The 34% Problem: When REITs Are a Third of Your Universe

Thirty-Four Percent Is REIT

Of 106 production issuers in the Apex Singapore universe^[1], 36 are REITs or Business Trusts. They constitute 34.0% of the scored set — the largest type cohort, narrowly ahead of locally-incorporated family and corporate issuers (SGP-L, 33.0%) and well above Government-Linked Companies (SGP-G, 14.2%) or foreign-domiciled issuers (SGP-F, 14.2%).

The 36 SGP-R issuers are not a marginal segment. At a third of the universe, they are foundational to the shape of every distribution Note 1 reported. They are also why Singapore is the only Apex coverage market in which REITs are scored as first-class issuers under the same framework as corporate equities.

How Other Markets Read REITs

Across the eight Apex coverage markets, REIT inclusion in the main G-Score universe is the exception, not the norm.

Other markets segregate REITs into separate scored universes or exclude them altogether on the grounds that REIT governance is structurally distinct: a REIT does not have a corporate board overseeing operating management; it has a manager licensed by the regulator and a trustee enforcing fiduciary duties on behalf of unitholders. The conventional view is that a corporate-governance ruler does not measure the same thing on a REIT.

The Apex framework's response is to keep REITs in the universe and adapt the ruler. Variant indicators — explicit REIT-specific sub-versions of corporate indicators that replace the corporate version where it does not apply — score the manager's board, the sponsor relationship, and the leverage trajectory in a way that preserves comparability with corporate issuers on a TBR-axis composite. The framework treats REIT inclusion as a measurement choice that surfaces governance variation other markets simply do not score.

Why REITs Score Above Average

The SGP-R cohort mean composite is 76.4. The production overall mean is 71.8. REITs sit 4.6 points above the SG production average, not below it.

This is not what most readers expect. The intuition that REITs carry sponsor-conflict and pipeline-injection risk is correct as far as it goes — but it does not translate into a lower composite under the framework. The reason is structural and regulatory.

S-REITs are governed by the MAS Code on Collective Investment Schemes, which imposes a denser structural disclosure obligation on S-REITs than the SGX Listing Rules impose on ordinary corporate issuers^[2]. The Code mandates: distribution of at least 90% of taxable income to unitholders, gearing within a regulatory ceiling^[3], independent trustee oversight, manager licensing with capital adequacy and fit-and-proper requirements, and unitholder-circular requirements for related-party transactions above prescribed thresholds. Each of these obligations creates a measurable disclosure surface that the framework's T-axis registers — and that the corporate equivalent does not exist for SGP-L or SGP-S issuers.

The framework's reading is that the structural disclosure obligations imposed on S-REITs by the CIS Code create an upward governance bias relative to non-REIT corporate issuers in the same market. Including REITs raises the universe ceiling, not the floor.

Bimodal Within SGP-R

The cohort mean conceals a bimodal split inside SGP-R itself. Within the 36 REITs:

Two readings of the same cohort.

The first: 72.2% of SGP-R issuers are Celestial — almost three times the rate of non-REIT issuers (27.1%). The dense regulatory disclosure regime produces a concentration of high-composite issuers far above the corporate average.

The second: 13.9% of SGP-R issuers carry a Kill Switch override — three times the non-REIT KS rate (4.3%). When REITs fail, they fail through narrow concentrated mechanisms — gearing breach, manager replacement, sponsor pipeline failure — captured by REIT-specific Kill Switch triggers and surfaced through the framework's KS override layer rather than through gradual archetype drift. The five SGP-R Kill Switches in the current production set cluster in concentrated segments rather than distributing across the cohort. REIT failure, when it occurs, surfaces through narrow structural channels.

The cohort is therefore neither uniformly strong nor uniformly weak. It is bimodal: most REITs cluster Celestial because the structural disclosure regime works; the few that fail, fail concentrated and hard.

The Variant Indicator Architecture

The framework adapts to REIT structure through a set of explicit variant indicators^[4]. These are not separate scoring systems; they are sub-versions of corporate indicators that fire on SGP-R issuers in place of the corporate-equivalent versions.

The architecture follows the four-party structure of every S-REIT:

SPONSOR  →  REIT MANAGER  →  TRUSTEE  →  REIT (legal trust entity)
(corporate    (MAS-regulated   (independent  (holds underlying
 anchor)       licensed         fiduciary)    real estate assets)
               entity)

Each link in the chain is a potential conflict-of-interest surface. The sponsor typically owns the REIT manager and appoints its board, aligning the sponsor's interests with the manager's incentives — not necessarily with unitholders'. The manager makes investment, divestment, leverage, and acquisition decisions on behalf of unitholders, and earns fees that scale with assets under management and transaction volume. The sponsor sells assets to the REIT under right-of-first-refusal arrangements that constitute the recurring "pipeline injection" mechanism. The trustee is the legal owner of REIT assets and oversees the manager — but the trustee's fees are paid by the REIT itself, an embedded structural tension.

Manager fees are typically structured in four named components, public from any S-REIT trust deed: a base fee scaling with AUM or property value; a performance fee scaling with net property income or distribution growth; an acquisition fee scaling with acquisition price; and a divestment fee scaling with disposal proceeds. The fee mix matters. A high acquisition-fee weighting incentivizes sponsor pipeline injections regardless of unitholder benefit — a structural conflict that B-04R, R-01R, and T-07 jointly capture.

The variant indicator architecture exists because the governance board that matters in an S-REIT is the manager's board, not a "trust board." This is a categorical difference from corporate-issuer scoring, and the framework treats it as such rather than forcing a corporate-board template onto a structure that does not have one.

Sample REITs

Public-tier illustrations from the SGP-R cohort^[5]:

The intra-type range within SGP-R spans roughly 55 composite points — from the upper 30s (retrospective collapse) to 93.5 (Parkway Life). This is the widest intra-type range observed in any Singapore type cohort, and it reflects the structural diversity of the S-REIT segment itself: sponsor quality varies substantially, asset class mix varies (healthcare, logistics, office, retail, industrial, hospitality, US-office), and leverage strategy varies across managers.

A 1-point composite difference between two corporate issuers at the median may be noise. A 50-point composite difference between two REITs in the same segment is structural — and the framework's variant indicators identify which axis the difference loads onto.

The Counterfactual

Take the REITs out and Singapore looks more like the other markets in the Apex panel. Leave them in and the universe has a different shape.

Excluding REITs would remove 26 of 45 Celestials and 5 of 8 Kill Switches. The bimodal Celestial-Chameleon distribution that defines Singapore's Note 1 signature^[6] would flatten substantially. The Celestial cell would no longer hold 42.5% of the universe; it would hold roughly 27% — within the band observed across other Apex markets but no longer a defining Singapore signature.

The implication is that the SG signature is partly a consequence of REIT inclusion, not an independent feature of the corporate market. The other Apex markets' decision to exclude REITs is itself a measurement choice, not a neutral default. By excluding REITs, those frameworks describe a corporate-governance market that does not exist in pure form on those exchanges either — they simply chose not to score the REIT segment.

The 34% number is what shows up when the exclusion is not made.

What Inclusion Reveals

Three implications follow.

The first is methodological. Most governance frameworks treat REITs as a special case to be excluded or scored separately. The Apex framework treats Singapore's S-REITs as first-class issuers with a parallel set of variant indicators. The result is empirical, not assumed: REITs sit above the production mean, not below it. The structural disclosure obligations that the CIS Code places on S-REITs produce a measurement signature that the framework registers cleanly. The decision to include REITs is the decision to measure that signature.

The second is about where REIT failure surfaces. The 13.9% Kill Switch rate inside SGP-R is concentrated, not gradual. The framework does not produce a distribution of slowly-deteriorating REITs trending toward failure. It produces Celestials with high composites and a discrete subset that tripped a REIT-specific Kill Switch — gearing breach, manager forced replacement, or sponsor pipeline collapse. Eagle Hospitality Trust's 2019-2021 collapse remains the cleanest demonstration of how the SGP-R variant indicators read sponsor-pipeline risk: R-01R (sponsor-related transactions) and KS-09 (manager replacement) jointly captured a structural failure that the trust's own annual filings disclosed in plain language^[7]. Case 2 in this series will examine the timing of that signal in detail.

The third is about how to read the Singapore universe. A reader who comes to Singapore expecting it to look like the other Asian markets in the Apex panel will be surprised by the bimodal Celestial-Chameleon distribution, the empty Poison Apple and Time Bomb cells, and the 4.6-point cohort premium that REITs enjoy over the corporate average. None of these features are noise. They are the visible residue of a market structure in which a third of the listed universe operates under a denser regulatory disclosure regime than the corporate listings around them — and a framework that scores both on the same composite scale.

When the Choice Is the Finding

Korea's Research Note 2 contrasted two existing exchange segments — KOSPI and KOSDAQ — that the Korean market itself maintains as separate listing tiers^[8]. Singapore's Research Note 2 is structurally different. It contrasts one universe with itself under inclusion versus exclusion. The 34% problem is not a feature of SGX. It is a feature of how the framework chose to read SGX.

That choice is the finding. The same framework, applied to seven other markets that exclude REITs, produces a uniform corporate-governance reading. Applied to Singapore with REITs included, it produces a market with two empty archetype cells, a bimodal Celestial-Chameleon distribution, the widest type-cohort spread in the Apex coverage panel, and a third of the universe scoring above the corporate mean. The framework is the same. The universe definition is different. That difference is what the rest of this series unpacks.

The Apex G-Score Singapore production universe of 106 includes 36 SGP-R issuers (REITs and Business Trusts) under the variant indicator architecture described above. Sample-issuer figures are public-tier disclosures; variant indicator weights, gearing-breach thresholds, manager-fee scoring formulas, and grade boundaries remain non-public.