The thesis
channel-pullback is a disciplined "buy the dip, but only in a confirmed uptrend" strategy. It waits for price to pull back to the lower regression channel or a volume-support level inside an established uptrend, then buys, aiming to exit at the upper channel or overhead resistance. It trades a universe of 24 US large caps spanning tech, financials, healthcare, staples, and energy — AAPL, MSFT, NVDA, JPM, V, UNH, XOM, CVX and peers. It is live and running scheduled daily sessions.
Recent activity
The last week of runs has been orderly: every scheduled session executed cleanly with zero rejections. On 2026-11-04 it rotated into energy (XOM at 168.81, CVX at 216.40) and re-entered AAPL and MSFT on subsequent pullbacks. By 2026-11-11 it was trimming — selling 5 AAPL at 354.62 and 11 XOM at 165.94 — consistent with the "exit at the upper channel" rule after AAPL rallied from its ~346 entry.
Worth noting: the live book sits around $9,856 total equity against a $10,000 seed. In other words, since inception the live account is modestly underwater, even as the strategy manages risk sensibly trade-to-trade.
Backtest and validation
The full backtest over 451 days returned 7.62% (CAGR 4.19%), ending at $10,761 with a Sharpe of 0.40 and a 14.83% max drawdown. The win rate is a low 39.4% across 137 trades — this is a strategy that loses more often than it wins and relies on its winners being larger than its losers, which is characteristic of pullback/trend hybrids. Turnover is extremely high at 2,311%, so execution quality and fees matter; recorded fees were a modest $137 with no FX cost.
Walk-forward testing tells the real story. Three of four folds were positive, but the dispersion is severe:
- Fold 1: +6.53% (Sharpe 1.25)
- Fold 2: −11.42% (Sharpe −1.70)
- Fold 3: +20.68% (Sharpe 3.86)
- Fold 4: +3.63% (Sharpe 0.74)
Out-of-sample return was 3.63% with a 0.74 Sharpe — respectable, but the single catastrophic fold shows the strategy can break badly when uptrends fail and pullbacks keep falling. Our validation did not pass. The probabilistic Sharpe (PSR 0.702) is borderline, but the deflated Sharpe ratio of 0.196 — which penalizes for the 7 trials run — is the killer. Adjusted for how many variants we tried, we cannot confidently distinguish this edge from luck.
Verdict
Strengths: a coherent, rules-based thesis, clean execution, and genuine positive expectancy in three of four periods. Risks: a sub-40% win rate, Fold 2's −11% blow-up, heavy turnover, and — most importantly — a DSR that flags likely overfitting. The live drawdown reinforces caution. channel-pullback is worth keeping on the leaderboard as a live experiment, but it has not earned a capital-allocation promotion until it clears the validation gate on an overfitting-adjusted basis.