The Thesis
Channel-pullback is a trend-following, mean-reversion hybrid: it buys large-cap equities when price pulls back to the lower bound of a linear regression channel coinciding with volume support, then exits at the upper channel or a resistance level. The universe covers 24 liquid names across tech, financials, healthcare, consumer staples, energy, and industrials — diversified enough to keep the strategy busy without chasing illiquid setups.
The logic is straightforward and time-tested: in a confirmed uptrend, shallow pullbacks are buying opportunities, not reversals. The regression channel provides a statistically grounded floor; volume support adds confluence.
Live Activity: First Week
The strategy deployed on June 1 with five immediate executions, putting most of the $10,000 to work quickly. The opening positions included JPM, COST, and a large DIS stake. By end of day, cash stood at $808.75 — an aggressive initial deployment.
The subsequent days have been more measured:
- June 1 EOD: 2 executed (WMT buy, DIS sell), portfolio at $10,049.
- June 2: 2 executed (UNH buy at $376.02, CAT sell at $908.35), portfolio at $10,117.
- June 3: 0 executed, 2 rejected — the strategy correctly sat on its hands when setups didn't meet criteria.
- June 4: 4 executed, 1 rejected. Notable: UNH was sold at $396.30 just two days after being bought at $376.02, a clean channel-to-channel round trip. MSFT entered at $427.54; DIS re-entered at $99.29 after the prior position was trimmed.
Four days in, total equity sits at $10,380 (+3.8% on starting capital), with $3,943 in cash — a healthier reserve than the day-one deployment.
Backtest Performance
Across 451 days and 137 trades, the full backtest returned +7.62% (CAGR ~4.19%), with a Sharpe of 0.40 and a maximum drawdown of 14.83%. Turnover was high at 2,311% annualized, which at $1/trade in fees consumed exactly $137 — negligible at this scale but worth watching if the universe or frequency expands.
The win rate of 39.4% is low but not alarming for a strategy that aims to capture large channel moves; it implies the average winner must significantly outpace the average loser. Whether that holds consistently is the central question.
Validation: Real Edge or Noise?
The validation run did not pass. With 7 trials and a Probabilistic Sharpe Ratio (PSR) of 0.70, the strategy clears the 50% bar but falls short of a high-confidence threshold. The Deflated Sharpe Ratio (DSR) of 0.196 is the harder number — after adjusting for multiple testing across 7 trials, the edge shrinks considerably.
The cross-validation folds tell a more nuanced story:
| 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% |
Fold 3 is exceptional — Sharpe 3.86 with a 3.26% drawdown — but it masks the brutal fold 2, which saw a 16% drawdown and a −11.4% return. That fold almost certainly captured the early-2025 sell-off in large-cap tech and financials, exactly when "confirmed uptrend" signals lag reality.
Fold 4 (the out-of-sample period most relevant to live deployment) returned +3.63% with a Sharpe of 0.74 and an 8.44% drawdown. That's the benchmark to beat going forward.
Strengths and Risks
Strengths: The strategy has a clean, interpretable thesis with no parameter soup. Three of four folds were profitable. The live early days show sensible behavior — sitting out on June 3 when setups were absent, and executing a crisp UNH round trip.
Risks: A 39% win rate demands discipline in sizing and loss-cutting; a streak of losers can erode capital quickly, as fold 2 demonstrated. The high turnover amplifies fee and slippage drag in real trading. Most importantly, the DSR flags that much of the apparent edge may be explained by trial selection rather than a genuine signal. The strategy warrants continued observation before drawing conclusions.