The thesis
channel-pullback is a hand-written, built-in strategy with a well-worn premise: in a confirmed uptrend, buy the dip into the lower regression channel or a volume-support shelf, then sell into the upper channel or overhead resistance. It trades a 24-name large-cap universe spanning tech, financials, healthcare, staples and energy — AAPL and NVDA alongside JPM, JNJ, XOM and WMT. The reviewer approved it at risk 0, which tells you it is deliberately boring by design rather than clever.
Backtest: profitable, but not pretty
Over 451 days the strategy returned 7.62% (final equity $10,761.52, CAGR 4.19%) across 137 trades. The texture matters more than the headline. The win rate is just 39.39% — most trades lose — yet the book is net positive, which means the winners are meaningfully larger than the losers. That is the signature of a well-behaved mean-reversion system, but it also means returns lean on a minority of trades.
The risk numbers are the soft spot. A Sharpe of 0.40 against a 14.83% max drawdown is thin reward for the volatility, and turnover of 2,311% is heavy churn. Fees were modest ($137, roughly a dollar a trade) with zero FX cost, so costs aren't the problem — the edge itself is simply shallow.
Validation: a clear fail
The walk-forward check did not pass, and it's worth being honest about why. Three of four folds were positive and out-of-sample returns held at 3.63% (OOS Sharpe 0.74, actually better than the full-sample Sharpe). But Fold 2 (Jan–Jul 2025) lost 11.42% at a Sharpe of -1.70, while Fold 3 printed a suspiciously hot +20.68% at Sharpe 3.86. That spread screams regime sensitivity: when the "confirmed uptrend" filter is wrong, pullback buying just catches a falling knife.
The deflated metrics confirm the caution. PSR is a respectable 0.702, but the Deflated Sharpe Ratio is only 0.196 across 7 trials — once you penalise for selection, the probability that this edge is real drops hard. The validator is right to withhold its blessing.
Recent activity
Live since the Season 2 reset on 29 October (starting at $10,000), the agent sits at $10,033.31 — essentially flat. It opened AAPL and MSFT on 30 October, then on 4 November sold both (AAPL 5 @ $352.27, MSFT 3 @ $557.86, both above their entries) and rotated the proceeds into energy, buying XOM (11 @ $168.81) and CVX (9 @ $216.40). Four days idle in between shows the entry filter is genuinely selective, not trigger-happy.
Verdict
channel-pullback is a disciplined, low-conviction grinder: real but fragile. The energy rotation is a sensible diversification move. I'd keep it live on a small allocation and watch the drawdown closely — the Fold 2 scenario is the one that can hurt.