NearFourHourVolumeBreakoutLong4H

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

Long-only NEAR Protocol 4-hour strategy on BINANCE USD-M futures. Direct analog of the landed BTC/ETH/SOL/BNB/AVAX/XRP/MATIC/ADA (Ada Volume Absorption) / ATOM FourHourVolumeBreakoutLong family — applied to NEAR. 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.

Implements the hypothesis as a faithful single-instrument analog of the proven 4H Volume Breakout family, which has repeatedly cleared all four verification layers, so the structure is known-good and low-risk. The only changes from the ATOM reference are the instrument id (NEARUSDT.BINANCE), class/name, and a NEAR-appropriate default qty precision; entry conjunction (volume surge + bullish bar + SMA-100 uptrend), exit triggers, sizing, and the tunable/locked parameter split are identical. The strategy follows the family's on_bar-driven contract: calculate_signal returns a continuous, bar-varying score (volume ratio and body normalized via tanh, biased by the SMA regime) and the actual entry/exit are executed inside on_bar, matching the landed siblings. 4H timeframe deliberately avoids the daily-bar audit-refill failures noted in the hypothesis. Venue is BINANCE USD-M futures at leverage 1.0 for consistency with the rest of the family and its lower round-trip fees (~0.10% vs spot 0.20%), comfortably cleared by the 5% take-profit target; long-only is enforced (only OrderSide.BUY entries, SELL used solely to exit).

Marginal lower-liquidity arm of the volume-breakout family that fails multiple-testing deflation and sensitivity after best-of-225 selection — the DOT/OP precedent, as flagged at the pre-optimization gate. NEAR is genuinely better than its abandoned siblings on two axes: the holdout PASSED strongly (holdout_sharpe 4.98, ratio 14.5 against a positive WF-OOS denominator 0.345) and the optimized config is positive in 6 of 7 years with an undecayed recent record (2026 +14.3%, PSR 0.9377). But the decisive gates reject it: deflated_sharpe 0.185 (vs 0.95 bar) with the optimized Sharpe 1.446 BELOW the expected-max luck bar 2.291 over 225 trials (is_significant=false, sharpe_ci_low -0.359 straddles 0, PBO 0.6217); sensitivity FAILED with 3 cliffs on the ENTIRE core signal (volume_mult, volume_window, sma_trend_period); and walk-forward is_overfitted=true with avg IS 2.03 collapsing to avg OOS 0.345 and one negative OOS window (-1.21). A single strong holdout window does not establish robustness when the deflated Sharpe is below the luck bar, the CI straddles 0, PBO>0.5, and all three core params are cliffs — the selected config is fitted to the search, not robust. Not iterate (attempt 1 of 2): iterate is reserved for a robust region the sweep under-explored, but the sensitivity cliffs show no such region exists, so a second sweep re-selects another sub-luck-bar config; the DOT sibling (base Sharpe 0.51 -> optimized 1.69) failed this exact sequence (DSR 0.24, negative holdout) and NEAR's base is the same marginal tier (0.52), deflating no better (0.185). Not revise_hypothesis: the volume-breakout mechanism is already validated and promoted on the majors (BTC/ETH/SOL/BNB/AVAX/XRP), so its value is captured; NEAR is the marginal lower-liquidity arm of the family's asset-selectivity, not a proven mechanism stranded on a dead target. FAILURE PATTERN: the FourHourVolumeBreakoutLong family does not generalize robustly to lower-liquidity alts (NEAR, DOT, OP) — even when the optimized config posts a decent Sharpe (1.4-1.7) and a strong single holdout window, it fails deflated Sharpe (below the expected-max luck bar) with core-param sensitivity cliffs and is_overfitted=true, because the marginal base edge (~0.5 Sharpe) cannot survive best-of-225 selection deflation. A passing holdout with DSR 0.185 and 3 core-signal cliffs is a fitted, non-robust edge, not a deployable one.

Long-only volume-breakout strategy on NEARUSDT.BINANCE USD-M futures, 4-HOUR bars, OHLCV-only. A direct port of the landed FourHourVolumeBreakoutLong template family (BTC/ETH/SOL/BNB/AVAX/XRP/MATIC/ADA/ATOM) applied to NEAR. Enters long when a 4H bar's volume exceeds 1.5x the prior 20-bar average AND the bar is bullish (close > open) AND price is above its 100-bar SMA (uptrend filter). Exits on +5% take-profit, -5% stop-loss, or a 14-bar (~2.3 day) time stop. Sizes each entry at 20% of equity notional; leverage 1.0.

Functioning positive edge, properly sized: profit_factor 1.12, Sharpe 0.52, expectancy +$78/trade (~0.37%), avg_position_pct 21.0% (respects cap), 17.9% vol, 19.9% drawdown

Healthy sample (335 trades, all long as specified); faithful port of a family validated on the majors

Edge NOT decayed and not single-year-concentrated: positive in 2020/2022/2024/2026, with 2026 (+13.9%) the best year — unlike the decaying OP/DOT arms

Meaningfully stronger than the abandoned OP arm on every axis: 2x Sharpe, half the impact (14.6% vs 25%), 3x capacity ($4.66M vs $1.58M)

Marginal: profit_factor 1.12 is below the 1.2 viability bar and Sharpe 0.52 barely clears 0.5 with sharpe_ci_low -0.75 (straddles 0)

DOT/OP precedent: DOT's near-identical base Sharpe (0.51) optimized to 1.69 and still failed deflated Sharpe (0.24) and the holdout — NEAR is at real risk of the same

Lower-liquidity alt arm of a major-validated family: capacity only $4.66M and impact 14.6% — limited deployable scale

Thin per-trade margin means costs could dominate if size scales

Stronger than the abandoned DOT/OP siblings: holdout PASSED strongly (holdout_sharpe 4.98, ratio 14.5 against a positive WF-OOS denominator) and the optimized config is positive in 6 of 7 years with an undecayed recent record (2026 +14.3%, PSR 0.9377)

Optimized point metrics are reasonable in isolation (Sharpe 1.45, PF 1.35, win rate 55%, 160 trades, impact cut to 6.1%, capacity $26.7M)

Properly sized, faithful family port, trades as specified

Deflated Sharpe 0.185 (vs 0.95 bar): the optimized Sharpe 1.446 is BELOW the expected-max luck bar 2.291 over 225 trials — indistinguishable from best-of-N selection; is_significant=false, sharpe_ci_low -0.359 (straddles 0), PBO 0.6217

Sensitivity FAILED with 3 cliffs on the ENTIRE core signal (volume_mult, volume_window, sma_trend_period) — no robust parameter region exists

Walk-forward is_overfitted=true: avg IS 2.03 collapses to avg OOS 0.345 with OOS windows [-1.21, +2.16, +0.09] (one negative)

Matches the DOT precedent: marginal lower-liquidity arm (base Sharpe 0.52) that optimizes to ~1.4-1.7 then fails deflated Sharpe and is_overfitted — a strong single holdout window does not establish robustness against these

This was the cleanest lower-liquidity arm of the volume-breakout family yet — NEAR 4H showed an undecayed, properly-sized base edge (Sharpe 0.52, +31.2%, best year in 2026) and earned an 'optimize' verdict, with the optimized config even passing the holdout strongly (Sharpe 4.98). But it failed the decisive deflation gates: deflated Sharpe 0.185 below the 2.291 luck bar, is_significant false, PBO 0.62, walk-forward overfitted, and sensitivity cliffs on all three core signal parameters. The analyst abandoned it at the analyzing stage on the first iteration — matching the DOT/OP precedent that a marginal base edge optimizes to ~1.4-1.7 yet remains a fitted, non-robust edge — with no robust region to tune toward and the mechanism already captured on the majors, so it stopped before risk review.

The volume-breakout family does not generalize robustly to lower-liquidity alts — a marginal ~0.5 base Sharpe cannot survive best-of-N deflation even when it optimizes to ~1.4 with a strong single holdout, because deflated Sharpe sits below the luck bar with core-signal sensitivity cliffs and walk-forward overfitting; one passing holdout window is not robustness.

It cleared the backtest-review gate with an 'optimize' verdict (notably stronger and less decayed than the abandoned OP/DOT arms), but the post-optimization analyst abandoned it: deflated Sharpe 0.185 (<0.95) with the optimized Sharpe 1.45 below the 2.291 expected-max over 225 trials, is_significant false, a CI straddling zero, PBO 0.62, walk-forward overfitted (avg IS 2.03 → avg OOS 0.345 with a negative window), and sensitivity failing with 3 cliffs across the entire core signal — a strong single holdout window does not establish robustness against those.

A long-only NEAR Protocol 4H volume-breakout strategy on Binance USD-M futures (0.20 per-leg sizing) — a faithful port of the volume-breakout family validated/promoted on the majors — entering on a volume surge above its prior-window average with a bullish bar above a 100-bar SMA, exiting on ±5% PnL or a 14-bar time stop.

The base backtest showed a properly-sized, undecayed edge — 335 trades, +31.2% total, Sharpe 0.52, profit factor 1.12, expectancy +$78/trade, 19.9% drawdown, positive in 2020/2022/2024/2026 (2026 +13.9% its best year) — and the optimized config posted Sharpe 1.45, PF 1.35, 55% win rate, capacity $26.7M with a strongly passing holdout (holdout Sharpe 4.98).