Use Case: Not Selling Even After the Thesis Has Broken

Framework for when your investment thesis weakens: name the broken assumption, choose exit/trim/re-underwrite, and avoid loss-aversion paralysis.

When the thesis breaks, “patience” quietly turns into avoidance. This page gives you a decision workflow to (1) separate normal volatility from thesis deterioration, (2) write the broken assumption in one sentence, (3) choose an explicit action (exit, trim, or re-underwrite), and (4) document why you acted for future review. Use it when fundamentals or your valuation band meaningfully change—unit economics, balance-sheet strength, management behavior, or competitive advantage. Don’t use it as an excuse to panic-sell on headlines; if your original thesis is intact, treat the drawdown as a stress test, not a thesis break.

Visual Playbook

Step 1

Name the broken assumption directly

Avoid vague wording. Identify the exact assumption that failed and how that changes expected return or downside.

Step 2

Choose between full exit, trim, or re-underwrite

Once the thesis breaks, the next action should be explicit. Drift and indecision usually make the outcome worse.

Step 3

Document why action was delayed

The learning value is not only in the thesis break but also in why the sell decision was postponed.

Use-Case Playbook

1) Name the broken assumption directly

Avoid vague wording. Identify the exact assumption that failed and how that changes expected return or downside.

2) Choose between full exit, trim, or re-underwrite

Once the thesis breaks, the next action should be explicit. Drift and indecision usually make the outcome worse.

3) Document why action was delayed

The learning value is not only in the thesis break but also in why the sell decision was postponed.

4) Use a post-break action matrix

If the core driver is invalidated, default to exit. If risk increased but the thesis still holds, consider a trim and tighten your review date. If the thesis changed but might still be true, pause new adds and re-underwrite from today’s price as if you did not own the position.

Action Checklist (Shareable)

Write your decision objective in one sentence before reading price action.

Run at least one relevant case in KeepRule Scenarios (/scenarios).

Tie the action to one principle and one invalidation trigger (/prompts).

Set position size from downside tolerance first, then expected upside.

Schedule a 7-day post-mortem using the same checklist before any new change.

Share Kit

FAQ

How do I tell a thesis break from normal volatility?

A thesis break is about fundamentals and probabilities, not price movement. Compare new evidence to your original key assumptions (business quality, balance-sheet safety, competitive position, valuation band). If the assumptions are still intact, treat volatility as a stress test, not an exit signal.

What is a simple action matrix after a thesis break?

Make one explicit choice quickly: exit (core driver invalidated), trim (thesis intact but risk up), or re-underwrite (thesis changed; treat it like a new decision). The key is to stop “doing nothing by default.”

Should I wait to “get back to break-even” before selling?

Break-even is not an investment thesis. Decide based on expected return and risk from today’s price. If you would not buy the position again today under your rules, holding only to erase regret usually compounds mistakes.

How do taxes and trading costs fit into the decision?

They matter, but they are secondary to thesis and downside. If risk has materially worsened, saving taxes is rarely worth staying in a broken thesis. For complex situations, confirm implications with a qualified professional.

What should I record so the thesis break improves my future decisions?

Log the broken assumption, the first signal you missed, why you hesitated (ego, loss aversion, identity), and one rule update. The goal is to convert a painful exit into a repeatable trigger for next time.

Use Case: Not Selling Even After the Thesis Has Broken

Turn this page into one decision step

Quick Take