defeat-beta/defeatbeta-api/defeatbeta-earnings-analysis

Equity Research Earnings Update

Create professional EARNINGS UPDATE REPORTS analyzing quarterly results for companies already under coverage, following institutional standards (JPMorgan, Goldman Sachs, Morgan Stanley format).

Key Characteristics:

• Length: 8-12 pages
• Word Count: 3,000-5,000 words
• Tables: 1-3 summary tables (NOT comprehensive)
• Figures: 8-12 charts
• Audience: Clients already familiar with the company
• Focus: What's NEW — beat/miss, updated estimates, thesis impact
• Font: Times New Roman throughout (unless user specifies otherwise)

When to Use

Use when the user requests:

• "Create an earnings update for [Company] Q3 2024"
• "Analyze [Company]'s quarterly results"
• "Post-earnings report for [Company]"
• "Q1/Q2/Q3/Q4 update for [Company]"

Do NOT use if:

• User requests "initiation report" → Use different skill
• User requests "flash note" or "quick take" → Different format
• Company is not already covered → Need initiation first

Critical Requirements

1. Language Matching

• Match the user's language for all user-facing output, including the DOCX report, delivery summary, replies, visible reasoning notes, assumptions, methodology explanations, and clarification questions
• If the user mixes languages, use the language that dominates the user's request unless the user explicitly specifies the report language
• Keep company names, tickers, financial terms, source names, accounting labels, and quoted source text in their original language when translation would reduce precision
• Do not expose hidden chain-of-thought; provide concise visible rationale, assumptions, and methodology in the user's language when useful
• If the report language differs from the source language, preserve source citations, MCP tool names, ticker symbols, and financial labels accurately

2. Focus

• Focus on what's NEW from this quarter — don't rehash company background
• Assume reader has seen the prior coverage (initiation report or prior earnings update on this name)

3. Beat/Miss Analysis

• Lead with whether company beat or missed estimates
• Quantify variances (e.g., "Revenue beat by $120M or 3%")
• Explain WHY results differed from expectations

4. Summary Format

• Keep tables to 1-3 (summary only, not comprehensive)
• No full P&L/Cash Flow/Balance Sheet (just key metrics)
• Assume reader has seen initiation report

5. Data Source Tiers ⭐⭐⭐ MANDATORY

Every data point in the report falls into exactly one of three tiers. Each tier has a fixed source policy — there is no "default to MCP, sometimes web" ambiguity.

Tier 1 — MCP only (strict)

Reported, authoritative financial data. No web fallback. If MCP returns nothing, state the data gap in the report and proceed; do not patch from 10-Q, press release, IR site, or news snippets.

Tier 2 — MCP preferred, web fallback allowed (must be labeled)

MCP is the primary source; web fallback is allowed when MCP returns nothing or coverage is incomplete. Fallback rows must be explicitly labeled in figures, tables, and the Sources section — never blended with MCP data.

Tier 3 — Web only (MCP does not cover)

Data MCP does not provide. Cite source name and "as of" date. No pretense of MCP coverage.

Where each tool's cache file lives (Call-Then-Write — Phase 1)

After Phase 1, every T1 and T2 MCP tool call has produced one cache file under ./<TICKER>_<PERIOD>/cache/. This is the map: pick the right file to Read when drafting each section.

Two helpful conventions for the multi-tool bundles (valuation_multiples, margins, growth, capital_efficiency, industry):

• Within one bundled file, store each tool's return under a key named after the tool, e.g.

``json { "ttm_pe": {...}, "ev_to_ebitda": {...}, "ps_ratio": {...}, "pb_ratio": {...}, "peg_ratio": {...} } ``

• This still satisfies the verbatim rule — you concatenate raw returns under labeled keys; you don't summarize.

Tier 3 data lands in ./<TICKER>_<PERIOD>/cache/web_<topic>.txt (web excerpts written verbatim from the search result). Phase 1's MCP fetch does not include Tier 3.

6. Citations & Source Attribution ⭐⭐⭐ MANDATORY

Every data point cites its source. The format depends on tier (see Section 5).

Citation formats by tier

Tier 1 (MCP only):

Source: DefeatBeta MCP get_stock_quarterly_income_statement, AAPL,
        fiscal period 2026-03-31, retrieved May 13, 2026.

For transcript-derived commentary, also cite speaker and paragraph number where available:

Source: DefeatBeta MCP get_stock_earning_call_transcript, AAPL FY2026 Q2,
        report date April 30, 2026, CFO prepared remarks, paragraph 14.

Tier 2 (MCP preferred, web fallback labeled):

• If sourced from MCP, use the Tier 1 format above.
• If sourced from web fallback, prefix with "Fallback:" and cite source name + retrieval date:

Fallback source: Apple Q2 FY2026 earnings release supplementary table,
                 https://investor.apple.com/..., retrieved May 13, 2026.
                 (DefeatBeta MCP get_quarterly_revenue_by_segment did not cover this period.)

Tier 3 (web only):

Source: Yahoo Finance analyst estimates, as of April 29, 2026 (pre-earnings close).

Source: TipRanks analyst price targets, retrieved May 13, 2026 (43 analysts, 3-month window).

General rules

• Cite the exact tool name (T1/T2) or source name + "as of" date (T3) — never a generic "Bloomberg" or "company filings"
• Do not display raw URLs for T1/T2 MCP sources; do display URLs for T2 fallback and T3 web sources
• For derived metrics that combine multiple MCP tools, cite each tool used
• For T2, never blend MCP rows and fallback rows in the same column without a fallback label

Sources section layout (end of report)

Group by tier so readers can immediately see what is authoritative vs. fallback vs. web:

SOURCES & REFERENCES
 
Tier 1 — DefeatBeta MCP (authoritative):
• get_stock_quarterly_income_statement, AAPL, fiscal period 2026-03-31, retrieved May 13, 2026
• get_stock_quarterly_balance_sheet, AAPL, fiscal period 2026-03-31, retrieved May 13, 2026
• get_stock_quarterly_cash_flow, AAPL, fiscal period 2026-03-31, retrieved May 13, 2026
• get_stock_earning_call_transcript, AAPL FY2026 Q2, report date April 30, 2026, retrieved May 13, 2026
• get_stock_dcf_analysis, AAPL, retrieved May 13, 2026
 
Tier 2 — DefeatBeta MCP + fallback (where applicable):
• get_quarterly_revenue_by_segment, AAPL, fiscal period 2026-03-31, retrieved May 13, 2026
• Fallback: AAPL Q2 FY2026 earnings press release, https://..., retrieved May 13, 2026
  (used for geographic split — MCP did not cover this period)
 
Tier 3 — Web (MCP does not cover):
• Yahoo Finance consensus estimates, as of April 29, 2026
• TipRanks analyst price targets, retrieved May 13, 2026 (43 analysts)
• Apple Q2 FY2026 user metrics — Apple IR site, https://..., retrieved May 13, 2026