The thesis
channel-pullback is a mean-reversion-within-trend strategy: it buys pullbacks to the lower regression channel or volume support inside confirmed uptrends, then exits at the upper channel or overhead resistance. It trades a universe of 24 US large caps spanning tech, financials, healthcare, staples, energy and industrials — AAPL, MSFT, NVDA, JPM, V, UNH, XOM, CAT and the like. The design intent is clear: let the trend provide the tailwind, and use the channel to time entries and exits rather than chase momentum.
Recent activity
The strategy is live and was reset to $10,000 on 29 October (Season 2). Since then it has stayed close to the line: total equity read $9,964.63 on 30 October, peaked at $10,033.31 on 4 November, and sat at $9,989.09 after the 5 November run — essentially flat, a hair below the start.
The cleanest illustration of the thesis in action is an AAPL/MSFT round-trip. On 30 October it bought MSFT at $522.07 and AAPL at $344.84; on 4 November it sold both into strength at $557.86 and $352.27 — a lower-channel entry and an upper-channel exit, exactly as advertised. The same session it rotated into energy, buying XOM at $168.81 and CVX at $216.40, then re-entered MSFT ($538.81) and AAPL ($346.49) on 5 November. Several scheduled runs executed nothing, which is appropriate for a strategy that waits for pullbacks rather than forcing trades.
Backtest and validation
Over 451 days the backtest returned 7.62% (4.19% CAGR) to a final equity of $10,761.52, with a Sharpe of 0.40 and a 14.83% max drawdown across 137 trades. The win rate is just 39.39% — most trades lose — yet the strategy is net positive, which tells us the payoff is asymmetric: the upper-channel exits capture winners large enough to outweigh frequent small losers. Turnover of 2,311% is high, though fees stayed modest ($137, no FX cost).
The harder truth is that validation failed. Three of four walk-forward folds were positive, but fold 2 (Jan–Jul 2025) lost 11.42% at a −1.70 Sharpe, while fold 3 ripped +20.68% at a 3.86 Sharpe. That spread is the warning sign: performance is heavily regime-dependent. Out-of-sample return was a thinner 3.63% (OOS Sharpe 0.74), and while the Probabilistic Sharpe Ratio is a respectable 0.702, the Deflated Sharpe is only 0.196 after adjusting for 7 trials — below any reasonable confidence bar.
Verdict
channel-pullback has a coherent edge and is behaving on-thesis in live trading, with disciplined entries and exits. But the uneven folds and weak Deflated Sharpe mean we should treat the backtest edge as unproven, not merely modest. The strategy deserves to keep running on paper to gather genuine out-of-sample evidence — its real test is the next adverse regime, not another fold-3.