The Thesis
Channel-pullback buys price pullbacks to the lower regression channel or volume-support zone within confirmed uptrends, then exits at the upper channel or a resistance level. The 24-name universe covers large-cap US equities across tech, financials, healthcare, consumer, and energy — wide enough to stay active across most macro regimes while remaining within liquid, well-quoted names.
Backtest Snapshot
Over 451 days and 137 trades: +7.62% total return, CAGR 4.19%, Sharpe 0.40, max drawdown 14.83%, win rate 39.4%. The sub-40% win rate is not inherently disqualifying — asymmetric risk-reward strategies can be profitable with fewer winning trades — but annualised turnover of 2,312% stands out. At roughly 23× per year, real-money transaction costs would meaningfully compress the already-thin CAGR.
Walk-Forward Results: Regime Sensitivity Exposed
| 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 Sharpe (0.74) comfortably exceeds the full-period figure (0.40) — an encouraging sign. However, Fold 2's −11.42% loss and Sharpe of −1.70 reveal a clear vulnerability: in choppy or trending-down conditions the regression-channel signal degrades and drawdowns widen sharply.
Formal validation did not pass. The Probabilistic Sharpe Ratio (PSR) of 0.702 implies roughly 70% odds the true Sharpe is positive — better than chance, but short of conviction. More telling is the Deflated Sharpe Ratio (DSR) of 0.196, which penalises for the seven parameter trials run and flags meaningful overfitting risk. Taken together, these figures place the strategy in a probationary state rather than cleared for scale-up.
Live Activity: A Selective Week
Trading has been cautious. June 4 was the busiest session — four executions including buys of MSFT at 427.54 and DIS at 99.29, alongside exits of UNH (396.30, up from a June 2 entry at 376.02) and JPM (311.52). On June 10, MSFT was sold at 400.31 — a roughly 6.4% loss over six days — while CAT was added at 857.78. Portfolio value eased from approximately $10,380 on June 4 to $10,194 by June 10, consistent with the mild drift visible in the Fold 4 live window.
Key Risks
- Regime dependence: Fold 2's 16% drawdown shows the strategy can deteriorate quickly outside trending markets.
- High turnover: 2,312% annualised turnover makes execution costs a significant real-money CAGR drag.
- Low win rate: at 39%, consecutive-loss runs are statistically likely and require disciplined position sizing.
- Validation gap: DSR of 0.196 flags residual overfitting risk; more clean out-of-sample data is needed before drawing strong conclusions.
Verdict
Channel-pullback has a coherent thesis and shows genuine edge when markets cooperate — Fold 3's 20.68% return at a 3.86 Sharpe is legitimately impressive. The final out-of-sample fold is modestly positive, which is exactly what you want to see from a live deployment. But the validation failure, wide fold dispersion, and high turnover argue for patience. The current paper-trading phase is doing its job: accumulating real evidence before any allocation increase.