Markets rate creditworthiness and audit financial statements. The reasoning behind a company's most consequential commitments — where capital is actually won or lost — has never had an independent measure. CREI is building it: a published standard and an independent rating for decision quality, assessed before the outcome is known.
A bond gets a rating. A balance sheet gets an audit. A control environment gets tested. But the quality of the judgment that commits the capital — which assumption carries the outcome, whether the downside was modeled honestly, whether the reasoning was sound before anyone knew how it turned out — is left to confidence and hindsight.
The costliest losses in capital markets are rarely failures of information. They are failures of decision architecture that no one measured in advance. As AI floods every committee with more recommendations than it can examine, an independent measure of decision quality stops being merely useful and becomes necessary.
The one input that determines all the others — the quality of the commitment — has no independent standard. CREI exists to set it.
The rating is an independent assessment of how well a consequential decision is reasoned, structured, and governed, expressed on a defined scale — the way credit quality is expressed in a credit rating. It does not promise that a decision will succeed; outcomes turn on luck as much as judgment. It measures something more durable: whether the decision was made well, given what could be known at the time.
Three independent dimensions, combined into one grade. Two can be assessed today, before any outcome is in; the third accrues as a record forms.
A high rating is no guarantee. A low rating is a warning that the reasoning is structurally exposed — before the capital is committed.
CREI's authority does not rest on private judgment. The standard is published, versioned, and open to challenge — anyone can read how a rating is derived, contest it, and apply it. The Decision-Quality Standard composes three independent instruments, two of which can be assessed at the moment of decision, before any outcome exists.
Assesses the structure of the reasoning: whether the load-bearing assumption is identified and evidence-rated, whether genuine alternatives were weighed, whether the adverse case was modeled honestly, and whether a falsifying signal and an accountable owner were named.
Measures consistency: whether the same decision, put to the same organization, yields the same judgment — or whether the verdict swings with who is in the room and what day it is. An organizational measure, assessable without waiting for outcomes.
Measures whether stated confidence tracks reality: across resolved predictions, does "70% likely" occur about 70% of the time? Brier-based, drawn from the established science of forecast calibration. It accrues — the rating sharpens as the record of resolved predictions grows.
The three combine as a product, not an average: the chain is only as strong as its weakest link, so a polished process cannot paper over an inconsistent organization. As a decision-maker's record of resolved predictions grows, calibration carries more of the weight — one formula carries a rating from its first assessment to a fully outcome-validated grade. Read the full methodology →
The value of an independent standard is precisely its independence. CREI is built so that the body setting the measure has no interest in the result. It does not advise on, build, or implement the decisions it rates — those functions sit in separate hands, at arm's length. The methodology is maintained by a standards committee, not a single author; conflicts are disclosed; the rating opinion is the institution's, governed by published policy.
These structures are being established as CREI moves from a published methodology to a fully governed institution. Founding participants help shape them — and hold them to the standard CREI asks of everyone else.
The Decision Before the Decision sets out the principles the standard formalizes: that a serious decision is a hypothesis about reality, that confidence without architecture is dangerous, and that the quality of a decision can be judged before its outcome is known. The book is the argument; the standard is its instrument.
The quality of the decision is often determined before the recommendation reaches the room.
A rating means something only if it tracks reality — if better-rated decisions, in aggregate and over time, produce more reliable outcomes. That evidence is built, not asserted. CREI maintains a growing record of assessed decisions and their resolutions, and recalibrates the standard against it.
The grounding is established science: the measurement of forecasting calibration, the auditing of organizational noise, and the decision-quality literature each carry decades of validated research. CREI's work is to bring them into a single standard for the enterprise decision — and to prove it, in the open. Validation is the active frontier, not a finished claim.
Govern the reasoning before governing the result. Ask not only whether management is confident, but whether the decision earns a rating.
An independent read on decision quality across a portfolio — diligence on how a management team decides, not only on what they have decided.
A standard to hold a commitment to before it hardens — and a credential to carry once the reasoning is shown to be sound.
Decision quality as a forward signal — an early read on how reliably the commitments behind a relationship are likely to be met.
Creditworthiness was once a matter of reputation and relationship — until it was rated, and the rating became infrastructure capital could not function without. Decision quality is at the same threshold.
It begins with a published standard and the first independently rated decisions. As the record accrues and the rating proves out, the question moves from the margin to the center of how capital is governed: not only what was decided, but how well — and what CREI rated it.
A published methodology and grade scale, open to scrutiny.
Founding participants put real decisions on the record.
An accruing dataset proves the rating predicts.
The independent measure capital governs decisions by.
Read the methodology in full, or join the founding participants helping establish the standard and the first body of independently rated decisions.