# Korea Is Where Cheap Screens Go to Die

Most stock-research agents can summarize cheap stocks. Fewer can delete weak ideas before inventing a thesis.

I turned a Korean equities research piece into a small agent-readable benchmark: can an AI research agent move from a cheap screen to a disciplined kill list using local source checks?

The checks are simple but unforgiving:

- verify ticker and entity identity through KRX/KIND
- verify financial claims through DART
- test liquidity, float, spread, and practical access
- separate theme exposure from actual P&L
- flag trapped cash, related-party economics, and minority-shareholder leakage

The public note has a core survivor stack of Poongsan, BHI, and Woojin, plus a separate moonshot drawer of Innospace, Mobiis, CellBion, IntoCell, and SBB Tech. Those are research prompts, not recommendations.

The broader point: a stock-research LLM is not useful because it can find cheap stocks. It is useful if it can explain which cheap stocks should be deleted.

Agent-readable benchmark: https://barneywohl.github.io/korea-research-map-agents/

GitHub: https://github.com/barneywohl/korea-research-map-agents

Full article: https://barneywohl.substack.com/p/the-korea-research-map-where-cheap?r=7xa6e3

Research commentary only. Not investment advice.