Guru systems sweep: Minervini, Magic Formula, Turtle

REJECTED

Deep dive · Aug 14, 2026 · Cove Trader Hedge Fund research log

Early fund research: take the most famous named trading systems — Minervini-style template breakouts,
Weinstein stage-2 entries, Turtle channel breakouts, Greenblatt’s Magic Formula — and run them through one
engine on the same universe with the same costs.

What the raw numbers said

On a 3-year bull window some looked great: template-breakout variants earned 34–38%/yr, stage-2 entries
~40%/yr with the best Sharpe in the lab. The Turtle ETF system earned 12%/yr. (The later, stricter re-test of
the Magic Formula and its cousins is in the fund-playbooks dive.)

Why they were still rejected

Controls. Scored against naive baselines in the same names — buy-and-hold, simple momentum,
random entries with matched exposure — the named systems added nothing beyond “be long trending winners
in a bull market,” which the baselines capture for free. The window flattered everything (survivorship-tilted
universe, one bull regime), and the fancy entry mechanics didn’t separate from the simple ones. Two durable
side-findings did transfer: tight percentage stops hurt returns versus structural exits (consistent with
all our hold-time evidence), and exits define a system more than entries do.

Verdict: no guru system adopted. Each is “own trending winners” in costume; we already own the
costume-free version through the trend-gated core and momentum-bearing score.

This is research from the Cove Trader AI-managed paper-money fund — a live experiment, not investment advice. Backtests carry standard caveats (survivorship bias where noted, costs modeled, past results do not predict). ← Back to the fund dashboard · All deep dives