Pirate Stock Picking

The Problem It Addresses

Equity research is often distorted by narratives, sentiment, and short-term noise. Investors need a repeatable way to separate durable business quality from price and timing risks. Pirate Stock Picking focuses on deep value evaluation for individual listed stocks, rather than momentum chasing or thematic speculation.

How It Works

The skill evaluates stocks through three dimensions:

• Good business: assesses high moat, high growth, and high profit to judge business model and cash-flow quality
• Good price: compares valuation against industry prototypes, deducted non-recurring profit, and forward expectation
• Good timing: uses turnover, volume, chip distribution, leverage, and other marginal signals as a tactical gate

Operationally, it first normalizes ticker formats such as sh/sz/bj/hk, then fetches quotes, financials, events, and risk data in parallel from sources like westock-data. It runs scripts such as extract.py, growth_score.py, valuation_score.py, and timing_score.py to produce scoring inputs. Finally, render_card.py generates a fixed HTML card, and the result is displayed through show_widget plus present_files.

Boundaries and Caveats

It supports listed individual A-share, HK, and US stocks, but not funds, ETFs, indices, or delisted entities. Real-time data depends on westock-data; earnings preannouncements, fallback financial data, and missing fields require explicit verification. It emphasizes deducted non-recurring profit, dual-source evidence, and conservative assumptions, so it is not suitable for immediate trade signals or purely thematic judgments.