Quantitative Portfolio Optimization · MIT Licensed

Given any set of stocks,find the allocationthat maximises returnper unit of risk.Then prove it holds up.

A research tool for people who think seriously about portfolio construction.

Launch app →View on GitHub ↗

6 algorithms  ·  Rust + WASM  ·  Runs in your browser

The frontier, interactive.

Drag along the curve to find any portfolio on the efficient set. Weights, return, volatility, and Sharpe ratio update in real time at sub-15ms. Overlay HRP, Risk Parity, CVaR, or Robust solutions on the same chart to compare approaches.

VOLATILITY →RETURN ↑TLTGLDSPYIWMQQQMSFTAAPLMax Sharpe
Markowitz MVOThe starting point. Traces the full efficient frontier via mean-variance optimisation.
Hierarchical Risk ParityWhen you have many assets or short history. Clusters the correlation matrix instead of inverting it.
Equal Risk ContributionDefensive allocation across mixed asset classes. Each asset contributes equally to variance.
CVaR MinimisationWhen tail risk matters more than average risk. Minimises expected loss in the worst 5%.
Robust MVOConservative and real-world oriented. Holds up even when your return estimates are wrong.
Black-LittermanWhen you have conviction but want the math to keep you honest. Blends equilibrium with your subjective views.

Built for speed.

<15ms

per efficient frontier, 6 assets, 5yr daily data

<5ms

Monte Carlo, 500 paths

~100ms

Walk-forward backtest, 5yr

60fps

Frontier drag interaction