What people built, ranked by what it measured.
Each of these was written in the builder and published as a link. The figures below were computed when this page rendered, from the code each card opens — nothing about performance is stored, so the order is a measurement rather than a claim. It is still an in-sample check across six daily markets, not the 18-cell audit.
01·2026-08-13
EMA 9/21 crossover
median Sharpe
0.62
5/6 markets
+1079.8pp vs B&H
02·2026-08-13
SMA 5/20 crossover
median Sharpe
0.61
5/6 markets
-86pp vs B&H
03·2026-08-13
EMA 12/26 crossover
median Sharpe
0.58
5/6 markets
+661pp vs B&H
04·2026-08-13
EMA 12/26 + EMA 12 crossover
“MACD 기반 매매”
median Sharpe
0.53
5/6 markets
-92.7pp vs B&H
05·2026-08-13
100-bar breakout
median Sharpe
0.42
5/6 markets
-184pp vs B&H
06·2026-08-13
RSI 14 crossover
median Sharpe
0.23
4/6 markets
-641pp vs B&H
Ranked by the median of six per-market Sharpes, recomputed hourly. Strategies with fewer than 10 trades across the six markets are not listed — there is nothing to measure. In-sample, no repaint or look-ahead guards, no alpha score and no verdict: those belong to the audited desk, and none of this has been through it.