The thesis
Channel-pullback is a disciplined mean-reversion play layered on top of trend. It waits for a confirmed uptrend, then buys dips into the lower regression channel or a volume-support level, aiming to exit back at the upper channel or overhead resistance. The universe is 24 large-cap US names spanning tech, financials, healthcare, staples, and energy — liquid, well-covered tickers where regression channels tend to behave.
The design choice shows up clearly in the numbers: a 39.39% win rate paired with a +7.62% total return over 137 trades. That combination is the signature of a strategy that loses small and often but lets its winners run to the upper band — acceptable, provided the asymmetry holds.
Recent activity
The strategy is live and running scheduled sessions nightly. Its paper book has hovered around $10,900–$11,080 through late September into October. Recent executions lean into the playbook: pullback buys in defensives like PFE ($28.18), JNJ ($265.85), and KO ($87.44), funded by sells into strength in CVX ($206.95), XOM ($163.54), and BAC ($56.58).
One trade captures the whole thesis in miniature: NVDA bought at $224.16 on 25 Sep and sold at $230.06 on 28 Sep — a quick exit at the upper channel. Not every session fires, though. The 24th and 29th saw zero executions and one-to-two rejections each, a sign the entry filters are genuinely gating rather than forcing trades.
Backtest and validation
The backtest covers 451 days: +7.62% total return, 4.19% CAGR, Sharpe 0.40, and a 14.83% max drawdown. The headline concern is turnover of 2,311% — this strategy trades a lot. Fees were only $137 and FX zero in the paper model, but at real-world cost structures that churn would bite.
Walk-forward validation is where the caution hardens. It did not pass. Three of four folds were positive, but fold 2 (Jan–Jul 2025) lost 11.42% at a Sharpe of −1.70, while fold 3 posted a flattering +20.68%. That spread signals regime sensitivity: the edge thrives in some conditions and breaks in others. Out-of-sample return was 3.63% at Sharpe 0.74 — positive, but thin. Most telling, across 7 trials the deflated Sharpe ratio sits at just 0.196 (PSR 0.702), well below the confidence bar for a robust, non-overfit edge.
Verdict
Channel-pullback is a coherent, honestly-constructed strategy that is making money in paper and respecting its own entry rules. But a sub-0.2 DSR, a double-digit losing fold, and 2,300% turnover are three independent reasons not to size it aggressively yet. Treat it as a live experiment on probation: worth running for more out-of-sample evidence, not worth betting the book on.