Skip to content

View original

InjFourHourVolumeBreakoutLong

Hypotheses

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

Hypotheses

Long-only INJ (Injective) 4-hour strategy on BINANCE USD-M futures. Direct analog of the landed XRP (Sharpe 1.84, paper_stage) / BTC / ETH / SOL / BNB / AVAX / ARB / OP FourHourVolumeBreakoutLong family — applied to INJ. The 2.0x volume surge trigger produces 30-60 trades/year (statistically valid for walk-forward), explicitly avoiding the low-frequency channel-breakout pathology that killed ADA daily. Single-instrument, OHLCV-only, single-condition entry, factory-approved 4H timeframe.

Hypotheses

Iteration 2 fixes the Layer 2 frozen-signal failure. Previously calculate_signal() returned 0.0 on every bar unless the full three-part entry condition fired, so on smooth synthetic scenarios (steady_uptrend/downtrend, flat_ranging, zero_volume, price_gap) the signal was constant 0.0. It now returns a CONTINUOUS conviction score every bar = trailing volume ratio + normalized breakout margin, which varies bar-to-bar in all scenarios (the price-derived breakout margin keeps it moving even under zero volume, where vol_ratio is defined as 0.0). The exact discrete entry gate (surge AND breakout AND bullish) is still evaluated in calculate_signal and stored in self._long_ok, which should_enter() and position_size() read — preserving the identical entry semantics that already passed Layer 1. Imports, class structure, exit logic and sizing (genuine 2.0x leverage consumption) are unchanged, so earlier passing layers do not regress.

Hypotheses

Functioning volume-breakout strategy with NO edge on INJ — Sharpe -2.57, profit_factor 0.71, negative expectancy (-384), total return -57.8%, negative across most of the 2022-2026 sample, and an 11% commission-of-gross fee drag from 4h-average churn. This is asset-selectivity, not a code bug (nothing to iterate) and not a tunable (a Sharpe of -2.57 with PF<1 across 4 years is an absence of edge that parameter optimization cannot manufacture — the optimizer would only overfit noise and fail walk-forward/holdout). The FourHourVolumeBreakoutLong family is landed on XRP/BTC/ETH/SOL/BNB/AVAX/ARB/OP; INJ simply does not exhibit a volume-confirmed breakout-continuation edge and the fast-churn fees dominate. Not worth 2 hours of optimization compute.

Implementation

Long-only INJ (INJUSDT) 4-hour volume-surge breakout on Binance USD-M futures at 2.0x leverage. Enters long when the current 4H bar's volume exceeds vol_mult x trailing average volume, the close breaks above the prior N-bar high, and the bar is green. Exits on take-profit, stop-loss, or a max-hold time stop. Pure OHLCV, single-instrument, single-condition entry.

Verification Results

Verification failed (Layer 2 — synthetic scenarios): Parameters used: ['vol_mult', 'min_notional', 'position_pct', 'vol_lookback', 'max_hold_bars', 'stop_loss_pct', 'take_profit_pct', 'breakout_lookback'] Check that __init__ sets all attributes from self.parameters.get(). - steady_uptrend: Frozen signal: all 476 signals are identical (value=0.0). calculate_signal() likely has a bug — the signal never varies. - steady_downtrend: Frozen signal: all 476 signals are identical (value=0.0). calculate_signal() likely has a bug — the signal never varies. - flat_ranging: Frozen signal: all 476 signals are identical (value=0.0). calculate_signal() likely has a bug — the signal never varies. - zero_volume: Frozen signal: all 476 signals are identical (value=0.0). calculate_signal() likely has a bug — the signal never varies. - price_gap: Frozen signal: all 476 signals are identical (value=0.0). calculate_signal() likely has a bug — the signal never varies.

Backtest Review

Strategy functions correctly: 145 trades signaled and submitted, no drops, no sizing/MTM artifacts, leverage sized as designed (~1.2x).

Backtest Review

Mechanism is proven — the FourHourVolumeBreakoutLong family is landed on 8 other assets (XRP paper_stage at Sharpe 1.84).

Backtest Review

No edge on INJ and strongly negative: Sharpe -2.57, Sortino -3.58, probabilistic_sharpe 0.034, total_return -57.8%, max_drawdown 61.4%.

Backtest Review

profit_factor 0.71 (<1), expectancy -384 (negative) — the signal loses money on average.

Backtest Review

Negative in the bulk of the sample (2023 -35.9%, 2024 -6.4%, 2026 -35.8%); the two positive years are small.

Backtest Review

Severe fee drag: commission_pct_of_gross 11.03% with 4h16m average holds — rapid churn on a fee-magnet trigger with no compensating edge.

Outcome Summary

InjFourHourVolumeBreakoutLong applied the proven FourHourVolumeBreakoutLong family — a volume-surge, breakout, green-bar long — to INJ 4H futures, expecting the ~30-60 trades/year to support valid walk-forward. It ran correctly (145 trades, no drops, leverage sized as designed) but had no edge on INJ: -57.8% return, Sharpe -2.57, profit factor 0.71, negative expectancy, a 61.4% drawdown, and negative years across most of 2022–2026, with an 11% commission-of-gross drag from rapid ~4-hour churn. The reviewer abandoned it at the backtest-review gate without optimization, judging a deeply negative Sharpe with PF below 1 across four years an absence of edge that tuning could only overfit. The decisive lesson was asset-selectivity: INJ simply does not exhibit the volume-confirmed breakout-continuation edge the family captures elsewhere, and its fast-churn fees dominate.

Outcome Summary

A breakout mechanism landed on eight other assets (XRP reached paper_stage at Sharpe 1.84) does not transfer to INJ — this is asset selectivity, and a high-frequency volume trigger with no compensating edge lets fee drag (here 11% of gross) dominate.

Outcome Summary

The backtest-review verdict was 'abandon' before optimization ran (optimization_report is null): the strategy executed correctly but had no edge on INJ — strongly negative across most of the 2022–2026 sample, with fast-churn fees compounding the loss, and a Sharpe of -2.57 with PF<1 cannot be recovered by parameter tuning.

Outcome Summary

A long-only INJ 4H strategy on BINANCE USD-M futures (leverage 2.0) that enters on a single-condition volume-confirmed breakout — current bar volume ≥ 2.0x its trailing average, close breaking above the prior 10-bar high, and a green bar — porting the landed FourHourVolumeBreakoutLong family to INJ.

Outcome Summary

The backtest took 145 long trades and lost heavily: -57.8% total return, Sharpe -2.57, Sortino -3.58, profit factor 0.71, expectancy -384 per trade, 46.2% win rate, 61.4% max drawdown, probabilistic Sharpe 0.034, and an 11.03% commission-of-gross fee drag from ~4h16m average holds.
Strategy report

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