Stock Selector

The Specific Problem It Solves

In stock investment analysis, investors often need to screen stocks based on multiple financial and technical indicators. Manually performing these screening tasks is tedious, error-prone, and inefficient for handling large datasets. For example, when simultaneously applying roe profitability and macd technical indicators, separate calculations and comparisons are required, slowing down the process. Moreover, complex strategy combinations, such as requiring "low valuation and high growth," are difficult to iterate and verify quickly through manual operations, limiting the timeliness and accuracy of investment decisions.

Core Capabilities and Workflow

This skill aggregates 11 stock selection strategies into a unified entry point, supporting flexible combinations and efficient execution. Its key features include:

• Strategy Diversity: Covers fundamentals (e.g., roe, dividend), technical indicators (e.g., macd, trend), and pattern analysis (e.g., pattern, bollinger). Each strategy has defined core conditions and scoring dimensions; for instance, the valuation strategy assesses low valuation based on metrics like PE, PB, and PEG.
• Combination Support: Allows independent single-strategy runs or combinations via AND, OR, and SCORE modes. For example, using AND logic to screen stocks that meet both "cashflow quality" and "shareholder concentration," or using SCORE mode to rank strategy results with weights.
• Concurrent Acceleration: Implements concurrency with ThreadPoolExecutor, particularly suitable for IO-intensive technical strategies (e.g., macd, volume_surge). The workers parameter controls concurrency levels, with a recommendation not to exceed 16 to avoid API rate limits, thereby improving data processing speed.
• Visualization Reports: Generates interactive HTML reports for intuitive analysis of screening results. Outputs include JSON, CSV, and HTML files, automatically saved to a specified directory (e.g., an external hard drive), with customizable paths.

The core workflow involves:

1. Strategy Selection: Choose one or more strategies based on investment goals, such as the growth strategy for focusing on growth stocks.
2. Parameter Configuration: Adjust strategy-specific parameters, like ROE thresholds (ROE≥15%, ROA≥5%) for the roe strategy or moving average settings for trend.
3. Execution: Launch the screening process, supporting serial (workers=1) or concurrent modes to adapt to different data loads.
4. Result Output: Save raw data (screener_{strategy}_{time}.json), tabular data (CSV), and visual reports (HTML) for further analysis or archiving.

Applicable Boundaries and Caveats

While this skill enhances screening efficiency, users should be aware of the following limitations:

• Concurrency Optimization: Fundamental strategies (e.g., roe, dividend) involve smaller datasets, so concurrency benefits are limited; technical strategies (e.g., trend, pattern) are recommended for concurrent execution to improve speed, but trade-offs with API stability must be considered.
• API Rate Limits: High concurrency may trigger rate limiting from data source APIs; it's advisable to set workers no higher than 16 and monitor call frequency.
• Data Dependency: The skill relies on external data sources (e.g., real-time market APIs), requiring assurance of data availability and accuracy to avoid screening errors due to missing data.
• Output Management: The default output directory is an external hard drive path, customizable via the output_dir parameter to prevent insufficient storage space or path conflicts.
• Strategy Applicability: Some strategies (e.g., K-line patterns in pattern) depend on technical analysis tools and should be used in conjunction with market conditions to avoid over-reliance on single indicators.