Minerva
Sign inSign up
← The Minerva Research Letter

One edition, exactly as it lands in your inbox.

N° 000 · 2026-07-31

Carver EWMAC Trend on TLT

carver_trend@1.0.0 · 1day · WF+CPCV
MinervaScore23Seal · Fail5 gates short.
DSRFAIL0.07
PBOFAIL0.53
SPAFAIL1.000
MinTRLFAIL0.00
Window stabilityFAIL0.33
0.72
OOS Sharpe
48,720
Backtests
-0.00%
Net return
7
Trades

Carver's EWMAC is the closest thing systematic trend has to a reference implementation, and trend made its money on rates for a decade. We run the rule on TLT now that the cut cycle is behind us.

Robustness battery0 OF 5 PASSED
Deflated Sharpe ratio0.07need >= 0.95FAIL
Backtest overfitting (PBO)0.53need <= 0.50FAIL
Predictive ability (SPA p)1need <= 0.10FAIL
Track record length (MinTRL)0.00seal gateFAIL
Regime stability0.33need >= 0.60FAIL

FROM THE REPORT · DSR 0.068 < 0.5 · SPA p 1.0000 ≥ 0.05 · Too few trades (7 < 10)

METHOD: WF+CPCV · 2024-06-162026-06-16 · 10 WF windows · 36 CPCV paths · verbatim from the report

This edition was produced by Minerva itself. Want the verdict on your own idea?

Run your own idea →