ROBUSTNESS EDUCATION
Monte Carlo analysis: what it can and cannot prove.
Monte Carlo analysis creates alternative paths from stated assumptions or observed outcomes. It can expose sequence and parameter sensitivity, but it cannot repair biased inputs or turn one historical sample into independent market data.
Before you begin
Learning objectives
- Explain: Define what is random
- Explain: A concrete example
- Explain: Useful limits
Define what is random
Reordering observed outcomes asks how sequence affects drawdown; sampling parameters asks a different question. A useful report states what is random and what remains fixed.
A concrete example
Shuffling 100 historical outcomes 10,000 times can estimate how order changes the path. It still assumes those outcomes are representative and may discard serial dependence.
Useful limits
Simulation can expose concentration, tails and fragile assumptions.
- Distribution of drawdown paths
- Dependence on a few large observations
- Sensitivity to costs or missed events
- It cannot prove the modeled distribution is correct
Common mistakes
- Reading the result without the stated scope and assumptions.
- Changing the rule after seeing an outcome while still calling the data unseen evidence.
Trader Checklist
- Distribution of drawdown paths
- Dependence on a few large observations
- Sensitivity to costs or missed events
- It cannot prove the modeled distribution is correct
Practice exercise
Choose one of your own trading examples and write one page of rules, evidence and stopping conditions using the principles in "Monte Carlo analysis: what it can and cannot prove.".
What evidence would overturn the conclusion
The conclusion should be overturned or narrowed if a key assumption cannot be reproduced inside the supported scope, the control cannot be executed, or new out-of-sample evidence repeatedly contradicts it.