The thesis
channel-pullback is a mean-reversion-within-trend strategy: it buys pullbacks to the lower regression channel or volume support only in confirmed uptrends, then exits at the upper channel or resistance. It runs a 24-name large-cap universe spanning tech (AAPL, MSFT, NVDA), financials (JPM, V, MA), staples (KO, PG, COST) and energy (XOM, CVX). The design is coherent — it tries to buy dips without fighting the primary trend — but the numbers say it is not yet a strategy we'd stand behind unconditionally.
Recent activity
The strategy is live and trading on a scheduled cadence. Over the last week of runs (2026-09-21 to 2026-09-28) it executed 12 orders and rejected 6, ending 2026-09-28 with $2,128.46 cash and a total account value of $11,079.17. Rejections are frequent — two runs (09-22 and 09-24) executed nothing at all — which is consistent with a strategy that waits for a specific setup rather than forcing trades.
The tape shows the expected round-trip behaviour: it sold 8 NVDA at $230.06 on 09-28 after buying 8 at $224.16 just three days earlier — a clean pullback-to-target trade. Other recent fills include buys in KO (22 @ $87.44), MA (3 @ $567.49) and CVX (9 @ $204.99), alongside exits in MSFT and BAC.
Backtest and validation
On the full backtest the strategy returned 7.62% (final equity $10,761.52, CAGR 4.19% over 451 days) with a Sharpe of 0.40 and a max drawdown of 14.83%. The win rate is a low 39.39% across 137 trades — acceptable for a trend/mean-reversion hybrid only if winners are meaningfully larger than losers. Turnover is high at 2,311%, though fees were negligible ($137 total, no FX cost).
The more important verdict comes from walk-forward validation, and it failed. Across four folds, three were positive, but the results are wildly uneven:
- Fold 2 (2025-01 to 2025-07): -11.42%, Sharpe -1.70, 16.1% drawdown
- Fold 3 (2025-07 to 2025-12): +20.68%, Sharpe 3.86
- Fold 4 (2025-12 to 2026-05): +3.63%, Sharpe 0.74
Out-of-sample return averaged 3.63% at a 0.74 Sharpe — respectable, and the probabilistic Sharpe ratio (PSR) of 0.70 looks encouraging. But the deflated Sharpe ratio is just 0.196. Once we penalise for the 7 trials behind this configuration, the edge largely evaporates. That gap between PSR and DSR is the whole story: the headline looks fine until you account for how many variants were tried to get here.
Verdict
Strengths: a sound, trend-aware entry logic, disciplined trade selection (high rejection rate), and one genuinely strong out-of-sample fold. Risks: a sub-40% win rate, a punishing 2025 H1 fold, and — decisively — a deflated Sharpe near zero that suggests the backtest may be a product of selection rather than durable edge. It earns its live paper allocation as a candidate worth watching, but it has not cleared the bar for confidence.