finance-industry-research

Start by building an investment logic chain from macro trend to beneficiary stocks, then map the supply chain into upstream, midstream, and downstream segments. Screen listed companies per segment and apply moat, risk, and valuation frameworks to the leaders.

It applies four investor frameworks: Duan Yongping for business quality, Buffett for moat scoring across five types, Munger for risk checklists, and Li Lu for long-term civilization trend judgment. Each company gets a 1-5 star recommendation rating.

Yes, the global scan covers A-shares, Hong Kong stocks, US exchanges, and other international markets like Japan, Korea, and Europe. It also flags key unlisted companies as future IPO candidates and includes sector ETFs.

The report_audit.py script extracts financial data points from the Markdown report, randomly samples 15 percent, and compares each against reliable sources like macrotrends or eastmoney. Reports pass with under 1 percent deviation or get rejected for correction.

AI analysis carries biases toward mature industries, large-cap leaders, and English-language sources, potentially overlooking small caps and unlisted players. The skill mitigates this with explicit anti-bias checks and information-sufficiency grading per company.

No, the SVG charts are generated with a pure Python standard library script requiring zero external dependencies. It supports bar and pie charts for margin comparisons, market share, and portfolio allocation visualization.