Multi-Company Analysis

Use This Skill When

• The user wants to compare multiple companies within a sector, theme, or watchlist.
• The task is to rank candidates for deeper work, portfolio inclusion, or near-term monitoring.
• The analysis should surface trade-offs across business quality, valuation, and risk rather than crown a winner by vibes.
• The output needs to be consistent across names while still acknowledging imperfect comparability.

Required Inputs

• company list with tickers if available
• comparison objective: best business, best stock, best risk/reward, best short candidate, or best watchlist candidate
• common dimensions: growth, margins, balance sheet, valuation, management, cyclicality, moat, or other user-prioritized factors
• time basis: LTM, NTM, current cycle, or multi-year history
• context: sector, geography, end market, theme, and why these names belong together
• If some data is missing for certain names, keep the framework consistent, flag gaps, and avoid false ranking precision where the evidence is incomplete.

Workflow

1. Define the scope of comparison and explain why each company belongs in the set.
2. Select a compact, decision-relevant set of dimensions and apply them consistently across all names.
3. Summarize each company on the same framework:
   • business quality
   • growth durability
   • margin and cash profile
   • balance-sheet quality
   • valuation
   • principal risks
4. Identify where comparability breaks down because of size, geography, business model, or cycle exposure.
5. Rank or tier the names only after explaining the main trade-offs.
6. End with clear next-step diligence for the top candidates and the uncertain cases.

Output Requirements

• Keep dimensions consistent across the group.
• Explain why the ranking matches the stated objective; "best company" and "best stock" are not always the same.
• Use tiers if the data does not support a strict rank order.
• Distinguish Fact, Assumption, and Inference.
• Note where missing information prevents a high-confidence conclusion.

Output Template

Quality Checks

• Verify each company is being judged on the same stated dimensions.
• Check whether the ranking reflects the user's objective rather than generic quality preference.
• Confirm that differences in business model or cycle exposure are acknowledged before ranking.
• Ensure low-confidence calls are tiered or caveated rather than overstated.
• Remove any numerical scoring system that lacks clear support.

Guardrails

• Keep dimensions consistent across the group.
• Separate facts, assumptions, and inferences explicitly.
• Avoid fake precision in rankings when the evidence only supports rough tiers.
• Do not force comparability where business models are meaningfully different.
• Do not invent metrics for weaker-covered names just to complete a table.

Example Prompts

• Compare these companies and rank them by business quality and investment attractiveness.
• Build a multi-company analysis for this theme and recommend which names deserve deeper research.
• Compare this watchlist for best stock rather than best business, and explain the trade-offs clearly.