Isolate Marginal Attrition
Track how each condition removes candidates step-by-step to expose assumptions that prune 90% of candidates silently.
Detect Phantom Redundancies
Identify secondary filters that duplicate an earlier boundary, adding formula overhead without filtering new outliers.
Verify Threshold Stability
Sweep parameters across plus and minus ten percent increments to confirm candidates do not fall off an arbitrary cliff.
Screening Assumption Validation Protocol
Researchers frequently construct compound multi-factor screeners under the unverified belief that every additional condition tightens focus on authentic signal. In practice, layered parameters interact unpredictably. A volume filter intended to exclude illiquid securities might accidentally remove high-growth assets whose turnover happens to cluster below an arbitrary round-number threshold. Assumption Tester isolates each criterion into an atomic step, measuring total candidate loss and revealing whether your filter isolates real traits or merely trims candidate lists at random.
The tool produces an exportable diagnostic profile containing pairwise correlation indicators, incremental candidate drops, and a stability variance coefficient. By systematically nudging boundary thresholds across pre-selected universes, you can pinpoint exactly which metric dictates outcome composition and which variables serve solely as decorative complexity.
Technical Parameters & Execution Profiles
| Specification Parameter | Value / Configuration | Application Context |
|---|---|---|
| Max Condition Layers | Up to 24 simultaneous rules | Compound screeners containing technical, fundamental, and liquidity tests. |
| Sensitivity Delta Window | ±5% to ±30% step sweeps | Stress-testing rigid parameters to distinguish robust trends from curve-fit edges. |
| Redundancy Metric | Jaccard Similarity & Rank Correlation | Exposing conditions whose candidate lists overlap by 95% or higher. |
| Target Platform Compatibility | TC2000, Python Pandas, QuantConnect | Plug-and-play formula scripts with standard CSV/JSON input-output structures. |
Included Diagnostic Assets & Documentation
Sensitivity Evaluation Script
Python automation file that ingests rule syntax and iterates numeric cutoffs automatically.
TC2000 Condition Parser Snippets
Pre-formatted PCF formula blocks designed to separate compound formulas into individual step columns.
Overlap & Pruning Matrix Template
Structured spreadsheet workbook for logging baseline universes against successive filter stages.
Framework Reference Manual
Comprehensive guide highlighting baseline stability benchmarks and common screening fallacies.
Three-Phase Audit Execution Cycle
Condition Disaggregation
Deconstruct long-form logical strings into isolated clauses. Each metric receives its own calculation track against your target historical baseline or active watchlist.
Marginal Attrition Sweep
Execute automated perturbation routines that shift numeric parameters across specified bands, capturing candidate universe shifts at every intermediate interval.
Synthesis and Redundancy Flagging
Review the generated summary report. Remove duplicate filtering criteria and adjust boundary parameters so results reflect persistent characteristics rather than arbitrary cutoffs.
Frequently Asked Questions
Audit Your Screening Assumptions Today
Eliminate hidden filter bias, isolate fragile cutoffs, and document criterion impact with our standalone analytical module.