earnings risk signal detector

Earnings reports often mask deteriorating fundamentals behind adjusted metrics and management optimism. This skill systematically compares current results against prior periods to surface material warning signs that standard summaries miss. It provides a structured framework for auditing guidance, margins, and cash conversion without the manual overhead of cross-referencing multiple PDFs.

What it does

• Comparative analysis identifies shifts in KPIs, margins, and liquidity between reporting periods.
• Signal auditing rates the evidence strength and potential impact of detected red flags.
• Alternative sourcing provides the strongest benign explanations to counter confirmation bias.
• Liquidity monitoring tracks changes in cash flow and debt disclosures.
• KPI tracking defines specific, measurable items to watch in the subsequent reporting cycle.

How it works

1. Source audit to map available earnings releases, transcripts, and 10-Q/K filings.
2. Delta detection to compare management wording and financial guidance period-over-period.
3. Signal synthesis to categorize findings into a risk table with evidence ratings.
4. Counter-argumentation to generate offsetting facts for every identified risk.
5. Summary generation to isolate the high-priority signals requiring active monitoring.

Frameworks & tools

This skill is designed for analysts using LLMs for financial document synthesis. It works best when provided with raw text or OCR outputs from investor relations portals and SEC EDGAR filings.

Why this beats prompting it yourself

Standard prompts often hallucinate negativity or miss subtle changes in management language. This skill enforces a contrarian check by requiring benign explanations for every risk, ensuring a balanced view that avoids typical AI bias toward alarmism.

Use cases

• Auditing quarter-over-quarter changes in margin guidance and bridge commentary.
• Identifying subtle shifts in risk disclosure language within SEC filings.
• Preparing for Q&A by surfacing potential holes in management's narrative.
• Establishing baseline watch items for a new stock in a coverage universe.

Known limitations

Does not provide personalized investment advice or buy/sell recommendations. Requires the user to provide the source text or comparable filings to perform the analysis.