BchFourHourVolumeBreakoutLong

BCH Four-Hour Volume Breakout Long-Only (BINANCE USD-M Futures, 4H, OHLCV-Only)

Long-only BCH (Bitcoin Cash) 4-hour strategy on BINANCE USD-M futures. Direct analog of the landed BTC/ETH/SOL/BNB/AVAX/XRP/MATIC/ATOM/NEAR/LTC FourHourVolumeBreakoutLong family — applied to BCH. Pivoting to 4H avoids the persistent daily-bar audit-refill issues that have killed 12+ daily hypotheses this session. Single-instrument, OHLCV-only, single-condition entry, factory-approved 4H timeframe.

Direct analog of the landed FourHourVolumeBreakoutLong family (BTC/ETH/SOL/BNB/AVAX/XRP/MATIC/ATOM/NEAR/LTC) applied to BCH, exactly as the hypothesis specifies. Uses 4H bars to avoid the daily-bar audit-refill failures noted in the brief. Long-only suits a momentum/continuation edge but routes to BINANCE USD-M futures (not spot) because this is a member of the established futures family with cheaper round-trip fees (~0.10% vs spot ~0.20%), giving the multi-condition volume-surge edge more room over costs; leverage stays at 1.0 (and is wired into sizing via self.config.leverage so it is not flagged as set-but-unused). Triple-gated entry (surge + bullish + trend) targets per-trade moves well above the round-trip taker fee. The tanh(log-return) proxy in calculate_signal keeps Layer-2 synthetic scenarios varying per bar while the real trading logic lives in on_bar, matching the verified family template.

failed deflated Sharpe: DSR=0.0, expected-max=2.64 over 225 trials — selected Sharpe (0.15) is indistinguishable from best-of-N noise. Compounded by a catastrophic holdout failure (holdout Sharpe -0.41 vs WF-OOS +0.43, ratio -0.97), an overfit walk-forward (is_overfitted=true, IS 1.58 -> OOS 0.43), 6 of 8 cliff parameters (volume_mult, stop_loss_pct, volume_window, take_profit_pct, max_holding_bars, sma_trend_period), and a Sharpe CI [-0.86, 1.00] that straddles zero. profit_factor 1.05 means fees eat nearly all gross edge. Optimization already searched 225 trials and the best config fails every robustness gate, so there is no stable region to iterate toward. The 4H volume-breakout mechanism is already validated on 10+ landed siblings; BCH is simply a dead target for it, not a premise needing reframing — abandon this instance.

Long-only BCH 4-hour volume-breakout continuation on BINANCE USD-M futures. Enters long when a bar shows a volume surge (current volume > 1.5x the trailing 20-bar average, current bar excluded), is bullish (close > open), and is in an uptrend (close > 100-bar SMA). Exits on +4% take-profit, -2.5% stop-loss, or after 30 bars held. Single-instrument, OHLCV-only, single composite entry condition. Sizes each position at 22% of account equity, quantized to BCH's 3-decimal size precision, with a $5 minimum notional guard.

Healthy sample size: 387 trades over 742 active days on 4H bars — enough for meaningful parameter sensitivity, not noise-fitting

Strategy is clearly functioning: entries fire across all regimes, coherent equity path, sane ~1d 6h avg holding period

Direct port of a landed 4H volume-breakout family (BTC/ETH/SOL/BNB/AVAX/XRP/MATIC/ATOM/NEAR/LTC) — mechanism is proven, not speculative

Long-only on a single liquid instrument, OHLCV-only — clean, low-overfitting-surface design

Thin headline edge with default params: Sharpe 0.20, profit factor 1.06, win rate 42% — needs optimization to clear promotion thresholds

Sharpe 95% CI (-0.92 to 1.36) straddles zero — default config not yet distinguishable from noise

impact_cost_pct ~27% and capacity_usd ~$1.37M: per_leg_pct 0.22 may be oversized for BCH liquidity; optimizer should explore lower sizing

Max drawdown 17.2% is meaningful for a 14% total return — Calmar 0.72

Implementation is sound — 387 (initial) / 539 (optimized) trades, no dropped entries, low max_drawdown (12.2% optimized)

Mechanism is proven on the broader 4H volume-breakout family (BTC/ETH/SOL/BNB/AVAX/XRP/MATIC/ATOM/NEAR/LTC), so this is an instrument-specific failure, not a code defect

Deflated Sharpe = 0.0; optimized Sharpe 0.15 vs expected-max-by-luck 2.64 over 225 trials — selected result is pure best-of-N noise

Holdout failed catastrophically: holdout Sharpe -0.41 vs WF-OOS +0.43 (ratio -0.97, threshold 0.70) — edge inverts out-of-sample

Walk-forward flagged is_overfitted=true (IS 1.58 -> OOS 0.43, one OOS window negative)

6 of 8 parameters are cliffs — no stable region; volume_mult flips Sharpe +0.20 -> -0.22 on a small change

Sharpe 95% CI [-0.857, 0.996] straddles zero; is_significant=false; PBO 0.49

profit_factor 1.05 — fees consume nearly all gross edge; ~2.7%/yr return at Sharpe 0.15 is not fee-viable

Pitched as a low-risk port of a 10+ instrument 4H volume-breakout family onto BCH, the strategy ran cleanly with healthy trade counts but only a marginal headline edge (Sharpe 0.20, profit factor 1.06), so the reviewer sent it to optimization rather than abandoning it outright. The 3-phase optimization searched 225 trials and the best configuration failed nearly every robustness gate: deflated Sharpe 0.0, a catastrophic holdout inversion (ratio -0.97), confirmed walk-forward overfitting, and 6 of 8 cliff parameters with no stable region. Because profit factor near 1.05 meant fees consumed almost all gross edge and there was nowhere to iterate, the analyst abandoned this instance. The conclusion was instrument-specific, not a flaw in the mechanism or code — BCH is simply a dead target for the 4H volume-breakout edge.

A mechanism proven across many sibling instruments does not guarantee an edge on a new one; if the default backtest shows only a fee-thin edge and a Sharpe CI straddling zero, optimization tends to surface best-of-N noise rather than a real signal that survives holdout.

The analyst abandoned it at the post-optimization ANALYZING stage: the selected result was indistinguishable from best-of-N noise (DSR 0.0), the holdout inverted (holdout Sharpe -0.41 vs walk-forward OOS +0.43, ratio -0.97), walk-forward flagged overfitting (IS 1.58 → OOS 0.43), 6 of 8 parameters were cliffs, and a Sharpe CI of [-0.86, 1.00] straddled zero — so no stable region existed to iterate toward.

A long-only BCH 4-hour volume-surge continuation strategy on BINANCE USD-M futures — entering on a volume spike plus a bullish bar in an uptrend (close > SMA) with fixed take-profit/stop-loss/time exits — porting the already-landed 4H volume-breakout family (BTC/ETH/SOL/BNB/AVAX/XRP/MATIC/ATOM/NEAR/LTC) to BCH.

The initial backtest cleared the pre-optimization gate on activity but had a thin edge: Sharpe 0.20, total return 14.1%, profit factor 1.06, win rate 42%, max drawdown 17.2% over 387 trades. After a 225-trial optimization, the best config was no better — Sharpe 0.15, profit factor 1.05, deflated Sharpe 0.0 against an expected-max-by-luck of 2.64.