mlsanityData to 7 Oct 2026

A Three-Phase Foundation Model for Tax-Aware Personalized Portfolio Management

arXiv 2606.30997 ↗First posted 30 Jun 2026cs.AIcited by 2

On the map: Mixture-of-experts → MoE for time series and graphs

Abstract (arXiv)

We present a three-phase deep reinforcement learning system for personalized portfolio management that addresses three limitations shared by all prior financial RL work: 1) ticker lock-in, 2) monolithic objectives , and 3) static user models. Phase 1 pretrains a ticker-identity-free cross asset encoder via self-supervised learning on a multi-asset corpus, augmented by a frozen parallel branch using Chronos, a T5-based time series foundation model, fused via a learned gating mechanism. To our knowledge, this is the first application of a time series foundation model to portfolio management RL. The encoder generalizes to any publicly traded asset via a 50-dimensional observable metadata vector that requires no retraining for new tickers. Phase 2 fine-tunes a MoE (Mixture of Experts) portfolio actor critic with PPO under an objective-conditioned reward that simultaneously serves six distinct investment goals sampled per episode: short-term alpha, short-term gain, long-term gain, capital preservation, tax-loss harvesting, and long-term-gains-only. A MoE architecture assigns each objective to a specialized expert head (momentum, growth, defensive, tax-aware), and a learned intent router blends experts based on the active objective and current market regime, which eliminates cross-objective gradient conflict. Phase 3 adds a lightweight personalization layer further adapted at inference time to each individual via a 76-parameter LoRA module fine-tuned on real brokerage transaction history, inferring investment objectives from revealed trading behavior rather than questionnaires. A natural language intent parser converts free-form goals directly into structured investment objective parameters.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Placed in MoE for time series and graphs, one of 12 directions in Mixture-of-experts. Builds on 13 of its 20 references that are in the map. First posted on arXiv 30 Jun 2026.

Builds on

13 of 20 references.

References · 13

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  3. 2024Chronos: Learning the Language of Time Seriescited by 712
  4. 2023A decoder-only foundation model for time-series forecastingcited by 650
  5. 2024Unified Training of Universal Time Series Forecasting Transformerscited by 588
  6. 2016Successor Features for Transfer in Reinforcement Learningcited by 331
  7. 2021A Practical Guide to Multi-Objective Reinforcement Learning and Planningcited by 240
  8. 2023PIXIU: A Large Language Model, Instruction Data and Evaluation Benchmark for Financecited by 217
  9. 2014From Predictive to Prescriptive Analyticscited by 142
  10. 2017A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problemcited by 123
  11. 2018Dynamic Weights in Multi-Objective Deep Reinforcement Learningcited by 96
  12. 2020FinRL: A Deep Reinforcement Learning Library for Automated Stock Trading in Quantitative Financecited by 86
1 more
  1. 2020Reinforcement-Learning based Portfolio Management with Augmented Asset Movement Prediction Statescited by 55

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026An Emerging Retail Portfolio Management Application: Personalized, Tax-Aware Reinforcement Learning with Natural Language Goalscited by 0
  2. 2026Retrieval-augmented generation vs. deterministic tax computation in multi-agent financial advisory: A 2x2 factorial experimentcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Learning to Manage Investment Portfolios beyond Simple Utility Functions
  2. 2026Regime-Aware Portfolio Management via Retrieval-Augmented LLM-Guided Expert Switching
  3. 2026High-Stakes Personalization: Rethinking LLM Customization for Individual Investor Decision-Making
  4. 2022MetaTrader: An Reinforcement Learning Approach Integrating Diverse Policies for Portfolio Optimization
  5. 2025Fin-R1: A Large Language Model for Financial Reasoning through Reinforcement Learning
  6. 2022Combining Reinforcement Learning and Inverse Reinforcement Learning for Asset Allocation Recommendations
  7. 2025Trading-R1: Financial Trading with LLM Reasoning via Reinforcement Learning
  8. 2026Eliciting ESG Preferences for Reinforcement Learning-Based Portfolio Optimization