Newest papers

STOCK-JEPA: Prior-Anchored Latent Revision Representation Learning in Equity Markets

Yizhi Luo, Jiahe Yi, Jianhui Zhang, Shuo Sun · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.07006 · cited by 0 works (OpenAlex) · open-access PDF

Indonesian

Pembelajaran representasi konteks yang efektif membantu untuk memprediksi struktur dan dinamika sinyal pasar ekuitas dari data keuangan dengan rasio sinyal-to-noise yang rendah. Model dalam kotak hitam dapat menangkap pola yang kompleks tetapi dapat melampaui sampel kebisingan yang sama dan tidak memiliki struktur ekonomi yang eksplisit. Sementara itu, model keuangan linear klasik memberikan referensi yang dapat ditafsirkan, tetapi asumsi-asumsi mereka yang terlalu sederhana meninggalkan sinyal non-linear yang tidak tertangkap. Untuk menggabungkan kekuatan dari dua arah ini, kami mengusulkan Stock-JEPA, kerangka prediksi gabungan yang menggabungkan kebisingan dan tidak memiliki struktur ekonomi eksplisit. Pada saat yang sama, model keuangan linear klasik memberikan referensi yang dapat dipahami, tetapi asumsi-asumsi mereka yang terlalu sederhana untuk meninggalkan sinyal non-linear yang tidak tertangkap. Untuk menghasilkan statistik kesalahan tetap yang mer

Machine translation (Meta M2M100); the English original follows.

Abstract

Learning effective representations helps characterize the structure and dynamics of equity markets from financial data with a low signal-to-noise ratio. Black-box deep models can capture complex patterns but may overfit sample noise and lack explicit economic structure. Meanwhile, classic linear financial models provide interpretable references, but their oversimplified assumptions leave non-linear signals uncaptured. To combine the strengths of these two directions, we propose Stock-JEPA, a joint-embedding predictive framework that learns predictable incremental revisions relative to a point-in-time financial prior. First, we leverage a low-complexity financial model to produce fixed statistics summarizing multi-horizon return and risk. A prior projector then maps these statistics into the target encoder's latent space as an anchor. Second, we design a context-conditioned revision predictor to estimate the future representation's predictable displacement from the anchor. Separate losses update the two branches: the anchor learns from prior statistics, while the revision captures additional predictable information from historical context. Third, we freeze all representation modules and train a downstream readout, evaluating its forecasts through cross-sectional ranking and portfolio performance. Theoretically, we prove that optimal revision reduces the prior anchor's expected squared error for the same future representation by exactly $\mathbb{E}[\|\boldsymbolΔ\|_2^2]$. This non-negative gain is the expected squared magnitude of the additional signal predictable from historical context. Experimentally, Stock-JEPA outperforms 13 strong baselines across large-scale China and U.S. equity universes on 5 key evaluation metrics. Ablation studies and representation analysis further demonstrate the value of the learned revisions for representation learning in equity markets.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Overfitting, Econometrics, Computer science, Leverage (statistics), Representation (politics), Equity (law)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange