one place the agency moat is real: developers need scale to clear stock before the 5-year ABSD deadline, and pay 2–5%. It’s a scale game — join a firm that has the mandates, or build toward them.

Survive the cycle

Cooling measures can halve volumes. The agents who last keep a cash buffer, a rental book for steady cash flow, and don’t over-commit fixed marketing they can’t sustain in a quiet year.

The honest AI edge

AI genuinely earns its place on the low-value 80%: listing & ad copy, multilingual translation, photo staging, a fast comparables draft, lead-timing on MOP/TOP triggers, CRM follow-up. It does NOT create your moat — trust, niche and brand do. Automate the admin so your time goes where humans still win.

Questions founders ask

How do I become a property agent in Singapore?

Four steps, governed by the Council for Estate Agencies (CEA): meet the entry bar (at least 4 GCE O-Level passes or the WPLN equivalent, and be 21+), complete the RES (Real Estate Salesperson) course at a CEA-approved provider with ≥75% attendance, pass the RES examination, then register through a licensed estate agency — you cannot operate solo. Realistically ~6 months end-to-end. The exam fee is S$512.30 for both papers (incl. GST); the course runs roughly S$700–900.

How hard is the RES exam and what is the pass rate?

It is two papers of 2.5 hours each (60 marks of MCQs + 40 marks of case studies; you need 60/100 to pass each paper), held three times a year. It is widely described as notoriously hard, with an industry-circulated pass rate "under half" (commonly quoted around 45%). Be careful: that figure is repeated everywhere as "from CEA's annual report", but it does not actually appear in the CEA Annual Report — treat ~45% as an unofficial estimate, not an official statistic.

How much do property agents earn in Singapore?

There is no official income distribution, and the dispersion is extreme — a small minority of agents earn most of the commissions. Agents are self-employed and 100% commission (no base salary, no employer CPF). A new agent often makes S$30,000–60,000 a year gross; mid-career S$80,000–150,000; the top tail S$200,000+. The often-quoted "median S$132,000 + bonus" figure is from a recruiter survey of salaried real-estate roles — not commission-only agents — so do not read it as the typical agent's take.

What commission do property agents charge in Singapore, and is it fixed?

Commissions are not regulated — CEA does not set or cap them, and they are negotiable (it should be documented in a Prescribed Estate Agency Agreement). Market conventions: HDB resale seller ~2% and buyer ~1%; private resale ~1–2%; rental about half a month's rent for a one-year lease; new launches are paid by the developer (commonly ~2–5%, occasionally higher to clear slow stock). Co-broking typically splits a side roughly 50/50 between the two agents, and GST (9%) applies if the agency is GST-registered.

Should I join PropNex or ERA, or start my own boutique agency?

PropNex is the largest agency by far (~14,200 agents, ~38% of all registered salespersons), ahead of ERA (~8,400). The big franchises give you brand, leads, training, AI tools and new-launch (developer) access in exchange for a slice of your commission; top producers keep 80–90%. The catch is structural: the client relationship and the licence follow you, not the firm, so agents and entire teams switch firms freely. A boutique or your own agency lets you keep more of your own production — but you give up the brand and the developer panel, and you carry your own marketing. Specialise (luxury/GCB, landed, rental, relocation) and build a personal brand the firm can't take with it.

Want a report like this for your company or sector?

Tell us what you're trying to figure out. We'll build you a Deep-Context report — the same depth, focused on your business.

A human replies within a business day. No sales spam.

About this report. Built with SGAI's Deep-Context Engine — human-directed, AI-accelerated. Figures draw on CEA (industry statistics, the public register, the Annual Report 2024/25), URA and HDB transaction data, IRAS/MAS cooling measures, and audited SGX filings (PropNex, APAC Realty) plus Ohmyhome's SEC 20-F, with reporting from EdgeProp, Stacked Homes, The Edge and others (2023–2026). The widely-quoted RES pass rate (~45%) is an unofficial, industry-circulated figure — it does not appear in CEA's Annual Report; treat it as directional. Commissions are not regulated; the rates shown are conventions, not law. The MOAT Score is a transparent SGAI judgement on economic quality, not a verdict on an owner-operated livelihood or a successful agent's career. Verify fees and regulatory steps with CEA before acting.

We're not a grant office — we build AI and intelligence for Singapore SMEs. Talk to us.