What we saw
The information and market context visible at the time.
SELECTED CASE LIBRARY
Market decisions
preserved for learning
Cases enter this library for the quality of the logic, the decision path, Human–AI collaboration, validation, and learning value—not simply for returns.
3 selected casesHuman approvedReturns are secondary
Case principle
Three questions keep every case grounded
What we decided
The opening-side judgment and how it changed when verified evidence changed.
How the market validated it
Observed evidence is appended without rewriting the original decision.
01
Selected caseINTC
Sep 16 → Sep 21, 2026
AI kept a cautious watch → human identified the opportunity → human-led confirmation
01
What we saw
On Sep 16, AI surfaced INTC and SK hynix as an event-driven narrative but kept it at a cautious C-grade watch. The human investor recognized a larger opportunity in the cross-country industrial coordination that the system had not ranked as the day's best setup.
02
What we decided
Drawing on professional experience, the human judged that this was more than a short event trade. With rate-decision uncertainty close to resolution, the setup also offered a trend entry, so the human expanded both the decision level and intended holding horizon.
03
How the market validated it
On Sep 21, the human proactively introduced MUSE as a follow-on driver, used it to reinforce the CPU-demand thesis, and confirmed an add-on opportunity.
02
Selected caseDELL
Sep 2 → Sep 10, 2026
Expectation gap → evidence strengthened → trend logic upgraded
01
What we saw
An expectation gap with improving industry feedback.
02
What we decided
Entered on Sep 2. Further confirmation on Sep 10 upgraded the case to stronger ongoing trend-following logic.
03
How the market validated it
Subsequent evidence strengthened the original thesis.
03
Selected caseMRNA
Aug 19, 2026
Major catalyst → rapid repricing → no unsupported re-entry
01
What we saw
A major event catalyst created rapid repricing potential.
02
What we decided
Entered quickly and took profit after the large same-day repricing.
03
How the market validated it
No sufficient new reinforcement appeared afterward, so the position was not re-entered.
Admission criteria
Learning value comes before performance headlines
This library publishes only verified case narratives. Missing dialogue, market data, or validation details are not reconstructed from later price action.
1Logic representativeness2Decision-path representativeness3Human–AI collaboration value4Outcome validation value5Learning and review value