Who this is for — Anyone who wants to know whether their edge is truly transferable or depends on a setting too fragile for the real world.
Robustness measures how well a strategy holds up to small changes without losing its identity: slightly different parameters, higher costs, different market phases, less-than-ideal execution.
In plain terms — If a small tweak makes everything collapse, you do not have a reliable machine: you have a delicate prototype.
Bronze prerequisite — Before this lesson: trading-journal, trade-result, trading-mistake. See bronze-path.
How to test robustness
A robust test does not seek the best result, but stability:
- moderate variations of main parameters;
- prudent increase in costs and slippage;
- segmentation by market regime;
- comparison across similar instruments;
- checking curve shape, not only the final total.
If the strategy stays understandable and positive in realistic scenarios, operational confidence grows.
Example — A trend-following system stays profitable with stops between 1.5× and 2.2× ATR and costs +30%. Performance is less brilliant, but behaviour is consistent: a sign of robustness.
Practical decisions to make
- Define acceptable parameter ranges, not a single value.
- Build a safety margin into costs.
- Keep active only what holds across plausible scenarios.
- Revalidate after structural market changes.
Card
- What it is: resilience of the method to realistic input and context variations.
- When to use it: before going live and after every relevant system revision.
- Typical mistake: choosing the parameter that maximises the past instead of the most stable one.
Silver path — Module: Validation. Part of silver-path.