Selected Stocks: Automated Daily Signal Generation & Stock Selection Pipeline

An institutional reference guide to the automated daily scanning, scenario sweeping, and candidate ranking engine.

Target Audience: Execution Traders, Equity Research Analysts & Systems Engineers

Key Takeaways for Investors & Traders

Automated End-of-Day Execution

Runs automatically post-close, executing vectorized sweeps across thousands of US equities in under 90 seconds.

35-Scenario Parametric Matrix

Tests 7 liquidity tranches (A through G) across 5 lookback horizon tiers (Tiers 1 to 5) to isolate robust candidates.

15-Minute EMA Trend Gatekeeping

Employs an intraday EMA triad (50, 200, 800) to confirm that daily signals align with prevailing intraday momentum.

Complete Multi-Format Artifacts

Generates master interactive dashboards (daily_signal_report.html), per-scenario diagnostics (results.html), enriched candidate CSVs (promising_symbols.csv), and compressed signal Parquet archives.

Executive Summary

The Selected Stocks framework is an automated quantitative pipeline executed each weekday evening following US market close. It scans multi-thousand liquid equity universes across six systematic strategy families, evaluates 35 distinct liquidity and parameter scenarios, confirms technical trends using intraday 15-minute Exponential Moving Averages, and synthesizes multi-factor candidate scores enriched with fundamental valuation metrics.

Pipeline Overview & Navigation Guide

The Selected Stocks documentation provides an exhaustive reference to the daily signal generation architecture, divided into ten operational modules:

graph TD
    A[1. Executive Summary] --> B[2. Operational Lifecycle & Workflow]
    B --> C[3. 15m EMA Regime Confirmation]
    C --> D[4. Liquidity Tranche Segregation]
    D --> E[5. Parameter Tier Sweeps]
    E --> F[6. Strategy Catalog & Profiles]
    F --> G[7. Input Schemas & CLI]
    G --> H[8. Interactive Dashboards]
    H --> I[9. Output Data Structures]
    I --> J[10. Monitoring & Best Practices]

Select any module from the Table of Contents sub-menu to inspect its mathematical formulas, configuration schemas, operational runbooks, and data dictionaries.

Summary of Generated Artifacts per Strategy

On each trading day, the pipeline executes across all enabled strategies and compiles a standardized directory of reporting and data artifacts under selected_stocks/{strategy_name}/result/{YYYY-MM-DD}/:

Artifact Filename	File Format	Operational Purpose & Primary Contents	Downstream Consumer
`daily_signal_report.html`	Standalone HTML	Master Daily Signal Dashboard: Top-level KPI cards, 15m EMA regime distribution charts, the 11-column Top Ranked Promising Symbols table, and 35-scenario comparison table.	Portfolio managers, research analysts, web portal iframe.
`results.html` (or `result.html`)	Standalone HTML	Per-Scenario Diagnostic Tearsheet: Individual scenario diagnostics (e.g. `scenario_A1/results.html`) displaying equity curves, trade logs, and parametric metrics.	Quantitative researchers, scenario audit.
`promising_symbols.csv`	Tabular CSV	Enriched Candidate Universe: Top 50 Long and Top 50 Short candidates (max 100 symbols) formatted with 8 common prefix columns, strategy indicators, and 12 fundamental/consensus suffix columns.	Execution desk, trading algorithms, spreadsheet export.
`signals.parquet` / `signals.csv`	Parquet & CSV	Full-Universe Vectorized Signals: Complete historical signal arrays for every analyzed symbol across all 35 sweep scenarios.	Quantitative backtesting engine, machine learning pipelines.
`sweep_summary.csv` / `.parquet`	Tabular & Columnar	Parameter Sweep Summary: Cross-scenario aggregate metrics, including total triggers, confirmation rates, and mean scores across all 35 scenarios.	Strategy performance monitoring, scenario heatmaps.
*`logs/cron_daily_.log`**	Plain Text Log	Pipeline Execution Audit Trail: Complete timestamped execution log including worker pool allocation, database timings, and FMP API HTTP status codes.	Site Reliability Engineers (SREs), DevOps.