Guijiang Value Investing Framework

What It Addresses

Investment discussions often fall into predictable traps: reading only the income statement, mistaking cycle peaks for growth, answering from stale data, or prioritizing short-term timing over asset quality. Guijiang Value Investing Framework makes those constraints explicit: before answering, it classifies whether fresh facts are needed, then applies mid-to-long-term lenses such as PB-ROE, grand cycles, and the social balance sheet, reducing the tendency to produce generic or ungrounded analysis.

How It Works

The skill responds from a “Guijiang perspective,” but it is positioned as a framework distilled from public statements, not a representation of the person. The core workflow has three stages:
- Question triage: if the request involves specific companies, industries, or financial data, research first; pure investment-philosophy questions can go straight to mental models; mixed questions combine facts with framework analysis.
- Guijiang-style research: prioritize key variables such as ROE, ROIC, PB, cash flow, dividends, and industry-cycle position, cross-checking local knowledge sources when available.
- Framework-driven answers: start with the broader cycle position, use the PB-ROE framework to compare market pricing against operating quality, and use historical analogies to flag uncertainty.

Recurring lenses include energy-block thinking for asset release, the social balance sheet for payer constraints, potato investing for dividends and duration, and “daughter stocks” versus “son stocks” to distinguish wealth preservation from high-risk innovation.

Scope and Limits

It is useful for long-horizon investment research, industry-cycle reasoning, value-investing retrospectives, and turning scattered financial metrics into a structured discussion. It is not suited for short-term timing, high-frequency trading, technical analysis, crypto, frontier-tech due diligence, or high-leverage strategies. Cycle calls are qualitative, and ROE and PB remain historical data; if real-time data or local references are missing, confidence in the conclusion drops.