ILLUSTRATIVE SCENARIO — All names, figures and findings are illustrative. No actual specialist work was executed, MAX produced no grade and no Guardian review occurred.
Sample 04 / 06 · Investment Manager Search
$450M Pension Manager Search
Pension trustees compare three sample finalists responding to an RFP for a proposed $450M mandate. No actual manager has been researched.
Example client request
“Pension trustees compare three sample RFP finalists for a proposed $450M mandate.”
$450M
Proposed mandate
Illustrative allocation under review
3
RFP finalists
Sample Manager A, B and C
10 yrs
Horizon
Hypothetical strategic objective
Executive Summary
Illustrative committee readout
No manager selection yet. A top historical return is insufficient. Obtain independently reconciled performance, holdings and fee files; compare risk-adjusted results to mandate objectives; assess operational due diligence; and preserve a committee audit trail.
Research questions AI Team One would investigate
Mandate fit
Confirm investment policy, benchmark, tracking-error budget, liquidity needs and implementation restrictions.
Performance attribution
Net-of-fee time series, alpha significance, factor exposures, regime sensitivity and risk-adjusted return.
Manager organization
People, ownership, succession, incentives, turnover, assets under management and capacity.
Portfolio construction
Holdings overlap, concentration, liquidity, stress losses and interaction with existing managers.
Operational risk
Custody, trade controls, valuation, cybersecurity, compliance, business continuity and service providers.
RFP governance
Equal questions, score definitions, conflict declarations, committee records and implementation milestones.
Illustrative diagnostic flags
Not verified findings
- Headline return biasA higher synthetic return could be explained by sector or factor exposure rather than skill.
- Fee dragCompare all-in implementation cost, not just stated investment-management fees.
- Portfolio overlapA standalone attractive strategy may add little diversification to the existing plan.
Socratee's Critical Thinking Perspective
Board question · sample text, not generated
“Which finalist meets the pension mandate after controlling for fees, factor exposures, drawdowns and operational risk?”
Research & Review Record
Trustee RFP Evaluation Blueprint
All sample manager statistics are fabricated for display. Actual selection requires audited/validated evidence and fiduciary review.
Evidence plan: what must actually be collected
- MandateIPS; funded-status goals; liabilities; approved benchmark and risk budget.
Output: fixed evaluation criteria prior to reviewing bids - PerformanceManager data to independent custodian; time series; composites; calculation method.
Output: consistent net-of-fee and risk-adjusted comparisons - PortfolioHoldings history; factor model; stress and liquidity analytics.
Output: marginal impact on the total pension portfolio - People / operationsRFP responses; regulatory materials; DDQ; independent operations references.
Output: control assessment, conflicts and key-person issues - ImplementationFee proposals; capacity; transitions; legal documents; counsel and trustees.
Output: documented shortlist, conditions and approval record
Example: fabricated manager comparison
All statistics synthetic; 5-year annualized net returns do not imply future results.
ManagerA
- 5Y net return
- 7.5%
- Fee
- 0.42%
- Tracking error
- 3.2%
- Decision
- Diligence
ManagerB
- 5Y net return
- 7.9%
- Fee
- 0.58%
- Tracking error
- 5.8%
- Decision
- Diligence
ManagerC
- 5Y net return
- 7.3%
- Fee
- 0.36%
- Tracking error
- 2.1%
- Decision
- Diligence
What independent challenge and safeguards would examine
SOCRATEE + RED
Would question benchmark choice, survivorship bias, changing factor exposure, liquidity under stress and optimistic manager explanations.
LIBBY + MAX
Would preserve procurement fairness, fiduciary documentation and independent checking. MAX grades the analysis record, not the investment outcome.
Proposed role routing · Not executed
TED coordinates • RES conducts manager/market research • KNOLLY cross-checks data • SOCRATEE probes attribution • LIBBY checks mandates • RED stress-tests • MAX evaluates completeness.
Actual record
- Actual agent executions: 0
- MAX independent grade: No grade. Not run.
- 18 Guardians: No review occurred. Built · Integration Pending.
- Final decision: Always a human decision.
Evidence labels
Sample scenario · Unverified illustrative content · 0 actual agent executions