equity-valuation-framework

Files (3)

• SKILL.md
• _meta.json
• openai.yaml · agents/

SKILL.md (first lines)

# Equity Valuation Framework

Use this skill as the "rules of the game" for valuation decisions and report standardization.

## Scope and role
- Purpose: transform already-fetched data into a professional valuation view.
- This skill does **not** fetch data.
- Upstream data should come from:
  - `vnstock-free-expert` for company/price/ratio inputs
  - `nso-macro-monitor`, `us-macro-news-monitor`, `vn-market-news-monitor` for macro/news context

## When to trigger
- User asks: "value this stock", "is it cheap/expensive", "best stock between A/B/C", "give me bull/base/bear", "build an investment memo".
- User requests a decision-ready report, not only raw metrics.

## Required input contract
Accept an input bundle with these sections (missing fields allowed, but must be flagged):

```json
{
  "ticker": "HPG",
  "as_of_date": "YYYY-MM-DD",
  "currency": "VND",
  "financials": {
    "income_statement": {},
    "balance_sheet": {},
    "cash_flow": {},
    "ratios": {}
  },
  "price_history": {
    "daily": [],
    "returns": {
      "1m": null,
      "3m": null,
      "6m": null,
      "12m": null
    }
  },
  "peer_set": ["AAA", "BBB"],
  "macro_snapshot": {},
  "news_digest": {},
  "metadata": {
    "source": "kbs|vci",
    "data_quality_notes": []
  }
}
```

## Execution workflow (ordered)
1. Validate input bundle completeness and freshness.
2. Run the data quality gate and assign initial confidence.
3. Select valuation modules based on available data (`Multiples`, `DCF`, sector adaptation).
4. Build bull/base/bear scenarios with explicit assumptions.
5. Triangulate fair value, define safety zone, and list key risks.
6. Apply confidence rubric and disclose gaps that can change conclusions.
7. Return the report using the required section order.

## Data quality gate (must run first)
1. Check freshness: state report periods and price cutoff date.
2. Check completeness: identify missing key lines (revenue, EBIT, net income, CFO, debt, equity, shares).
3. Check consistency: basic identity checks (assets = liabilities + equity if available).
4. Mark confidence tier:
- `High`: complete + recent + internally consistent.
- `Medium`: minor gaps, valuation still usable.
- `Low`: major gaps; only directional view allowed.

## Shared confidence rubric (required)
Use this standardized interpretation:
- `High`: valuation triangulation is valid (>= 2 robust methods), assumptions are explicit, and key inputs are complete.
- `Medium`: only one robust method is usable or moderate gaps require wider valuation ranges.
- `Low`: major input gaps/quality issues force directional valuation only (no precise fair-value claim).

Always report:
1. Confidence level.
2. Which modules were actually run (`Multiples`, `DCF`, sector adaptations).
3. Critical missing inputs that would most likely change fair value.

## Valuation modules
Run modules based on available data. Prefer triangulation (2+ methods).

### 1) Relative valuation (Multiples)
Use when at least one of earnings/book/EBITDA is reliable.

- Core multiples:
  - `P/E` (earnings-based)
  - `P/B` (capital-intensive, banks/financials)
  - `EV/EBITDA` (operating comparison)
  - Optional: `EV/Sales`, `P/CF`
- Compare across:
  - peer median / percentile
  - company 3-5y own history
- Normalize for one-off items when possible.
- Output:
  - implied value range per multiple
  - weighted relative-value estimate

### 2) DCF valuation
Use only when cash-flow visibility is acceptable.

- Model setup:
  - Forecast horizon: 5-10 years (default 5 if uncertain)
  - Revenue growth path by scenario
  - Margin path (EBIT/FCF margin)
  - Reinvestment assumptions
  - WACC with explicit inputs (risk-free, ERP, beta, debt cost)
  - Terminal value: Gordon or exit multiple (state choice)
- Mandatory sensitivity grid:
  - WACC ±100 bps
  - terminal growth ±50 bps
- Output:
  - base/bull/bear fair value
  - sensitivity table

### 3) Sector-specific adaptation
#### Banks / Insurance / Financials
- Prioritize: `P/B`, `ROE`, asset quality proxies, capital adequacy proxies, funding cost/NIM proxies.
- De-emphasize EV/EBITDA.
- Evaluate sustainability of ROE and provisioning pressure.

From claude-office-skills/claude-office-plugin, MIT.

More from claude-office-skills/claude-office-plugin