Asymmetry & Convexity
Updated 2026-08-11
INTRODUCTION
English translation pending.
CORE DEFINITION
A structural idea in Charlie Munger's investing framework, closely tied to the convexity argument later formalised by Nassim Taleb. The core claim is that the shape of the payoff distribution, not the accuracy of a forecast, determines long-run compounding: when losses are bounded and gains are not, repeated participation produces a positive geometric outcome even without predictive skill. The key qualification is that convex positions usually bleed small amounts for long periods before paying off, so they require both a durable budget and patience. Munger described the search for such setups as the centre of his approach to investing.
SCAFFOLDING EFFECT
Reduce cognitive load
- Two-sided map: write the worst and best outcome and their rough probability orders first - Floor test: check whether the worst case has a hard floor such as limited liability - Structure filter: prefer capped loss with open gain and reject capped gain with open loss
Anchor fast decisions
The shape of the payoff distribution decides the direction of long-run compounding. When losses are capped while gains stay open, the same volatility produces a higher geometric mean return, because gains are compounded while losses are truncated; volatility itself becomes a source of return. The reverse structure, capped gains with open losses, drives the participant to ruin through repetition no matter how good the average forecast is. The practical consequence is that the first question is not which way the price will move but where the two boundaries sit.
MINIMUM ACTION
In progress 0/1Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- mungermodels.comhttps://mungermodels.com/models/asymmetry-convexityverified
PRIVATE NOTES · Only visible to you
SAVED Q&A
ENTRY Q&A · Private saving available
Ask with a clear boundary
thinkingmodels answers from published entry context only.
Your question is sent to thinkingmodels. The answer uses public entry context only.
RELATED MODELS