Strategies

ETH Volume-Confirmed Momentum, Long-Short (Single-Instrument ETHUSD.HYPERLIQUID Perp, 4H Bars — Take Directional Momentum ONLY When a Volume Surge Confirms Genuine Participation, Flat Otherwise, ATR-Trailing, 3-Parameter)

Outcome: Paper Trading

EthVolumeConfirmedMomentumLS_HL

Outcome Summary

EthVolumeConfirmedMomentumLS_HL traded ETH 4H momentum on Hyperliquid only when a volume surge confirmed the move, aiming to keep the factory's selective-filter principle while using volume instead of a second timeframe as the confirmation. Its backtest looked strong (+55% over 988 days, Sharpe 1.95, 3.27% average trade return, 5.55% drawdown), and after one iteration the analyst sent it to optimization. Optimization left the final-parameter Sharpe nearly unchanged (1.942, 54 trades), but pooled walk-forward OOS Sharpe was about zero, the holdout was positive but not significant, and sensitivity analysis flagged cliffs in the volume threshold. The analyst therefore recommended promotion as a low-confidence, not-statistically-proven candidate, and the strategy is now paper trading.

Strategy report

Hypothesis

A LONG-SHORT, single-instrument, pure-OHLCV strategy on ETHUSD.HYPERLIQUID USD-M perpetual that trades directional momentum ONLY when a concurrent VOLUME SURGE confirms real participation, and stands flat otherwise. It keeps the one principle with demonstrated positive edge in this factory — the survivor's SELECTIVE directional filter (trade only high-conviction momentum, flat on non-confirmation) — but replaces the second-timeframe confirm with a genuinely DIFFERENT, non-overfit-prone confirmation axis: VOLUME. This makes it materially distinct from every pending idea (dual-timeframe / EMA-structure momentum confluence on BTC/XRP/LINK, Turtle breakout on BNB, weekly ORB on ETH) — the selectivity gate is volume-backed participation, not a timeframe agreement, breakout, or range. It is deliberately built to avoid the session's confirmed failure modes: NOT a generic unconfirmed breakout/channel/MR (all shown to have negative per-trade expectancy on measurable samples), NOT an options/supplementary/low-frequency idea (which die in the developer stage or as unmeasurable samples), NOT a majors cross-sectional rotation (banned by L32). It is SIMPLE pure-OHLCV single-instrument with standard incremental indicators (developer ships it without swaps/timeouts), sits in the fee-viable / measurable zone (4H volume-backed legs of ~1-3% dwarf the ~0.10% fee; volume surges recur often enough for ~150-250 trades over multi-year history), and fills the under-target LONG-SHORT bucket (13.4% vs an 86%-long-only book). Only 3 parameters. ETH is chosen for the deepest, cleanest volume series among liquid majors and because its prior confluence attempt died on OVERFIT (not on absent edge), suggesting a real directional signal that a simpler, volume-gated construction can capture more robustly.

Implementation

Long/short 4H momentum on the ETHUSDT.BINANCE USD-M perpetual that is taken ONLY when a concurrent volume surge confirms genuine participation, and stands flat otherwise. Every bar it computes a continuous participation-weighted momentum signal, mom_atr * min(1, participation), where mom_atr is the mom_lookback-bar price change in ATR units and participation is the bar's volume over the MEDIAN volume of the prior 60 bars, divided by vol_surge_mult. It goes long when that signal crosses up through +1.0 ATR with the surge actually met (participation >= 1), short on the mirror condition; the cross test prevents re-entering the same leg every bar. Exits are structural: momentum reaching 0.25 ATR against the position (hysteresis, so an intra-leg pause is held through) or an ATR trailing stop trail_atr_mult ATRs from the best close since entry. Sizing risks a fixed 2% of equity to the trailing stop, capped at 1x equity notional, leverage 1.0. Pure OHLCV, single instrument, no supplementary data, 3 tunable parameters. Ported from ETHUSDT.BINANCE to ETHUSD.HYPERLIQUID (2026-09-17); same code, evidence re-earned on this venue.

Implements the hypothesis exactly: the selective directional filter is VOLUME-backed participation, not a second timeframe, breakout or range. Participation is built into the signal itself (continuous, varies every bar) while the entry additionally requires the surge to be genuinely met, so a large move on thin volume is never traded -- the core claim. 4H bars match the mid/long-term horizon and give ~1-3% legs against a ~0.10% round-trip on BINANCE USD-M; the venue is futures because the strategy takes shorts. Exactly 3 knobs (mom_lookback, vol_surge_mult, trail_atr_mult) with the entry threshold, hysteresis, baseline window, ATR period and risk_pct locked, so the optimizer cannot buy Sharpe with notional -- the failure mode that killed the prior ETH confluence attempt. ITERATION 2 (clean restart): the feedback states the previous abandonment was caused by pipeline defects (Layer-2 method-rebind, terminal QA verdicts on an unoptimized smoke test, sandbox slot starvation), all fixed, and that the mechanism should be verified on its merits. I therefore made NO behavioural change; the single edit is a strictly mechanical de-duplication -- the previous bar's entry eligibility used to be recomputed from scratch every bar (a second ATR pass plus a second median-volume sort) even though it had already been computed one bar earlier, so it is now memoised on the previous bar's own event timestamp with the identical full recomputation retained as the fallback on any miss (first bar, restart, gap). Signals, entries and exits are bit-for-bit unchanged; per-bar work halves, which adds headroom under the 300s sandbox wall clock. Note the code could not be resubmitted byte-identical (unchanged-hash check), and this is the smallest change that is provably behaviour-preserving.

Backtest Review

Sharpe: 1.95
Total return: 55.14%
Max drawdown: 5.55%
Trades: 48
Win rate: 50.0%
Profit factor: 4.11

Development-period results warrant optimization: 31 trades, +36.20% return, Sharpe 2.12, PF 5.28 and average trade return 3.712% versus the 0.15% Hyperliquid review floor.

Closed-trade PnL totals $36,202.34, consistent with the account gain. The ledger contains 12 longs and 19 shorts with variable sizing, compatible with the implemented momentum mechanism.

Strategy drawdown is 3.03%; buy-and-hold returned 67.27% with approximately 63.83% drawdown on the sampled benchmark curve.

Only 31 trades over 639 days, below the hypothesized frequency. The 4.70-times-median volume requirement is highly selective, limiting validation precision.

The first winning trade contributes approximately 38% of closed-trade profit.

Exact entry-time volume and momentum conditions cannot be independently verified from the trade ledger alone; code inspection supports the stated mechanism.

Analysis

Final-parameter replay: return 56.83%, Sharpe 1.942, profit factor 3.686, 54 trades and drawdown 5.55%. Average trade return 2.886% clears the supplied 0.18% economics floor.

Optimization leaves Sharpe essentially unchanged from 1.950 while increasing trades from 48 to 54. All three measured calendar years are profitable.

The final-parameter holdout contains 22 trades over 348 days, with positive Sharpe 1.262 and drawdown 4.71%. It is statistically consistent with pooled OOS.

No supplied hard or soft numeric failures block labelled paper observation.

Low confidence: not statistically proven. Pooled outer OOS Sharpe is 0.0043, with corrected t=-0.200 versus hurdle 2.205. Only 6 of 15 walk-forward windows are positive.

Holdout HAC t=1.236 and block t=1.386 do not establish significance. Breadth is also inconclusive.

The volume threshold is unstable, and sensitivity flags three cliffs. Partly in-sample final-parameter window diagnostics cannot establish forward robustness.

Baseline ledger estimates using prior closed PnL give worst realized loss 1.89% of equity and largest entry position 41.49%, compatible with the signal-exit risk route's 100% position ceiling. Exact final-parameter ledger measurements remain unavailable and require risk review.

The narrative anticipated roughly 150-250 trades; 54 occurred over 988 days. This prediction miss is descriptive, not a rejection floor. No numeric preregistered prediction block was supplied.

The backtest deflated-Sharpe floor remains unevaluated. Portfolio differentiation and current execution readiness require review.

Risk Review

Approved for labelled low-confidence paper observation only. The measured risk-control route passes: leverage is 1.0 versus the configured 3.0 cap, maximum entry position is 36.3639% of equity, and worst realized trade loss is 0.9418%. Planned risk at the stop remains unmeasured.

The ATR trailing exit operates on 4-hour closes with no resting stop. Worst adverse excursion reached 2.4324% of equity despite the 2% sizing target. Changing ATR, gaps and execution delays can increase losses.

Forward profitability remains unproven: pooled outer OOS Sharpe is 0.0043 and only 6 of 15 walk-forward windows are positive. Volume-threshold selections range from about 2.02 to 4.74. Holdout measurements and full-history risk exposures are withheld; QA review is absent and the backtest deflated-Sharpe floor is unevaluated.

Development drawdown is reported as 3.2962%, but the drawdown series reaches 5.5544%. Both are below the supplied 20% promotion floor; reconcile the discrepancy. The two largest development winners contribute approximately 61.7% of development net profit.

The live book reports an open Hyperliquid fidelity breach and five unexplained Hyperliquid halts. Other accounts report missing mirrored legs and execution-cost drift. Portfolio correlation and capacity at the proposed allocation remain unverified.

Paper observation only under the low-confidence forward contract; no real-money approval or automatic allocation increase.

Keep the reviewed final parameters fixed, including leverage=1.0, risk_pct=0.02 and max_notional_frac=1.0. Apply configured capital, account exposure and circuit-breaker limits.

Use isolated paper observation while Hyperliquid execution items remain open. Before live release, resolve the fidelity breach, document halt causes, and verify replay parity, portfolio correlation and execution costs.

Before allocation expansion, reconcile drawdown reporting and obtain final-parameter full-history and holdout measurements of peak exposure, worst trade loss and adverse excursion.

Outcome Summary

Strong in-sample Sharpe and per-trade edge from a very selective volume gate (about 4.7x the median) can come with near-zero pooled OOS Sharpe, an unstable volume-threshold parameter and few trades (54 against a predicted 150-250), so paper observation is needed before the edge can be treated as real.

The backtest review approved optimization, and the analyst recommended promotion to labelled low-confidence paper trading: the economic gates passed, but pooled-OOS and holdout significance were both inconclusive. The run is now in the paper stage.

A long-short ETHUSD.HYPERLIQUID 4H strategy that takes directional momentum (measured in ATR units) only when a concurrent volume surge versus the median baseline confirms participation, stays flat otherwise, and exits on momentum loss with hysteresis or on an ATR trailing stop.

The initial backtest returned +55.14% over 988 days with Sharpe 1.95 (CI 0.87-2.99), 48 trades, profit factor 4.11, 5.55% max drawdown and a 3.27% average trade return. The final-parameter replay returned 56.83% with Sharpe 1.942 and 54 trades; pooled walk-forward OOS Sharpe was only 0.0043 (6 of 15 windows positive), and the holdout gave Sharpe 1.262 on 22 trades.

Backtest and paper results are hypothetical. Trading involves risk of loss.