Choosing what deserves capital

The questions

• What objective, eligible option set and commitment authority define the capital decision before candidates are scored?
• Which scales, tradeoff weights and performance evidence support the preference model, and which changes reverse the ranking?
• When do budget limits, dependencies or unacceptable performance require a different feasible programme from the ranked list?
• What evidence, challenge record and reopening event justify the next bounded commitment or require another decision method?

The argument

A capital committee cannot delegate its decision to a preference ranking without first defining which proposals are admissible, whose objectives govern the comparison and what commitment is being authorised. This methodological study develops an eligibility-first architecture for transparent scoring, range-aware weighting and decision-directed sensitivity. Its illustrative studies reproduce the Portfolio compass engine and expose both useful preference reversals and failures that a score alone cannot repair. The analysis distinguishes evidence about performance from judgement about tradeoffs, and individual option preference from the selection of a feasible funded programme. It supplies records for option development, evidence review, stakeholder disagreement, committee challenge and learning after commitment. The central claim is conditional: transparency improves the ability to challenge and revise a recommendation when the model represents a legitimate decision problem. It does not make an inappropriate aggregation valid or establish that the selected investment will succeed.

Scope and boundaries

Evidence cut-off: 2026-10-10. Cross-sector analytical methods; every source retains its original jurisdiction and scope

Included

• Portfolio prioritisation and strategic options
• Transparent illustrative examples and reproducible methods
• Counterarguments, decision conditions and practical implementation records

Limits

• Illustrative inputs explain mechanisms and are not actual project data or empirical estimates.
• A deterministic model cannot establish implementation feasibility, legal effect, lender acceptance or future outcomes.
• Cited public guidance is used only within explicitly stated source scope; institutional frameworks require local adaptation.