Portfolio Management

portfolio-management

Purpose

This skill enables the AI to manage investment portfolios using quantitative models, calculate risk metrics like Value at Risk (VaR), and apply optimization algorithms such as mean-variance optimization. It processes portfolio data to generate actionable insights, supporting decisions in finance by integrating with data sources and executing trades based on predefined strategies.

When to Use

Use this skill for tasks involving portfolio rebalancing, risk assessment, or performance analysis in financial contexts. Apply it when handling user queries about investment strategies, such as diversifying assets or responding to market volatility. Ideal for scenarios with real-time data feeds or when optimizing allocations under constraints like budget limits.

Key Capabilities

• Quantitative Models: Implement models like CAPM or Black-Litterman; e.g., calculate expected returns with claw portfolio model capm --assets AAPL,GOOG.
• Risk Metrics: Compute VaR or Sharpe ratio; use API endpoint GET /​api/​portfolios/​risk/​var?confidence=0.95 to get 95% VaR for a portfolio.
• Optimization Algorithms: Run mean-variance optimization; configure via JSON file: {"assets": ["AAPL", "MSFT"], "weights": [0.5, 0.5]}.
• Data Integration: Pull market data from external APIs; supports formats like CSV or JSON for portfolio inputs.
• Performance Tracking: Generate reports on portfolio returns; e.g., claw portfolio track --period monthly --metric sharpe.

Usage Patterns

Always initialize with authentication via $PORTFOLIO_API_KEY environment variable. For CLI usage, pipe data inputs directly; e.g., start with claw portfolio load --file portfolio.json then chain commands like claw portfolio optimize --risk-level high. In API patterns, use POST requests for modifications and GET for queries; handle asynchronous operations by polling endpoints. For scripts, wrap in try-catch blocks to manage API failures, and use config files for reusable parameters like asset lists.

Common Commands/​API

• CLI Commands: Use claw portfolio manage --action optimize --config config.json to optimize a portfolio; flags include --action (optimize, analyze), --config (path to JSON), and --verbose for detailed logs.
• API Endpoints: Send requests to POST /​api/​portfolios/​create with body {"name": "my-portfolio", "assets": ["AAPL", "TSLA"]}; authenticate via header Authorization: Bearer $PORTFOLIO_API_KEY.
• Code Snippets:

  import requests
  response = requests.post('https://api.openclaw.ai/​api/​portfolios/​optimize', headers={'Authorization': f'Bearer {os.environ["PORTFOLIO_API_KEY"]}'}, json={'assets': ['AAPL', 'GOOG']})
  print(response.json()['optimized_weights'])
  

  claw portfolio analyze --assets AAPL,MSFT --metric var --confidence 0.99
  

• Config Formats: Use JSON for inputs, e.g., {"portfolio": {"assets": [{"symbol": "AAPL", "quantity": 100}], "constraints": {"max_risk": 0.05}}}; validate with claw portfolio validate --file config.json.

Integration Notes

Integrate by setting $PORTFOLIO_API_KEY in your environment before running commands. For external systems, use webhooks to sync data; e.g., connect to a brokerage API by mapping endpoints like POST /​api/​portfolios/​update to trigger updates. Ensure compatibility with financial libraries like NumPy for calculations; import as a module in Python scripts and handle rate limits by adding delays, e.g., time.sleep(1) between API calls. Test integrations in a sandbox environment using mock data.

Error Handling

Check for authentication errors by verifying $PORTFOLIO_API_KEY is set; if missing, prompt user with os.environ.get('PORTFOLIO_API_KEY') or raise ValueError("API key required"). For API failures, catch HTTP errors like 401 or 429 using try-except in code:

try:
    response = requests.get('https://api.openclaw.ai/​api/​portfolios/​risk')
except requests.exceptions.HTTPError as e:
    print(f"Error: {e.response.status_code} - {e.response.text}")

Handle invalid inputs by validating configs first with claw portfolio validate; log errors to file with --log-file errors.log flag, and retry transient errors up to 3 times with exponential backoff.

Graph Relationships

• Related to cluster: financial
• Connected via tags: finance, investments, risk-management, quant-analysis
• Links to other skills: depends on data-analysis for data processing; enhances trading-execution for automated trades