USStockToken: AI Multi-Factor Stock Analysis on Telegram

#python #stocks #telegram #ai

Multi-Factor Stock Analysis with AI: Technical + Fundamental + Sentiment

Most "AI stock analysis" tools are just GPT wrappers that summarize news. We built something different.

The Problem

Single-factor analysis fails:

Technical only: Ignores fundamentals (Enron looked great on charts)
Fundamental only: Ignores momentum (value traps)
Sentiment only: Easily manipulated (meme stocks)

The answer? Ensemble analysis combining all three.

Our Three-Factor Model

class MultiFactorAnalyzer:
    def analyze(self, ticker):
        # Factor1: Technical
        technical = self.technical_analysis(ticker)
        # RSI, MACD, Bollinger, Volume Profile

        # Factor2: Fundamental  
        fundamental = self.fundamental_analysis(ticker)
        # P/E, P/B, Revenue Growth, Margins, Debt

        # Factor3: Sentiment
        sentiment = self.sentiment_analysis(ticker)
        # News NLP, Social media, Options flow

        # Ensemble signal
        signal = self.combine_factors(technical, fundamental, sentiment)
        return signal

Technical Layer

RSI, MACD, Bollinger Bands
Volume profile analysis
Support/resistance levels
Pattern recognition (head & shoulders, flags)

Fundamental Layer

P/E ratio vs sector average
Revenue growth trajectory
Margin trends
Debt-to-equity ratio
Free cash flow

Sentiment Layer

News article NLP (FinBERT)
Social media sentiment
Options flow analysis
Institutional ownership changes

Real Example: AAPL

Technical: RSI58 (neutral), MACD bullish crossover
Fundamental: P/E28x (premium but justified), Revenue +8% YoY
Sentiment:73% bullish news, strong institutional buying

Ensemble: BUY (confidence:72%)

Results (Backtested)

Period:2023-01 to2026-07
Strategy: Long-only, rebalance weekly
Return:+47% (vs S&P500 +32%)
Sharpe:1.42
Max Drawdown:-12%

Try It

Free Telegram bot: @USStockToken

GitHub: [Coming Soon]

What factors do you think matter most for stock analysis?

DEV Community