Decision process · 02
A decision should be auditable.
The record should show what was believed, which evidence mattered, what could invalidate the thesis and why the view changed.
One manager.
One written record.
Most observations should not become positions. A structured rejection funnel protects attention, tests the quality of evidence and keeps the portfolio from becoming a collection of interesting stories.
Observe
Begin with a concrete shift in the economics or operating structure of a business — not with a price move alone.
What changed, why it matters and why the market may still be looking backwards.
Reject
Remove ideas without a testable mechanism, sufficient evidence, financial staying power or a credible path to becoming visible.
A short record of why the idea stopped — so a future change can be judged against the original concern.
Underwrite
Map the operating mechanism, normalised cash economics, capital needs, balance-sheet constraints and valuation context.
A compact account of the business, the expectation gap and the variables that carry the thesis.
Define
State what must improve, what can break and which evidence would require the thesis or position to change.
Fundamental conditions — not emotional thresholds — that make the original decision no longer valid.
Monitor
Review new evidence against the original map. Pay attention to changes in cash generation, timing, financing and execution.
A dated record of what strengthened, weakened or left the thesis unchanged.
Reassess
Conviction can rise, fall or disappear. The original record remains visible so the quality of the decision can be reviewed separately from its outcome.
Hold, add, reduce, exit or continue watching — with the new evidence and reasoning attached.
The decision record
Writing preserves the difference between what was known at the time and what only became obvious afterwards.