DBbun/100K_Buffetts_Alpha_v1.0

这是一个由合成数据构成的金融数据集，受到论文《Buffett’s Alpha》的启发。它模拟了伯克希尔哈撒韦风格的回报、杠杆、融资、因子暴露、投资组合分解和系统性复制绩效。数据集包含100个独立合成公司的数据，每个公司有大约83年的月度数据（1,000个观测值）。所有数据存储在单个统一文件中，适用于特征提取、总结、表格到文本转换、文本检索等任务，聚焦于金融和投资领域。

This is a finance dataset composed of synthetic data, inspired by the paper Buffett’s Alpha. It simulates Buffett-style returns, leverage, financing, factor exposures, portfolio decomposition, and systematic replication performance. The dataset includes data for 100 independent synthetic companies, each with approximately 83 years of monthly data (1,000 observations). All data is stored in a single unified file and is suitable for tasks such as feature extraction, summarization, tabular-to-text conversion, text retrieval, focusing on finance and investing.