For most asset owners, sustainability commitments are executed by someone else. The pension fund sets the climate policy, but external managers pick the stocks, cast the votes, and run the engagements. That delegation chain is where ESG strategy either becomes real or quietly dissolves, which is why manager selection is the moment the two questions that matter most get asked: will this mandate succeed in this manager’s hands, and how will we track progress against it each cycle.
The pressure on that chain is growing from both ends. FTSE Russell’s 2025 asset owner survey of 415 asset owners found 73% now apply sustainability considerations, with 85% significantly concerned about climate risk, and their top motivation has shifted to better risk-adjusted performance. Meanwhile Cerulli’s research across 200 institutional investors found 58% either require or plan to require managers to report portfolio-level exposure to financially material ESG risks. Asset owners are not just asking managers to care about sustainability. They are starting to check.
Key takeaways
- In manager selection, ESG is the lens asset owners use to judge whether a mandate will succeed in an external manager’s hands, and to track progress against that mandate each cycle after appointment.
- The reliable signal is the manager’s disclosure record (policies, stewardship reports, voting records, holdings), not the RFP narrative; 37% of asset owners name greenwashing as a top concern.
- Monitoring is the weak link: 58% of institutional investors require or plan to require portfolio-level ESG reporting, but only 23% require reporting on engagement activities, and the biggest oversight challenges are collecting and normalizing manager data.
- AI-assisted assessment lets asset owner teams review every manager’s full disclosure record against consistent criteria, at selection and every review cycle after.
What does ESG mean in manager selection?
In manager selection, ESG is the lens asset owners apply to evaluate an external manager’s sustainability practices before awarding a mandate: how ESG factors are integrated into the investment process, how stewardship is resourced and executed, and whether the manager’s evidence supports its claims. It sits inside the broader select, appoint, and monitor cycle that the PRI’s asset owner guides formalized, and it continues after appointment as ongoing monitoring against the expectations set in the mandate.
Done well, every stage of the process returns to those same two questions: will this manager deliver the mandate, and how will progress be tracked. The answers live in the manager’s record: policies, stewardship and voting disclosures, engagement outcomes, and fund-level reporting, assessed against the asset owner’s own framework.
Selection: read the record, not the pitch
Every manager RFP now arrives fluent in sustainability. The FTSE Russell survey helps explain why asset owners have become skeptical readers: 37% name greenwashing risk as a top barrier, with ESG data concerns close behind at 36%. The pitch and the practice can diverge, and the only way to see the divergence is to compare the manager’s claims against its public record: stewardship reports, proxy voting disclosures, engagement case studies, fund documents, and regulatory filings. What that reading is really testing is mandate fit, whether the practices in the record can carry the mandate you are about to hand over.
That is a reading problem, and it compounds fast. A serious search shortlist might be six managers, each with hundreds of pages of relevant disclosure, assessed across dozens of criteria. Sampling it invites exactly the surprises the process exists to prevent, which is why this stage increasingly looks like AI-assisted due diligence: the full record, read consistently, with evidence linked to every judgment.
Appointment: write the expectations down
The PRI’s guidance is blunt on this point: expectations that live in meeting notes do not survive contact with quarterly performance pressure. Whatever the selection process surfaced (integration approach, stewardship commitments, reporting cadence, escalation triggers) belongs in the mandate documentation, in terms specific enough to monitor against. The Global State of Investor Climate Action makes the case with data: investors with explicit commitments are dramatically more active on every downstream practice than those without, and the same logic applies to the asset owner and manager relationship. The mandate is where the success question becomes measurable: it defines what delivering looks like, and it hands every future review its yardstick.
Monitoring and stewardship reporting: the weak link in the chain
Here is where the delegation chain usually frays. Cerulli found 58% of institutional investors require or plan to require portfolio-level ESG risk reporting, but only 23% require managers to report on their engagement activities. The 2026 Glass Lewis Stewardship Survey fills in why: almost half of asset owners use external engagement providers, most want to monitor and report on stewardship carried out on their behalf, and their top-ranked challenges are collecting information from managers and normalizing inconsistent data and formats.
In other words, asset owners largely know what good oversight looks like. The blocker is operational: every manager reports differently, on different cycles, in different formats, and someone has to turn that pile into a consistent, decision-useful view. Annual reviews shrink to a questionnaire, the questionnaire grades its own homework, and progress against the mandate, the thing the whole exercise exists to track, never quite gets measured.
How AI changes the oversight equation
Purpose-built AI turns manager oversight from a sampling exercise into a full-record one. At selection, it assesses every shortlisted manager’s disclosures against the asset owner’s criteria, consistently, with each finding linked to its source. At monitoring, it re-runs the same assessment every cycle, so each review measures progress against the mandate baseline instead of starting from a blank questionnaire. And because the criteria are applied identically across managers, the results are comparable in exactly the way Glass Lewis’s respondents say manager data currently is not.
The asset owner’s team stays in charge of the judgments: what is material, what warrants escalation, when a manager’s practice has drifted from its mandate. The AI removes the part that made those judgments rare, which was the reading.
💡 Manifest Climate assesses your external managers’ disclosure records against your criteria, with source-linked evidence behind every finding, at selection and every review after. Explore our Benchmarking solution.
A practical cadence for asset owner teams
- Define the framework once. Selection criteria, appointment expectations, and monitoring indicators drawn from the same document, so the thing you hired for is the thing you review.
- Assess the full record at selection. Every shortlisted manager, every relevant disclosure, same criteria. Gaps and unsupported claims are findings, not footnotes.
- Put expectations in the mandate. Specific, monitorable, with reporting cadence and escalation paths agreed up front.
- Re-assess on a fixed cycle. Annual at minimum. Compare against the mandate baseline and track the same indicators over time so progress (or drift) shows up as a trendline, not a surprise.
- Close the loop. Findings feed the next review conversation, the watch list, and eventually the next selection process.
Oversee managers with Manifest Climate
Manifest Climate is the AI-powered assessment engine for sustainability, and manager oversight is a natural fit for it: your selection and monitoring framework, applied across every external manager’s full disclosure record, returning consistent, source-linked, audit-ready results, so you always know where each mandate stands. The same engine that asset managers use to assess companies works one level up the chain, for the asset owners assessing them.
If you want to know whether your mandates are on track before the review cycle ends, book a demo and see a manager assessment run on real disclosures.
Frequently asked questions
How does ESG factor into manager selection?
Asset owners apply ESG as an evaluation lens during manager selection: assessing an external manager’s sustainability integration, stewardship capabilities, and supporting evidence before awarding a mandate. The goal is to judge whether the mandate will succeed in that manager’s hands, and to set the baseline for tracking progress against it, inside the select, appoint, and monitor cycle formalized by the PRI’s asset owner guidance.
How do asset owners monitor external managers on ESG?
Through mandate-linked stewardship reporting requirements and periodic reviews of the manager’s disclosures, voting records, and engagement activity, tracking progress against the mandate each cycle. In practice, 58% of institutional investors require or plan to require portfolio-level ESG reporting, but only 23% require engagement reporting, and inconsistent manager data remains the biggest oversight challenge.
What is stewardship reporting?
Stewardship reporting is how investment managers account for the engagement, voting, and escalation activity they carry out on behalf of asset owners. Good stewardship reporting ties activity to outcomes, follows a consistent format the asset owner can compare across managers, and arrives on a cadence agreed in the mandate.
What should asset owners look for when selecting a manager on sustainability?
Evidence over narrative: how ESG factors demonstrably shape the investment process, how stewardship is resourced and executed, voting and engagement records that match stated policy, and reporting the asset owner can actually verify and compare.
How does AI help with manager selection and monitoring?
AI-assisted assessment reads every manager’s full disclosure record against the asset owner’s own criteria, links each finding to its source, and repeats the identical assessment every review cycle, which makes results consistent across managers and comparable over time.
