Decision Records are how CREI makes its intelligence public and accountable. Each record captures a consequential corporate decision — the prediction it represents, the assumptions it rests on, CREI's independent view and stated confidence, and the specific evidence that would change that view — and is tested prospectively as reality arrives, so CREI's judgment can be measured against what actually happens.
Most corporate history is rewritten after the fact. A decision that looked reckless becomes "bold" once it works, and a sound decision becomes "a mistake" once luck runs against it.
CREI writes the decision down while it is still live — with an explicit view and confidence — so the quality of the decision can be judged separately from the noise of the outcome. Records are dated, versioned, and revisited on a stated schedule.
Every CREI Decision Record uses the same fields. The example below is illustrative — a demonstration of the format, not a statement about any real company or security.
This is a format demonstration only. It is not a real company, not a real decision, and not a recommendation, opinion, or investment advice regarding any security.
The original view is preserved exactly as written. When evidence changes, CREI adds a dated update — the confidence moves, the reasoning is recorded, and the original remains visible beneath it.
This is what makes a track record honest: the first view and every revision stay on the record, so the judgment can be measured against what actually happened.
Confidence 61%. Defensible on strategy; sensitive to timing. Load-bearing assumption: demand pace.
Confidence 68%. Two periods of demand indicators tracked projection; timing risk reduced. Original view retained below.
Confidence 54%. Early discounting language appeared against a softening order book — an early-warning signal named at record open.
The specific consequential commitment, and when it was made — not a vague theme.
The company's own stated reason for the decision, in its own terms.
The scale of what is being risked, relative to the company's capacity.
The few things that must be true for the decision to work. If one breaks, the thesis breaks.
The public case on both sides, held side by side rather than argued in one direction.
What real failure looks like, and whether the decision can be unwound if it does.
An explicit, stated position and a number — a prediction that can be checked, not a hedge.
The specific evidence that would move CREI's confidence up or down. Named in advance.
The observable public markers that the thesis is starting to break.
When the record is revisited, and where it stands now: open, holding, revised, or closed.
The point of a stated view and a confidence number is that both can be wrong in public.
CREI records a view before the outcome is known, names in advance what would change that view, and revisits each record on a set schedule. When the evidence moves, the record is updated — not quietly rewritten. Over time, the accumulation of dated records is what makes CREI's judgment measurable rather than merely asserted. Decision Records are research built on public evidence; they are not investment advice.
CREI's first public Decision Records are in preparation. Each will be published in the format above, versioned as evidence changes, and listed here — original view preserved, updates dated. To be notified when the register opens, contact CREI.
Decision Records are one expression of how CREI works: an explicit view, a stated confidence, and a public account of what would change its mind.