DEMO — NOT A USER SUBMISSION

A real strategy that failed unseen data.

This is not a mockup. Strategy Arena's Rust engine executed the public EMA 9/21 template on 8,640 BTC 5-minute bars, kept the final 20% outside selection, and returned REJECT.

Check my Pine for free Read the methodology

EngineRust CPU reference

DatasetBTC · 5m · 8640 bars

Holdout20% · 1728 bars

Measured variants1

What the engine measured

The figures below come from the content-addressed Rust report. They are historical diagnostics, not a forecast or trading recommendation.

Train PnL-49.64%

Holdout PnL-18.02%

Buy & Hold3.73%

Holdout drawdown18.02%

Holdout trades136

Monte Carlo10000 paths

Loss probability100.00%

Holdout used in selectionNo

Training was already negative after the execution-cost contract.
The chronological holdout lost 18.02% while Buy & Hold gained 3.73%.
10,000 native Monte Carlo paths estimated a 100% loss probability for the observed trade distribution.

Meaning
Do not continue with this exact design as validated evidence. This does not prove that every EMA strategy will fail.

Exact evidence identities

The page is blocked if its artifact fingerprint or any mandatory identity fails validation at load time.

Open reproducibility details

Artifact SHA-256165137cfad1d06c148e5e35cd7d2a9ae67c1f0e59573afba67f50ee7a492be85

Report SHA-2564bdcae70cd7177e474fbf8c9adb1973a5586d6f6175563d76be7af5929d5f67e

Pine SHA-256206316b408bffc9026543a7f879f5eb3a9a6db4e96d0a4e19b223e57d8df3015

Dataset SHA-256dfaefcc0c95dc8f46f15841b588718b58aec7abf1599e4bfde631af30a9c03e0

Native dataset SHA-256261cf9db41fde31a6671a14b9a3a6f6b72e1aed040600e51dad782002a00c75f

Holdout SHA-256d1379bfe3acc10ae3de3400a1a769cb9951a36450446049e2d98813b9a901fa5

Engine SHA-2561d269ab0507c73fc280b3ed80e346bd102856051d888639725565ecc37eec078

Monte Carlo input SHA-256b7ea89056f43fff4824fd51e5f27a0e42d5c33592cdb444042b68ecc3b10cf27

//@version=5
strategy("Strategy Arena Demo EMA 9/21", overlay=true)
fast = ta.ema(close, 9)
slow = ta.ema(close, 21)
if ta.crossover(fast, slow)
    strategy.entry("Long", strategy.long)
if ta.crossunder(fast, slow)
    strategy.close("Long")

Why another strategy stayed inconclusive

INCONCLUSIVE · Strategy Arena Demo MACD Momentum

Only 1 closed holdout trade was observed. Strategy Arena requires at least 10, so Monte Carlo stayed at zero paths and no green conclusion was manufactured.

Holdout3.73%

Closed trades1/10

No fabricated survivor

No KEEP verdict was observed in the fixed public template benchmark; Strategy Arena does not fabricate a survivor.

A genuine KEEP example will appear only after the same Rust contract produces one. Even then, it will remain a demonstration unless it comes from a formally locked independent audit.