What is decision integrity?

The structural definition

Most software in private markets stores something: relationships (CRM), documents (data room), KPIs (portfolio monitoring), accounting state (fund admin), models (Excel + add-ins), narrative (board portal). None of those tools holds the decision as a structured object. The investment thesis lives in a memo on a shared drive. The conditions the IC relied on at approval are written in prose. The forward indicators the team committed to monitor are never tested against the operator data that arrives quarter after quarter.

Decision integrity is the discipline that closes that structural gap. It requires three primitives that don’t exist in the dominant stack:

• →An anchored record of what was known at decision time — including the structured assumptions, the explicit downside scenarios tested, the forward indicators the team committed to monitor. Cryptographically committed at the moment of IC approval; verifiable years later.
• →Continuous drift detection against that anchored record — surfacing what has changed since approval in human-readable language, so the mid-cycle question 'is this decision still defensible at current conditions' has a structured answer rather than depending on analyst memory.
• →A timeline of decision-commit events — preserving the chain of decisions made over the life of the position, with the evidence each was based on. Inspectable rather than reconstructed; survives team turnover and committee transition.

Why this is a separate stack layer

Portfolio monitoring tools (iLEVEL, Allvue, Chronograph, Cobalt) render current KPIs. They don’t hold what was approved or test against it. Fund admin (eFront, FundCount, Juniper Square) keeps the books. CRMs (DealCloud, Affinity) hold the relationship. None of these layers attempts to govern the investment decision through the life of the position. The structural problem isn’t any one tool — it’s that the decision-integrity layer wasn’t built.

Capital Refinery’s research on the private markets stack documents this as the missing 7th layer in a 6-layer landscape. The dominant private markets software stack has been built bottom-up: deal-source layer, diligence layer, accounting layer, monitoring layer, reporting layer, governance layer. Decision integrity sits on top of all of them.

What decision integrity replaces

• →The IC memo as a static document — replaced by an anchored record that re-tests every assumption against every operator update
• →Quarterly reconstruction exercises — replaced by structured diff detection against the IC anchor in human-readable language
• →Institutional memory that walks out the door with the analyst — replaced by an inspectable Decision Timeline per position that survives team transition
• →'We already underwrote this' as a justification — replaced by structured evidence of what was explicitly tested vs implicitly accepted
• →Hindsight-shaped post-mortems — replaced by evidence-based decision-quality review, separated structurally from decision-outcome evaluation

Why it matters now

Three trends make decision integrity necessary rather than optional:

• →Hold periods extended — Bain Global Private Equity Reports document holds stretching into year six and seven. The IC memo that closed the deal is no longer the document anyone opens at year four. The decision basis must survive longer than the team that built it.
• →LP review professionalization — ILPA Principles establish institutional LP expectations for evidence-based governance review. Sophisticated LPs increasingly demand structured decision records, not narrative quarterly letters.
• →Regulatory and fiduciary attention — under Delaware corporate law (Caremark, Marchand, Boeing), board oversight requires demonstrable information systems for material risks. Audit-defensible decision records are structural infrastructure for fiduciary discharge.

What decision integrity is not

• →Not a workflow tool — it doesn't move documents through approval stages; it preserves the decision basis after approval
• →Not portfolio monitoring — monitoring renders current KPIs; decision integrity tests them against IC-approved assumptions
• →Not a reporting layer — reports describe state; decision integrity governs whether prior decisions are still defensible at current state
• →Not an AI assistant — assistants accelerate authoring; decision integrity is the structural discipline that survives independent of who authored the original artifact
• →Not D&O insurance — insurance covers outcomes; decision integrity preserves the evidence of how decisions were made

How Capital Refinery delivers decision integrity

Capital Refinery’s platform is composed of five primitives that together produce the decision-integrity layer:

• →Investment Decision Ledger — the structured record of every IC decision, with a tamper-evident anchor at decision-commit
• →Continuous IC Memo — drift detection between IC cycles; 'Since last IC' panel surfaces what changed against the anchored snapshot in human-readable language
• →Risk Signals scoreboard — observed values vs firm policy thresholds on every IC memo, every export, every portfolio rollup
• →Anchored Exports — every docx/xlsx carries a deterministic fingerprint linking the artifact to the engine state at issuance
• →Decision Timeline — per-position chain of every decision-commit anchor, inspectable rather than reconstructed

The same primitives produce the seller-side Institutional Readiness Assessment — the operating-business mirror of the buy-side IC. One engine, five readers (operator, banker, buyer, lender, LP), every reader underwrites against the same 10-axis framework. See the same-engine architecture for the unifying framing.

Sources cited

• →Bain Global Private Equity Report (annual) — hold-period extension, top-quartile vs bottom-quartile process discipline → https://www.bain.com/insights/topics/global-private-equity-report/
• →ILPA Principles — Institutional Limited Partners Association governance and reporting standards → https://ilpa.org/principles/
• →In re Caremark International Inc. Derivative Litigation, 698 A.2d 959 (Del. Ch. 1996); Marchand v. Barnhill, 212 A.3d 805 (Del. 2019); In re The Boeing Company Derivative Litigation, 2021 WL 4059934 (Del. Ch. 2021) — Delaware board oversight standard
• →Daniel Kahneman, Olivier Sibony, Cass R. Sunstein — Noise: A Flaw in Human Judgment (2021); decision-discipline framework
• →Atul Gawande — The Checklist Manifesto (2009); structural discipline as primitive
• →Dan Lovallo & Olivier Sibony — McKinsey Quarterly 'The Case for Behavioral Strategy' (2010); HBR 'Are You Solving the Right Problem?' (2017)
• →Capital Refinery — research/private-markets-decision-integrity (the 7th-layer thesis)
• →NACD (National Association of Corporate Directors) — board governance principles → https://www.nacdonline.org