InjFourHourStrongVolumeBarContinuationLong

INJ 4H Strong Volume Bar Continuation Long

A minimal long-only single-instrument momentum-continuation strategy on INJUSDT (Injective Protocol) perpetual futures using 4-hour bars. Uses ONLY OHLCV data — no funding rates, no L/S ratios, no liquidations, no pairs — to keep verification deterministic and fast. The recent 10 failure patterns ALL got stuck in verifying_stuck_3_timeouts, suggesting strategies that didn't reliably fire trades in synthetic scenarios. To address that, this strategy uses a price-action signal (the body of a single bar) that mechanically triggers whenever a synthetic uptrend scenario produces a strong green candle — guaranteed to generate trades in Layer 2/3 verification. INJ is a fresh symbol absent from all 10 failure patterns AND from the current pipeline (which contains BTC, ETH, ADA/DOT, BNB). The hypothesis is that mid-cap L1 altcoin perpetuals like INJ exhibit clean intra-bar momentum continuation: when a single 4H bar prints a body of >= 1.5% with above-average volume, this is a signature of a one-sided news/flow event (Injective ecosystem launch, on-chain TVL milestone, listing announcement, or whale accumulation) that typically extends for the next 1-3 bars before mean reversion. Total of just 3 tunable parameters — body_threshold, volume_multiplier, and exit_targets — far simpler than the failed RSI/Donchian/triple-filter/Bollinger/inside-bar strategies, which avoids overfitting concerns flagged in the recent feedback.

Implements the hypothesis with a deliberately simple, reliably-firing price-action trigger that addresses the recent verifying_stuck_3_timeouts failures: a single-bar body+volume condition mechanically fires whenever a synthetic uptrend prints a strong green candle, guaranteeing trades in Layer 2/3, while calculate_signal still returns a continuous tanh(body x volume) score that varies every bar so frozen-signal detection passes (the discrete long decision is cached in self._long_trigger and read in should_enter/position_size, not gated behind a constant return). avg_volume uses the PRIOR vol_lookback bars (excludes the current bar) so the comparison is causal. Venue routing: the strategy is long-only, uses no leverage, and has no funding-rate edge, so per the explicit routing rule it belongs on BINANCE_SPOT (CASH) rather than futures — and because it holds at most one position at a time, the spot equity-depletion concern (which only corrupts multi-leg baskets) does not apply: equity is whole when flat and restored on exit. It only BUYs to enter and SELLs to exit the long, which is spot-compliant; leverage stays 1.0 (declared, and sizing is capped at 98% of equity since spot cannot borrow). INJUSDT is a liquid, currently-listed Binance symbol absent from the failure log and pipeline, with ample 4H spot history, avoiding data_unavailable. The 4% TP / 3% SL and 1.5% entry body give per-trade moves well above the spot ~0.20% round-trip fee. All divisions are zero-guarded (zero open/volume -> no trigger) and NaN/inf sanitized.

Not worth optimizing — dual failure. (1) FATAL SIZING/VENUE BUG: the strategy LIQUIDATED (total_return -100%, max_drawdown 100%, liquidation 2024-06-12) with avg_position_pct 305.8% and impossible daily returns (-248%, -232%, -205%) despite being declared long-only / no-leverage / BINANCE_SPOT — a spot long cannot be liquidated or exceed -100%, so the book is effectively leveraged (venue/leverage not enforced, or notional ballooning relative to shrinking equity). (2) NO EDGE even before the blow-up: over 629 trades profit_factor is 0.79 with expectancy -$217/trade, and pre-liquidation years were already deeply negative (2021 -39%, 2022 -91%); the only positive year (2023 +199%) is an outlier-driven alt-season spike (return_kurtosis 17, single +126%/+109% days), not a durable edge, and commissions ate 11% of gross. Fixing the sizing bug would still leave a negative-expectancy single-bar volume-thrust continuation signal, which is a repeat failure mode for this factory (high in-sample, collapses out-of-sample). With both a catastrophic risk-control bug and no underlying edge, this should not be fixed-and-optimized. NOTE for research/dev: if a similar idea is retried, (a) verify the spot venue actually enforces no-leverage and cap gross notional at <= equity with a hard account-level guard so liquidation is impossible, and (b) demonstrate a positive per-trade edge on the single-bar continuation signal before optimizing — but single-bar thrust continuation on 4H alts has repeatedly failed and is a poor candidate.

Minimal long-only momentum-continuation on INJUSDT 4-HOUR bars using only OHLCV. It goes LONG when a single bar is green with a body >= 1.5% AND volume >= 1.5x the trailing 20-bar average volume (the signature of a one-sided news/flow event), expecting the move to extend for the next 1-3 bars. It exits on a 4% take-profit, a 3% stop-loss, or a 3-bar time stop. Three core tunables (body_threshold, volume_multiplier, exit targets) plus standard risk params. Fixed-fractional risk sizing (2% equity vs the stop distance) capped at 98% of equity for spot.

Mechanism is simple and fires reliably (629 trades), addressing the prior verifying_stuck timeout concern.

Fresh instrument (INJ) absent from prior failure patterns; clean OHLCV-only design.

LIQUIDATED (total_return -100%, max_drawdown 100%, liquidation 2024-06-12) on a strategy that is supposed to be long-only / no-leverage / spot — which cannot be liquidated. Fatal sizing/venue bug.

avg_position_pct 305.8% and impossible daily returns (-248%, -232%, -205%) show the book runs like a leveraged futures position, not unleveraged spot; annualized_volatility 315%.

No edge even pre-blowup: profit_factor 0.79, expectancy -$217/trade over 629 trades, deeply negative in 2021 (-39%) and 2022 (-91%); only 2023 (+199%) positive and that is outlier-driven (kurtosis 17, +126%/+109% days).

Heavy cost drag: commission_pct_of_gross 11.2%, total_commission $56k + impact $66k.

Single-bar volume-thrust continuation on a volatile 4H alt is a known repeat failure (high IS, collapses out-of-sample).

InjFourHourStrongVolumeBarContinuationLong deliberately used a simple single-bar body+volume trigger on a fresh symbol (INJ) to dodge the prior session's verification-timeout failures, and it did fire reliably (629 trades). But it failed catastrophically on two fronts: the supposedly unleveraged spot book ran at 305% average position size and was liquidated in June 2024 (-100% total return), and even before that it had no edge (profit factor 0.79, -$217/trade, with only an outlier-driven 2023 positive). The analyst abandoned it at backtest review after one iteration, citing both the fatal sizing/venue bug and a negative-expectancy signal that is a known repeat failure mode. The recorded guidance is to enforce a hard account-level no-leverage guard and to first prove a positive per-trade edge before optimizing — though single-bar thrust continuation on 4H alts is a poor candidate regardless.

Single-bar volume-thrust continuation on volatile 4H alts is a repeat negative-expectancy failure mode, and a strategy declared long-only spot must hard-enforce no-leverage and cap gross notional at ≤ equity so liquidation is mathematically impossible — neither edge nor risk control can be assumed from the venue label.

The analyst issued an 'abandon' verdict at the pre-optimization backtest-review gate on a dual failure: a fatal sizing/venue bug (a spot long cannot be liquidated or exceed -100%, so leverage/notional was not enforced) plus no underlying edge — a negative-expectancy single-bar volume-thrust signal that the factory has seen repeatedly collapse out-of-sample.

A minimal long-only momentum-continuation strategy on INJUSDT 4H bars (declared no-leverage spot, OHLCV-only) that entered long when a single green bar printed a body of at least 1.5% with volume at least 1.5× its 20-bar average — a signature of a one-sided news/flow event — and exited on a 4% take-profit, 3% stop, or a 3-bar time-stop.

It blew up: the account was liquidated on 2024-06-12 (total return -100%, max drawdown 100%), with an average position size of 305.8% of equity and impossible daily returns (-248%, -232%) showing it ran like a leveraged book despite being declared unleveraged spot. Even pre-liquidation it had no edge across 629 trades — profit factor 0.79, expectancy -$217/trade, deeply negative in 2021 (-39%) and 2022 (-91%) with only an outlier-driven 2023 (+199%) positive — and commissions ate 11.2% of gross.