Strategy Thesis
Channel-pullback operates on a time-tested premise: in confirmed uptrends, price frequently retreats to the lower boundary of a linear regression channel or a volume-weighted support zone before resuming higher. The strategy buys those dips across 24 blue-chip U.S. equities—spanning technology, financials, healthcare, consumer staples, energy, and industrials—and exits at the upper channel boundary or nearby resistance. The logic is sound: it combines trend confirmation with a disciplined entry trigger, sidestepping the trap of buying into broken trends.
Recent Activity
The strategy has been notably quiet this week. Daily scheduled runs on June 15, 16, and 17 each reported zero executions, suggesting the current tape offers few pullbacks that satisfy the entry criteria. The most recent action came on June 12 (two executed, one rejected) and June 10 (two executed, one rejected).
The recent trade log tells a tidy story. The strategy completed a round-trip on Caterpillar (CAT): buying 2 shares at $857.78 on June 10 and selling them at $912.06 on June 12—a 6.3% move in two sessions. It also closed a MSFT position at $400.31 against a $427.54 entry, booking a loss consistent with its low-hit-rate design. NVDA was added on June 12 at $204.82 and remains open. As of June 17, the paper book holds $3,739 in cash against a $10,261 total portfolio value.
Backtest Performance
Over 451 trading days (roughly 15 months), the strategy returned 7.62%—a 4.19% annualized CAGR—with a Sharpe of 0.40 and a peak-to-trough drawdown of 14.83%. Across 137 trades at a 39.39% win rate, this is a classic low-hit-rate, positive-expectancy system: most individual trades lose, but winners are sized to produce a net positive outcome over a full cycle.
Walk-Forward Validation
The four-fold walk-forward results paint a more nuanced picture:
| Fold | Period | Return | Sharpe | Max DD |
|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +6.53% | 1.25 | 6.16% |
| 2 | Jan 2025 – Jul 2025 | −11.42% | −1.70 | 16.09% |
| 3 | Jul 2025 – Dec 2025 | +20.68% | 3.86 | 3.26% |
| 4 | Dec 2025 – May 2026 | +3.63% | 0.74 | 8.44% |
Three of four folds are positive, and the out-of-sample fold (Fold 4, used as the live-trading benchmark) posts a respectable 0.74 Sharpe. That OOS Sharpe exceeding the full-period Sharpe of 0.40 is an encouraging signal—it suggests the strategy is not severely overfit. However, the strategy did not pass formal validation (PSR: 0.702; DSR: 0.196). With seven optimization trials on record, the deflated Sharpe ratio falls well short of the acceptance threshold, meaning confidence that the edge is real rather than luck remains insufficient.
Strengths
- Positive OOS evidence. A 0.74 OOS Sharpe on the most recent, unseen fold is a meaningful result for a live-deployed system.
- Diversified universe. 24 liquid, large-cap names limit idiosyncratic exposure and enable consistent entry-signal generation.
- Coherent thesis. Regression-channel pullback entries are a well-studied technique with a behavioral finance rationale.
Risks
- Regime sensitivity is the central vulnerability. Fold 2's −11.42% return and −1.70 Sharpe—coinciding with a period of trend disruption—demonstrates the strategy can turn sharply negative when the broader uptrend assumption breaks down. Without a market-regime overlay, these episodes are structurally unavoidable.
- Low win rate demands patience. A 39% hit rate means the majority of trades close at a loss. In a live account, emotional pressure to override the system during drawdown periods is a real operational risk.
- High turnover. The backtest logged 2,311% annualized turnover. Paper-trading fees are minimal here, but real-world slippage on 137 trades per cycle could meaningfully compress net returns.
Bottom line: Channel-pullback is a coherent, executable strategy with genuine out-of-sample support—but it needs a regime filter before it earns a passing validation score. A simple breadth or volatility gate could dampen the Fold 2-type losses without sacrificing the strong Fold 3 upside.