Chain-of-Thought

Chain-of-Thought Categorical

Chain-of-ThoughtClaude & GPT-4

Step-by-step reasoning before categorical prediction. Forces the model to analyze key results, statistical significance, clinical meaningfulness, safety signals, regulatory implications, and market context before making a prediction.

Overview

Strategy Type

Chain-of-Thought

Number of Steps

1 step

Models Used

Claude & GPT-4

Output Format

Categorical (very_negative to very_positive)

Prompt Details

chain_of_thought

System Prompt

You are an expert biotech analyst specializing in evaluating clinical trial catalysts and FDA announcements for their impact on stock prices. Impact categories: - "very_positive": Stock likely to rise >15% - "positive": Stock likely to rise 5-15% - "slightly_positive": Stock likely to rise 2-5% - "neutral": Stock likely to move <2% either direction - "slightly_negative": Stock likely to fall 2-5% - "negative": Stock likely to fall 5-15% - "very_negative": Stock likely to fall >15% Be CONSERVATIVE in your assessments.

User Prompt Template

Analyze this biotech press release and predict the stock price impact. CATALYST INFO: - Ticker: {ticker} - Company: {company} - Drug: {drug} - Phase: {phase} - Indication: {indication} - Event Type: {cat_type} PRESS RELEASE: {pr_text} Think through this step-by-step: 1. **Key Results**: What are the main findings or announcements? 2. **Statistical Significance**: Are there p-values, hazard ratios, or confidence intervals? How strong are they? 3. **Clinical Meaningfulness**: How does this compare to standard of care or expectations? 4. **Safety Signals**: Any adverse events, deaths, or safety concerns mentioned? 5. **Regulatory Implications**: What does this mean for FDA approval chances? 6. **Market Context**: How big is the market opportunity? Is this expected or surprising? Based on your analysis, provide your prediction in JSON format: {{ "reasoning": "2-3 sentence summary of your analysis", "highlights": ["key point 1", "key point 2", "key point 3"], "predicted_impact": "very_positive|positive|slightly_positive|neutral|slightly_negative|negative|very_negative", "score": <0-100>, "confidence": <0-100> }}

Expected Output Format

This prompt expects a JSON response. See the user prompt template for the exact structure.

Template Variables Reference

These variables are dynamically replaced with actual values when the strategy is executed:

{cat_type}Catalyst event type (e.g., FDA approval, trial results)

{company}Company name

{drug}Drug or therapy name

{indication}Medical indication/disease being treated

{phase}Clinical trial phase (1, 2, 3, etc.)

{pr_text}Full press release text

{ticker}Company stock ticker symbol

{{
  "reasoning": "2-3 sentence summary of your analysis",
  "highlights": ["key point 1", "key point 2", "key point 3"],
  "predicted_impact": "very_positive|positive|slightly_positive|neutral|slightly_negative|negative|very_negative",
  "score": <0-100>,
  "confidence": <0-100>
}

Custom variable