mc-portfolio-simulatorlisted

# mc-portfolio-simulator You hand over a book (weights per ticker) and a horizon. The skill fits the covariance matrix on the historical window, simulates N correlated return trajectories forward, and reports the distribution of outcomes. Companion to `position-sizer` and to `risk-report --mc`. Same math underneath: shrunk correlation × per-name vols → covariance → Cholesky-factored path simulation. The three tools differ in framing: - `position-sizer` produces target weights under a target vol. - `risk-report` includes MC as one lens alongside historical VaR, drawdown, stress days. - `mc-portfolio-simulator` is the standalone MC lens: you already have weights, you want the P&L distribution. ## When to invoke - "Given my proposed weights, what's the 5th percentile 60-day outcome?" - Comparing two candidate books by tail severity - Answering "how bad can this get" for a small book without needing full risk-report output - The user says "monte carlo my book", "simulate this portfolio", "P(loss > 10%)", "forward P&L distribution" Not for: predicting the direction (MC doesn't pick winners; it fans the future out). Not for options portfolios (payoffs are non-linear; this simulates linear returns). ## What you need - A book: `--positions T=w,T=w,...` - `MASSIVE_API_KEY` exported - Stocks Basic plan minimum Optional: - `--simulation-days` (default 60): forward horizon in trading days. - `--n-paths` (default 10000): Monte Carlo path count. - `--tail {normal, stude