ml4t-deflated-sharpe skillA
ml4t-deflated-sharpe is agent-read markdown (skill) from ml4t/skills: Adjust the Sharpe ratio for multiple testing bias when selecting from many trials. Use when reporting strategy performance after parameter or model search..
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
# Deflated Sharpe Ratio
Reporting the best Sharpe ratio from N trials is misleading. The Deflated Sharpe Ratio corrects for the number of trials, non-normal returns, and sample size to test whether observed performance reflects genuine skill.
## The Problem
If you test 100 strategy variants and report the best Sharpe, you are performing selection bias. Under the null of zero skill, the expected maximum Sharpe across N trials grows with `sqrt(2 * log(N))`, measured in units of the spread of Sharpes across those trials: at 100 trials the best result is about 2.5 spreads above zero before any alpha exists. Without correction, most "discovered" strategies are artifacts that fail out of sample.
## The Pattern
### WRONG
```python
import numpy as np
# Test 50 parameter combos, report the best
sharpes = []
for params in param_grid: # 50 configurations
returns = run_backtest(params)
sr = returns.mean() / returns.std() * np.sqrt(252)
sharpes.append(sr)
best = max(sharpes)
print(f"Strategy Sharpe: {best:.2f}") # Inflated by selection
```
### CORRECT
```python
import numpy as np
from scipy import stats
def deflated_sharpe_ratio(observed_sr, n_trials, sr_std, n_obs,
…Read the whole file at its exact version.
How to install
mdr add ml4t/skills/ml4t-deflated-sharpe@git:20260901.2c5ab85mdr add ml4t/skills/ml4t-deflated-sharpe@sha256:0bdd5d2edd8821c3Pin 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.
[](https://markdownregistry.com/a/art_tokusjodzvo4vfc2)
0 badge views in 30 days
Versions
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20260901.2c5ab85 latest | 2026-09-01 | 2c5ab85 | 4,619 B | A | view · diff |
| git:20260901.c415df0 | 2026-09-01 | c415df0 | 4,458 B | A | view · diff |
| git:20260528.303089e | 2026-05-28 | 303089e | 4,374 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (4619 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
ml4t/skills · 19 stars · license Apache-2.0 · pushed 2026-09-24 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_tokusjodzvo4vfc2 GET https://markdownregistry.com/api/v1/resolve?ref=ml4t/skills/ml4t-deflated-sharpe GET https://markdownregistry.com/api/v1/blob/0bdd5d2edd8821c37845a26c6dd14aa65e2abc277d3320be2d145d9ee4bd7adf
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.