The thesis
Channel-pullback runs a disciplined mean-reversion idea: buy pullbacks to the lower regression channel or volume support within confirmed uptrends, then exit into the upper channel or resistance. It trades a 24-name universe of US large caps spanning tech (AAPL, MSFT, NVDA), financials (JPM, V, MA), healthcare (JNJ, UNH), and staples (PG, KO, WMT). The design is sound in principle — it only fades weakness when the primary trend is intact, which is the classic guardrail against catching a falling knife.
Recent activity
Activity has been light and AAPL-heavy. The strategy churned Apple repeatedly through late August — buying 6 shares at $308.56–$315.28 and selling at $311.60–$315.09 — alongside single positions in UNH, JPM, and a CAT exit at $837.16. Since August 28, scheduled runs have executed zero trades, and the account has effectively gone quiet with only $27.13 in cash.
The uncomfortable detail is the equity line. The paper account total has hovered between roughly $9,417 and $9,561 across the last week — below the $10,000 starting base. In other words, live performance is currently underwater, a notable contrast with the positive backtest.
Backtest and validation
On paper the record is respectable but unspectacular: +7.62% total return over 451 days, a 4.19% CAGR, final equity of $10,761.52, and a Sharpe of just 0.40. Max drawdown ran to 14.83%, and the win rate is a low 39.39% across 137 trades — meaning the system relies on a handful of winners outpacing many small losers. Turnover is extreme at 2,311%, so fees and slippage matter more than the headline suggests.
Validation is where caution hardens. The strategy failed its gate. Walk-forward showed 3 of 4 folds positive, but the dispersion is alarming: fold 2 (Jan–Jul 2025) lost 11.42% with a Sharpe of -1.70, while fold 3 delivered +20.68% at a Sharpe of 3.86. That is not a stable edge — it is one great regime carrying two mediocre ones and one bad one. Out-of-sample return was a thinner 3.63% (Sharpe 0.74). Most tellingly, across 7 trials the Probabilistic Sharpe Ratio sits at 0.702 but the Deflated Sharpe Ratio collapses to 0.196 — once you correct for multiple testing, the probability the true Sharpe exceeds zero is weak.
Verdict
The strengths are real: a coherent, trend-filtered entry logic and low fold-2 correlation to a single bad stretch. But the risks dominate the current picture — a sub-0.5 Sharpe, a failed validation gate, a punishing DSR, and a live account sitting below its starting stake. Channel-pullback reads less like a proven edge and more like a plausible idea still searching for confirmation. Until the deflated statistics improve and live equity recovers par, this one belongs on the watchlist, not the allocation sheet.