Strategy Thesis
Channel-pullback is a mean-reversion-within-trend system: it enters when price retraces to the lower regression channel or a volume-supported floor inside a confirmed uptrend, then exits at the upper channel or a resistance level. The logic has a well-documented pedigree — disciplined pullback entries in trending markets offer a repeatable structural edge — but execution across a 24-stock large-cap universe demands rigorous filtering to distinguish genuine retracements from the early stages of trend reversals.
Recent Activity
The strategy has been running on its daily schedule through early June 2026, with the paper portfolio currently sitting near $10,300. Execution has been episodic: after a more active stretch on June 1–4 — four trades on June 4, two on each of the prior two sessions — the system found no qualifying setups on June 5 or June 8, rejecting one candidate each day.
The June 4 session stood out with a UNH rotation (sold at $396.30 having bought at $376.02 two days earlier) alongside a JPM round-trip (bought at $295.58 on June 1, sold at $311.52 on June 4). DIS and MSFT were added as fresh entries on the same day. The rising rejection rate over the most recent sessions suggests the signal filter is working as intended — the system is not forcing trades when setups do not qualify.
Backtest and Cross-Validation
Over 451 days and 137 trades, the strategy returned +7.62% (CAGR 4.19%). Those headline numbers look modest but plausible for a conservative, large-cap approach. The full-period Sharpe of 0.40 and a max drawdown of 14.83% reflect meaningful volatility relative to returns.
The four-fold walk-forward split reveals a more textured picture:
| Fold | Period | Return | Sharpe | Max DD |
|---|---|---|---|---|
| 1 | Aug 2024 – Jan 2025 | +6.53% | 1.25 | 6.16% |
| 2 | Jan – Jul 2025 | −11.42% | −1.70 | 16.09% |
| 3 | Jul – Dec 2025 | +20.68% | 3.86 | 3.26% |
| 4 | Dec 2025 – May 2026 (OOS) | +3.63% | 0.74 | 8.44% |
Three of four folds are positive, and the out-of-sample fold delivered a Sharpe of 0.74 — higher than the full-period figure, which is a constructive data point. Fold 3 was exceptional by any measure. Fold 2, however, is a live-fire stress test: a 16.09% drawdown and a Sharpe of −1.70 during the volatile early-2025 environment demonstrate what happens when a pullback extends into a reversal and the trend-confirmation filter is too slow to react.
Strengths and Risks
Strengths: The thesis is grounded in established technical analysis. OOS performance remained positive with a respectable Sharpe, and the system's increasing selectivity in June — rejecting marginal setups — indicates the entry filter is functioning rather than fishing for trades.
Risks: The validation gate did not pass (PSR 0.702, DSR 0.196). With seven parameter trials over the backtest window, the low Deflated Sharpe Ratio signals a meaningful probability that observed performance reflects curve-fitting rather than genuine edge. A win rate of just 39.4% means the strategy depends structurally on outsized winners; any regime that compresses winners while preserving losers will erode the edge quickly. Annualized turnover of 2,311% also means friction costs compound aggressively at scale, even before slippage.
Bottom line: Channel-pullback shows a coherent thesis and encouraging out-of-sample behaviour, but the failed validation and the severity of Fold 2 argue for continued observation before any scaling decision. The next meaningful test will be how the strategy navigates the next high-volatility window.