Build a Stock Analysis Agent with AutoGPT (Fundamentals + News)

Financial research is one of the clearest use cases for autonomous AI agents. The task structure is perfect — gather data from multiple sources, synthesize it, apply consistent analytical frameworks, and produce a structured output. That's exactly what AutoGPT does well.

This guide walks through building a stock analysis agent that fetches fundamental data from Yahoo Finance, summarizes recent news, applies basic analytical frameworks, and outputs a structured investment research report. We'll also be clear about where this approach falls short and what you should never rely on it for.

Disclaimer: Nothing in this article constitutes financial advice. The agent described here is a research tool. Automated analysis has significant limitations and can be wrong. Always consult a licensed financial advisor before making investment decisions.

What the Agent Will Do

Before writing any code, define the exact scope. This agent will:

Accept a stock ticker symbol as input
Fetch current price, fundamentals, and historical data via Yahoo Finance
Pull and summarize recent news articles about the company
Apply basic valuation checks (P/E, P/B, debt ratios)
Generate a structured markdown report with a summary, data tables, and analyst notes

What it will explicitly not do: give buy/sell recommendations, predict prices, or access paid financial data sources.

Setting Up the Environment

pip install yfinance autogen-agentchat openai newspaper3k requests beautifulsoup4 pandas

Create your project structure:

stock_agent/
├── main.py
├── tools/
│   ├── __init__.py
│   ├── yahoo_finance.py
│   └── news_fetcher.py
├── prompts/
│   └── analyst_system.txt
└── output/

Building the Yahoo Finance Tool

The core data fetching module handles fundamentals, price history, and basic metrics:

# tools/yahoo_finance.py
import yfinance as yf
import pandas as pd
from typing import Dict, Any, Optional
import json


def get_stock_fundamentals(ticker: str) -> Dict[str, Any]:
    """Fetch key fundamental data for a stock ticker."""
    try:
        stock = yf.Ticker(ticker)
        info = stock.info

        # Extract key metrics with safe fallbacks
        fundamentals = {
            "ticker": ticker.upper(),
            "company_name": info.get("longName", "Unknown"),
            "sector": info.get("sector", "Unknown"),
            "industry": info.get("industry", "Unknown"),
            "market_cap": info.get("marketCap"),
            "current_price": info.get("currentPrice") or info.get("regularMarketPrice"),
            "52_week_high": info.get("fiftyTwoWeekHigh"),
            "52_week_low": info.get("fiftyTwoWeekLow"),
            "pe_ratio": info.get("trailingPE"),
            "forward_pe": info.get("forwardPE"),
            "pb_ratio": info.get("priceToBook"),
            "ps_ratio": info.get("priceToSalesTrailing12Months"),
            "ev_to_ebitda": info.get("enterpriseToEbitda"),
            "profit_margin": info.get("profitMargins"),
            "revenue_growth": info.get("revenueGrowth"),
            "earnings_growth": info.get("earningsGrowth"),
            "debt_to_equity": info.get("debtToEquity"),
            "current_ratio": info.get("currentRatio"),
            "return_on_equity": info.get("returnOnEquity"),
            "return_on_assets": info.get("returnOnAssets"),
            "dividend_yield": info.get("dividendYield"),
            "analyst_target_price": info.get("targetMeanPrice"),
            "recommendation": info.get("recommendationMean"),
            "number_of_analysts": info.get("numberOfAnalystOpinions"),
            "business_summary": info.get("longBusinessSummary", "")[:500],
        }

        return {"status": "success", "data": fundamentals}

    except Exception as e:
        return {"status": "error", "message": str(e)}


def get_price_history(ticker: str, period: str = "1y") -> Dict[str, Any]:
    """Get historical price data and calculate basic metrics."""
    try:
        stock = yf.Ticker(ticker)
        hist = stock.history(period=period)

        if hist.empty:
            return {"status": "error", "message": "No price data found"}

        # Calculate returns
        start_price = float(hist["Close"].iloc[0])
        end_price = float(hist["Close"].iloc[-1])
        total_return = ((end_price - start_price) / start_price) * 100

        # Volatility (annualized)
        daily_returns = hist["Close"].pct_change().dropna()
        annualized_volatility = float(daily_returns.std() * (252 ** 0.5) * 100)

        # Average volume
        avg_volume = float(hist["Volume"].mean())

        summary = {
            "period": period,
            "start_price": round(start_price, 2),
            "end_price": round(end_price, 2),
            "total_return_pct": round(total_return, 2),
            "annualized_volatility_pct": round(annualized_volatility, 2),
            "average_daily_volume": int(avg_volume),
            "data_points": len(hist),
        }

        return {"status": "success", "data": summary}

    except Exception as e:
        return {"status": "error", "message": str(e)}


def get_financial_statements(ticker: str) -> Dict[str, Any]:
    """Get simplified income statement and balance sheet data."""
    try:
        stock = yf.Ticker(ticker)

        # Income statement (annual)
        income_stmt = stock.income_stmt
        balance_sheet = stock.balance_sheet

        result = {}

        if not income_stmt.empty:
            latest_year = income_stmt.columns[0]
            result["income_latest_year"] = str(latest_year.date())
            result["total_revenue"] = _safe_float(income_stmt, "Total Revenue", latest_year)
            result["gross_profit"] = _safe_float(income_stmt, "Gross Profit", latest_year)
            result["operating_income"] = _safe_float(income_stmt, "Operating Income", latest_year)
            result["net_income"] = _safe_float(income_stmt, "Net Income", latest_year)

        if not balance_sheet.empty:
            latest_quarter = balance_sheet.columns[0]
            result["total_assets"] = _safe_float(balance_sheet, "Total Assets", latest_quarter)
            result["total_debt"] = _safe_float(balance_sheet, "Total Debt", latest_quarter)
            result["cash"] = _safe_float(balance_sheet, "Cash And Cash Equivalents", latest_quarter)

        return {"status": "success", "data": result}

    except Exception as e:
        return {"status": "error", "message": str(e)}


def _safe_float(df, row_name: str, col) -> Optional[float]:
    try:
        val = df.loc[row_name, col]
        return float(val) if pd.notna(val) else None
    except (KeyError, TypeError):
        return None

Building the News Fetcher

# tools/news_fetcher.py
import yfinance as yf
from typing import List, Dict


def get_stock_news(ticker: str, max_articles: int = 10) -> Dict:
    """Fetch recent news headlines for a stock ticker."""
    try:
        stock = yf.Ticker(ticker)
        news = stock.news

        articles = []
        for item in (news or [])[:max_articles]:
            articles.append({
                "title": item.get("title", ""),
                "publisher": item.get("publisher", ""),
                "link": item.get("link", ""),
                "published": item.get("providerPublishTime", 0),
                "type": item.get("type", ""),
                "summary": item.get("summary", "")[:300] if item.get("summary") else "",
            })

        return {
            "status": "success",
            "ticker": ticker.upper(),
            "article_count": len(articles),
            "articles": articles,
        }

    except Exception as e:
        return {"status": "error", "message": str(e)}

Configuring the AutoGen Analyst Agent

Now build the actual AutoGen agent that uses these tools:

# main.py
import autogen
import json
import os
from datetime import datetime
from tools.yahoo_finance import get_stock_fundamentals, get_price_history, get_financial_statements
from tools.news_fetcher import get_stock_news


def analyze_stock(ticker: str, output_dir: str = "output") -> str:
    os.makedirs(output_dir, exist_ok=True)

    llm_config = {
        "config_list": [
            {
                "model": "gpt-4o",
                "api_key": os.getenv("OPENAI_API_KEY"),
            }
        ],
        "temperature": 0.1,
        "functions": [
            {
                "name": "get_fundamentals",
                "description": "Fetch fundamental financial data for a stock ticker",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "ticker": {"type": "string", "description": "Stock ticker symbol"}
                    },
                    "required": ["ticker"]
                }
            },
            {
                "name": "get_price_history",
                "description": "Get 1-year price history and return metrics",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "ticker": {"type": "string"}
                    },
                    "required": ["ticker"]
                }
            },
            {
                "name": "get_news",
                "description": "Get recent news articles about the stock",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "ticker": {"type": "string"},
                        "max_articles": {"type": "integer", "default": 8}
                    },
                    "required": ["ticker"]
                }
            },
            {
                "name": "get_financials",
                "description": "Get income statement and balance sheet data",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "ticker": {"type": "string"}
                    },
                    "required": ["ticker"]
                }
            },
        ]
    }

    analyst = autogen.AssistantAgent(
        name="StockAnalyst",
        system_message="""You are a financial research analyst. Your job is to produce
structured research reports based on publicly available financial data.

IMPORTANT DISCLAIMER: You must always include a prominent disclaimer that your analysis
is for informational purposes only and does not constitute investment advice.

When analyzing a stock, follow this process:
1. Fetch fundamental data using get_fundamentals
2. Get 1-year price history using get_price_history
3. Get financial statements using get_financials
4. Fetch recent news using get_news
5. Synthesize all data into a structured markdown report

Your report must include:
- Company overview
- Key metrics table (P/E, P/B, debt ratios, margins)
- 1-year price performance summary
- Recent news summary (3-5 key themes)
- Basic valuation context (compare P/E to sector average if known)
- Risks and considerations
- Disclaimer

End your response with ANALYSIS_COMPLETE when done.""",
        llm_config=llm_config,
    )

    # Function map for the user proxy
    function_map = {
        "get_fundamentals": lambda ticker: get_stock_fundamentals(ticker),
        "get_price_history": lambda ticker: get_price_history(ticker),
        "get_news": lambda ticker, max_articles=8: get_stock_news(ticker, max_articles),
        "get_financials": lambda ticker: get_financial_statements(ticker),
    }

    user_proxy = autogen.UserProxyAgent(
        name="ResearchCoordinator",
        human_input_mode="NEVER",
        max_consecutive_auto_reply=15,
        is_termination_msg=lambda msg: "ANALYSIS_COMPLETE" in (msg.get("content") or ""),
        function_map=function_map,
        code_execution_config=False,
    )

    user_proxy.initiate_chat(
        analyst,
        message=f"Please conduct a comprehensive research analysis of {ticker.upper()} stock. Use all available tools to gather data and produce a structured report.",
    )

    # Extract the final report
    messages = analyst.chat_messages.get(user_proxy, [])
    final_report = ""
    for msg in reversed(messages):
        if msg.get("role") == "assistant" and msg.get("content"):
            final_report = msg["content"].replace("ANALYSIS_COMPLETE", "").strip()
            break

    # Save the report
    timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
    output_file = os.path.join(output_dir, f"{ticker.upper()}_analysis_{timestamp}.md")

    with open(output_file, "w", encoding="utf-8") as f:
        f.write(f"# Stock Analysis: {ticker.upper()}\n")
        f.write(f"*Generated: {datetime.now().strftime('%Y-%m-%d %H:%M UTC')}*\n\n")
        f.write("---\n\n")
        f.write(final_report)

    print(f"Analysis saved to: {output_file}")
    return output_file


if __name__ == "__main__":
    import sys
    ticker = sys.argv[1] if len(sys.argv) > 1 else "AAPL"
    analyze_stock(ticker)

Understanding the Output Format

A well-configured run produces output like this:

# Stock Analysis: AAPL
*Generated: 2026-05-31 14:30 UTC*

---

## Company Overview
Apple Inc. (AAPL) is a consumer electronics and software company...

## Key Metrics

| Metric | Value | Context |
|--------|-------|---------|
| P/E Ratio (TTM) | 28.4 | Above tech sector avg (~25) |
| Forward P/E | 25.1 | More reasonable on growth basis |
| P/B Ratio | 42.1 | Premium brand valuation |
| Debt/Equity | 1.73 | Moderate leverage |
| Profit Margin | 25.3% | Strong operating efficiency |
| Revenue Growth (YoY) | 4.2% | Moderate growth |
| Analyst Target | $198.50 | Based on 42 analyst estimates |

## 1-Year Price Performance
- Start: $165.23 → Current: $182.41
- Total Return: +10.4%
- Annualized Volatility: 22.1%

## Recent News Themes
1. **AI Integration:** Multiple reports on Apple Intelligence rollout...
2. **Services Growth:** App Store revenue up in Q1 2026...

## ⚠️ DISCLAIMER
This report is generated by an AI system for informational purposes only.
It does not constitute investment advice...

Limitations You Need to Know

Limitation	Impact	Mitigation
15-minute data delay	Price data not real-time	Note timestamp in output
LLM hallucination risk	Agent may misstate ratios	Cross-check key figures manually
No analyst consensus detail	Shallow competitive analysis	Supplement with earnings call transcripts
No sentiment scoring	News summary is qualitative	Add NLP sentiment layer manually
No historical comparison	Can't benchmark against sector	Add sector ETF comparison
Token context limits	Long news articles get truncated	Summarize before passing to agent

This type of structured research automation is discussed extensively in the AI research agent build guide, which covers more advanced patterns for handling data quality and fact verification.

Running the Agent

export OPENAI_API_KEY=sk-...
python main.py MSFT

Expected runtime: 2-4 minutes per ticker, depending on data availability and LLM response speed. Token usage typically runs 3,000-6,000 tokens per analysis using GPT-4o.

For batch analysis across a watchlist:

# Create a watchlist file
echo "AAPL\nMSFT\nGOOGL\nAMZN" > watchlist.txt

# Run analysis on each
while read ticker; do
  python main.py $ticker
  sleep 10  # Rate limit courtesy delay
done < watchlist.txt

The same principles that make this agent work well connect directly to Build AI agent with LangChain patterns — structured tool use, clear termination conditions, and output validation are universal regardless of which framework you use.

Frequently Asked Questions

Is AutoGPT stock analysis accurate enough for real investment decisions? No. AutoGPT-generated stock analysis should be treated as a research starting point, not investment advice. The agent can make factual errors, misinterpret financial data, or miss critical context. Always verify outputs with authoritative sources and consult a licensed financial advisor before making investment decisions.

Can this agent access real-time stock prices? The Yahoo Finance integration via yfinance provides near-real-time quotes with a 15-minute delay for free tier access. For actual real-time data, you'd need a paid data provider like Alpha Vantage premium, Polygon.io, or a direct exchange data feed integrated as a custom AutoGPT plugin.

How do I run this agent on multiple stocks at once? Use a loop that creates a separate agent instance for each ticker, or build a multi-agent setup where a coordinator agent spawns individual analyst agents per stock. Running more than 5-10 simultaneous analysis tasks will hit API rate limits and increase costs significantly.

What the Agent Will Do

Before writing any code, define the exact scope. This agent will:

Accept a stock ticker symbol as input
Fetch current price, fundamentals, and historical data via Yahoo Finance
Pull and summarize recent news articles about the company
Apply basic valuation checks (P/E, P/B, debt ratios)
Generate a structured markdown report with a summary, data tables, and analyst notes

What it will explicitly not do: give buy/sell recommendations, predict prices, or access paid financial data sources.

Setting Up the Environment

pip install yfinance autogen-agentchat openai newspaper3k requests beautifulsoup4 pandas

Create your project structure:

stock_agent/
├── main.py
├── tools/
│   ├── __init__.py
│   ├── yahoo_finance.py
│   └── news_fetcher.py
├── prompts/
│   └── analyst_system.txt
└── output/

Building the Yahoo Finance Tool

The core data fetching module handles fundamentals, price history, and basic metrics:

# tools/yahoo_finance.py
import yfinance as yf
import pandas as pd
from typing import Dict, Any, Optional
import json


def get_stock_fundamentals(ticker: str) -> Dict[str, Any]:
    """Fetch key fundamental data for a stock ticker."""
    try:
        stock = yf.Ticker(ticker)
        info = stock.info

        # Extract key metrics with safe fallbacks
        fundamentals = {
            "ticker": ticker.upper(),
            "company_name": info.get("longName", "Unknown"),
            "sector": info.get("sector", "Unknown"),
            "industry": info.get("industry", "Unknown"),
            "market_cap": info.get("marketCap"),
            "current_price": info.get("currentPrice") or info.get("regularMarketPrice"),
            "52_week_high": info.get("fiftyTwoWeekHigh"),
            "52_week_low": info.get("fiftyTwoWeekLow"),
            "pe_ratio": info.get("trailingPE"),
            "forward_pe": info.get("forwardPE"),
            "pb_ratio": info.get("priceToBook"),
            "ps_ratio": info.get("priceToSalesTrailing12Months"),
            "ev_to_ebitda": info.get("enterpriseToEbitda"),
            "profit_margin": info.get("profitMargins"),
            "revenue_growth": info.get("revenueGrowth"),
            "earnings_growth": info.get("earningsGrowth"),
            "debt_to_equity": info.get("debtToEquity"),
            "current_ratio": info.get("currentRatio"),
            "return_on_equity": info.get("returnOnEquity"),
            "return_on_assets": info.get("returnOnAssets"),
            "dividend_yield": info.get("dividendYield"),
            "analyst_target_price": info.get("targetMeanPrice"),
            "recommendation": info.get("recommendationMean"),
            "number_of_analysts": info.get("numberOfAnalystOpinions"),
            "business_summary": info.get("longBusinessSummary", "")[:500],
        }

        return {"status": "success", "data": fundamentals}

    except Exception as e:
        return {"status": "error", "message": str(e)}


def get_price_history(ticker: str, period: str = "1y") -> Dict[str, Any]:
    """Get historical price data and calculate basic metrics."""
    try:
        stock = yf.Ticker(ticker)
        hist = stock.history(period=period)

        if hist.empty:
            return {"status": "error", "message": "No price data found"}

        # Calculate returns
        start_price = float(hist["Close"].iloc[0])
        end_price = float(hist["Close"].iloc[-1])
        total_return = ((end_price - start_price) / start_price) * 100

        # Volatility (annualized)
        daily_returns = hist["Close"].pct_change().dropna()
        annualized_volatility = float(daily_returns.std() * (252 ** 0.5) * 100)

        # Average volume
        avg_volume = float(hist["Volume"].mean())

        summary = {
            "period": period,
            "start_price": round(start_price, 2),
            "end_price": round(end_price, 2),
            "total_return_pct": round(total_return, 2),
            "annualized_volatility_pct": round(annualized_volatility, 2),
            "average_daily_volume": int(avg_volume),
            "data_points": len(hist),
        }

        return {"status": "success", "data": summary}

    except Exception as e:
        return {"status": "error", "message": str(e)}


def get_financial_statements(ticker: str) -> Dict[str, Any]:
    """Get simplified income statement and balance sheet data."""
    try:
        stock = yf.Ticker(ticker)

        # Income statement (annual)
        income_stmt = stock.income_stmt
        balance_sheet = stock.balance_sheet

        result = {}

        if not income_stmt.empty:
            latest_year = income_stmt.columns[0]
            result["income_latest_year"] = str(latest_year.date())
            result["total_revenue"] = _safe_float(income_stmt, "Total Revenue", latest_year)
            result["gross_profit"] = _safe_float(income_stmt, "Gross Profit", latest_year)
            result["operating_income"] = _safe_float(income_stmt, "Operating Income", latest_year)
            result["net_income"] = _safe_float(income_stmt, "Net Income", latest_year)

        if not balance_sheet.empty:
            latest_quarter = balance_sheet.columns[0]
            result["total_assets"] = _safe_float(balance_sheet, "Total Assets", latest_quarter)
            result["total_debt"] = _safe_float(balance_sheet, "Total Debt", latest_quarter)
            result["cash"] = _safe_float(balance_sheet, "Cash And Cash Equivalents", latest_quarter)

        return {"status": "success", "data": result}

    except Exception as e:
        return {"status": "error", "message": str(e)}


def _safe_float(df, row_name: str, col) -> Optional[float]:
    try:
        val = df.loc[row_name, col]
        return float(val) if pd.notna(val) else None
    except (KeyError, TypeError):
        return None

Building the News Fetcher

# tools/news_fetcher.py
import yfinance as yf
from typing import List, Dict


def get_stock_news(ticker: str, max_articles: int = 10) -> Dict:
    """Fetch recent news headlines for a stock ticker."""
    try:
        stock = yf.Ticker(ticker)
        news = stock.news

        articles = []
        for item in (news or [])[:max_articles]:
            articles.append({
                "title": item.get("title", ""),
                "publisher": item.get("publisher", ""),
                "link": item.get("link", ""),
                "published": item.get("providerPublishTime", 0),
                "type": item.get("type", ""),
                "summary": item.get("summary", "")[:300] if item.get("summary") else "",
            })

        return {
            "status": "success",
            "ticker": ticker.upper(),
            "article_count": len(articles),
            "articles": articles,
        }

    except Exception as e:
        return {"status": "error", "message": str(e)}

Configuring the AutoGen Analyst Agent

Now build the actual AutoGen agent that uses these tools:

# main.py
import autogen
import json
import os
from datetime import datetime
from tools.yahoo_finance import get_stock_fundamentals, get_price_history, get_financial_statements
from tools.news_fetcher import get_stock_news


def analyze_stock(ticker: str, output_dir: str = "output") -> str:
    os.makedirs(output_dir, exist_ok=True)

    llm_config = {
        "config_list": [
            {
                "model": "gpt-4o",
                "api_key": os.getenv("OPENAI_API_KEY"),
            }
        ],
        "temperature": 0.1,
        "functions": [
            {
                "name": "get_fundamentals",
                "description": "Fetch fundamental financial data for a stock ticker",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "ticker": {"type": "string", "description": "Stock ticker symbol"}
                    },
                    "required": ["ticker"]
                }
            },
            {
                "name": "get_price_history",
                "description": "Get 1-year price history and return metrics",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "ticker": {"type": "string"}
                    },
                    "required": ["ticker"]
                }
            },
            {
                "name": "get_news",
                "description": "Get recent news articles about the stock",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "ticker": {"type": "string"},
                        "max_articles": {"type": "integer", "default": 8}
                    },
                    "required": ["ticker"]
                }
            },
            {
                "name": "get_financials",
                "description": "Get income statement and balance sheet data",
                "parameters": {
                    "type": "object",
                    "properties": {
                        "ticker": {"type": "string"}
                    },
                    "required": ["ticker"]
                }
            },
        ]
    }

    analyst = autogen.AssistantAgent(
        name="StockAnalyst",
        system_message="""You are a financial research analyst. Your job is to produce
structured research reports based on publicly available financial data.

IMPORTANT DISCLAIMER: You must always include a prominent disclaimer that your analysis
is for informational purposes only and does not constitute investment advice.

When analyzing a stock, follow this process:
1. Fetch fundamental data using get_fundamentals
2. Get 1-year price history using get_price_history
3. Get financial statements using get_financials
4. Fetch recent news using get_news
5. Synthesize all data into a structured markdown report

Your report must include:
- Company overview
- Key metrics table (P/E, P/B, debt ratios, margins)
- 1-year price performance summary
- Recent news summary (3-5 key themes)
- Basic valuation context (compare P/E to sector average if known)
- Risks and considerations
- Disclaimer

End your response with ANALYSIS_COMPLETE when done.""",
        llm_config=llm_config,
    )

    # Function map for the user proxy
    function_map = {
        "get_fundamentals": lambda ticker: get_stock_fundamentals(ticker),
        "get_price_history": lambda ticker: get_price_history(ticker),
        "get_news": lambda ticker, max_articles=8: get_stock_news(ticker, max_articles),
        "get_financials": lambda ticker: get_financial_statements(ticker),
    }

    user_proxy = autogen.UserProxyAgent(
        name="ResearchCoordinator",
        human_input_mode="NEVER",
        max_consecutive_auto_reply=15,
        is_termination_msg=lambda msg: "ANALYSIS_COMPLETE" in (msg.get("content") or ""),
        function_map=function_map,
        code_execution_config=False,
    )

    user_proxy.initiate_chat(
        analyst,
        message=f"Please conduct a comprehensive research analysis of {ticker.upper()} stock. Use all available tools to gather data and produce a structured report.",
    )

    # Extract the final report
    messages = analyst.chat_messages.get(user_proxy, [])
    final_report = ""
    for msg in reversed(messages):
        if msg.get("role") == "assistant" and msg.get("content"):
            final_report = msg["content"].replace("ANALYSIS_COMPLETE", "").strip()
            break

    # Save the report
    timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
    output_file = os.path.join(output_dir, f"{ticker.upper()}_analysis_{timestamp}.md")

    with open(output_file, "w", encoding="utf-8") as f:
        f.write(f"# Stock Analysis: {ticker.upper()}\n")
        f.write(f"*Generated: {datetime.now().strftime('%Y-%m-%d %H:%M UTC')}*\n\n")
        f.write("---\n\n")
        f.write(final_report)

    print(f"Analysis saved to: {output_file}")
    return output_file


if __name__ == "__main__":
    import sys
    ticker = sys.argv[1] if len(sys.argv) > 1 else "AAPL"
    analyze_stock(ticker)

Understanding the Output Format

A well-configured run produces output like this:

# Stock Analysis: AAPL
*Generated: 2026-05-31 14:30 UTC*

---

## Company Overview
Apple Inc. (AAPL) is a consumer electronics and software company...

## Key Metrics

| Metric | Value | Context |
|--------|-------|---------|
| P/E Ratio (TTM) | 28.4 | Above tech sector avg (~25) |
| Forward P/E | 25.1 | More reasonable on growth basis |
| P/B Ratio | 42.1 | Premium brand valuation |
| Debt/Equity | 1.73 | Moderate leverage |
| Profit Margin | 25.3% | Strong operating efficiency |
| Revenue Growth (YoY) | 4.2% | Moderate growth |
| Analyst Target | $198.50 | Based on 42 analyst estimates |

## 1-Year Price Performance
- Start: $165.23 → Current: $182.41
- Total Return: +10.4%
- Annualized Volatility: 22.1%

## Recent News Themes
1. **AI Integration:** Multiple reports on Apple Intelligence rollout...
2. **Services Growth:** App Store revenue up in Q1 2026...

## ⚠️ DISCLAIMER
This report is generated by an AI system for informational purposes only.
It does not constitute investment advice...

Limitations You Need to Know

Limitation	Impact	Mitigation
15-minute data delay	Price data not real-time	Note timestamp in output
LLM hallucination risk	Agent may misstate ratios	Cross-check key figures manually
No analyst consensus detail	Shallow competitive analysis	Supplement with earnings call transcripts
No sentiment scoring	News summary is qualitative	Add NLP sentiment layer manually
No historical comparison	Can't benchmark against sector	Add sector ETF comparison
Token context limits	Long news articles get truncated	Summarize before passing to agent

This type of structured research automation is discussed extensively in the AI research agent build guide, which covers more advanced patterns for handling data quality and fact verification.

Running the Agent

export OPENAI_API_KEY=sk-...
python main.py MSFT

Expected runtime: 2-4 minutes per ticker, depending on data availability and LLM response speed. Token usage typically runs 3,000-6,000 tokens per analysis using GPT-4o.

For batch analysis across a watchlist:

# Create a watchlist file
echo "AAPL\nMSFT\nGOOGL\nAMZN" > watchlist.txt

# Run analysis on each
while read ticker; do
  python main.py $ticker
  sleep 10  # Rate limit courtesy delay
done < watchlist.txt

Build a Stock Analysis Agent with AutoGPT (Fundamentals + News)

What the Agent Will Do

Setting Up the Environment

Building the Yahoo Finance Tool

Building the News Fetcher

Configuring the AutoGen Analyst Agent

Understanding the Output Format

Limitations You Need to Know

Running the Agent

Frequently Asked Questions

Frequently Asked Questions

AiTechWorlds Team

Related Articles

10 AutoGPT Command Line Arguments (Continuous Mode, Speak)

10 AutoGPT Configuration Tweaks for Better Performance

Build a Content Research Agent with AutoGPT (Trends, Outlines)

Build a Data Analysis Agent with AutoGPT (CSV, SQL, Plots)

Get Free AI Notes Daily

Build a Stock Analysis Agent with AutoGPT (Fundamentals + News)

What the Agent Will Do

Setting Up the Environment

Building the Yahoo Finance Tool

Building the News Fetcher

Configuring the AutoGen Analyst Agent

Understanding the Output Format

Limitations You Need to Know

Running the Agent

Frequently Asked Questions

Frequently Asked Questions

AiTechWorlds Team

Related Articles

10 AutoGPT Command Line Arguments (Continuous Mode, Speak)

10 AutoGPT Configuration Tweaks for Better Performance

Build a Content Research Agent with AutoGPT (Trends, Outlines)

Build a Data Analysis Agent with AutoGPT (CSV, SQL, Plots)

Get Free AI Notes Daily