The thesis
Channel-pullback is a textbook mean-reversion play with a trend filter. It buys pullbacks to the lower regression channel or volume support — but only inside confirmed uptrends — and exits at the upper channel or resistance. In principle this is a disciplined "buy the dip, sell the rip" approach that leans on trend to keep it out of falling knives. The strategy is currently live, trading a 24-name large-cap universe spanning tech, financials, healthcare, staples and energy.
Recent activity
The last week has been dominated by AAPL. The strategy round-tripped Apple repeatedly: buying at $308.56 and $315.28, selling at $311.60, $314.66 and $315.09 — small, frequent scalps rather than big directional bets. It also rotated through UNH (bought 5 at $390.72), JPM (bought 5 at $352.96), BAC (bought 31 at $64.24) and trimmed CAT. Scheduled runs have executed cleanly with zero rejections, and total equity has hovered in a tight $9,446–$9,611 band over the period. Cash swings from ~$60 to ~$4,200 between runs, showing the book is often close to fully deployed.
Backtest performance
Over 451 days the strategy returned 7.62% (final equity $10,761), a 4.19% CAGR, with a Sharpe of 0.40 and a max drawdown of 14.83%. Two numbers stand out. First, the win rate is just 39.39% across 137 trades — this is a system that wins less than half the time and relies on winners outrunning losers. Second, turnover is 2,311%, extremely high; the strategy is churning the book many times over, which raises real-world sensitivity to fees and slippage. In backtest, fees were a modest $137 total, but live frictions could erode a 4% CAGR quickly.
Validation: the sceptic's view
The walk-forward suite did not pass, and it is worth understanding why. Three of four folds were positive, and the probabilistic Sharpe ratio (PSR 0.702) looks respectable. But the deflated Sharpe ratio — which penalises for the 7 trials run — collapses to 0.196, signalling the edge may not survive multiple-testing scrutiny. The folds also tell a story of inconsistency: fold 2 lost 11.42% (Sharpe −1.70) while fold 3 gained 20.68% (Sharpe 3.86). That is a wide dispersion for a strategy meant to grind out steady mean-reversion gains. Out-of-sample return averaged a thinner 3.63% with a 0.74 Sharpe.
Verdict
Channel-pullback has a coherent thesis and a real, if modest, positive expectancy in backtest. Its strengths are a sensible trend filter, controlled drawdowns in three of four folds, and clean live execution. The risks are equally clear: a sub-40% win rate leaves little margin, punishing turnover exposes it to friction, and the failed validation — especially the deflated Sharpe — warns that the historical edge may be partly luck. This is a strategy worth watching in live paper trading, but not one to size up on until it demonstrates the edge holds out-of-sample.