It holds a decision.
CheapMoat measures four locks on a company, shows how far its price sits from its own long trend, and asks you to write the thesis before you act.
The rule
The price never votes.
What a file looks like
Specimen. Not a real company, and not a real measurement.
- ROIC ABOVE COST OF CAPITAL +2.4 ptsHELD
- PIOTROSKI F-SCORE 7 / 9HELD
- ALTMAN Z 1.42✕ BROKEN
- TEN YEAR SLOPE —NOT MEASURED under ten years of prices
One broken link and the chain is gone. There is no three-out-of-four.
The reading
−1.5σ below its own trend.
A distance, nothing more. It arms nothing and triggers no order.
Why it is built this way
What it is
A private journal for a single investor.
A file that cannot be measured is never presented as rejected, and it never leaves the list.
What it does not do
- No advice, no score, no ranking. There is no list of best ideas.
- The price never votes.
- It hides nothing: a line a filter removes is counted on screen, with the criterion that removed it.
- Missing data stays missing. Nothing is filled in with a neutral value.
The method, in three lines
Quality first: three locks read from filed accounts, none weighted against another.
Price second: a regression over a long window, and a deviation from it.
Horizon last, and it is the only part the tool does not measure: the thesis, its invalidation criteria, and the date they will be reviewed.
Where the choices live
Every choice has a dated decision file, written in the context, decision, consequences form.
The journal is private
Accounts are created by hand. There is no sign-up.