skill-backtest-overfit

Detect backtest overfitting and selection bias from multiple testing. Use when a user has a backtest / factor result and asks whether the Sharpe is real, whether a strategy is overfit, or wants to validate results before trusting them. Computes Deflated Sharpe Ratio, Probability of Backtest Overfitting (PBO), multiple-testing haircut, and Minimum Track Record Length.

Jump to install

Repository: quantskills/skill-backtest-overfit
Last source activity: July 16, 2026 at 09:36
Detected SKILL.md language: Chinese
Stars: 26
Forks: 2

Install options

The review-first prompt is selected by default. You can switch to a direct command or download a local copy.

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

A direct command skips the review prompt. Inspect the source before running it.

npx skills add https://github.com/quantskills/skill-backtest-overfit --skill skill-backtest-overfit

The command stays on one line. Scroll horizontally to inspect it before copying.

Prefer a local copy? Download the files currently available to SkillsMP.

Review the source files

Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.

SKILL.md

Source instructions · Read-only preview

Back to search Back to installation

name	skill-backtest-overfit
description	Detect backtest overfitting and selection bias from multiple testing. Use when a user has a backtest / factor result and asks whether the Sharpe is real, whether a strategy is overfit, or wants to validate results before trusting them. Computes Deflated Sharpe Ratio, Probability of Backtest Overfitting (PBO), multiple-testing haircut, and Minimum Track Record Length.
license	GPL-3.0
category	工具
metadata	{"organization":"QuantSkills","organization_url":"https://github.com/quantskills","repository":"skill-backtest-overfit","repository_url":"https://github.com/quantskills/skill-backtest-overfit","project_type":"skill","collection":"portfolio-risk-validation"}

{
  "version": 1,
  "task": {
    "placeholder": "说明待检验的回测、收益序列、试验次数及数据文件；可上传 selected_returns 和 trials_matrix",
    "required": true
  },
  "fields": [
    {
      "key": "n_trials",
      "label": "试验次数",
      "type": "number",
      "placeholder": "例如 200",
      "help": "请如实填写所有尝试过的参数或策略配置数量"
    },
    {
      "key": "periods_per_year",
      "label": "年化期数",
      "type": "number",
      "default": "252",
      "help": "日频通常为 252"
    },
    {
      "key": "haircut_method",
      "label": "多重检验方法",
      "type": "select",
      "default": "holm",
      "options": [
        { "value": "holm", "label": "Holm" },
        { "value": "bonferroni", "label": "Bonferroni" },
        { "value": "bhy", "label": "BHY" }
      ]
    }
  ],
  "prompt_template": "{{#task}}任务与材料：\n{{task}}\n\n{{/task}}{{#attachments}}用户上传的材料（已放入工作区）：\n{{attachments}}\n\n{{/attachments}}检验该回测的过拟合与多重选择偏差。{{#n_trials}}按真实试验次数 {{n_trials}} 计算。{{/n_trials}}{{#periods_per_year}}年化期数采用 {{periods_per_year}}。{{/periods_per_year}}{{#haircut_method}}多重检验校正采用 {{haircut_method}}。{{/haircut_method}}计算 DSR、PBO、Haircut Sharpe 和最小样本长度，缺少试验矩阵时明确降级项及局限，给出可追溯的通过/失败结论，输出中文报告。"
}