← Ground Truth report library

Ross Cameron's Momentum Rules — Full Test Report

Episode A01 · watch the episode · report format v1
VERDICT — FAITHFUL NO-EDGE / TRANSFER-NEGATIVE

Ross Cameron's published rules encode cleanly into a machine — and priced faithfully at 1-minute resolution, on survivorship-free data with real costs, they have no positive expectancy. The edge, if it exists, lives in a layer the published rules don't capture.

What we granted vs. what we tested

Granted, not re-audited: that Ross Cameron makes money trading — taken at face value from his own published, broker-verified statements. That was never the question.

Tested: only whether his published rules, handed to a machine, reproduce that result. They don't. The violinist is real; the sheet music doesn't reproduce the performance.

How the test was run (the Five Gates applied)

Gate 1 — the claim in writing: we encoded the published momentum day-trading rules; the large majority of the rule set mechanizes cleanly. Gate 2 — real data: survivorship-free, point-in-time US equity data — including the tickers that later died. Gate 3 — real costs: commissions, slippage, and honest fills at 1-minute resolution. Gate 4 — pre-registered pass/fail: written down before results. Gate 5 — reopen bounty: published below.

The results ledger

FigureWhat it isStatus
41 days, 0 trades, equity flat at $99,802.50 The live clone on a real-time feed (paper account) traded nothing for its entire run Real, but inconclusive on profitability — zero trades is zero P&L signal. We use it as the hook, never as the verdict.
−$20,584.83 net · profit factor 0.632 · 34 of 37 trades stopped out The faithful rebuild of the best-documented entry pattern, priced over years of honest data MEASURED — the evidence seat of the verdict
0 for 12 · profit factor 0.000 The twelve setups the rules themselves call textbook — the faithful robot lost all twelve MEASURED
99.95% of gross profit = one lucky all-day hold What little the backtest made came from a single position held all day — not breakout edge MEASURED
−$35,791 · 31 of 32 stopped A second detector we built produced this scarier number — and failed our own fidelity audit (it dropped rule components and hardened one filter to always-pass) EXCLUDED — our artifact, not his edge. It appears nowhere in the episode or this verdict. Publishing a number your own audit killed is the exact sin this channel exists to expose.
Why the exclusion row is printed at all: a test that only shows you its surviving numbers is grading its own homework (Gate 4). The graveyard of our own instruments is part of the report.

What the verdict does NOT claim

It does not claim Ross Cameron doesn't trade profitably (granted above). It does not claim the 41-day zero was "discipline" — zero trades supports no such story, in either direction. It does not claim no human can trade this style — only that the published rules, executed mechanically at the resolution a machine can see, don't carry the edge.

What would change our mind (the reopen bounty)

Pre-registered before publication: (1) instruments that can see below the one-minute bar, tested under the identical honest-friction floor; (2) a clean-label re-run of the faithful detector; (3) a real scanner-fed ticker universe replacing hand-labelled setups. If any of these clears the floor, the update ships here first.