Home / ml4t / skills · concepts/backtest-overfitting/SKILL.md · GitHub

ml4t-backtest-overfitting skillA

ml4t-backtest-overfitting is agent-read markdown (skill) from ml4t/skills: Detect and prevent overfitting to historical data via multiple testing corrections and pre-registration. Use when evaluating strategy variants to ensure performance is not a data-mining artifact..

Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.

What the file says

# Backtest Overfitting

Testing many strategies on the same data guarantees finding one that looks profitable by chance. With 100 independent trials at p < 0.05, you expect five false positives.

## The Problem

Every parameter you tune, every feature you try, and every universe filter you adjust is an implicit trial. A researcher who reports a Sharpe ratio of 2.0 after exploring 200 configurations has not found alpha - they have found the luckiest draw from a noise distribution. Correcting for the number of trials, by haircut here and by Deflated Sharpe Ratio in `ml4t-deflated-sharpe`, is what separates the two. Without it, most published backtests are statistically meaningless.

## The Pattern

### WRONG

```python
# Tune until something looks good
best_sharpe = 0
for lookback in [5, 10, 21, 63, 126, 252]:
    for top_k in [5, 10, 20, 50]:
        result = backtest(lookback=lookback, top_k=top_k)
        sharpe = result["sharpe"]
        if sharpe > best_sharpe:
            best_sharpe = sharpe
            best_params = (lookback, top_k)

print(f"Best Sharpe: {best_sharpe:.2f}")  # meaningless without correction
```

### CORRECT

```python
import numpy as np
…

Read the whole file at its exact version.

How to install

Latest version
mdr add ml4t/skills/ml4t-backtest-overfitting@git:20260901.1323bf4
Exact content
mdr add ml4t/skills/ml4t-backtest-overfitting@sha256:be65e56bddcbd414

Pin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_ejvcx4toailuvhde.svg)](https://markdownregistry.com/a/art_ejvcx4toailuvhde)

1 badge views in 30 days

Versions

versioncommittedcommitsizeaudit
git:20260901.1323bf4 latest2026-09-01 1323bf4 5,054 BA view · diff
git:20260901.f84f8252026-09-01 f84f825 5,015 BA view · diff
git:20260901.32123e72026-09-01 32123e7 4,995 BA view · diff
git:20260901.2c5ab852026-09-01 2c5ab85 4,942 BA view · diff
git:20260901.c415df02026-09-01 c415df0 4,773 BA view

Audit of the latest version

A  17 of 17 checks passed. Deterministic, no model, same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (5054 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

ml4t/skills · 19 stars · license Apache-2.0 · pushed 2026-09-24 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_ejvcx4toailuvhde
GET https://markdownregistry.com/api/v1/resolve?ref=ml4t/skills/ml4t-backtest-overfitting
GET https://markdownregistry.com/api/v1/blob/be65e56bddcbd414ae8b9049a0ad5bf7eda87fdab9118fdc637ff8ba01d4372a

Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

More from ml4t/skills

AGENTS.md agents
ml4t/skills · AGENTS.md
git:20260901.e8c1b9a · audit A · 19 stars
CLAUDE.md claude
ml4t/skills · CLAUDE.md
git:20260901.c415df0 · audit A · 19 stars
ml4t-agent-governance skill
ml4t/skills · advanced-ai/agent-governance/SKILL.md · Security and governance controls for autonomous financial agents. Use when agents can call tools, read untrusted…
git:20260901.3c85a54 · audit A · 19 stars
ml4t-agent-state-memory skill
ml4t/skills · advanced-ai/agent-state-memory/SKILL.md · Durable agent state, memory, and replay for autonomous research workflows. Use when an agent must resume, audit, or…
git:20260901.c415df0 · audit A · 19 stars
ml4t-agent-tool-contracts skill
ml4t/skills · advanced-ai/agent-tool-contracts/SKILL.md · Typed tool contracts for autonomous research agents. Use when exposing files, search, databases, or execution tools to…
git:20260901.c415df0 · audit A · 19 stars
ml4t-multi-agent-forecasting skill
ml4t/skills · advanced-ai/multi-agent-forecasting/SKILL.md · Multi-agent probability forecasting with diversity, aggregation, and debate controls. Use when combining several agent…
git:20260901.c415df0 · audit A · 19 stars
ml4t-research-operator skill
ml4t/skills · advanced-ai/research-operator/SKILL.md · Thin autonomous research operator pattern for ML4T experiments. Use when an agent should inspect artifacts, read…
git:20260901.c415df0 · audit A · 19 stars
ml4t-cost-model skill
ml4t/skills · backtest/cost-model/SKILL.md · Commission, slippage, and market-impact cost models for realistic strategy simulation. Use when backtesting to ensure…
git:20260901.297789f · audit A · 19 stars
ml4t-rl-execution skill
ml4t/skills · backtest/rl-execution/SKILL.md · Reinforcement learning for trade execution and hedging. Use when optimizing execution algorithms or dynamic hedging…
git:20260901.c415df0 · audit A · 19 stars
ml4t-run-backtest skill
ml4t/skills · backtest/run-backtest/SKILL.md · Event-driven backtesting with realistic order execution, position tracking, and performance measurement. Use when…
git:20260901.c415df0 · audit A · 19 stars
ml4t-sensitivity-analysis skill
ml4t/skills · backtest/sensitivity-analysis/SKILL.md · Test strategy robustness to parameter variation and detect overfitting cliffs. Use when validating that performance is…
git:20260901.85fc79b · audit A · 19 stars
ml4t-tearsheet skill
ml4t/skills · backtest/tearsheet/SKILL.md · Generate comprehensive performance reports from backtest returns. Use when summarizing backtest results for review or…
git:20260901.32123e7 · audit A · 19 stars

Every file in ml4t/skills

Browse by kind, by grade A, or by owner.