AdaFourHourVolumeAbsorptionBreakoutLong

ADA 4H Volume Absorption Breakout Long

Mechanism family (a) from analyst-recommended untried list: volume/flow imbalance derivable from OHLCV bars. The strategy detects 'absorption bars' — bars with anomalously high volume but small body (close-open) relative to the bar range. This pattern indicates aggressive directional flow was met by passive liquidity without price resolution, suggesting hidden accumulation. When followed by a confirming breakout bar that closes above the absorption bar's high, the absorbing party (buyers in this setup) has demonstrated control and the imbalance typically resolves in their direction. Trades long-only on ADA 4H, the only consistently-fresh bar feed. Distinct from prior single-instrument breakouts (which used Donchian channel breaks) and prior cross-instrument signals (BTC-leads-ADA, BTC-ETH-decorrelation) — this is the first volume-flow mechanism tried on ADA 4H this session.

Implements mechanism family (a) volume/flow imbalance purely from OHLCV. The pattern is a genuine two-bar sequential event: an absorption bar (high vol + small body) arms a level, and a later close above that level confirms — distinct from Donchian-channel breakouts (which key off N-bar extremes regardless of volume/body) and from prior cross-instrument ADA signals. calculate_signal returns a CONTINUOUS, bar-varying score (absorption strength = vol_ratio x (1-body_ratio) plus breakout proximity to the armed level, tanh-bounded) so Layer-2 frozen-signal detection passes, while the discrete confirmation is gated in should_enter via the _breakout flag set that bar. The breakout check runs BEFORE arming a new level so an absorption bar can never be its own confirmation (it arms only for future bars), and the window disarms after confirm_window bars to keep the trigger crisp. Volume and ranges use PRIOR bars (exclude current) so the comparison is causal. Routed to BINANCE_SPOT because the edge is long-only, uses no leverage, and is not funding-based (per the venue-selection rule); single-instrument long-only on a CASH account is safe — sizing only runs while flat so get_account_equity() reflects full USDT, and notional is capped at 0.98x equity to respect the no-margin constraint. The 10% take-profit dwarfs the ~0.20% spot round-trip fee, giving ample per-trade edge above costs. All divisions are zero-guarded (zero range -> body_ratio 1.0 disqualifies; zero avg-vol -> vol_ratio 0) and NaN/inf sanitized; ADAUSDT spot has deep, fresh 4H history avoiding data_unavailable.

Correctly-implemented volume-absorption breakout with NO edge on ADA 4H: profit_factor 0.90, Sharpe -0.79, expectancy -$68/trade, total return -6.0% over 8 years, negative in 5 of 8 years (124-trade multi-regime sample, so this is a fair read, not noise). Win rate 42.7% with avg_win only marginally above avg_loss means the 2:1 TP/SL still nets negative — the signal does not pay, and fees (6.3% of gross) add drag. The failure is in the signal, not the parameters: there is no positive base or robust region for optimization to build on, so tuning to 124 trades would overfit and fail DSR/walk-forward/holdout. Not worth 2 hours of optimization. NOTE for research: the absorption/hidden-accumulation flow pattern derived from OHLCV does not produce a tradable long-only edge on ADA 4H — a different mechanism family or genuine order-flow data would be needed rather than re-tuning this one.

Long-only volume-absorption breakout on ADAUSDT 4H spot bars. Detects 'absorption bars' — bars with anomalously high volume (>= 2x the trailing 20-bar average) but a small body relative to range (|close-open|/(high-low) <= 0.30), indicating aggressive flow absorbed by passive liquidity (hidden accumulation). The absorption bar's high is armed as a breakout level for up to 3 subsequent bars; if a later bar closes above that level, the strategy enters long (buyers demonstrated control). Positions exit on a 10% take-profit, 5% stop-loss, or an 18-bar (3-day) time-stop. Capital-relative fixed-fractional sizing (1% equity risk vs the stop distance), capped at deployable equity for the spot CASH account.

Mechanism correctly implemented and matches the hypothesis: 124 long-only absorption-breakout trades, clean two-bar arm/confirm logic, clean entry diagnostics (0 dropped).

First volume-flow mechanism tried on ADA 4H — informative negative result for the mechanism family.

Low exposure (8%) and contained max drawdown (30%), so the negative outcome is a clean read on the edge rather than a blow-up.

No edge: profit_factor 0.90, Sharpe -0.79 (CI [-3.67, 1.97]), expectancy -$68/trade, total return -6.0% over 8 years.

Negative in 5 of 8 years with no consistent positive regime — not a single-period artifact.

Win rate 42.7% with avg_win ($1,481) only marginally above avg_loss ($1,224) despite a 2:1 TP/SL — the signal does not pay.

commission_pct_of_gross 6.34% adds fee drag on top of an absent gross edge.

124-trade sample is sufficient to conclude the absorption-breakout signal has no edge on ADA 4H; optimizing it would overfit and fail downstream robustness gates.

AdaFourHourVolumeAbsorptionBreakoutLong was the session's first volume-flow attempt on ADA 4H, inferring hidden accumulation from high-volume small-body bars and entering on a confirming breakout above the absorption high. The two-bar arm/confirm logic worked cleanly across 124 trades, but the signal simply had no edge — profit factor 0.90, Sharpe -0.79, expectancy -$68/trade, -6.0% over eight years, and losses in 5 of 8 years even with a 2:1 reward/risk setup. The analyst abandoned it at backtest review after one iteration, concluding the failure was in the signal rather than the parameters and that optimizing would only overfit. The note for research is that OHLCV-derived absorption patterns don't yield a long-only edge on ADA 4H, and genuine order-flow data or a different mechanism family would be needed.

An absorption / hidden-accumulation flow pattern derived purely from OHLCV does not produce a tradable long-only edge on ADA 4H; capturing genuine order-flow imbalance would require real order-flow data or a different mechanism family rather than re-tuning this one.

The analyst issued an 'abandon' verdict at the pre-optimization backtest-review gate: the mechanism was correctly implemented over an adequate multi-regime sample, so the negative result was judged a clean read of a signal that doesn't pay — there was no positive base or robust region for optimization to build on, and tuning 124 trades would overfit and fail downstream robustness gates.

A long-only OHLCV volume-flow strategy on ADAUSDT 4H spot bars that detected 'absorption bars' — anomalously high volume (≥2× the trailing 20-bar average) with a small body relative to range (≤0.30) — as a sign of hidden accumulation, then entered long when a subsequent bar within a short confirmation window closed above the absorption bar's high, exiting on a 10% take-profit, 5% stop, or time-stop.

Over ~8 years and 124 trades it had no edge: profit factor 0.90, Sharpe -0.79 (CI -3.67 to 1.97), expectancy -$68/trade, -6.0% total return, and a 42.7% win rate where average wins ($1,481) barely exceeded average losses ($1,224) despite a 2:1 take-profit/stop ratio. It was negative in 5 of 8 years with no consistently positive regime, and fees added 6.34%-of-gross drag; drawdown stayed contained at 30% on low 8% exposure.