Knowledge Tree
mean-variance optimization
Facts (10)
Sources
Wealthfront Classic Portfolio Investment Methodology White Paper research.wealthfront.com Mar 9, 2026 9 facts
claimMean-Variance Optimization produces a collection of portfolios that form the efficient frontier, aiming to generate the maximum return for a given level of risk or minimize risk for a given expected return.
procedureWealthfront determines optimal portfolios by using mean-variance optimization with inputs including the variance-covariance matrix of asset class returns and net-of-fee, after-tax expected returns for each asset class.
claimThe Black-Litterman model derives expected return parameters from market equilibrium allocations and manager views on the expected return of assets, which mitigates the sensitivity problem of mean-variance optimization and enables the production of diversified and intuitive portfolios.
procedureWealthfront uses a mean-variance optimization approach to create California-specific portfolios, accounting for the specific after-tax and after-fee expected returns, volatility, and correlations of California municipal bonds with other asset classes.
referenceFischer Black and Robert Litterman proposed the Black-Litterman model in 1992 while working at Goldman Sachs to overcome the difficulty of applying mean-variance optimization in practice.
procedureWealthfront uses net-of-fee, after-tax rates of return as inputs to mean-variance optimization to determine the efficient frontier.
claimMean-variance optimization requires estimates of each asset class’s expected return, volatility, and pairwise correlations as inputs, but the method is sensitive to these parameters and tends to produce concentrated and unintuitive portfolios if the parameters are naively specified.
referenceWealthfront determines the optimal mix of asset classes using Mean-Variance Optimization, a method introduced by Harry Markowitz in 1952 that serves as the foundation of Modern Portfolio Theory.
claimThe constraints in the mean-variance optimization problem, which include lower and upper bounds on individual asset weights and constraints on pairs of assets, ensure that resulting portfolios are long-only, fully invested, do not use leverage, and are not overly concentrated in a small number of asset classes.
Risk Factors, Expected Returns, and Investment Instruments analystprep.com Aug 5, 2024 1 fact
claimMean-variance optimization, which relies on standard deviation, is often unsuitable for alternative investments because these assets frequently suffer from infrequent valuation, erratic correlations, and extended lock-up periods.