Paper: arXiv 2401.05080
Authors: Stephen Boyd, Kasper Johansson, Ronald Kahn, Philipp Schiele, Thomas Schmelzer
Abstract
More than seventy years ago Harry Markowitz formulated portfolio construction as an optimization problem that trades off expected return and risk, defined as the standard deviation of the portfolio returns. Since then the method has been extended to include many practical constraints and objective terms, such as transaction cost or leverage limits. Despite several criticisms of Markowitz’s method, for example its sensitivity to poor forecasts of the return statistics, it has become the dominant quantitative method for portfolio construction in practice. In this article we describe an extension of Markowitz’s method that addresses many practical effects and gracefully handles the uncertainty inherent in return statistics forecasting. Like Markowitz’s original formulation, the extension is also a convex optimization problem, which can be solved with high reliability and speed.
Complexity vs Empirical Score
- Math Complexity: 7.0/10
- Empirical Rigor: 6.5/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper features advanced convex optimization theory and robust formulations (high math), while providing extensive numerical experiments and back-testing frameworks (high rigor).
Research Flowchart
flowchart TD A["Research Goal: Extend Markowitz portfolio construction<br>to handle practical constraints & statistical uncertainty"] --> B["Key Methodology: Convex Optimization Framework"] B --> C["Inputs: Asset Return Forecasts<br>& Covariance Matrices"] C --> D["Computational Process:<br>Convex Optimization with Constraints"] D --> E["Key Outcomes:<br>Robust, Fast, & Practical Portfolio Solution"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes an extension of the classic Markowitz portfolio construction called Markowitz++ that adds practical constraints and handles uncertainty in return forecasts.
Source sentence
When we need to distinguish the extension of Markowitz’s portfolio construction that we recommend from the basic Markowitz method, we refer to it as Markowitz++.
Assumes 6
- Asset returns are modeled as Gaussian for the utility analysis. model
Source sentence
Specifically if we model the returns as Gaussian, and use the exponential utility U(R) = 1 −exp(−γR), then the expected utility is the risk-adjusted return, up to an additive constant
- Synthetic return predictions are used to simulate a proprietary mean prediction method. data
Source sentence
we use synthetic return predictions to simulate a proprietary mean prediction method
- Bid‑ask spread transaction costs are linear in trade size and proportional to half the spread. execution
Source sentence
(κspread)T|z|, where κspread ∈Rn is the vector of one‑half the asset bid‑ask spreads
- Short‑selling incurs an additional 5% annualized spread over the effective federal funds rate. execution
Source sentence
When shorting an asset we add a 5% annualized spread over the effective federal funds rate to approximate the shorting cost in our simulation
Show 2 more
- Trade limits are enforced as bounds on the trade vector relative to forecasted trading volumes. execution
Source sentence
Trade limits impose lower and upper bounds on trades, as zmin ≤z ≤zmax
- Cash weight is constrained between –5% and 100% to guarantee feasibility. other
Source sentence
We limit the cash weight to lie between −5% and 100% (which guarantees feasibility)
Data
- S&P 100 daily adjusted close price data market-prices public
Source sentence
We use daily adjusted close price data from 2000-01-04 to 2023-09-22. … We
- Bid‑ask spread data market-prices public
Source sentence
In addition to the price data, we use bid‑ask spread data to estimate the trading costs
- Effective federal funds rate macro public
Source sentence
the effective federal funds rate [Fed23] for short term borrowing and lending
Costs
Modeled
Bid‑ask spread cost = (κspread)T|z|; short‑selling cost = 5% annualized spread over federal funds rate; forecast spread = average of last 5 days; κshort =7.5% annualized, κborrow = federal funds rate.
Source sentence
For forecasting, we set κshort to 7.5% annualized, and κborrow to the effective federal funds rate.
Finds 4
- Basic Markowitz yields poor performance: 3.7% return, 14.5% volatility, Sharpe 0.19, leverage 9.3, drawdown 78.9%. performance
Source sentence
Basic Markowitz 3.7% 14.5% 0.19 1145.2 9.3 78.9%
- Tuned Markowitz++ further improves Sharpe to 4.65 with 41.8% return and lower turnover. performance
Source sentence
Tuned Markowitz++ 41.8% 8.8% 4.65 38.6 1.6 6.4%
- Adding simple weight limits (10% long, –5% short) raises Sharpe to 1.69 compared to basic Markowitz. performance
Source sentence
Weight‑limited 20.2% 11.5% 1.69 638.4 5.1 30.0%
- Parameter tuning converges after 45 back‑tests to specific γ values, increasing Sharpe while keeping leverage, turnover, and volatility reasonable. performance
Source sentence
This tuning process converged after 45 back‑tests to the parameter values γrisk = 4 × 10−2, γhold = 0.64, γtrade = 0.64, γlev = 5 × 10−4, γturn = 1.6 × 10−3.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 16 related papers). Coverage is partial, so this is neither replication nor refutation.
10 papers in the archive : Hopfield Networks for Asset Allocation
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.
Cited References (161)
- Leverage aversion, efficient frontiers, and the efficient region (2013)
- 60 years of portfolio optimization: Practical challenges and current trends (2014)
- Optimization methods in finance (2006)
- Robust portfolio optimization and management (2007)
- Multicriteria portfolio construction with python (2020)
- YALMIP : A toolbox for modeling and optimization in MATLAB (2004)
- Efficient portfolio selection in a large market (2016)
- Portfolio choice problems (2010)
and 153 more
- Inside the Black Box: The Simple Truth about Quantitative Trading (2009)
- Anomalous price impact and the critical nature of liquidity in financial markets (2011)
- On the Markowitz Enigma for Minimum Variance (2023)
- The dispersion bias (2022)
- Il problema dei “Pieni” (1940)
- Bruno de F inetti and mean-variance portfolio selection (2006)
- Approximating expected utility by a function of mean and variance (1979)
- Risk-return analysis: T he theory and practice of rational investing (2014)
- Theory of games and economic behavior, 2nd rev (1947)
- Optimization of conditional value-at-risk (2000)
- Here are 9 fascinating facts to know about Blackrock, the world’s largest asset manager (2021)
- MATLAB Financial Toolbox Documentation (2023) — link
- Numerical methods in finance and economics: a MATLAB-based introduction (2013)
- Portfolios L ab website (2023)
- Hudson & T hames website (2023)
- A xioma by Q ontigo (2023)
- BARRA ’s risk models (1996)
- Portfolio123 website (2023)
- skfolio (2023)
- Scikit-learn: Machine Learning in P ython (2011)
- Aladdin Platform Overview (2023)
- QuantLib: a free/open-source library for quantitative finance (2023) — link
- Portfolio Visualizer (2023)
- An open-source implementation of the critical-line algorithm for portfolio optimization (2013)
- Computation of mean-semivariance efficient sets by the critical line algorithm (1993)
- Mean-variance analysis in portfolio choice and capital markets (2000)
- Building diversified portfolios that outperform out of sample (2016)
- PyPortfolioOpt: portfolio optimization in P ython (2021)
- V12. 1: User’s Manual for CPLEX (2009)
- Performance bounds and suboptimal policies for multi-period investment (2013)
- Multi-period portfolio optimization using model predictive control with mean-variance and risk parity frameworks (2022)
- Stochastic control with affine dynamics and extended quadratic costs (2021)
- Advances in Active Portfolio Management: New Developments in Quantitative Investing (2020)
- The behavior of stock-market prices (1965)
- Portfolio selection: E fficient diversification of investments (1959)
- The theory of investment value (1938)
- The intelligent investor (1973)
- Security analysis (2009)
- Portfolio construction with G aussian mixture returns and exponential utility via convex optimization (2023)
- Empirical tests of biases in equity portfolio optimization: What do ex-ante mean-variance optimal portfolios look like ex-post? (1993)
- Dynamic programming and optimal control (2012)
- Dynamic programming (1966)
- Algorithms for decision making (2022)
- Decision making under uncertainty: theory and application (2015)
- Introduction to applied linear algebra: V ectors, matrices, and least squares (2018)
- Gurobi Optimizer Reference Manual (2023) — link
- Trading cost: T he critical link between investment information and results (1983)
- Autonomous precision landing of space rockets (2016)
- Lectures on stochastic programming: M odeling and theory (2021)
- The elements of statistical learning: D ata mining, inference, and prediction (2009)
- Risk M etrics: Technical Document (1996)
- Covariance prediction via convex optimization (2022)
- Dynamic Conditional Correlation (2002)
- Asset allocation: C ombining investor views with market equilibrium (1990)
- Harry Markowitz: S elected works (2009)
- Basel III: an overview (2011)
- Convex Optimization Additional Exercises (2023)
- SDPT3 —a MATLAB software package for semidefinite programming, version 1.3 (1999)
- Using SeDuMi 1.02, a MATLAB toolbox for optimization over symmetric cones (1999)
- Interior-Point Polynomial Algorithms in Convex Programming (1994)
- Applications of second-order cone programming (1998)
- Quantitative portfolio management: T he art and science of statistical arbitrage (2021)
- Interpretable sparse proximate factors for large dimensions (2022)
- State-Varying Factor Models of Large Dimensions (2022)
- Factors that fit the time series and cross-section of stock returns (2020)
- Estimating latent asset-pricing factors (2020)
- Inferential theory for factor models of large dimensions (2003)
- Large dimensional factor analysis (2008)
- The cross-section of expected stock returns (1992)
- Common risk factors in the returns on stocks and bonds (1993)
- Robust and optimal control (1996)
- Essentials of robust control (1998)
- Theory and applications of robust optimization (2011)
- Robust optimization—methodology and applications (2002)
- Robust and convex optimization with applications in finance (2000)
- Robust optimization—a comprehensive survey (2007)
- Recent advances in robust optimization: A n overview (2014)
- Robust optimization (2009)
- Embedded Code Generation With CVXPY (2022) — link
- CVXGEN : A code generator for embedded convex optimization (2012)
- Conic optimization via operator splitting and homogeneous self-dual embedding (2016)
- OSQP : A n operator splitting solver for quadratic programs (2020)
- ECOS : A n SOCP solver for embedded systems (2013)
- Clarabel Documentation (2024) — link
- Harry M arkowitz and the philosopher’s stone (2023)
- Active portfolio management (2000)
- CVXR : A n R package for disciplined convex optimization (2017) — link
- Convex optimization in J ulia (2014)
- A rewriting system for convex optimization problems (2018)
- CVX : MATLAB software for disciplined convex programming, version 2.1 (2014)
- MOSEK modeling cookbook (2020)
- MOSEK Portfolio Optimization Cookbook (2023)
- The M arkowitz optimization enigma: I s `optimized’ optimal? (1989)
- Optimal Execution of Portfolio Transactions (2000)
- Measuring and modeling execution cost and risk (2012)
- Agent-based models for latent liquidity and concave price impact (2014) — link
- Convex Optimization (2004)
- Toward maximum diversification (2011)
- CVXMARKOWITZ (2023) — link
- CVXPY : A P ython-Embedded Modeling Language for Convex Optimization (2016) — link
- Coherent Measures of Risk (2002) — link
- Handbook of Portfolio Construction (2010)
- Methods of Conjugate Gradients for Solving Linear Systems (1952)
- A Simple Method for Predicting Covariance Matrices of Financial Returns (2023) — link
- Modern Portfolio Management: Active Long/Short 130/30 Equity Strategies (2009)
- Portfolio selection (1952)
- The Optimization of Quadratic Functions Subject to Linear Constraints (1955)
- Nobel P rize Lecture (2023) — link
- The Early History of Portfolio Theory: 1600–1960 (1999) — link
- From Portfolio Theory to Practice: An AQR interview with H arry M arkowitz (2019)
- Avoiding the Downside: A Practical Review of the Critical Line Algorithm for Mean-Semivariance Portfolio Optimization (2020) — link
- Efficient asset management: a practical guide to stock portfolio optimization and asset allocation (2008)
- Numerical recipes in C : the art of scientific computing (1992)
- Seven Sins in Portfolio Optimization (2013) — link
- Portfolio construction as linearly constrained separable optimization (2022) — link
- Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers (2011) — link
- A Simplified Model for Portfolio Analysis (1963)
- Portfolio selection (1970)
- Multi-Period Trading via Convex Optimization (2017) — link
- Hedge fund leverage (2011) — link
- Maximization of a linear function of variables subject to linear inequalities (1951)
- The Simplex Method for Quadratic Programming (1959)
- A Light-Weight Multi-Objective Asynchronous Hyper-Parameter Optimizer (2022) — link
- Differentiable Convex Optimization Layers (2019)
- Robust asset allocation (2004)
- Applications of second-order cone programming (1998) — link
- Semidefinite Programming (1996) — link
- Conic formulation of a convex programming problem and duality (1992) — link
- A general system for heuristic minimization of convex functions over non-convex sets (2018) — link
- The arbitrage theory of capital asset pricing (1976) — link
- Common risk factors in the returns on stocks and bonds (1993) — link
- On Persistence in Mutual Fund Performance (1997) — link
- Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk (1964) — link
- The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets (1965)
- Risk Reduction in Large Portfolios: Why Imposing the Wrong Constraints Helps (2003)
- Solutions of ill-posed problems (1977)
- Federal Funds Effective Rate ( DFF ) (2023)
- JuMP 1.0: R ecent improvements to a modeling language for mathematical optimization (2023) — link
- Pyomo– O ptimization modeling in P thon (2021)
- Pyomo: M odeling and solving mathematical programs in P ython (2011)
- A review and comparison of solvers for convex MINLP (2018) — link
- Embedded Convex Optimization for Control (2020)
- Global optimization: Deterministic approaches (2013)
- Prediction of common stock investment risk (1984)
- The Properties of Equally Weighted Risk Contribution Portfolios (2010) — link
- Honey, I Shrunk the Sample Covariance Matrix (2004) — link
- Block splitting for distributed optimization (2013) — link
- POGS - Proximal Operator Graph Solver (2023)
- Parameter Selection and Preconditioning for a Graph Form Solver (2018) — link
- Solving High-Order Portfolios via Successive Convex Approximation Algorithms (2021) — link
- Machine Learning Optimization Algorithms & Portfolio Allocation (2020) — link
- Convex Optimization of Portfolio Kurtosis (2022) — link
- Application of Robust Statistics to Asset Allocation Models (2007) — link
0 of 161 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.