Finance Toolkit

Finance Toolkit is Jeroen Bouma's open-source Python library for financial analysis, and this is its MCP server. Instead of leaving the arithmetic to the model, your agent calls the Finance Toolkit's 500+ documented methods: ratios, valuation models, risk and performance metrics, technical indicators and macroeconomic series, calculated from financial statements and market data. Every formula is in the documentation, so a number from a chat can be traced back to how it was calculated. What you can do with it - Compare companies' profitability, efficiency, liquidity and solvency: margins, ROE, ROIC, asset turnover, debt-to-equity and interest coverage - Calculate valuation ratios and models: P/E, EV/EBITDA, free cash flow yield, DCF and intrinsic value, WACC, DuPont, Altman Z-Score and Piotroski F-Score - Measure performance and risk: Sharpe and Sortino ratios, alpha and beta, CAPM, Fama-French factors, Value at Risk, CVaR, GARCH and maximum drawdown - Check technical indicators such as RSI, MACD, moving averages and Bollinger Bands - Pull macroeconomic and rates data by country: GDP, inflation, unemployment, government debt, central bank rates and the US Treasury yield curve - Run econometrics on prices, returns or any indicator: regressions, cointegration and causality tests, ARIMA and VAR forecasts, event studies - Work on portfolios, ETFs, options and bonds: optimal weights, the efficient frontier, ETF holdings and overlap, Black-Scholes pricing and Greeks, duration and yield to maturity - Explore the market: a stock screener, sector performance, earnings and IPO calendars, 8-K filings and insider trades How it connects - Hosted: add https://financetoolkit.jeroenbouma.com/mcp to your agent. The first time it's used, an OAuth page asks for your Financial Modeling Prep (FMP) API key. The maker says the key is sealed into a signed token your agent holds and is never written to disk on the server. - Local: run uvx --from "financetoolkit[mcp]" financetoolkit-mcp-setup, which writes the config for your agent, including your key. Or run uvx --from "financetoolkit[mcp]" financetoolkit-mcp yourself with FINANCIAL_MODELING_PREP_API_KEY set. An optional FRED_API_KEY adds a few US-only economic indicators, and Claude Desktop users can install the MCPB bundle instead. - It has 22 categorical tools, each taking an indicator that picks the exact metric, plus three search tools for finding one. You ask in plain English and your agent picks the tool. Good to know - You need an FMP API key. The maker says FMP's free plan is enough to try it, and paid FMP plans add full history, all exchanges and higher request limits. The server and the library are free and MIT licensed. - Company data comes from Financial Modeling Prep, subject to your FMP plan, and the library falls back to Yahoo Finance when FMP can't supply it. Macroeconomic data comes from the OECD and, optionally, FRED. - Every tool reads and calculates; none of them changes anything. - It's an independent project: Jeroen Bouma isn't part of Financial Modeling Prep, so questions about the data go to FMP.

• “Which is more profitable, Apple or Microsoft, and does that change when you look at return on capital?”

  1. profitability
  Copy

• “What growth is priced into Nvidia's share price, based on a discounted cash flow?”

  1. valuation
  2. models
  Copy

• “Compare the Value at Risk and maximum drawdown of Berkshire Hathaway, Visa and Costco since 2020.”