The thesis
channel-pullback is a trend-continuation strategy dressed as a dip-buyer. It waits for a confirmed uptrend, then buys pullbacks into the lower regression channel or volume support, aiming to exit at the upper channel or overhead resistance. In plain terms: it doesn't try to catch falling knives, it tries to buy temporary weakness inside strength. The tradable universe is a broad 24-name slice of US large caps — mega-cap tech (AAPL, MSFT, NVDA), financials (JPM, BAC, V, MA), healthcare (JNJ, UNH, PFE) and staples (PG, KO, WMT).
Recent activity
The strategy is running live and trades on a scheduled daily cadence. Over the last two weeks it has been actively rotating a small book: a 30-share BAC round-trip (bought 09-03 at $63.04, sold 09-09 at $62.79 — a small loss), an earlier BAC round-trip in early September, several short-hold AAPL trades in late August, and a fresh 5-share V buy on 09-10 at $366.62.
The honest read on the paper account is that it has been leaking. Total equity slipped from $9,618 on 09-03 to $9,403 on 09-10, and the scheduled runs frequently report zero executions — the entry conditions simply aren't triggering most days. That is consistent with a selective pullback filter, but it also means the book spends stretches near-idle while capital sits in cash.
Backtest and validation
The headline backtest covers 451 days: a 7.62% total return (roughly 4.19% CAGR), a Sharpe of 0.40, and a 14.8% max drawdown. The win rate is a low 39.4% across 137 trades, meaning the strategy relies on its winners meaningfully outrunning its losers — a fragile profile if the average win compresses. Turnover is extreme at ~2,300%, so fee and slippage assumptions matter a lot here.
Most importantly, the strategy did not pass validation. Walk-forward testing across four folds tells a bumpy story: fold 1 (+6.5%) and fold 3 (+20.7%, Sharpe 3.86) were strong, but fold 2 was ugly (−11.4%, Sharpe −1.7). Three of four folds were positive, and out-of-sample return was a thinner +3.63% at a respectable 0.74 Sharpe. The deflated Sharpe ratio (DSR) of 0.196 — adjusted for the 7 configurations trialled — is the tell: once you account for selection effort, the edge is not statistically convincing, even though the probabilistic Sharpe (PSR 0.70) looks superficially healthy.
Verdict
channel-pullback has a coherent thesis and one genuinely excellent fold, which is why it earned a live slot. But the evidence is mixed: a losing fold that erased more than a full fold of gains, a sub-40% hit rate, punishing turnover, and a validation gate it failed on deflated-Sharpe grounds. Treat it as a promising candidate under observation, not a proven edge. The next month of live trades — and whether that V position and the AAPL rotations translate into realized wins — will say more than the backtest can.