Skip to main content

Discipline & sizing

Principles on paper; discipline in position numbers. First layer of risk control: survive, then seek excess return.

1. Total equity vs cash

Environment (subjective tag)Max equity exposureNotes
Risk-on, earnings revisions up70%–85%Still keep L0; no leverage
Neutral50%–70%Default
Credit tightening / extreme valuations / poor personal state30%–50%More cash and short duration
Major uncertainty (war, systemic shock)≤30%Core holdings + L0 only

Tag each quarter in the Decision journal: environment tier + one-line rationale.

2. Single-name and sector caps

TypeCap (% of total equity)Notes
Single stock15%Lower if less understood
Single sector (incl. ETFs)25%Cross-check Sector research
Single country non-index40%Rest via broad indices
L3 opportunity bucket total10%Matches L3 in framework

If over limit: trim first, then add—never stack into a favorite.

3. Building and adjusting positions

Initial entry

ConvictionSuggested tranchesSpacing
High (clear thesis, fair value)2–31–2 weeks or after key data
Medium3–4Around earnings / industry prints
Low (exploratory)Starter ≤3%, then decideAt least one validation event

Add only if all true

  • Move is not the only reason (thesis strengthened or cheaper on forward metrics)
  • Still under single-name cap
  • Journal updated: which assumption was validated
  • Sleep test passes after add

Trim / stop

SituationAction
Falsifier hitCut to zero or watch size; don’t “wait for bounce”
Valuation at upper bandTrim in tranches; optional tracking stub
Emotional gap down, no new factsNo forced stop; rerun checklist
Liquidity crisis / fundamentals brokenReduce quickly; accept slippage

Hard stop (optional): max loss on one position ≤ 25% of that position’s cost or 3% of total equity—pick one, write it in your constitution, don’t flip weekly.

4. Rebalancing

CadenceTask
MonthlyLimits breached? L0 intact?
QuarterlyMacro tier vs total equity; sector vs benchmark
AnnuallyReturn attribution: selection, timing, or allocation

Trigger example: any sleeve ±5% off target weight, or any single name over cap.

5. Never list

  • No margin or opaque leveraged structures (personal boundary—adjust for your jurisdiction).
  • No L3 decisions during major life stress (job change, family crisis).
  • No averaging down because of “get even” psychology.
  • No order without a written thesis.

6. Discipline self-check (monthly)

  • Trade count unusually high? If >2× normal, write why
  • Any limit breaches? Plan to fix when
  • Cash + L0 still ≥ 6 months necessary spending
  • Max drawdown acceptable? If not, drop equity tier one notch
  • Re-read triggered items in Biases
tip

Discipline is not “never wrong”—it’s making errors visible, measurable, and improvable. Name every exception in the journal (e.g. “2025Q2 raised sector cap to 30% for policy bet”) and review next quarter whether to repeat it.