发表机构
University of Chicago; Johns Hopkins University; Brown University（芝加哥大学; 约翰霍普金斯大学; 布朗大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究提出Market-1T数据集、严格评估协议及大规模比较18种编码器策略，为金融世界模型建立训练与评估基础，发现相似预测性能的编码器可能组织市场状态不同。
AI 中文摘要
构建世界模型需要一个对规划和决策有用的状态表示——可能涉及训练时未知的任务。在金融市场背景下，规划和决策可能需要模型推理市场整体状况、资产特定预期收益、流动性、波动性以及跨资产关系。然而，金融表示学习主要是在单个预测任务上进行评估，通常是在单一时间段使用相对狭窄的数据集。我们通过三项主要贡献来解决这一问题。首先，我们引入了Market-1T，这是一个包含从2008年到2025年美国股票近一万亿条观测值的数据集，分辨率为1赫兹。其次，我们开发并实施了一个严格的评估协议。第三，我们进行了金融表示学习的系统性大规模研究，比较了18种编码器训练策略，覆盖了近二十年的市场体制。我们通过常见金融任务上的预测效用和潜在结构探针来评估学习到的表示。我们发现，具有相似预测性能的编码器可以非常不同地组织市场状态。总的来说，我们为训练和评估金融市场表示建立了基础，以支持诸如DINO-WM、V-JEPA 2和LeWM等世界模型。
英文摘要
Building a world model requires a state representation useful for planning and decision-making---potentially over tasks unknown at training time. In the context of financial markets, planning and decision-making may require a model to reason about market-wide conditions, asset-specific expected returns, liquidity, volatility, and cross-asset relationships. Yet financial representation learning has largely been evaluated on individual predictive tasks, oftentimes on a single time period using comparatively narrow datasets. We address this through three primary contributions. First, we introduce Market-1T, a dataset containing nearly one trillion observations across U.S. equities from 2008 to 2025 at 1 Hz resolution. Second, we develop and implement a rigorous evaluation protocol. Third, we conduct a systematic large-scale study of financial representation learning, comparing 18 encoder-training strategies across nearly two decades of market regimes. We evaluate learned representations both by their predictive utility on common finance tasks and through probes of latent structure. We find that encoders with similar predictive performance can organize market state very differently. Collectively, we establish a foundation for training and evaluating financial market representations in support of world models such as DINO-WM, V-JEPA 2, and LeWM.
