FolioForecast MCP Server

Portfolio optimization tools for AI agents via the Model Context Protocol.

Connects AI assistants (Claude, GPT, etc.) to FolioForecast's portfolio optimization engine — 20 optimization methods across 165,000+ tickers.

Tools

| Tool | Description | Cost | |------|-------------|------| | optimize_portfolio | Run mean-variance optimization with 20 methods (Sharpe, min-vol, risk parity, Kelly, CVaR, etc.) | 1-2 units | | search_tickers | Search 165,000+ ticker symbols by name or symbol | Free | | stress_test | Crisis scenario analysis (2008 crash, COVID, custom shocks) | 1 unit/scenario | | get_asset_metadata | Fundamental data, sector classification, market cap | 1 unit |

Quick Start

1. Get an API Key

Sign up at folioforecast.com/developers and create an API key.

2. Install

npm install -g @folioforecast/mcp-server

3. Configure Claude Desktop

Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "folioforecast": {
      "command": "folioforecast-mcp",
      "env": {
        "FOLIOFORECAST_API_KEY": "ff_live_your_key_here"
      }
    }
  }
}

Or run from source:

{
  "mcpServers": {
    "folioforecast": {
      "command": "node",
      "args": ["path/to/mcp-server/dist/index.js"],
      "env": {
        "FOLIOFORECAST_API_KEY": "ff_live_your_key_here"
      }
    }
  }
}

Environment Variables

| Variable | Required | Description | |----------|----------|-------------| | FOLIOFORECAST_API_KEY | Yes | Your API key (starts with ff_live_) | | FOLIOFORECAST_API_URL | No | Custom API base URL (default: https://www.folioforecast.com/api/v1) |

Example Prompts

Once configured, try asking Claude:

• "Optimize a portfolio of AAPL, GOOGL, MSFT, and AGG for maximum Sharpe ratio"
• "What's the minimum volatility portfolio using SPY, TLT, GLD, and VNQ?"
• "How would a 60/40 portfolio perform during the 2008 financial crisis?"
• "Search for gold ETFs and find the best risk-parity allocation"
• "Compare my portfolio (50% AAPL, 30% GOOGL, 20% AMZN) against the optimal allocation"

Optimization Methods

Basic (1 compute unit)

• max_sharpe — Maximize risk-adjusted returns
• min_volatility — Minimize portfolio variance
• risk_parity — Equal risk contribution
• equal_weight — Simple 1/N allocation

Target-Based (1 compute unit)

• min_vol_target_return — Minimum risk for a target return
• max_return_target_vol — Maximum return for a risk budget

Advanced (2 compute units)

• min_cvar — Minimize tail risk (Conditional VaR)
• max_kelly — Kelly criterion (geometric growth)
• min_tracking_error — Track a benchmark closely
• max_information_ratio — Best risk-adjusted excess return
• max_sortino_target_return — Downside-risk adjusted
• max_omega_target_return — Omega ratio optimization
• min_drawdown_target_return — Minimize maximum drawdown

Robust (2 compute units)

• robust_max_sharpe — Sharpe, robust to estimation error
• robust_min_volatility — Min-var, robust to estimation error
• robust_min_vol_target_return — Robust target-return
• robust_max_return_target_vol — Robust target-vol

Development

cd mcp-server
npm install
npm run build
npm start

License

MIT