MCP Server: Security Performance Report

What It Does

A portfolio reporting tool using dual calculation methodology, combining custom performance calculations with FFN library analytics. FastAPI endpoints, Model Context Protocol (MCP) integration, and a clean web interface:

Dual methodology: custom calculations for core metrics + FFN library for comprehensive analytics
Core performance metrics (Total Return, CAGR, Sharpe, Sortino) use custom implementations based on QuantStats methodology
Additional analytics (drawdowns, monthly returns, statistical analysis) powered by FFN
Selective validation against QuantStats for key performance metrics
Professional HTML reports with TIGZIG AI branding and matplotlib visualizations
Exports multiple CSV files with detailed data for further analysis

Methodology Detail

Full methodology, validation results and calculation details: SPR vs QuantStats Comparison.

How to Use

Endpoints

Base URL: https://ffn.tigzig.com/ MCP Endpoint: https://ffn.tigzig.com/mcp API Endpoint: https://ffn.tigzig.com/analyze

Feel free to try out these endpoints, subject to rate limits and server load.

Access Methods

Web Interface: visit / for the clean web form
Direct API: make HTTP POST requests to /analyze with required parameters:
- symbols - comma-separated Yahoo Finance symbols (e.g. AAPL,MSFT,GOOG)
- start_date - analysis start date (YYYY-MM-DD)
- end_date - analysis end date (YYYY-MM-DD)
- risk_free_rate - optional risk-free rate percentage (default: 0.0)
MCP/LLM: connect to /mcp using any MCP-compatible client

How It Works

Dual Calculation Methodology

Core Performance Metrics: custom implementations based on QuantStats methodology for Total Return, CAGR, Sharpe, and Sortino ratios. Additional Analytics: FFN library for drawdown analysis, monthly returns, and statistical metrics. Historical price data sourced from Yahoo Finance with preprocessing applied including removal of zero/NaN values and date alignment.

Validation & Accuracy

Selective validation against QuantStats library shows: perfect matches for Total Return (100%) and CAGR (100%), near-perfect matches for Sharpe (97%+) and Sortino (97%+) ratios. The 3% variance is attributed to data quality filters and precision differences.

Integration & Output

FastAPI backend with efficient dual-methodology processing
Professional HTML reports with matplotlib charts and TIGZIG AI branding
6 different CSV files: price data, returns, correlations, and statistics
MCP protocol integration for AI/LLM interactions

How to Replicate

Clone the repository: shared-fastapi-mcp-ffn (now SPR - Security Performance Report)
Create virtual environment and install dependencies
Configure environment variables for development/production
Run the FastAPI server with uvicorn

Key Dependencies

Custom calculation engine - core performance metrics
FFN library - additional analytics and drawdown analysis
FastAPI - high-performance web framework
yfinance - Yahoo Finance data integration
fastapi-mcp - Model Context Protocol integration

Detailed methodology, calculations, and validation results are documented in the GitHub repository README and the SPR vs QuantStats methodology comparison.

Resources

MCP Server: SPR Portfolio

FastAPI + MCP with dual methodology (custom calculations + FFN)

SPR vs QuantStats Methodology

Detailed comparison, validation results, methodology documentation

API Docs: SPR Server

Interactive Swagger UI for the SPR FastAPI server

FFN Library

Open-source Python library used for additional analytics (drawdowns, monthly returns)

FastAPI-MCP

Tadata's package to convert any FastAPI into an MCP server

Fine Print: This is for informational purposes only and should not be considered as investment advice. Dual methodology: core performance metrics (Total Return, CAGR, Sharpe, Sortino) calculated using custom implementations based on QuantStats methodology, while additional analytics are powered by the open-source FFN library. Selective validation against QuantStats shows 97%+ accuracy for key metrics, with perfect matches for Total Return and CAGR. Input data sourced from Yahoo Finance with preprocessing applied. For multi-security analysis, date mismatches due to differing exchange calendars are forward-filled for up to five days - an industry-accepted practice with minimal impact on results. For complete methodology, validation results, and detailed explanations of any variations, see the SPR vs QuantStats methodology comparison and the linked GitHub repository. Always validate outputs.