Gold Spot Price (XAU/USD) Minute-Level OHLCV Dataset with Realized Volatility and Regime Indicators: 2003–2025

This dataset accompanies the research paper "Gold Volatility Regime Analysis Across Global Crises: Evidence from 22 Years of Minute-Level OHLCV Data (2003–2025)." Contents: daily_ohlcv_clean.csv — 6,763 daily observations of gold spot price (XAU/USD) including open, high, low, close, volume, log returns, and five OHLCV-based range volatility estimators: close-to-close, Parkinson (1980), Garman-Klass (1980), Rogers-Satchell (1991), and Yang-Zhang (2000). All volatility measures are provided in both raw daily variance and annualized form. Crisis period labels are included. realized_volatility.csv — 6,761 daily observations of model-free intraday volatility measures constructed from 7.4 million minute-level observations, including Realized Variance (RV), Bipower Variation (BV), Jump Component, Jump Ratio, and annualized Realized Volatility. Used in HAR-RV estimation following Corsi (2009). clean_gold_data.csv --- ~7.5 million one minute observations of gold spot price (XAU/USD) including open, high, low, close, volume Coverage: June 2003 – August 2025 Asset: Gold spot price, XAU/USD Frequency: Daily (aggregated from minute-level source data)本数据集随附于研究论文《全球危机下的黄金波动率制度分析：基于22年分钟级OHLCV（Open, High, Low, Close, Volume）数据的实证证据（2003–2025年）》。 数据集内容： `daily_ohlcv_clean.csv`：包含6763条黄金现货价格（XAU/USD）的日度观测数据，涵盖开盘价、最高价、最低价、收盘价、成交量、对数收益率，以及5种基于OHLCV的区间波动率估计量：收盘-收盘估计量、帕金森（Parkinson, 1980）估计量、加曼-克拉斯（Garman-Klass, 1980）估计量、罗杰斯-萨切尔（Rogers-Satchell, 1991）估计量以及杨-张（Yang-Zhang, 2000）估计量。所有波动率指标均提供原始日方差与年化形式两种版本，并附带危机时期标签。 `realized_volatility.csv`：基于740万条分钟级观测数据构建的6761条日度无模型日内波动率测度数据，涵盖已实现方差（Realized Variance, RV）、双幂变差（Bipower Variation, BV）、跳跃成分、跳跃比率以及年化已实现波动率。该数据集可用于遵循科尔西（Corsi, 2009）提出的HAR-RV模型开展估计。 `clean_gold_data.csv`：包含约750万条黄金现货价格（XAU/USD）的一分钟级观测数据，涵盖开盘价、最高价、最低价、收盘价与成交量。 数据覆盖范围：2003年6月至2025年8月；标的资产：黄金现货价格（XAU/USD）；数据频率：日度（由分钟级源数据聚合得到）。