Portfolio Analysis

A financial portfolio analysis tool that uses AI agents to analyze stock portfolios and provide optimization strategies for selling off holdings to avoid over-leverage, plus daily technical analysis for individual stocks.

Overview

This tool uses OpenRouter-powered AI agents to:

• Analyze portfolio holdings from CSV files
• Gather real-time financial data and market analysis for specific stocks
• Generate comprehensive position analysis and detailed optimization strategies
• Provide actionable sell-off plans over time
• Perform daily technical analysis with RSI, MACD, and plain-language insights

Why OpenRouter? This implementation uses OpenRouter to give you access to multiple AI providers (Anthropic, OpenAI, Google, Meta, and more) through a single API. You can configure and switch between different AI models without changing code. Simply update the model IDs in your .env file to experiment with different models and find the best performance for your analysis needs.

Features

Portfolio Analysis (portfolio_analysis.py)

• Portfolio Agent: Reads and analyzes CSV files containing your stock holdings
• Finance Agent: Retrieves current stock prices, company info, historical data, analyst recommendations, and company news using YFinance
• Team Collaboration: Combines both agents to provide comprehensive analysis and optimization strategies

Daily Technical Analysis (daily_analysis.py)

• Daily Analysis Agent: Performs technical analysis on individual stocks
• RSI & MACD Calculations: Analyzes 100 days of historical data to calculate RSI(14) and MACD(12,26,9)
• Buy/Hold/Sell Signals: Generates actionable recommendations based on technical indicators
• Plain-Language Insights: Provides easy-to-understand market commentary without jargon

Requirements

• Python 3.x
• OpenRouter API key (get one at https://openrouter.ai/)
• Required Python packages (see requirements.txt)

Setup

1. Install dependencies:

pip install -r requirements.txt

2. Create a .env file with the following required variables:

# OpenRouter API Configuration
OPENROUTER_API_KEY=your_openrouter_api_key_here

# Portfolio Configuration
PORTFOLIO_CSVS_LOCATION=/path/to/your/csv/files
STOCK_TICKER=NFLX

# Model Configuration (OpenRouter model IDs)
FINANCE_AGENT_MODEL=anthropic/claude-sonnet-4.5
PORTFOLIO_AGENT_MODEL=anthropic/claude-sonnet-4.5
PORTFOLIO_ANALYSIS_TEAM_MODEL=anthropic/claude-sonnet-4.5
DAILY_ANALYSIS_MODEL=anthropic/claude-sonnet-4.5

OpenRouter Model ID Examples:

• anthropic/claude-sonnet-4.5 - Claude 4.5 Sonnet (latest, most capable)
• anthropic/claude-opus-4.5 - Claude 4.5 Opus (most intelligent)
• anthropic/claude-3.5-sonnet - Claude 3.5 Sonnet
• openai/gpt-4 - OpenAI GPT-4
• openai/gpt-4-turbo - OpenAI GPT-4 Turbo
• google/gemini-pro-1.5 - Google Gemini Pro 1.5
• meta-llama/llama-3.1-70b-instruct - Meta Llama 3.1 70B

For the full list of available models, visit: https://openrouter.ai/models

You can mix and match different models for each agent based on your performance and cost requirements.

3. Prepare your portfolio data:
   • Go to each of your brokerage accounts
   • Download your full portfolio in CSV format
   • Place all CSV files in the location specified by PORTFOLIO_CSVS_LOCATION

Usage

Portfolio Analysis

Run the portfolio analysis:

python portfolio_analysis.py

The tool will:

1. Validate all required environment variables
2. Load portfolio data from CSV files
3. Fetch current market data for the specified stock ticker
4. Generate a comprehensive position analysis
5. Provide a detailed optimization strategy with specific sell-off recommendations

Daily Technical Analysis

Run the daily analysis:

python daily_analysis.py

The tool will:

1. Validate required environment variables
2. Fetch 100 days of historical price data for the specified stock ticker
3. Calculate RSI(14) and MACD(12,26,9) technical indicators
4. Generate a Buy/Hold/Sell recommendation based on:
   • Buy Signal: RSI < 30 (oversold) AND MACD > Signal (bullish crossover)
   • Sell Signal: RSI > 70 (overbought) AND MACD < Signal (bearish crossover)
   • Hold Signal: All other conditions
5. Provide plain-language insights and helpful tips

Configuration

All configuration is managed through environment variables:

Required Variables

• OPENROUTER_API_KEY: Your OpenRouter API key (get one at https://openrouter.ai/)
• PORTFOLIO_CSVS_LOCATION: Path to folder containing portfolio CSV files
• STOCK_TICKER: Stock ticker symbol to analyze (e.g., NFLX for Netflix)

Portfolio Analysis Model Variables

• FINANCE_AGENT_MODEL: OpenRouter model ID for the finance agent
• PORTFOLIO_AGENT_MODEL: OpenRouter model ID for the portfolio agent
• PORTFOLIO_ANALYSIS_TEAM_MODEL: OpenRouter model ID for the team coordinator

Daily Analysis Model Variables

• DAILY_ANALYSIS_MODEL: OpenRouter model ID for the daily analysis agent

Dependencies

• agno: AI agent framework
• yfinance: Yahoo Finance data retrieval
• python-dotenv: Environment variable management
• openai: OpenAI SDK for Python (used to connect to OpenRouter API)
• duckdb: Database library used by CSV tools