Perform back-testing
How perform back-testing are reshaped as AGI capability advances.
Business-as-Code
Read as an executable program — the work decomposed into Code, Generative, Agentic, and Human.
Perform back-testing sits inside a larger value-flow — 1 parent structure it composes into. The hierarchy is grounding, not the story: it tells you which aggregate exposure Perform back-testing inherits.
Where Perform back-testing sits
Related articles
No articles yet for this entity.
Recent capability events
No capability events for this entity yet.
How the work flows
Trigger: A new predictive model is proposed, an existing model reaches its scheduled review date, or a regulatory mandate requires validation.
- Define the back-testing scope and extract relevant historical datasets
- Run the predictive model or strategy against the historical data
- Compare the model's generated predictions against actual historical outcomes
- Calculate error margins, variance, and confidence intervals
- Investigate underlying causes for significant predictive deviations or breaches
- Compile the validation report with findings and recalibration recommendations
Outcome: The model's predictive accuracy is quantified against historical data, resulting in a validation report that either approves the model for use or mandates recalibration.