The Thesis
channel-pullback bets on a structural observation: in a confirmed uptrend, price routinely reverts to its lower regression channel before resuming higher. The strategy identifies those pullbacks across a 24-stock universe of liquid large-caps — spanning tech, financials, healthcare, consumer staples, and energy — and exits when price reaches the upper channel or encounters resistance. It is a trend-following overlay with a mean-reversion entry trigger, which theoretically avoids chasing while staying aligned with the primary trend direction.
Recent Activity
The strategy has been largely patient heading into the July 4th holiday. Over the past six sessions it executed just two trades: a purchase of 19 shares of DIS at $96.32 and a sale of 7 WMT shares at $108.53, both on July 1. The preceding five sessions each logged a single rejected signal — a sign that candidate pullbacks failed to satisfy entry criteria such as channel depth or volume confirmation. Portfolio value has been stable, ranging from $10,006 to $10,226 over the past week with roughly $771 in cash.
Looking back further, recent activity has included AAPL added at $279.39, a round-trip on DIS in the $99–$104 range, a CAT flip bought at $857.78 and sold at $912.06, and a MSFT position entered at $427.54 and exited at $400.31 — the latter a meaningful loss that illustrates the strategy's willingness to cut positions at resistance even when the exit is unfavorable.
Backtest Performance
Across 137 trades over 451 days, channel-pullback grew a $10,000 starting portfolio to $10,761.52, a total return of 7.62% and a CAGR of 4.19%. The Sharpe ratio is 0.40 and maximum drawdown reached 14.83% — a drawdown-to-return ratio that demands scrutiny. Annualized turnover of 2,311% is the most pressing structural concern: in a live account, commissions and market impact would substantially erode these figures, and the $1 flat-fee assumption in the backtest captures only a fraction of that cost.
Win rate sits at 39.4%, meaning the strategy is a net loser on more than six in ten trades and depends entirely on winners being materially larger than losers. Whether that asymmetry is durable or a backtest artifact is the central question.
Validation: The Numbers Tell a Cautious Story
The strategy did not pass formal validation. Cross-validation across four folds reveals extreme regime sensitivity:
| Fold | Period | Return | Sharpe |
|---|---|---|---|
| 1 | Aug '24 – Jan '25 | +6.53% | 1.25 |
| 2 | Jan '25 – Jul '25 | −11.42% | −1.70 |
| 3 | Jul '25 – Dec '25 | +20.68% | 3.86 |
| 4 | Dec '25 – May '26 | +3.63% | 0.74 |
Fold 3's outsized performance inflates the aggregate figures considerably; remove it and the picture is far less flattering. The Probabilistic Sharpe Ratio of 0.702 suggests a better-than-coin-flip chance the Sharpe is genuinely positive, but the Deflated Sharpe Ratio of 0.196 — which penalises for seven trials — indicates the observed edge is unlikely to be robust after accounting for selection bias. Three of four folds finish positive, which is encouraging, but the severity of fold 2 (−11.42%, Sharpe −1.70) confirms the strategy has a real vulnerability in trending-down or volatile regimes.
Outlook
channel-pullback demonstrates a coherent entry logic and has produced a modest positive out-of-sample result in its most recent fold (+3.63%, Sharpe 0.74). That is not nothing. But the failed validation, high turnover cost, and fold-to-fold volatility argue against any increase in sizing or priority. The strategy is best treated as a watch candidate: continue paper trading through the next market regime, and revisit the validation gate once fold 5 data accumulates.