Paper: arXiv 2604.00415

Authors: Tan Chin Hong, Chung-Han Hsieh

Abstract

The Double Linear Policy (DLP) framework guarantees a Robust Positive Expectation (RPE) under optimized constant-weight designs or admissible prespecified time-varying policies. However, the sequential optimization of these time-varying weights remains an open challenge. To address this gap, we propose a Stochastic Model Predictive Control (SMPC) framework. We formulate weight selection as a receding-horizon optimal control problem that explicitly maximizes risk-adjusted returns while enforcing survivability and predicted positive expectation constraints. Notably, an analytical gradient is derived for the non-convex objective function, enabling efficient optimization via the L-BFGS-B algorithm. Empirical results demonstrate that this dynamic, closed-loop approach improves risk-adjusted performance and drawdown control relative to constant-weight and prescribed time-varying DLP baselines.

Complexity vs Empirical Score

  • Math Complexity: 7.5/10
  • Empirical Rigor: 6.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper presents a mathematically solid integration of Stochastic Model Predictive Control (SMPC) into the Double Linear Policy framework, supported by analytical gradient derivations for non-convex optimization. While the theoretical formulation is robust and novel, the empirical validation relies on basic baselines and would benefit from broader out-of-sample testing across diverse market regimes.

Research Flowchart

  flowchart TD
    A[Research Goal: Sequential Weight Optimization for Double Linear Policy DLP] --> B[Inputs: Equity Market Data & DLP Framework Constraints]
    B --> C[Formulate SMPC Receding-Horizon Control Problem]
    C --> D[Derive Analytical Gradient for Non-Convex Objective Function]
    D --> E[Optimize Weights via L-BFGS-B Algorithm]
    E --> F[Key Outcomes: Improved Risk-Adjusted Returns & Drawdown Control]

Cited References (54)

  1. On Robust Optimal Linear Feedback Stock Trading (2022) — link
  2. Spot Trading Fee Rate (2026)
  3. A Jump Start to Stock Trading Research for the Uninitiated Control Scientist: A Tutorial (2024)
  4. On Robustness of Double Linear Trading with Transaction Costs (2023) — link
  5. Analysis of Financial Time Series (2005)
  6. On Robustness of Double Linear Policy with Time-Varying Weights — → 2303.10806
  7. Robust Algorithmic Trading in a Generalized Lattice Market (2025)
  8. From Semi-Infinite Constraints to Structured Robust Policies: Optimal Gain Selection for Financial Systems (2025) — link
and 46 more
  1. On a New Paradigm for Stock Trading Via a Model-Free Feedback Controller (2016) — link
  2. On Trading of Equities: A Robust Control Paradigm (2008) — link
  3. On Arbitrage Possibilities via Linear Feedback in an Idealized Brownian Motion Stock Market (2011) — link
  4. On Performance Limits of Feedback Control-Based Stock Trading Strategies (2011) — link
  5. A Generalization of Simultaneous Long–Short Stock Trading to PI Controllers (2018) — link
  6. On Simultaneous Long-Short Stock Trading Controllers with Cross-Coupling (2020)
  7. On Robustness of Adaptive Feedback Control for Stock Trading With Time-Varying Price Dynamics (2025) — link
  8. Cross-Coupled SLS for Pairs Trading: an Adaptive Control Approach (2023) — link
  9. A Generalization of the Robust Positive Expectation Theorem for Stock Trading via Feedback Control (2018) — link
  10. Numerical Optimization (2006)
  11. Reinforcement Learning to Run… Fast (2018)
  12. Discretizing Continuous Action Space for On-Policy Optimization (2020)
  13. Learning dexterous in-hand manipulation (2020) — link
  14. Combinatorics (1996) — link
  15. Proximal Policy Optimization Algorithms (2017) — link
  16. Trust Region Policy Optimization (2015) — link
  17. Reinforcement Learning in Continuous State and Action Spaces (2012) — link
  18. High-Dimensional Continuous Control Using Generalized Advantage Estimation (2015) — link
  19. Asynchronous Methods for Deep Reinforcement Learning (2016) — link
  20. Truly Proximal Policy Optimization (2020) — link
  21. Mastering the game of Go with deep neural networks and tree search (2016)
  22. Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm (2017) — link
  23. Human-level control through deep reinforcement learning (2015)
  24. Reinforcement Learning for Robust Parameterized Locomotion Control of Bipedal Robots (2021) — link
  25. Deep Reinforcement Learning: A Survey (2024) — link
  26. Reinforcement learning for logistics and supply chain management: Methodologies, state of the art, and future opportunities (2022) — link
  27. Reinforcement learning for intensive care medicine: actionable clinical insights from novel approaches to reward shaping and off-policy model evaluation (2024)
  28. Performance functions and reinforcement learning for trading systems and portfolios (1998) — link
  29. Reinforcement Learning for Trading (1998) — link
  30. Q-learning (1992)
  31. A reinforcement learning approach to dynamic portfolio optimization (2025)
  32. Continuous Control with Deep Reinforcement Learning (2015) — link
  33. A $Q$ -Learning Approach to Derive Optimal Consumption and Investment Strategies (2008) — link
  34. A DDPG Algorithm for Portfolio Management (2020) — link
  35. Applications of MPC to Finance (2018)
  36. Stochastic Model Predictive Control and Portfolio Optimization (2007)
  37. MultiObjective Dynamic Optimization of Investment Portfolio Based on Model Predictive Control (2022) — link
  38. Pairs Trading under Transaction Costs using Model Predictive Control (2018) — link
  39. A Limited Memory Algorithm for Bound Constrained Optimization (1995)
  40. Continuous control with deep reinforcement learning (2020)
  41. Deterministic Policy Gradient Algorithms (2014)
  42. Optimization by Simulated Annealing (1983) — link
  43. Differential Evolution – A Simple and Efficient Heuristic for Global Optimization over Continuous Spaces (1997) — link
  44. Global Optimization by Basin-Hopping and the Lowest Energy Structures of Lennard-Jones Clusters Containing up to 110 Atoms (1997)
  45. An L-BFGS-B Approach for Linear and Nonlinear System Identification Under $\ell_1$ and Group-Lasso Regularization (2025)
  46. On the usage of differential evolution for function optimization (1996) — link

1 of 54 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.