SolImpulseBarContinuationTrendLS1H

SOL Intraday Large-Impulse-Bar Continuation (Single-Instrument USD-M, Long-Short, 1H, Ride Directional Impulse Bars, Tight Initial Stop, Trailing-Stop Winners, Discrete Capital-Capped, Pure OHLCV)

A LONG-SHORT, SINGLE-INSTRUMENT, pure-OHLCV intraday MOMENTUM-CONTINUATION strategy on SOLUSDT.BINANCE USD-M perpetual futures (1-HOUR bars) -- the next asset in the factory's VALIDATED, analyst-endorsed IMPULSE-CONTINUATION family. The ETH instance (EthImpulseBarContinuationTrendLS1H) just PASSED optimization at Sharpe 1.40 (risk_review), and the BTC instance is in pipeline; the analyst's explicit redirect after four basket-refinement failures was 'redirect ALL novelty to the generalizing intraday impulse-continuation family,' and that family is now empirically validated and being populated per-asset (exactly as single-asset TSM was a sanctioned series -- except impulse-continuation GENERALIZES forward where continuous trend/TSM/VWAP forms decayed). MECHANISM (mirrors the proven ETH template): an IMPULSE bar is a 1H bar whose directional body (close - open) exceeds impulse_mult x ATR(atr_period) AND closes in the strong third of its high-low range -- a high-conviction directional thrust (genuine flow/information, not noise). The strategy enters in the impulse direction to capture the short-horizon CONTINUATION before the move is absorbed, with a TIGHT initial stop (cut losers fast) and a TRAILING stop that lets continuation winners run. WHY SOL: among uncovered majors it is the strongest impulse-continuation candidate -- top-4 liquidity (clean signal) and the HIGHEST intraday volatility of the majors, so it produces the most frequent large directional impulse bars (densest walk-forward population, no event-sparsity) and the strongest intraday continuation character (SOL thrusts and follows through harder than BTC/ETH). WHY IT GENERALIZES WHERE OTHER SINGLE-ASSET INTRADAY DECAYED: impulse-continuation is EVENT-DRIVEN and SELECTIVE (only large-conviction bars), the higher-quality signal that made the ETH sibling pass its recent-regime holdout, unlike VWAP-continuation (rode noisy excursions, decayed) and single-asset trend/TSM (continuous, whipsawed, mania-beta). WHAT IT AVOIDS (every closed path): NOT a basket refinement (the prohibited axis -- 4 consecutive degradations); NOT funding/carry (uncreditable in the engine + HL-data-walled); NOT reversion (dead); NOT continuous-rebalance/netting-flip (the 316%/385% sizing-artifact cause -- this uses DISCRETE enter-then-exit-to-FLAT, like the cleanly-sized ETH sibling); NOT single-asset TSM (dead -- different mechanism). EVERY design choice respects the lessons: (1) DISCRETE entries, flat between signals, sized against STABLE realized equity (excl. unrealized PnL), hard-capped <=100% (leverage 1x). (2) FEE-AWARE: fires only on large-impulse bars; SOL's big continuation moves clear the ~0.10% RT floor easily -- NOT scalping. (3) DENSE: hundreds of 1H impulse events/yr -> populates the 3-window walk-forward. (4) PURE OHLCV: no ticks (2-day wall), options (unwired), funding (data ends 2024). Per the 'favor simpler' mandate, 3 tunables (impulse_mult, trail_mult, init_stop_mult) mirroring the proven ETH config.

Mirrors the empirically-validated ETH template (EthImpulseBarContinuationTrendLS1H, Sharpe 1.40 at risk_review) one-for-one, swapping only the asset (SOLUSDT.BINANCE) and lot precision (SOL size step 0.001). It encodes every analyst lesson: DIRECTION is continuation, not fade — the repeatedly-refuted anti-predictive family is closed and continuation is the validated direction; PAYOFF SHAPE is let-winners-run/cut-losers-fast via a tight initial ATR stop plus an ATR trailing stop, manufacturing the avg_win>avg_loss asymmetry the fades and the mis-built 1:3.3 flush-bounce lacked; SIZING is discrete enter-then-exit-to-flat against stable realized equity (get_account_equity, excludes unrealized PnL) and hard-capped at max_notional_pct·leverage, avoiding the continuous-rebalance/netting-flip sizing artifact; it is PURE OHLCV single-instrument, immune to the supplementary-stream coverage walls (OI/taker/L-S/liquidations/funding) and the basket data-collection wall. SOL is chosen for top-4 liquidity (clean signal) and the highest intraday volatility of the majors, giving the densest large-impulse population for the 3-window walk-forward and the strongest continuation character. The selective, event-driven trigger fires only on large-conviction bars so SOL's big continuation moves clear the ~0.10% round-trip fee floor (not scalping). Kept to 3 edge tunables per the favor-simpler mandate.

No edge and a structural capacity wall — not optimizable. At the config that mirrors the proven ETH sibling (impulse_z 3.0, vol_lookback 48, trail 3.0), SOL 1H produces Sharpe 0.01, profit_factor 1.016, omega 1.002, expectancy ~$12/trade, win_rate 34%, total_return +13% over 5.7yr — essentially break-even (sharpe_ci_low -1.29, PSR 0.51). The designed 2:1 avg_win/avg_loss asymmetry is present but the 34% hit rate nets it to zero. The decisive structural problem is impact_cost_pct 73.79% (modeled impact consumes ~3/4 of gross PnL) with capacity_usd only $183k — 100-1000x below the other daily strategies this batch (ADA $46M, DOGE $5.8B) — because the per-trade edge is so thin that impact (turnover 215) swamps it; even an optimized positive Sharpe would be undeployable at this capacity. This is a legitimate per-asset GENERALIZATION FAILURE: the impulse-continuation family is validated on ETH (Sharpe 1.40) but does NOT carry to SOL 1H, with three double-digit-negative years (2020 -12.6%, 2021 -11.8%, 2024 -20.1%) confirming it's not a regime fluke. NOT optimize: there is no edge to tune toward at the proven config, and the capacity/impact wall is structural (trading frequency vs near-zero edge), not a parameter the optimizer can resolve; a 2-hour sweep would curve-fit noise. NOT iterate: no code bug — execution is clean, balanced, and mirrors the validated template. The family's value is preserved in the ETH sibling; SOL 1H is simply a poor target. FAILURE PATTERN: the impulse-bar continuation family generalizes to ETH but not SOL at 1H — on SOL the per-trade continuation edge is too thin to clear realistic market impact (impact 74% of gross, capacity only $183k), yielding a break-even Sharpe (0.01) with multiple double-digit-negative years despite clean balanced long-short execution; per-asset validation correctly identifies SOL as a non-carrier of the edge.

SOL instance of the validated impulse-continuation family: a single-instrument, long-short, pure-OHLCV intraday momentum-continuation strategy on SOLUSDT.BINANCE USD-M perp (1H bars). It flags a large-impulse bar via a return z-score (current-bar return / std of the last vol_lookback returns) and enters WITH the impulse — long on a large up-impulse, short on a large down-impulse — to capture short-horizon continuation. Exits cut losers fast and let winners run: a tight initial ATR stop, an ATR trailing stop that ratchets with the move, and a time cap. Discrete enter-then-exit-to-flat; sized to fixed fractional risk, hard-capped at half equity notional, leverage 1x. Three edge tunables (impulse_z, vol_lookback, trail_atr_mult).

Clean execution and design: balanced 290 long / 307 short, 597/597 submitted, discrete enter-to-flat sizing (no netting-flip artifact), metrics_reliable=true, mirrors the validated ETH template

Genuinely market-neutral-ish (beta 0.024); the designed avg_win/avg_loss ~2:1 asymmetry is present

Impulse-continuation mechanism is validated on the ETH sibling (Sharpe 1.40) — the family's value is not lost

No edge at the proven config: Sharpe 0.01, profit_factor 1.016, omega 1.002, expectancy ~$12/trade, win_rate 34%, total_return +13% over 5.7yr (CAGR 1.25%); sharpe_ci_low -1.29

Structural capacity wall: impact_cost_pct 73.79% (impact ~3/4 of gross PnL), capacity_usd only $183k — undeployable, 100-1000x below sibling strategies

Generalization failure: SOL produces Sharpe 0.01 with the ETH-proven params despite being claimed 'the strongest candidate'; three double-digit-negative years (2020 -12.6%, 2021 -11.8%, 2024 -20.1%)

deployable (>~$1M)

SolImpulseBarContinuationTrendLS1H was the SOL rollout of the factory's one validated, analyst-endorsed intraday family — impulse-bar continuation — whose ETH sibling had just passed optimization at Sharpe 1.40, and it faithfully mirrored that proven template with clean, balanced, artifact-free execution. But the edge did not carry: at the identical config SOL produced a Sharpe of 0.01 and a profit factor of 1.016, with the designed 2:1 win/loss asymmetry undone by a 34% hit rate, and three double-digit-negative years confirming it was no fluke. Decisively, a structural capacity wall sealed it — modeled impact consumed ~74% of gross PnL at a deployable capacity of only $183k, 100-1000× below the batch's other strategies — so even an optimized positive Sharpe would be untradeable. The reviewer abandoned it as a legitimate per-asset generalization failure: the impulse-continuation edge lives on ETH, not SOL 1H, and neither iteration (the code is clean) nor optimization (no edge, structural impact wall) could change that.

A validated family does not generalize to every asset: SOL's intraday impulse-continuation edge is too thin to clear realistic market impact (74% of gross, capacity only $183k) and nets to break-even despite the correct 2:1 payoff and clean execution — so per-asset validation correctly rejects SOL as a non-carrier even though the family's value is preserved in the promoted ETH sibling.

The analyst abandoned it at the backtest-review gate, before optimization, as a per-asset generalization failure with a structural capacity wall: the impulse-continuation edge that validated on ETH (Sharpe 1.40) does not carry to SOL 1H (Sharpe 0.01, three double-digit-negative years), and modeled market impact consumes ~3/4 of gross PnL at a capacity of just $183k — a frequency-vs-thin-edge problem the optimizer cannot resolve.

The SOL instance of the validated, analyst-endorsed impulse-continuation family: a long/short, single-instrument, pure-OHLCV 1H strategy on SOLUSDT.BINANCE USD-M perp that enters in the direction of large-conviction impulse bars (return z-score >= impulse_z) to ride short-horizon continuation, cutting losers with a tight initial stop and letting winners run on a trailing stop, with discrete enter-to-flat sizing to avoid the netting-flip artifact.

At the ETH-proven config it was essentially break-even with no deployable capacity: Sharpe 0.01, profit factor 1.016, omega 1.002, expectancy ~$12/trade, 34% win rate, +13.2% total return over 5.7 years (CAGR 1.25%) across 597 balanced trades — the designed ~2:1 avg_win/avg_loss asymmetry present but netted to zero by the low hit rate — with impact_cost_pct 73.79% and capacity_usd only $183k; execution was clean (metrics_reliable=true, beta 0.024).