BtcFourHourBollingerUpperBandMomentumContinuationLong

BTC 4H Bollinger Band Upper-Band Breakout Momentum-Continuation Long with Daily Bull Regime Filter

A long-only single-instrument STATISTICAL-DISTANCE MOMENTUM-CONTINUATION strategy on BTCUSDT perpetual futures using 4-hour bars and OHLCV-only data, with a daily-bar bull regime filter. This proposal is the SYMMETRIC INVERSION of the failed BtcFourHourBollingerLowerBandMeanRevLong (Sharpe -0.38, PF 0.93). The analyst's failure analysis documented a CRITICAL ECONOMIC FINDING: 'BB-distance triggers tend to coincide with regime-change events (liquidation cascades, macro shocks) that take days to resolve, NOT noise that mean-reverts quickly.' This is mechanistically different from RSI(2) velocity (which captures noise overshoot that retraces). The corollary: if BB-LOWER signals downward CONTINUATION (failed as long mean-reversion), then BB-UPPER should signal upward CONTINUATION and succeed as a long momentum entry. The economic mechanism: on BTC, a 4H close above the 20-period upper Bollinger Band (more than 2 std-devs above the 20-period mean) typically reflects (a) institutional FOMO surges (ETF inflows accelerating), (b) macro flight-to-BTC events (Fed pivot, regulatory tailwind), (c) systematic momentum traders cascading-buying, (d) trapped shorts forced to cover. These regime-change UPWARD events typically take days to resolve, providing CONTINUATION over 4-24 hours rather than reversion. This is the SAME economic finding documented by the analyst (BB-distance = regime-change), applied symmetrically to the upside. The proposal explicitly satisfies the analyst's required filter: 'When proposing a new mean-reversion mechanism for an asset that already has a working mean-reversion strategy, REQUIRE the hypothesis to articulate why the new signal type would capture a fundamentally different but still positive-expectancy edge.' This is NOT a mean-reversion mechanism at all — it's a momentum-continuation that uses BB distance as the entry trigger. Mechanistically distinct from BtcFourHourRsi2OversoldMeanRevLong (different direction, different signal type) and from BtcFourHourVolumeBreakoutLong (BB-upper is std-dev-based, not Donchian+volume). This is the THIRD validated mechanism class on BTC 4H if it succeeds: velocity-reversion (RSI), volume-breakout (Donchian+vol), distance-continuation (BB-upper).

Iteration 2 fix for the Layer 2 'frozen signal' failure. The previous version gated calculate_signal behind `if not self._bull_regime: return 0.0`; because the synthetic Layer 2 scenarios do not feed the daily extra bars that drive the regime flag, _bull_regime stayed False and every signal collapsed to 0.0 (479 identical signals on steady_downtrend). The smallest correct change: calculate_signal now returns the continuous BB-distance z-score every bar (it varies bar-to-bar), and the bull-regime gate moved into should_enter (`signal >= entry_std AND self._bull_regime`). Live trading behavior is unchanged — entries still require a confirmed bull regime — but the signal is no longer frozen, satisfying the synthetic verifier. All imports, class structure, sizing, leverage usage, and exit logic from the previous (Layer-1/earlier-passing) code are preserved intact.

The BB-upper momentum-continuation thesis (symmetric inversion of the failed BB-lower mean-reversion) is not borne out: the strategy has no positive-expectancy edge after costs. Over ~6 years and 485 trades it returns +12.7% total with profit_factor 1.016, Sharpe -0.10, sortino -0.17 and win_rate 44.5%. avg_trade_return is ~$11 (~0.01% per trade) — an order of magnitude below the 0.15% futures viability threshold — and trading costs consume essentially all of the gross edge (commission 22.0% of gross plus impact_cost_pct 77.1%, with capacity only $168k). The trades correctly implement the hypothesis (no code/artifact issue), so this is a genuine economic finding: 4H closes above the 2-sigma upper Bollinger Band in a daily bull regime do not continue profitably net of fees and impact. No parameter optimization can raise a ~zero gross-of-cost edge above the cost floor; raising entry_std only thins an already cost-dominated signal. Not worth optimizing — abandon. (Secondary: avg_position_pct 135% vs the intended 1.2x indicates the same missing open-position guard seen in sibling strategies and should be fixed in future variants, but it is not the reason for this abandonment.)

Long-only BTC 4H Bollinger upper-band breakout momentum-continuation strategy on Binance USD-M perpetual futures with a daily bull-regime filter (daily close > daily SMA50). Enters long when the 4H close exceeds the 20-period mean by >= entry_std standard deviations while in a bull regime, expecting upside continuation over ~4-24h. Exits on take-profit, stop-loss, close back below the BB middle, or a max-hold time stop. Uses 2x leverage with capital-relative sizing (~1.2x equity notional).

Verification failed (Layer 2 — synthetic scenarios): Parameters used: ['bb_period', 'entry_std', 'min_notional', 'position_pct', 'max_hold_bars', 'regime_period', 'stop_loss_pct', 'take_profit_pct'] Check that __init__ sets all attributes from self.parameters.get(). - steady_downtrend: Frozen signal: all 479 signals are identical (value=0.0). calculate_signal() likely has a bug — the signal never varies.

Clean implementation matching the hypothesis: 485 long entries above the upper BB in a confirmed daily bull regime, no frozen signal, metrics_reliable=true, no compounding artifact.

Thoughtful economic framing (symmetric inversion of the failed BB-lower mean-reversion) and adequate sample (485 trades) for a clear read.

Risk contained relative to siblings (no leverage blow-up; max_drawdown 33.6%).

No edge after costs: profit_factor 1.016, Sharpe -0.10, sortino -0.17, win_rate 44.5%, +12.7% total over ~6 years (alpha ~0).

avg_trade_return ~0.01% (≈$11/trade) — an order of magnitude below the 0.15% futures viability threshold.

Costs dominate: commission 22.0% of gross + impact_cost_pct 77.1%; capacity only $168k — the edge exists only at toy size.

The symmetric-inversion thesis is falsified: BB-upper breakouts in a bull regime do not continue profitably net of costs; no parameter change lifts a ~0% gross edge above the cost floor.

Minor: avg_position_pct 135% vs the stated 1.2x suggests the same missing open-position guard as siblings, but it is not the cause of failure.

This strategy inverted a previously failed BB-lower mean-reversion long into a BB-upper momentum-continuation long on BTC 4H futures, gated by a daily bull regime, on the thesis that 2-sigma band breaks reflect regime-change events that continue over hours to days. It was implemented cleanly with 485 entries and even produced a positive +12.7% total return, but profit factor was only 1.016 with Sharpe -0.10 and an avg trade return near $11 (~0.01%) — far below viability — because commissions (22% of gross) and impact (77%) consumed the entire edge, leaving capacity at just $168k. The analyst abandoned it at the backtest-review gate on its second iteration as a genuine economic finding rather than a code bug, concluding no optimization could raise a zero gross-of-cost edge above the cost floor, so it never advanced to optimization or risk review.

A mechanistically reasonable thesis can be cleanly implemented and still produce a gross edge that is essentially zero after costs — at ~0.01% per trade with commissions and impact dominating, the BB-upper continuation signal had no exploitable alpha, and inverting a failed mean-reversion idea does not guarantee a profitable momentum mirror.

The backtest-review analyst issued an 'abandon' verdict: the implementation was faithful but costs consumed essentially the entire gross edge (commission 22.0% of gross plus 77.1% impact cost, capacity only $168k), so the symmetric-inversion thesis was falsified — upper-band breakouts in a bull regime do not continue profitably net of fees, and no parameter change can lift a ~zero gross-of-cost edge above the cost floor.

A long-only Bollinger-band momentum-continuation strategy on BTCUSDT 4H Binance futures (2.0x leverage) — the symmetric inversion of a failed BB-lower mean-reversion long — entering when a 4H close cleared the 2-sigma upper band during a confirmed daily bull regime, betting that BB-distance triggers are regime-change events that continue rather than revert.

Across 485 trades over ~6 years it returned +12.7% total but with no real edge: profit factor 1.016, Sharpe -0.10, Sortino -0.17, win rate 44.5%, and avg trade return of only ~$11 (~0.01%), an order of magnitude below the 0.15% futures viability threshold, with a 33.6% max drawdown.