The thesis
channel-pullback is a trend-following mean-reversion hybrid: it buys pullbacks to the lower regression channel or volume support within confirmed uptrends, then exits at the upper channel or resistance. The idea is sound and well-worn — you let an established trend do the directional work and use temporary weakness as your entry, aiming to buy low within a structure that is still rising. It trades a conservative universe of 24 large-cap US names spanning tech (AAPL, MSFT, NVDA), financials (JPM, V, MA), healthcare (JNJ, UNH, PFE), staples (KO, PG, WMT) and energy (XOM, CVX).
Recent activity
The strategy is live and running on schedule. Over the past week it executed a steady cadence of two trades per session on most days, with a couple of quieter runs: 2026-10-06 saw zero executions and one rejection, and 2026-09-29 logged two rejections. Recent fills lean defensive and energy-heavy — buys in XOM (10 @ $163.77), CVX (9 @ $207.21), PFE (70 @ $28.18), JNJ (7 @ $265.85) and KO (22 @ $87.44), with sells trimming MA, JPM and an earlier XOM position. The paper book sits around $10,915 total with roughly $1,767 in cash as of 2026-10-06, drifting slightly off its late-September high near $11,014.
Backtest performance
Over 451 days the strategy returned 7.62% (final equity $10,761.52), a 4.19% CAGR, on 137 trades. The texture is less reassuring than the top line: a 39.39% win rate means the strategy wins less than half the time and relies on winners outrunning losers, the Sharpe of 0.40 is modest, and max drawdown reached 14.83%. Turnover of 2,311% is extremely high — it churns its capital roughly 23 times — which in live trading amplifies slippage and fee drag well beyond the flat $137 modelled here.
Validation: the strategy did not pass
This is the headline risk, and we won't bury it. Across four walk-forward folds, three were positive but the spread is alarming:
- Fold 1 (Aug 2024–Jan 2025): +6.53%, Sharpe 1.25
- Fold 2 (Jan–Jul 2025): −11.42%, Sharpe −1.70
- Fold 3 (Jul–Dec 2025): +20.68%, Sharpe 3.86
- Fold 4 (Dec 2025–May 2026): +3.63%, Sharpe 0.74
Out-of-sample return was 3.63% at a Sharpe of 0.74 — respectable but far below fold 3's flattering run. The probabilistic Sharpe ratio (0.702) looks acceptable, but the deflated Sharpe ratio of just 0.196, adjusted for 7 trials, is what sinks it. In plain terms: once you account for how many variants were tried, the evidence that this edge is real rather than lucky is weak.
Verdict
channel-pullback has a coherent thesis and is behaving sensibly in live paper trading. But a negative fold deep enough to wipe out a full year of gains, a sub-coin-flip win rate, and a failing deflated Sharpe mean it has not earned a capital-weighted promotion. Treat it as a candidate under observation, not a validated edge — worth watching its live out-of-sample behaviour before trusting the backtest's green number.