Independent Singapore VCC guidance
Direct answer
Do not approve ESG data for a VCC merely because the provider is recognised or supplies a polished score. Define the exact use first, such as screening, research, portfolio construction, monitoring or investor reporting. Then test issuer and instrument coverage, source hierarchy, estimates, methodology, update timing, correction history, licensing and change notifications against that use. Reconcile the data to a representative portfolio and investigate missing or conflicting results. Approve it only with documented limitations, exception ownership and a route to reassess affected decisions and claims when the data changes.
At a glance
- Judge ESG data against one defined use rather than granting a general approval.
- Measure portfolio coverage and estimate reliance before interpreting scores.
- Trace methodology changes and corrections into controls, holdings and investor statements.
- Keep a human challenge route for conflicts, missing data and material outliers.
Who this is for
- VCC managers and directors using third-party ESG data in investment decisions, restrictions, monitoring, disclosures or reports.
Important exclusions
- An assurance opinion on sustainability outcomes or a substitute for the fund own mandate and disclosure analysis.
Define the use before reviewing the provider
Write the decision that the data will inform, the portfolio population, the timing of use and the consequence of an incorrect or missing value. Screening data may determine whether a security can be purchased, while research data may be one input among many. A portfolio characteristic used in investor materials carries a different evidence burden from an internal watchlist. Separate entity-level, instrument-level, sector and sovereign fields, and identify whether the VCC, sub-fund or share class is the reporting unit. Do not let a procurement label such as ESG platform become the approval scope. One data set may be acceptable for research exploration yet unsuitable for binding restrictions or public outcomes.
Sources: Monetary Authority of Singapore · Monetary Authority of Singapore · MoneySense| Use | Failure consequence | Minimum challenge |
|---|---|---|
| Pre-trade restriction | An ineligible holding enters the portfolio or an eligible one is blocked | Coverage, identifier matching, update time and exception route |
| Investment research | The thesis relies on unsupported or stale information | Source trace, estimate label, analyst challenge and conflicting evidence |
| Risk monitoring | Emerging exposure is missed or overstated | Population completeness, refresh cycle, outlier and change analysis |
| Investor reporting | A claim or metric cannot be supported for the stated scope | Method, period, denominator, limitations and reconciliation to holdings |
Measure coverage instead of accepting a headline score
Test the actual VCC portfolio, not a generic provider coverage statement. Match current and recent holdings by stable identifiers, legal entity, instrument and issuer hierarchy. Separate reported values, provider estimates, modelled proxies, inherited parent scores, missing fields and stale observations. Review concentration by exposure as well as item count because a small number of uncovered holdings may dominate risk or claims. For private assets, funds, derivatives and special-purpose vehicles, document how look-through, underlying exposure and unavailable data are treated. A high overall match rate can still fail the intended use when missing observations cluster in a strategy, geography, sub-fund or field that drives the decision.
Sources: Monetary Authority of Singapore · Monetary Authority of Singapore · MoneySensePortfolio coverage test
- Identifiers reconcile to the security, issuer and relevant parent or underlying entity without silent substitution.
- Reported, estimated, modelled, inherited, stale and missing values are counted separately.
- Coverage is measured by exposure and decision significance, not only by the number of holdings.
- Private assets, derivatives, funds and look-through positions have explicit treatment rules.
- Each affected VCC, sub-fund and share class is tested against its own mandate and claims.
- Material gaps have an exception route that prevents unsupported automatic decisions.
Related guidance: VCC pricing and market data provider review
Interrogate methodology and source hierarchy
Obtain enough methodology detail to understand what the field measures, which sources take priority, how estimates are produced, how controversies affect scores, how corporate actions are handled and when values expire. Distinguish observed company disclosure from provider judgement and modelled output. Check whether the methodology aligns with the VCC mandate wording and the way the manager explains the process to investors. Similar labels from different providers may represent different concepts, boundaries and time horizons. Ask how corrected source data propagates, whether history is overwritten, and whether prior decisions can be reconstructed. If the provider will not expose information needed to interpret a binding control, the data may be unsuitable for that use even if it remains useful for research.
Sources: Monetary Authority of Singapore · Monetary Authority of Singapore · MoneySenseMethodology acceptance decision
- Concept matchesContinue only if the field definition, boundary and time horizon align with the intended mandate, control or statement.
- Concept partly matchesLimit the use, add qualification or combine it with another control rather than treating the value as decisive.
- Estimate dominatesRequire an estimate label, confidence treatment and review route before the field affects binding investment action.
- Method is opaqueReject binding reliance when users cannot explain the basis, limitations and route by which the value may change.
- History cannot be reconstructedDo not use the field where the VCC must evidence why a past decision or investor statement was made.
Related guidance: approval controls for ESG claims in VCC investor materials
Run a representative portfolio test
Choose a test set containing ordinary holdings, recent additions, corporate actions, complex instruments, private assets, missing observations, large exposures and names with known controversy or provider disagreement. Reperform identifier mapping, field selection, transformation, restriction logic and the resulting portfolio or reporting outcome. Compare a sample with issuer disclosures and an alternative reliable source where feasible. Investigate large score movements and differences rather than averaging them away. Record whether the provider supplies a fact, estimate, opinion or composite. The test should show what the manager does when data conflicts with analyst evidence, when an observation changes after a decision, and when a holding has no acceptable value.
Sources: Monetary Authority of Singapore · Monetary Authority of Singapore · MoneySense| Sample type | Question to answer | Evidence to retain |
|---|---|---|
| Large or claim-driving exposure | Would an error change a restriction, metric or public statement? | Source trace, calculation and decision outcome |
| Missing or estimated holding | Is the fallback explicit and proportionate to the use? | Estimate label, alternative evidence and approval |
| Provider disagreement | Why does the selected signal better fit the defined purpose? | Comparison, analyst challenge and resolved treatment |
| Recent corporate action | Did mapping and history remain correct after the event? | Identifier lineage, before-and-after values and exception handling |
| Material methodology change | Which prior decisions, limits and reports may be affected? | Impact population, reassessment and communication decision |
Design change, correction and exception controls
Require a route for methodology releases, field additions, retired values, identifier changes, late source data and corrections. Classify which changes require technical testing, investment review, restriction reprocessing, holdings reassessment or investor communication. Preserve the prior value and method so a historic decision can be reconstructed. Set tolerance and escalation based on the use: a research view may permit a documented judgement, while a binding exclusion needs a controlled exception and authority. Monitor recurring missing data and manual overrides because they may reveal that the approved use is not operationally sustainable. Where the provider changes a result after publication, trace the effect through reports, claims and any action taken from the earlier value.
Sources: Monetary Authority of Singapore · Monetary Authority of SingaporeData-change response path
- DetectCapture provider releases, corrections, coverage shifts and unusual field movements through an owned monitoring route.
- ClassifyIdentify affected data fields, holdings, VCC uses, sub-funds, controls, reports and decision periods.
- ContainPause unsupported automated action or qualify outputs while material uncertainty and mapping errors are investigated.
- ReperformRecalculate affected restrictions, portfolio measures and disclosures using the approved corrected method and population.
- DecideRecord whether investment action, report correction, investor communication or provider remediation is needed.
Related guidance: VCC report data lineage from source to sign-off
Approve a bounded use and monitor reliance
The approval record should identify the provider, product, fields, portfolio population, permitted use, prohibited uses, methodology version, data latency, estimate treatment, coverage limitations, exception authority, monitoring and review triggers. Connect the approved use to the systems, spreadsheets, investment processes and reports that consume the data. A licence to receive the feed is not approval to use every field for every purpose. Assign an owner who can challenge both the provider and internal users, and define what would cause suspension. Review whether actual reliance has expanded through copied fields, new dashboards or repeated language in investor materials. Renew approval only when current evidence supports the same use and limitations.
Sources: Monetary Authority of Singapore · Monetary Authority of Singapore · MoneySenseBounded approval record
- The approved fields and exact VCC use are named, with prohibited uses stated just as clearly.
- Coverage, estimate, latency, methodology and history limitations are visible to decision-makers.
- Systems and reports consuming the data are inventoried with owners and change routes.
- Exceptions and manual overrides require reason, evidence, authority and expiry.
- Provider changes and corrections trigger an impact assessment rather than silent refresh.
- Investor-facing claims are withdrawn or qualified when the supporting data no longer fits their scope.
Related guidance: new investment instrument approval for a VCC mandate
Frequently asked questions
Can one ESG score support every VCC use?
No. A score may be useful for exploratory research but unsuitable for a binding exclusion, risk limit or investor claim. Approve the data against each defined use, portfolio population, method and consequence of error.
Are provider estimates automatically unacceptable?
No, but they must be identified and understood. Consider how the estimate is produced, how much portfolio exposure it covers, whether confidence is sufficient for the use and what alternative evidence or exception route applies.
What if two ESG providers disagree?
Treat disagreement as a signal for investigation, not a reason to average blindly. Compare definitions, sources, dates, entity mapping and methodology, then record which evidence better fits the VCC intended decision and why.
When should a methodology change trigger reassessment?
Reassess when the change can alter coverage, restrictions, portfolio construction, monitoring, reported metrics or investor language. Identify affected holdings and periods, preserve the prior method and decide whether earlier actions or statements require correction.
Who should approve ESG data for investment use?
The owner of the affected investment or control process should approve the bounded use with independent risk, compliance, product, data or legal challenge where relevant. Procurement completion and technical connection alone do not establish fitness.
Official sources and further reading
- Risk Management Practices for Fund Management Companies (Monetary Authority of Singapore)
- Disclosure and Reporting Guidelines for Retail ESG Funds (Monetary Authority of Singapore)
- Understanding Environmental, Social and Governance Funds (MoneySense)
- Understanding VCC Features, Eligibility and Requirements (Accounting and Corporate Regulatory Authority)
Discuss a Singapore VCC structure
For help coordinating a Singapore VCC setup or corporate administration, contact Raffles Corporate Services.
General information only. This article is not legal, tax, regulatory or investment advice and does not imply affiliation with or endorsement by ACRA, MAS or IRAS.