Balance Labs

EN · 8 min · Updated 2026-07-19

Berkshire Fundamentals Desk: Buffett, Munger, Duan & Li Lu Scores Explained

The Berkshire desk turns four investment lenses into one readable card: business quality, valuation, competition, and permanence of risk.

The four master panels

Duan · Business — margins, ROE, revenue quality, whether this is a “good business”.

Buffett · Valuation — PE, FCF yield, PEG/PB style cheapness relative to earnings power.

Munger · Competition — fragility from thin margins, leverage, and failure modes.

Li Lu · Risk — balance sheet stress, liquidity, and what can permanently impair capital.

Composite score vs verdict

Composite score averages the four masters (roughly 1–5). Verdict (BUY / HOLD / WATCH / AVOID) summarizes the desk output for quick scanning.

A high score with weak timing can still be a watchlist name. A strong screener signal with weak fundamentals is often a trap.

Why screener fields are timing only

Bull %, edge, and signal age help with when — not whether the business deserves capital.

Balance Labs keeps STM fields as an overlay so the desk does not inflate scores just because price is moving.

How to read bull / bear points

Bull points should cite durable advantages or valuation support. Bear points should name real failure modes — competition, leverage, regulation, or earnings quality — not generic fear.

If a card shows empty masters or “No bull points yet”, refresh from cache or re-run research; incomplete stubs are not a signal.

Try Balance Labs free

Stock Chat + Screener preview · 50 monthly AI credits.

Open Stock Chat

FAQ

Is the Berkshire desk the same as Warren Buffett’s letters?

No. It is a structured AI + fundamentals framework inspired by those lenses, not a copy of Berkshire Hathaway recommendations.

Where do the numbers come from?

Primarily live fundamental packets (Yahoo-style metrics) and optional AI summarization on top of those fields.

What score is “good enough”?

Context matters, but many workflows treat stronger composites (around 3.5+) and non-AVOID verdicts as a practical gate before sizing.

← All guides · Home