---
license: cc-by-4.0
task_categories:
- time-series-forecasting
- text-classification
tags:
- finance
- portfolio-management
- multi-asset
- market-data
- benchmark
language:
- en
size_categories:
- 10M<n<100M
---

# PortBench-Market: Multi-Asset Market Base Dataset

## Dataset Description

A ten-year (January 2015 – December 2025) multi-asset financial dataset covering 
**183 distinct instruments** across six heterogeneous asset classes. Designed as the 
foundation for evaluating LLMs on portfolio management tasks requiring cross-asset 
reasoning, correlation awareness, and stress robustness.

### Asset Class Composition

| Asset Class | Instruments | Sources | Examples |
|-------------|-------------|---------|----------|
| Equities | 126 | Yahoo Finance, Kaggle, SEC | SPY, QQQ, sector ETFs, factor ETFs, international |
| Commodities | 16 | Yahoo Finance, FRED | GLD, USO, DBA, commodity indices |
| Bonds | 15 | Yahoo Finance, FRED | Treasury ETFs (SHV→TLT), TIPS, corporate, EM |
| Cryptocurrency | 12 | Yahoo Finance, Kaggle | BTC, ETH, SOL, XRP, ADA, ... |
| Real Estate | 10 | Yahoo Finance, FRED | VNQ, IYR, Case-Shiller indices |
| Cash | 4 | Yahoo Finance, FRED | BIL, SGOV, T-bill rates |

### Data Fields

Each instrument includes:
- **Price data**: daily OHLCV (open, high, low, close, volume)
- **Returns**: daily log returns, rolling statistics
- **Macro indicators**: yield curves, credit spreads, inflation, volatility indices (60 FRED series)
- **News text**: financial news associated with asset classes (from Kaggle/SEC sources)
- **Correlation artifacts**: full cross-asset correlation matrix, covariance matrix, asset class mapping
- **Regime labels**: bull / bear / sideways / crisis per date (MA crossover + drawdown rules)

### Temporal Coverage

- **Full range**: 2015-01-01 to 2025-12-31
- **Stress windows**: 2015 China Shock (Aug 2015–Feb 2016), 2020 COVID Crash (Feb–May 2020), 2022 Crypto Collapse (May–Dec 2022)
- **Normal evaluation window**: January–December 2024

### Dataset Structure