What is the difference between mean-variance optimization and risk parity?

Mean-variance looks for the weights with the best trade-off between expected return and volatility, so it depends heavily on return estimates. Risk parity ignores expected returns and sets weights so that each holding contributes the same share of total risk.

Is there free portfolio optimization software?

Yes. Riskfolio-Lib and PyPortfolioOpt are free open-source Python libraries. Among web apps, MEDGE includes optimization on 12 objectives in its free plan, and PortfoliosLab and testfolio have free tiers with limits.

How is hierarchical risk parity (HRP) different from mean-variance?

HRP groups assets by how they move together and spreads risk across the groups, without inverting the covariance matrix. It tends to be more stable than mean-variance when estimates are noisy, but it does not target a return.

Guide · checked November 2026

Best portfolio optimization software in 2026

Portfolio optimization software calculates the weights that minimise or maximise an objective you choose, such as volatility, CVaR or the Sharpe ratio, within the limits you set. The options range from web apps with a point-and-click optimizer to open-source Python libraries that implement mean-variance, risk parity, hierarchical risk parity (HRP) and Black-Litterman.

Published by MEDGE Capital, which is one of the tools listed. Details of the other tools come from their public pages, checked in November 2026, with a link to each source below. Prices are list prices for one user, before tax; check with the vendor before you buy. The order is not a ranking.

Tool	Optimization	Price
MEDGE Capital	12 objectives, including Max Sharpe, Max Sortino, Min CVaR and Risk Parity. The optimized mix goes straight into the backtest with rebalancing and costs, the bear, neutral and bull regimes and, from Essential, the Stress Lab.	Optimization on every plan, Free included. Essential €14.99, Plus €39.99, Pro €59.99, Advisor €159 a month; 7-day Plus trial
PortfoliosLab	Mean-variance with eight objectives (including minimum volatility, CVaR or drawdown and maximum Sharpe, Sortino or Omega), plus risk parity, HRP, HERC and maximum diversification; new weights are backtested walk-forward.	Free (10 calculations a month); Plus $10, Pro $25, Max $90 a month, or $100, $250, $900 a year
Portfolio Visualizer	Optimization objectives including mean-variance, CVaR, risk parity, maximum drawdown, Sortino and Omega, with optional weight limits; a Black-Litterman option; rolling optimization.	Free plan (up to 15 assets, limited history); Basic $30 a month and Pro $55 a month, billed annually
testfolio	Four optimizers: a portfolio optimizer from expected returns, volatilities and correlations, a backtest optimizer, an efficient frontier and a Monte Carlo optimizer; advanced objectives on paid plans.	Free; Pro $12.50, Pro+ $25 and Max $80 a month, billed annually
Riskfolio-Lib	Mean-risk optimization with many risk measures including CVaR, risk parity, HRP and HERC, and three Black-Litterman variants.	Free, BSD 3-Clause licence
PyPortfolioOpt	Mean-variance efficient frontier, semivariance, CVaR and CDaR frontiers, a Black-Litterman model and HRP.	Free, MIT licence

MEDGE CapitalOur tool

Portfolio analytics for self-directed investors and independent advisors, on US and European listings.

Optimization: 12 objectives, including Max Sharpe, Max Sortino, Min CVaR and Risk Parity. The optimized mix goes straight into the backtest with rebalancing and costs, the bear, neutral and bull regimes and, from Essential, the Stress Lab.
Price: Optimization on every plan, Free included. Essential €14.99, Plus €39.99, Pro €59.99, Advisor €159 a month; 7-day Plus trial
Fits if: you want to optimize and then check the result on history, in regimes and under stress, without writing code.

How the backtesting lab works

PortfoliosLab

Web app for portfolio analysis, backtests and optimization.

Optimization: Mean-variance with eight objectives (including minimum volatility, CVaR or drawdown and maximum Sharpe, Sortino or Omega), plus risk parity, HRP, HERC and maximum diversification; new weights are backtested walk-forward.
Price: Free (10 calculations a month); Plus $10, Pro $25, Max $90 a month, or $100, $250, $900 a year
Fits if: you want many optimization methods, including cluster-based ones, in a web app.

Source: Portfolio optimization · Subscription

Portfolio Visualizer

Web-based backtesting and portfolio analysis tools.

Optimization: Optimization objectives including mean-variance, CVaR, risk parity, maximum drawdown, Sortino and Omega, with optional weight limits; a Black-Litterman option; rolling optimization.
Price: Free plan (up to 15 assets, limited history); Basic $30 a month and Pro $55 a month, billed annually
Fits if: you want a range of optimization objectives on US asset classes and funds.

Source: Optimize portfolio · Pricing

testfolio

Web app for backtesting and portfolio analysis.

Optimization: Four optimizers: a portfolio optimizer from expected returns, volatilities and correlations, a backtest optimizer, an efficient frontier and a Monte Carlo optimizer; advanced objectives on paid plans.
Price: Free; Pro $12.50, Pro+ $25 and Max $80 a month, billed annually
Fits if: you want to optimize on long backtested histories.

Source: Portfolio optimization · Pricing

Riskfolio-Lib

Open-source Python library for portfolio optimization.

Optimization: Mean-risk optimization with many risk measures including CVaR, risk parity, HRP and HERC, and three Black-Litterman variants.
Price: Free, BSD 3-Clause licence
Fits if: you write Python and want full control over the model and its constraints.

Source: Documentation

PyPortfolioOpt

Open-source Python library for portfolio optimization.

Optimization: Mean-variance efficient frontier, semivariance, CVaR and CDaR frontiers, a Black-Litterman model and HRP.
Price: Free, MIT licence
Fits if: you want a widely used Python library with clear documentation.

Source: Documentation

Frequently asked questions

What is the difference between mean-variance optimization and risk parity?: Mean-variance looks for the weights with the best trade-off between expected return and volatility, so it depends heavily on return estimates. Risk parity ignores expected returns and sets weights so that each holding contributes the same share of total risk.
Is there free portfolio optimization software?: Yes. Riskfolio-Lib and PyPortfolioOpt are free open-source Python libraries. Among web apps, MEDGE includes optimization on 12 objectives in its free plan, and PortfoliosLab and testfolio have free tiers with limits.
How is hierarchical risk parity (HRP) different from mean-variance?: HRP groups assets by how they move together and spreads risk across the groups, without inverting the covariance matrix. It tends to be more stable than mean-variance when estimates are noisy, but it does not target a return.

More comparisons

Optimize a portfolio, then test it on history and under stress.

Free plan with no time limit, no brokerage connection. Not investment advice.

Start free