The thesis
bollinger-reversion runs a classic mean-reversion play across a 24-name large-cap universe — AAPL, MSFT, NVDA, JPM, JNJ, WMT, XOM and the like. The rule is simple: buy when price closes below the lower Bollinger band, sell when it pushes above the upper band. It is a bet that stretched moves in liquid megacaps snap back, and simplicity is a genuine virtue here — there are few knobs to overfit, and the logic is easy to reason about.
Recent activity
The strategy is live and trading on schedule, but execution has been thin. Across the last six scheduled runs (16–23 Sep), it executed only on two days — two fills on the 22nd and two on the 18th — and logged at least one rejected order on every run. That is expected behaviour for a band-triggered system: when nothing is stretched far enough, or risk gates bite, it simply waits.
The fills it did take are on-thesis. On 18 Sep it sold GOOGL (6 @ $351.91) and rotated into NKE (70 @ $35.64); on 22 Sep it sold WMT (21 @ $109.83) and bought MCD (9 @ $250.63). Selling extended winners and buying beaten-down names is exactly what the description promises.
The uncomfortable number is the balance. As of 23 Sep the account totalled $9,654.21 with $2,398.39 in cash — below the $10,000 the backtest starts from. In paper trading, so far, this strategy is underwater.
Backtest vs. validation
On paper history the numbers look inviting: +17.55% over 451 days, a 9.46% CAGR, a 63.89% win rate over 76 trades, and a Sharpe of 0.66. The max drawdown of 20.57% is steep but not disqualifying.
Validation tells a harsher story, and it did not pass. All four walk-forward folds are positive, but the returns decay monotonically: 7.67% → 1.42% → 0.77% → 0.40%, with fold Sharpe collapsing from 1.78 to 0.14. Almost the entire edge lives in the first, oldest fold. The out-of-sample return of just 0.4% and OOS Sharpe of 0.14 confirm it. The Probabilistic Sharpe Ratio of 0.814 looks reassuring until you deflate it for the six trials run: the Deflated Sharpe of 0.342 says we cannot confidently reject luck.
Strengths and risks
Strengths: a transparent, low-parameter thesis; a high win rate; and disciplined, on-thesis execution.
Risks: the headline return is front-loaded and does not reproduce out-of-sample; turnover of 1,721% means it churns hard for a shrinking edge; and the live account is currently below cost. Until the OOS Sharpe recovers, this reads as a decaying edge, not a durable one — worth watching, not scaling.