HyperliquidSolVolatilitySqueezeBreakoutLS1H

Hyperliquid SOL Volatility-Squeeze Breakout, Long-Short (Single-Instrument Perp, 1H, Pure OHLCV, Range-Contraction Regime + Volume-Confirmed Directional Break, ATR-Trailed Winners)

A LONG-SHORT, SINGLE-INSTRUMENT, pure-OHLCV volatility-regime breakout strategy on SOLUSD.HYPERLIQUID perpetual futures (1-HOUR bars). It targets the Hyperliquid venue, which is severely under-represented in the portfolio (5.6% vs the >=20% live-trading quota) and where the factory is actively shifting live execution. The mechanism is deliberately DIFFERENT from the validated impulse-bar continuation family (SOL/ETH/BTC) already in the book: instead of riding a single large directional bar, it waits for a volatility CONTRACTION regime (range squeeze) and then trades the volume-confirmed expansion break in either direction. Volatility clusters and mean-reverts in crypto, so a local minimum in realized range reliably precedes an expansion; the direction of the first clean break out of the squeeze box carries short-term momentum because resting stops cluster just outside the range and breakout/stop-run flow produces follow-through. Pure OHLCV means zero supplementary-data dependency, sidestepping the HL funding/OI data-availability failures seen earlier. 1H bars (rather than 4H/1D) also produce far more bars per calendar day, keeping the strategy clear of the HL data-span insufficiency that previously failed slower-timeframe HL strategies. Kept intentionally low-parameter (4 core params) to resist the overfitting that has been the recurring abandonment cause.

Targets the under-represented Hyperliquid venue (5.6% vs the >=20% live quota, where the factory is shifting live execution) with a mechanism deliberately DISTINCT from the validated impulse-bar continuation family already in the book: instead of riding a single large directional bar, it conditions on the volatility-regime cycle (realized vol clusters and mean-reverts, so a local minimum in realized range reliably precedes an expansion) and trades the first clean break out of the squeeze box. Direction carries short-term momentum because resting stops cluster just outside the range and breakout/stop-run flow produces follow-through; the volume confirmation ensures the break is driven by real flow, not a low-volume drift. Pure OHLCV means ZERO supplementary-data dependency, sidestepping the HL funding/OI data-availability failures; 1H bars (not 4H/1D) produce far more bars per calendar day, clearing the HL data-span insufficiency that previously failed slower-timeframe HL strategies. Per-move targets are multi-percent vol-expansion legs, an order of magnitude above the ~0.09% HL round-trip fee. Engine-safe: discrete enter-once/exit-once, single position, hard-capped sizing, no continuous rebalancing. Kept to 4 core parameters to resist overfitting. leverage=1.0 with notional capped at equity, so no unused-leverage gate triggers.

Multiple independent, non-tunable disqualifiers. (1) CAPACITY-DESTROYED: capacity_usd is just $2,231 and impact_cost_pct is 669% — the modeled market impact is ~7x the gross PnL, so HL SOL 1H liquidity cannot absorb the strategy's sizing (avg_position_pct 51.7%); the edge is exhausted at toy scale and would never survive real execution. (2) NEGATIVE EDGE before impact: profit_factor 0.92, Sharpe -0.74, expectancy -$52/trade, win rate 32% — it loses. (3) INSUFFICIENT DATA: only ~7 months of HL SOL 1H history (start 2025-10-11, data_days 95) and 50 trades — far too little to optimize or validate via walk-forward/DSR; this is the HL data-span insufficiency the hypothesis claimed to sidestep by using 1H, but 1H bars do not manufacture more calendar history. (4) MECHANISM ALREADY REFUTED: the volatility-squeeze breakout is the same family as the ETH 4H squeeze abandoned this session (outlier-dependent, fails deflated-Sharpe), and porting it to a thinner-history, thinner-liquidity HL alt is strictly worse. None of these is fixable by tuning the 4 parameters: optimization cannot create liquidity (capacity), cannot turn a negative-expectancy losing edge positive, and cannot validate on 50 trades over 7 months. FAILURE PATTERN: single-instrument volatility-squeeze breakout on a Hyperliquid alt perp (SOL 1H) is capacity-destroyed (capacity ~$2k, impact 669% of gross) and data-starved (~7 months history), on top of carrying the same outlier-dependent squeeze-breakout edge that already failed deflation on ETH — the HL alt-perp venue lacks both the liquidity and the price history to support an intraday breakout strategy at any deployable size.

Long-short volatility-squeeze breakout on SOLUSD.HYPERLIQUID perpetual (1-HOUR bars), single-instrument, pure OHLCV. It waits for a volatility CONTRACTION regime -- a range squeeze where short-window realized vol (std of returns over coil_lookback bars) is below squeeze_ratio x the longer baseline vol (over coil_lookback*baseline_mult bars), both measured EXCLUDING the current bar so the expansion bar does not mask the contraction -- and then trades the VOLUME-CONFIRMED directional EXPANSION: when the current bar volume exceeds vol_mult x its trailing average AND price closes beyond the squeeze-box Donchian range (max-high/min-low of the prior coil_lookback bars), it enters in the breakout direction (BUY on an upside break, SELL on a downside break). calculate_signal returns a continuous signed breakout score = (close - coil_mid)/ATR(14); should_enter applies the squeeze + volume + channel-break gates. Exit is a pure ATR trailing stop (stop_atr_mult x ATR off the favorable extreme, seeded at entry so it also serves as the initial stop). Single position, flat between signals. Sizing risks risk_frac of equity over the stop distance, notional capped at equity. 4 core tunables (coil_lookback, squeeze_ratio, vol_mult, stop_atr_mult); atr_period, baseline multiplier and volume window are frozen.

Clean, low-parameter implementation; targets the under-represented Hyperliquid venue; pure-OHLCV (no supplementary-data dependency); balanced long-short.

Correct convex breakout shape when it wins (avg_win $1,760 vs avg_loss $905, skew +1.9) — the mechanism behaves as designed even though it doesn't profit.

Catastrophic capacity/impact: capacity_usd $2,231 and impact_cost_pct 669% (impact ~7x gross PnL) — untradeable at any real scale; HL SOL 1H liquidity cannot absorb the sizing (avg_position_pct 51.7%).

Negative edge: profit_factor 0.92, Sharpe -0.74, expectancy -$52/trade, win rate 32% — losing before impact is even considered.

Insufficient data: ~7 months (start 2025-10-11), data_days 95, only 50 trades — far too little for walk-forward/DSR validation; the HL data-span insufficiency the hypothesis claimed to avoid.

Mechanism already refuted: volatility-squeeze breakout failed on ETH 4H (outlier-dependent, DSR-failing); porting it to a thinner-history, thinner-liquidity HL alt is worse.

deployable scale

~7 months / 50 trades

multi-regime, enough for walk-forward

HyperliquidSolVolatilitySqueezeBreakoutLS1H applied a volatility-contraction-then-volume-confirmed-break mechanism to SOL on Hyperliquid, aiming to fill the under-represented HL venue with a low-parameter design distinct from the impulse-continuation family. The breakout behaved as designed (convex wins, +1.9 skew) but lost money — profit factor 0.92, Sharpe -0.74, -$52 expectancy — and, decisively, was untradeable: $2,231 capacity with impact cost at 669% of gross. The analyst stacked four non-tunable disqualifiers (capacity-destroyed, negative edge, ~7-month/50-trade data starvation, and a squeeze-breakout family already refuted on ETH), none fixable by the four parameters. It ended after one iteration as abandoned, never advancing to optimization or risk review.

A single-instrument intraday breakout on a Hyperliquid alt perp lacks both the liquidity and the price history to be deployable: capacity ~$2k with 669% impact means no parameter tuning can manufacture liquidity, dropping to 1H does not create more calendar history, and porting the already-DSR-failed squeeze-breakout family to a thinner, newer venue is strictly worse.

It was abandoned at the pre-optimization backtest-review gate (verdict: abandon) on multiple non-tunable disqualifiers — capacity-destroyed (impact ~7x gross), negative edge before impact, only ~7 months / 50 trades (too thin for walk-forward/DSR), and the same volatility-squeeze breakout family already refuted on ETH 4H — so optimization and all later stages were never reached.

A single-instrument, pure-OHLCV long/short volatility-squeeze breakout on SOLUSD.HYPERLIQUID 1H perps — waiting for a realized-volatility contraction regime (short-window vol below a fraction of its longer baseline) then trading the volume-confirmed Donchian-range break in either direction with an ATR trailing stop — targeting the under-represented Hyperliquid venue with a deliberately low-parameter (4 core) design distinct from the validated impulse-continuation family.

Over only ~95 days (start 2025-10-11) and 50 trades (25 long / 25 short) it had a negative edge even before liquidity costs — profit factor 0.92, Sharpe -0.74 (CI [-4.61, 2.28]), expectancy -$52/trade, win rate 32%, with correct convex breakout shape (avg_win $1,760 vs avg_loss $905, skew +1.9) — but was capacity-destroyed: capacity_usd just $2,231 and impact cost 669% of gross PnL (impact ~7x gross) at avg_position_pct 51.7%.