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

• URL: http://arxiv.org/abs/2606.30997v2
• Date: Thu, 02 Jul 2026 19:01:30 GMT
• Title: A Three-Phase Foundation Model for Tax-Aware Personalized Portfolio Management
• Abstract summary: We present a three-phase deep reinforcement learning system for personalized portfolio management.<n>Phase 1 pretrains a ticker-identity-free cross asset encoder via self-supervised learning on a multi-asset corpus.<n>Phase 2 fine-tunes a MoE (Mixture of Experts) portfolio actor with PPO under an objective-conditioned reward.<n>Phase 3 adds a lightweight personalization layer further adapted at inference time to each individual.
• Score: 0.8122270502556375
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: 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.

Related papers

• Designing Agentic AI-Based Screening for Portfolio Investment [1.7499351967216341]
  We introduce a new agentic artificial intelligence (AI) platform for portfolio management.<n>Our architecture consists of three layers. First, two large language model (LLM) agents are assigned specialized tasks.<n>Second, these agents deliberate to generate and agree upon buy and sell signals from a large portfolio, substantially narrowing the pool of candidate assets.
  arXiv  Detail & Related papers  (2026-03-24T15:03:40Z)
• Breaking User-Centric Agency: A Tri-Party Framework for Agent-Based Recommendation [59.362016745945375]
  We propose the first Tri-party LLM-agent Recommendation framework (TriRec)<n>TriRec explicitly coordinates user utility, item exposure, and platform-level fairness.<n>We show consistent gains in accuracy, fairness, and item-level utility.
  arXiv  Detail & Related papers  (2026-03-11T11:40:13Z)
• Rethinking Personalization in Large Language Models at the Token Level [89.64259958702165]
  Personalization is typically framed as an additional layer on top of a base NLP task.<n>From a token-level perspective, different tokens in a response contribute to personalization to varying degrees.<n>We propose PerContrast, a self-contrast method that estimates each output token's dependence on user-specific information.
  arXiv  Detail & Related papers  (2026-02-04T16:05:02Z)
• Learning to Manage Investment Portfolios beyond Simple Utility Functions [0.9999629695552193]
  We propose a generative framework that learns latent representations of fund manager strategies without requiring explicit utility specification.<n>We validate our framework on a dataset of 1436 U.S. equity mutual funds.<n>Our framework provides a data-driven approach for characterizing investment strategies for applications in market simulation, strategy attribution, and regulatory oversight.
  arXiv  Detail & Related papers  (2025-10-30T06:01:20Z)
• Attention Illuminates LLM Reasoning: The Preplan-and-Anchor Rhythm Enables Fine-Grained Policy Optimization [56.083511902353365]
  Reinforcement learning (RL) typically applies uniform credit across an entire generation of Large language models.<n>This work positions attention as a privileged substrate that renders the internal logic of LLMs as a mechanistic blueprint of reasoning itself.<n>We introduce three novel RL strategies that dynamically perform targeted credit assignment to critical nodes.
  arXiv  Detail & Related papers  (2025-10-15T13:49:51Z)
• Trade in Minutes! Rationality-Driven Agentic System for Quantitative Financial Trading [57.28635022507172]
  TiMi is a rationality-driven multi-agent system that architecturally decouples strategy development from minute-level deployment.<n>We propose a two-tier analytical paradigm from macro patterns to micro customization, layered programming design for trading bot implementation, and closed-loop optimization driven by mathematical reflection.
  arXiv  Detail & Related papers  (2025-10-06T13:08:55Z)
• From Headlines to Holdings: Deep Learning for Smarter Portfolio Decisions [4.288926547930663]
  We present an end-to-end framework that learns portfolio weights using deep learning.<n>We evaluate the framework on nine U.S. stocks spanning six sectors, chosen to balance sector diversity and news coverage.<n>Although the stock universe is limited, the results underscore the value of integrating price, relational, and sentiment signals for portfolio management.
  arXiv  Detail & Related papers  (2025-09-29T00:42:24Z)
• Your Offline Policy is Not Trustworthy: Bilevel Reinforcement Learning for Sequential Portfolio Optimization [82.03139922490796]
  Reinforcement learning (RL) has shown significant promise for sequential portfolio optimization tasks, such as stock trading, where the objective is to maximize cumulative returns while minimizing risks using historical data.<n>Traditional RL approaches often produce policies that merely memorize the optimal yet impractical buying and selling behaviors within the fixed dataset.<n>Our approach frames portfolio optimization as a new type of partial-offline RL problem and makes two technical contributions.
  arXiv  Detail & Related papers  (2025-05-19T06:37:25Z)
• Specialized Foundation Models Struggle to Beat Supervised Baselines [60.23386520331143]
  We look at three modalities -- genomics, satellite imaging, and time series -- with multiple recent FMs and compare them to a standard supervised learning workflow.<n>We find that it is consistently possible to train simple supervised models that match or even outperform the latest foundation models.
  arXiv  Detail & Related papers  (2024-11-05T04:10:59Z)
• Large-scale Time-Varying Portfolio Optimisation using Graph Attention Networks [4.2056926734482065]
  This study utilise 30 years of data on mid-cap firms, creating graphs of firms using distance correlation and the Triangulated Maximally Filtered Graph approach.<n>We show that the portfolio produced by the GAT-based model outperforms all benchmarks and is consistently superior to other strategies over a long period.
  arXiv  Detail & Related papers  (2024-07-22T10:50:47Z)
• Deep Graph Convolutional Reinforcement Learning for Financial Portfolio Management -- DeepPocket [6.85316573653194]
  Portfolio management aims at maximizing the return on investment while minimizing risk by continuously reallocating the assets forming the portfolio. A graph convolutional reinforcement learning framework called DeepPocket is proposed whose objective is to exploit the time-varying interrelations between financial instruments. DeepPocket is evaluated against five real-life datasets over three distinct investment periods.
  arXiv  Detail & Related papers  (2021-05-06T15:07:36Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.