EthDailyVolumeSurgeBullishLong

ETH Daily Volume Surge Bullish Bar Long-Only with Trend Filter (BINANCE USD-M Futures, Daily, OHLCV-Only)

Long-only ETH daily strategy on BINANCE USD-M futures. Enter when a daily volume surge confirms a bullish bar in an established uptrend. Mechanism is a direct analog of the BTC Daily Volume Surge strategy already in pipeline, applied to ETH to extend mechanism diversity to a second core instrument. OHLCV-only, single-instrument, single-condition entry, hand-tunable defaults to survive walk-forward.

Iteration 2 addresses the sole backtest-review issue: qualifying-trade density was too low (42 trades, ~7/yr) for a 225-trial optimizer to survive the deflated-Sharpe gate on an outlier-heavy sample. Per the analyst's named levers, I made the smallest change that raises density while preserving the volume-surge bullish-breakout mechanism exactly: volume_multiplier 1.5->1.3 and sma_lookback 100->50, with sma_break_pct kept at 0. Both loosen the two most restrictive AND-gates (surge threshold and trend lookback), so more bullish-breakout events qualify without altering the edge. Signal logic, gating structure, sizing, exits, and imports are untouched, so all previously-passing verification layers remain green. The validated ADA volume-absorption analog reached 133 trades at Sharpe 1.46 on a comparable loosening, indicating a denser ETH config >60 trades is achievable; if the re-backtest still cannot exceed ~60 trades or stays outlier-dominated, the mechanism is too sparse on ETH daily and should be abandoned rather than optimized.

The ETH volume-surge port is too weak to justify optimization and does not transfer from its validated ADA analog. Base Sharpe is 0.29 (CAGR 2.2%, total return +32% over 6.5y) with NEGATIVE alpha (-0.015 vs buy-hold) and profit_factor 1.08 — roughly 5x weaker than the ADA analog (Sharpe 1.46) this hypothesis is modeled on, the signature of a mechanism that does not transfer to this instrument. The edge is entirely outlier-carried (return_kurtosis 79.7, skew 7.5, tail_ratio 0.0033; the whole result rests on ~4 huge up days such as 2020-08-07 +26%, 2021-01-11 +26.4%, 2024-03-15 +27.3%), and the Sharpe CI straddles zero heavily (sharpe_ci_low -1.259, probabilistic_sharpe 0.686), so there is no distinguishable base edge for a 225-trial deflation to preserve. Not iterate: density was already the iteration-1 fix and is now adequate (71 trades), so the residual weakness is the mechanism failing to generalize to ETH, not a density/code lever the developer can change. Best-of-225 optimization on 71 outlier-dominated trades would only curve-fit the tail days and fail the deflated-Sharpe gate, so the 2-hour optimization is not warranted.

Long-only ETH daily (1D) strategy on BINANCE USD-M futures. Enters LONG when a daily volume surge (volume > 1.3x trailing 20-bar average) confirms a bullish bar (close > open) in an established uptrend (close > 50-bar SMA). Exits via a 5% trailing stop from the highest close since entry. Pure OHLCV, single-instrument, leverage 1.0, sized at 50% of equity notional per position.

Iter-1 density problem fixed: 71 trades (~11/yr), clean entry diagnostics (71 signaled -> 71 submitted, 0 dropped), long-only trend-gated as the hypothesis describes

Engine-clean single-instrument daily config; positive total return (+32%) and modest drawdown vs many siblings

Edge is very weak and does not transfer to ETH: Sharpe 0.29, CAGR 2.2%, negative alpha -0.015, profit_factor 1.08 (fees eat nearly all edge), win_rate 29.6% — ~5x weaker than the validated ADA analog (Sharpe 1.46) it copies

Entirely outlier-carried: return_kurtosis 79.7, skew 7.5, tail_ratio 0.0033, with the result resting on a handful of +22-27% single days — remove them and the edge is negative

No distinguishable base edge: sharpe_ci_low -1.259 (CI straddles zero), probabilistic_sharpe 0.686 — a profile that will not survive a 225-trial deflated-Sharpe gate, only curve-fit the outliers

Density was already the iter-1 lever (loosened this iteration); the remaining weakness is mechanism-doesn't-transfer, not a tunable, so there is no concrete iterate fix

Do not optimize yet — the sample is too thin. The backtest produced only 42 trades (~7/yr), below the ~62 walk-forward floor, because the triple entry gate (volume_multiplier 1.5 AND bullish bar AND close > SMA-100) fires too rarely. A 225-trial optimization on ~42 outlier-leaning events (kurtosis 59.6; headline carried by 2021-01 +25%, 2025-08 +25%, 2024-03 +20%) will very likely fail the deflated-Sharpe gate — the structurally-identical 37-trade DOGE daily breakout came back DSR 0.0, and I have flagged the same on the BNB/BTC-TSM/ETH-TSM thin samples this batch. FIX — increase qualifying-trade density to >60-100 while keeping the volume-surge bullish-breakout mechanism: lower volume_multiplier (test ~1.25-1.35), and/or shorten the trend SMA (test ~50 instead of 100) and/or keep sma_break_pct at 0. The validated analog (ADA volume-absorption) had 133 trades at Sharpe 1.46, so a denser ETH config is achievable. Re-backtest and confirm trade_count > 60 AND the edge is not reducible to a few outlier days (check return_kurtosis drops and the result survives removing the top 1-2 days) AND a positive sharpe_ci_low before resubmitting for optimization. If a denser config cannot exceed ~60 trades or stays outlier-dominated, abandon — ETH daily volume-surge is then too sparse to validate.

This strategy attempted to extend a volume-surge bullish-breakout mechanism — already validated on ADA at Sharpe 1.46 — onto ETH for mechanism diversity, and a second iteration loosened the volume and SMA gates to raise trade density from ~42 to an adequate 71 trades. But even with the density problem fixed, the ported edge simply did not transfer: Sharpe fell to 0.29 with negative alpha, a profit factor of 1.08 that fees nearly consumed, and returns resting on about four huge single up-days. Because density had been the only real lever and the remaining weakness was the mechanism not generalizing to ETH, the reviewer abandoned it at the backtest-review gate rather than spend optimization compute curve-fitting the outliers. It never reached the optimization or analyst stages, and the validated ADA analog remains the configuration where this mechanism clears the bar.

A volume-surge continuation mechanism validated on one instrument (ADA) does not automatically transfer to another core asset (ETH); an outlier-carried edge with negative alpha and a Sharpe CI through zero should be rejected before optimization, since best-of-N tuning would only curve-fit the handful of tail days.

It was abandoned at the pre-optimization BACKTEST_REVIEW gate: the analyst judged the edge roughly 5x weaker than its validated ADA analog with no distinguishable base edge to preserve, and since the density lever had already been spent in iteration 1, the residual weakness was the mechanism failing to transfer to ETH rather than a tunable — so the 2-hour optimization was not warranted.

A long-only, single-instrument daily strategy on ETHUSDT Binance USD-M futures that enters when a volume surge confirms a bullish bar in an established uptrend (above SMA), exiting via a percentage trailing stop — a direct ETH port of a volume-surge mechanism validated on ADA (Sharpe 1.46) and in-pipeline on BTC.

The backtest produced only a weak edge — Sharpe 0.287, CAGR 2.2%, +32.2% total return over ~6.5 years, negative alpha (-0.015), profit factor 1.08, and a 29.6% win rate across 71 long trades — with the result entirely outlier-carried (kurtosis 79.7, skew 7.5, tail_ratio 0.0033) and a Sharpe CI straddling zero (sharpe_ci_low -1.259).