DotImpulseBarContinuationTrendLS1H

DOT Intraday Large-Impulse-Bar Continuation, Long-Short (Binance USD-M, 1H, Deep-History High-Vol L1 Asset-Match, Clamped Impulse Threshold, Tight Initial Stop, Trailing-Stop Winners, Discrete Capital-Capped)

A LONG-SHORT, single-instrument, pure-OHLCV intraday CONTINUATION strategy on DOTUSDT.BINANCE USD-M perpetual using 1-HOUR bars — the next per-asset instance of the factory's VALIDATED and PROMOTED impulse-bar-continuation family (ETH-1H ~1.40 promoted, SOL-1H ~0.69 paper), applying the EXACT validated single-impulse trigger to the deep-history asset that best fits the now-legible winning discriminator. The family's boundary is mapped on two axes: it needs HIGH VOLATILITY (so intraday thrust moves are large enough to clear the ~0.10% round-trip fee — BTC failed here, fee-bound-zero across USD-M/HL/CoinM) AND CLEAN TREND-PERSISTENCE (so thrusts continue rather than mean-revert — LINK/XRP/DOGE failed here, news/legal-driven whipsaw). DOT satisfies both: a high-volatility L1 with the DEEPEST clean history of the untested candidates (Binance USD-M listing Aug 2020, ~5.5yr of dense 1H bars for a fully-populated walk-forward), consistently liquid, and structurally trend/flow-driven rather than single-event-spike-driven. It deliberately AVOIDS every dead/blocked path proven this session: NOT BTC (fee-bound, dead all triggers/venues — stop porting), NOT a funding/basis/carry/cross-venue strategy (entire family fee-dominated/event-sparse, confirmed across BTC/ETH/SOL), NOT relative/cross-sectional momentum (decayed), NOT a daily-trend basket refinement (correlated with promoted AbsoluteMomentum), NOT options (engine), NOT reversion (dead), NOT sub-1H (fee-dead), NOT HL intraday (data wall). It is distinct from the in-flight AVAX-1H (different asset) and from my ETH/SOL ORB and ETH-4H variants (different asset and the canonical 1H single-impulse trigger). To pre-empt the LINK overfit failure (its impulse_mult was left unbounded and the optimizer fled to a sparse corner), it bakes in the SOL-sibling hardening: impulse_mult CLAMPED [1.0, 1.8], strong_frac FROZEN, <=4 tunables, flat-between-signals. Engine-safe: discrete enter-once/exit-once, single position, hard-capped notional, no continuous rebalance/vol-scaling.

Implements the hypothesis as the next per-asset instance of the factory's VALIDATED and PROMOTED impulse-bar-continuation family (ETH-1H ~1.40 promoted, SOL-1H ~0.69 paper), applying the EXACT validated single-impulse trigger to DOT — a high-volatility L1 with the deepest clean Binance USD-M history of the untested candidates, satisfying both axes of the winning discriminator (high vol to clear fees; clean trend-persistence so thrusts continue). It is byte-for-byte the proven ETH/SOL/AVAX mechanism with only the instrument swapped (and QTY_PRECISION set to DOT's 0.1 lot step). It bakes in the SOL-sibling overfit hardening to pre-empt the LINK failure: impulse_mult HARD-CLAMPED to [1.0, 1.8] in code (not just the default), strong_frac frozen and capped at 0.70, exactly FOUR tunables (impulse_mult, trail_mult, init_stop_mult, atr_period) with risk_pct/max_notional_pct/min_notional frozen — keeping every walk-forward window populated. Engine-safe and capital-disciplined: discrete enter-once/exit-once flat-between-signals, single position, ATR-risk sizing capped at 25% notional, leverage 1.0 (no leverage referenced in sizing beyond the 1.0 multiplier, so no leverage_set_but_unused risk). The ~0.15%+ per-trade edge target is met because DOT's intraday impulse moves (>=1.5 ATR bodies) clear the ~0.10% round-trip taker cost on a high-vol asset. Futures venue (BINANCE USD-M) is correct: the strategy goes short, which a CASH spot account cannot do.

Base backtest has no edge and loses money before any optimization: profit_factor 0.90 (<1), Sharpe -0.69, Sortino -1.61, expectancy -$41.6/trade, total_return -7.8% over a deep 5.5-year, 901-trade sample, with sharpe_ci_high 0.18 (entire CI at/below zero) and losses every year since 2022 (2023 -17.5%, 2024 -20.6%). Optimization cannot rescue a net-negative, sub-1.0 profit-factor strategy over 901 trades — the four tunables shift entries but cannot flip a negative gross edge positive. The result empirically refutes the hypothesis's central claim that DOT has 'clean trend-persistence': DOT's impulse bars mean-revert/whipsaw rather than continue, placing it on the LINK/XRP/DOGE failure axis, not the ETH/SOL winning one. Not iterate: the strategy implements its premise correctly (901 balanced long/short trades matching the impulse direction), so there is no code defect — the premise itself is wrong for this asset. FAILURE PATTERN: the intraday 1H single-impulse-bar continuation family does not generalize beyond ETH/SOL; on DOT (as on BTC fee-bound, AVAX DSR-fail post-opt, LINK overfit, DOGE whipsaw) it produces a negative-expectancy PF<1 result on a full-history sample — asset-matching on 'high vol + deep history' does not confer the trend-persistence the edge requires.

DOTUSDT.BINANCE USD-M perpetual, 1-HOUR bars, pure-OHLCV long/short intraday impulse-bar continuation. Each bar computes Wilder ATR over atr_period and flags an IMPULSE when the candle body exceeds impulse_mult*ATR AND the close sits in the strong end of the bar's range (pos>=strong_frac up / <=1-strong_frac down). It enters WITH the impulse direction (BUY on up-thrust, SELL on down-thrust) to ride short-horizon continuation. Exits cut losers fast and let winners run: a tight initial stop at init_stop_mult*ATR (which also sizes the position), a trailing ATR stop ratcheting the best price by trail_mult*ATR, and a 48-bar time cap. Sizing is ATR-risk-based against realized equity (risk_pct), hard-capped at max_notional_pct of equity, with discrete flat-between-signals round-trips (one position at a time, 1x, no rebalance/reversal).

Mechanism functions correctly: 901 trades, balanced 451 long / 450 short, deep 5.5yr fully-populated sample, trades match the hypothesized impulse-continuation direction

Engine-safe design (discrete enter/exit, capped notional); fees are not the primary problem (7.2% of gross)

No edge: profit_factor 0.90 (<1), Sharpe -0.69, expectancy -$41.6/trade, total_return -7.8% over 5.5 years

sharpe_ci_high 0.18 — entire Sharpe CI at/below zero

Loses every year since 2022 (2023 -17.5%, 2024 -20.6%); positive only in the 2020-21 listing bull

Refutes the core premise: DOT impulse bars mean-revert/whipsaw rather than continue — the LINK/XRP/DOGE failure axis, not the ETH/SOL winner

positive

DotImpulseBarContinuationTrendLS1H ported the validated ETH/SOL impulse-continuation trigger unchanged to DOT, chosen as the deepest-history high-vol L1 thought to satisfy both the volatility and trend-persistence axes of the family's mapped boundary, with the SOL-sibling clamp/freeze hardening baked in. The code worked on a deep, balanced 901-trade sample, but there was simply no edge — profit factor 0.90, Sharpe -0.69, expectancy -$41.6, -7.8% return, profitable only in the 2020-21 listing bull and negative every year since. The analyst ruled the premise itself wrong for DOT (its impulses whipsaw, not continue), a no-code-defect negative-expectancy result optimization cannot rescue, confirming the family does not generalize past ETH/SOL. It ended after one iteration as abandoned, never advancing to optimization or risk review.

Asset-matching on 'high volatility + deep history' does not confer the trend-persistence the impulse-continuation edge requires: DOT's impulse bars mean-revert/whipsaw rather than continue, placing it on the LINK/XRP/DOGE failure axis, and the 1H single-impulse family now demonstrably does not generalize beyond ETH/SOL (failing on BTC fee-bound, AVAX DSR post-opt, LINK overfit, DOGE/DOT whipsaw).

It was abandoned at the pre-optimization backtest-review gate (verdict: abandon) because the base backtest is a statistically meaningful negative-expectancy, sub-1.0 profit-factor result over 901 trades that the four tunables cannot flip positive — empirically refuting the hypothesis's claim of clean DOT trend-persistence — so optimization and all later stages were never reached.

A single-instrument, pure-OHLCV long/short intraday impulse-bar continuation strategy on DOTUSDT.BINANCE USD-M 1H perps — entering with high-conviction directional impulse bars (body ≥ impulse_mult×ATR closing in the strong third of range) and riding continuation with a tight initial stop and ATR trailing stop — applying the exact validated ETH/SOL trigger to DOT, picked as the deepest-history (Aug 2020, ~5.5yr) high-vol L1 match to the winning discriminator, with impulse_mult clamped to [1.0, 1.8] and strong_frac frozen to pre-empt the LINK overfit.

Over 1,245 days and a deep, balanced 901 trades (451 long / 450 short) it had no edge: profit factor 0.90, Sharpe -0.69 (CI [-1.69, 0.18]), Sortino -1.61, expectancy -$41.6/trade, win rate 26%, total return -7.8%, max drawdown 37.9%, positive only in the 2020-21 listing bull and losing every year since 2022 (2023 -17.5%, 2024 -20.6%); fees were not the primary problem (7.2% of gross).