When Interoperability Becomes Measurable

EIS-019 holds one upstream evidence population constant and changes only the destination environment. The population is 128 canonical MME task positions anchored in IFRS ISSB and SASB. Routed across GRI, ESRS, TNFD, COSO, GHG Protocol Scope 3 and the United Nations Sustainable Development Goals, it creates 768 comparable positions.

Of those positions, 678 preserve a governed routing relationship and 90 preserve no default route. The observed routability density is therefore 88.28%. That percentage is not presented as a target, rating or proof of institutional equivalence. Its significance is that relationship presence and relationship absence are counted inside the same stable denominator.

The new layer is Distributional Interoperability: observing how governed relationships are distributed when a fixed upstream evidence architecture is tested against several institutional destinations. EIA-014 established why each relationship requires type, strength, activation conditions and boundaries. EIS-019 asks what becomes visible once those governed states form a complete measurable population.

The six destination environments produce different densities even though the upstream 128-task architecture does not change. GRI routes 127 of 128 positions, ESRS 128, TNFD 121, COSO 128, GHG Protocol Scope 3 77 and the UN SDGs 97. A lower count may reflect narrower purpose, stronger activation requirements or a defensible decision not to create a default route.

Task portability is also uneven. Sixty-one canonical tasks route across all six environments, 46 across five, 19 across four and two across three. This means 107 of 128 tasks, or 83.59%, route across at least five destinations, while 126 of 128, or 98.44%, route across at least four. Portability describes participation across governed routes, not interchangeable institutional meaning.

The measurement boundary is strict. This is a full census of EMJ.LIFE's six published Direct Mapping Guides, not an independent third-party benchmark or a census of global ESG activity. It does not establish compliance, materiality, taxonomy alignment, control effectiveness, inventory completeness, SDG contribution, assurance readiness or endorsement by any referenced institution.

Singapore should not import the 88.28% result into an issuer, portfolio or regulatory dashboard as a readiness score. SGX reporting, SFRS S2 preparation, Singapore-Asia Taxonomy classification, transition-planning supervision, credit review and assurance are distinct destinations. Each requires its own population definition, decision owner, period, authority and admissible relationship states.

A Singapore-listed group could begin with a stable set of material operational evidence objects and observe how each enters its disclosure, taxonomy and assurance workflows. A bank could hold a stable borrower-evidence population and compare its routes into onboarding, sector classification, environmental-risk review, transition-plan assessment, covenant monitoring and portfolio reporting. In both cases, the denominator must include rejected and held routes rather than only successful reuse.

The control question is whether the aggregate remains reconstructable. A board, risk committee, credit committee or assurance provider should be able to move from a density figure back to the underlying evidence identity, rule version, activation fact, threshold, reviewer and decision. Without that reversal path, a visually precise percentage can create false confidence in the Singapore capital-market process.

ASEAN is not a single destination environment. Member states differ in legal adoption, national taxonomy design, sector structure, grid factors, enterprise size, data systems and assurance capacity. The Foundation Framework, Plus Standard and national approaches may also produce different activation facts and thresholds. A regional routing measure must therefore distinguish the common evidence identity from every local classification, financing and disclosure test.

Consider one operational emissions or asset record originating from an ASEAN infrastructure operator or supply-chain borrower. It may enter a Singapore lender's due diligence, a national subsidiary's disclosure, an ASEAN Taxonomy assessment, a sustainable-finance instrument and the parent group's consolidated report. Reuse is defensible only when entity, facility, jurisdiction, activity, method, period, data quality and verification state travel with the record while each destination re-performs its own authorised decision.

The practical regional question is why a route is absent. Missing supplier data calls for collection or estimation controls. A missing legal classification calls for jurisdictional review. A failed technical threshold calls for a negative classification. An immature control environment may require enhanced diligence or pricing conditions. A correct no-default route requires the system to stop. Keeping these outcomes separate allows Singapore capital providers to compare ASEAN exposures without forcing unlike assets into one artificial regional score.

In global capital markets, the same evidence object can reach an issuer's sustainability disclosure, an arranger's due-diligence file, a sustainability-linked instrument's KPI package, a credit committee, an asset manager's mandate screen, a stewardship record, a data vendor and an assurance engagement. Distributional measurement can show repeated evidence demand and manual reconstruction points, but it cannot make those destinations institutionally equivalent. Each user still owns its materiality, eligibility, risk, pricing, mandate or assurance decision.

The central failure chain is measurable but dangerous: 88.28% corpus routability is promoted into issuer readiness, issuer readiness into portfolio alignment, and portfolio alignment into instrument eligibility or lower perceived risk. None of those promotions is supported by the corpus result. Similar errors occur when Scope 3 relevance becomes inventory completeness, an eligible activity becomes an aligned entity, a disclosed transition plan becomes verified performance, or one assured metric lends credibility to unaudited claims elsewhere.

The 90 no-default positions therefore carry capital-market value. They identify where an arranger should request transaction evidence, a lender should impose a condition or escalate review, an asset manager should exclude a claim from mandate reporting, a data provider should preserve an unknown state, or an assurance practitioner should restrict the conclusion. Institutions should measure collection efficiency alongside prevented claim promotion, exception resolution, route withdrawal and decision reproduction. Lower diligence cost is defensible only when these controls remain intact.

The minimum position-level record should include the canonical task and evidence identifiers, originating entity and facility, source architecture and version, destination framework and requirement, relationship type and strength, activation condition, applicable jurisdiction and period, method, evidence state, decision owner, reviewer and timestamp.

Execution fields should separately record whether the AI Agent routed, held, escalated or stopped; which claim was permitted; which stronger claim remained prohibited; what exception or missing fact controlled the decision; and whether the route was later corrected or superseded. Aggregate measures must be regenerated from these controlled position records rather than entered as standalone percentages.

Measurement governance also requires version discipline. When a framework, taxonomy, method, threshold or canonical task changes, institutions should preserve the previous population, identify affected positions, re-run only governed routes and explain changes in numerator and denominator. A time series is meaningful only when each distribution can be reproduced under its original rules.

The next test should use three bounded populations: evidence for a Singapore-listed or cross-listed group, evidence for an ASEAN borrower portfolio, and evidence for one cross-border green, transition or sustainability-linked transaction. Route each population through disclosure, taxonomy, underwriting, credit or investment approval, instrument reporting, portfolio monitoring and assurance. Preserve every positive, conditional, held, rejected and no-default position.

Each review cycle should retain the population definition, source and methodology versions, relationship rules, numerator, denominator, decision owner, exceptions, prohibited claims and changes from the prior cycle. The test succeeds only when a reviewer can reproduce both the distribution and the downstream capital decision. It fails when an issuer, instrument or portfolio receives a new label, eligibility state or risk treatment that cannot be traced to position-level evidence and authority.

Watch whether global disclosure adoption reduces evidence friction without creating automatic equivalence; whether sustainable debt KPI and use-of-proceeds records remain linked to operating evidence; whether Singapore institutions preserve activity, entity, instrument and portfolio levels across ASEAN; whether data vendors retain unknown and no-default states; and whether assurance teams can reproduce routed evidence under ISSA 5000 or applicable local requirements. A higher density is not automatically better. The decisive signal is whether capital allocation becomes more reconstructable without unsupported claims becoming easier to circulate.