Thesis
channel-pullback is a buy-the-dip strategy with discipline: it waits for a confirmed uptrend, buys pullbacks to the lower regression channel or volume support, and exits at the upper channel or resistance. In principle this is textbook mean-reversion inside a trend — collect the oscillation while the longer-term drift works in your favour. The universe is a sensible 24-name slice of large-cap US equities spanning tech, financials, healthcare, staples and energy.
Recent Activity
The strategy is live and trading on schedule. Over the last six runs it has been rotating defensively: buying XOM, CVX, JNJ, KO and a large PFE position (70 shares at $28.18) while trimming MA and JPM. Execution has been clean — most runs filled 2 of 2 orders — but not frictionless. The 2026-09-29 run saw both orders rejected, and the 2026-09-28 run had one rejection against three fills.
There is also visible churn. CVX was sold on 2026-10-01 at $206.95 and rebought the next day at $207.21; XOM was sold on 09-30 and a fresh 10-share lot bought on 10-05. Portfolio equity has drifted from $11,079.17 (09-28) down to $10,863.11 (10-05) — a mild pullback of its own, with cash hovering around $1,700–2,100.
Backtest and Validation
The headline backtest returns +7.62% over 451 days (CAGR 4.19%), with a Sharpe of 0.40 and a 14.83% max drawdown. That is a thin risk-adjusted return. Two numbers deserve attention. First, the win rate is only 39.39% across 137 trades — the strategy loses more often than it wins and leans entirely on winners outsizing losers, which is fragile. Second, turnover is 2,311%, generating a fee on essentially every trade; frictions compound on a strategy this active.
Most importantly, validation did not pass. Three of four walk-forward folds were positive, and the numbers look strong in benign regimes — fold 3 (2025-07 to 2025-12) returned +20.68% at a 3.86 Sharpe. But fold 2 (2025-01 to 2025-07) collapsed to −11.42% at a −1.70 Sharpe, exposing a clear regime dependence. The out-of-sample fold returned a tepid +3.63% (Sharpe 0.74). The statistical picture confirms the caution: PSR of 0.70 is borderline, but the deflated Sharpe ratio is just 0.196 after adjusting for 7 trials — low enough that we cannot confidently distinguish this edge from luck.
Verdict
channel-pullback has a coherent thesis and shines when trends are smooth, but it is regime-sensitive, trades too much, and wins too rarely to clear our bar. Treat the live allocation as an experiment, not a conviction position, and watch whether the current defensive rotation stabilises the equity curve.