Thesis risk
Define the evidence that invalidates the case. Review that evidence on a set schedule.
Risk discipline
Risk is a chain,
not a single limit.
The method controls failure at the thesis, position, portfolio, execution, and process levels.
Five layers
Each layer has a distinct control.
A stop price cannot replace research quality, portfolio judgment, or execution checks.
Position risk
Scale exposure to conviction and uncertainty. Keep the position small when the evidence remains incomplete.
Portfolio risk
Measure gross exposure, related themes, sector concentration, and liquidity before each change.
Execution risk
Separate the durable portfolio target from broker orders. Reject stale, inconsistent, or oversized plans.
Process risk
Keep every decision and outcome. Review errors in sourcing, judgment, construction, and execution.
Portfolio construction
Concentration follows evidence.
The portfolio process allocates only when a case improves the whole book.
Construction rule
Capital stays scarce
A new case competes with holdings and cash. The process does not fill positions to meet a target count.
Construction rule
Conviction earns concentration
The strongest supported cases receive more weight. Uncertainty receives less weight or no position.
Construction rule
Correlation counts
Different tickers can express the same risk. The portfolio treats shared drivers as one exposure.
Construction rule
Cash is a decision
The method accepts unused capacity when the available cases do not justify more exposure.
Review clock
The record closes the loop.
The process keeps the original case intact and appends each later judgment.
At decision
Store the thesis, evidence, price reference, risk, and review dates.
During the hold
Compare new facts with the original case. Record each keep, trim, add, or exit.
After the outcome
Grade the process and the result separately. Keep the complete set of eligible decisions.