Market-Bench - Dynamic Delta Hedging (Mean of 5)
Professional Work · 2025-12-13
The hard Market-Bench strategy asks a model to implement dynamic delta hedging using random-walk option deltas and MSFT order-book data, including persistent consumed liquidity, execution delay, stock and option exposure, profit and loss, equity, and drawdown. This definition reports mean absolute error averaged over five generated backtester runs.
Top models (lower is better)
| Model | Score |
|---|---|
| GPT-5.2 | 1370 |
| Grok 4 | 1482 |
| Gemini 3 Pro Preview | 4596 |
| GPT-5.1-Codex-Max | 10496 |
| DeepSeek-V3.2 | 14724 |
| Sonnet 4.5 | 16157 |
| Opus 4.5 | 18945 |
| Llama 4 Maverick | 21280 |
| Nova Premier | 21346 |
| Llama 3.1 Nemotron Ultra 253B V1 | 21346 |
| Command A | 21370 |
| Mistral Large 3 675B Instruct 2512 | 63683 |
| Qwen3-Max (2025-09-23) | 329700 |