Strategies

limin TqSdk Turtle: Faithful 20/10-Day Breakout and Frozen-N Pyramiding on LTC Hyperliquid

Outcome: Abandoned

LiminLtcTickTurtle

Outcome Summary

LiminLtcTickTurtle tried to replicate limin's Apache-2.0 TqSdk Turtle example on Hyperliquid LTC, executing on trade ticks and keeping N and unit size fixed for the length of each campaign. Over 3 iterations the code never got past verification. Every Layer-2 synthetic scenario crashed at setup on a self-imposed check that parameters.leverage must equal config.leverage. The pipeline abandoned the run in the verification loop, so backtesting, optimization, analyst review and risk review never ran.

Strategy report

Hypothesis

Translate limin's Apache-2.0 TqSdk Turtle demonstration at the supplied commit, preserving its actual rules rather than substituting textbook Turtle rules. Replace the original expired SHFE.au2006 gold contract with LTCUSD.HYPERLIQUID, a linear perpetual permitting both directions. Defaults are donchian_channel_open_position=20, donchian_channel_stop_profit=10, atr_day_length=20 and max_risk_ratio=0.5. Maintain a 100-slot daily candle buffer, including the developing current candle. Execution responds to changed last-trade prices, not daily closes. Start with target position=0 and last_price=NaN. Run parameter recalculation initially and subsequently only when a new daily candle starts WHILE FLAT. While invested, N and unit remain frozen, but the rolling 10-completed-day exit channel continues changing. Historical trade ticks are required to reconstruct the developing candle and execution sequence; daily OHLCV alone cannot faithfully reproduce this example. Missing adequate tick history makes the faithful experiment infeasible, rather than authorizing a close-only substitute.

The explicit request for faithful reproduction takes precedence over the preference against another daily trend strategy. Hyperliquid is underrepresented at 8.0% versus its 20% target, and both trading directions address the portfolio's 82.8% long-only concentration. LTC avoids adding another BTC or ETH experiment; corpus_concentration is empty, so its recent utilization cannot be verified. This preserves source-specific frozen-N sizing and quote-event sequencing, but remains in the broad Turtle family already represented by EtcTurtleDonchianBreakoutLS. It is a requested reproduction, not a claim of novel mechanism. Do not adopt performance claims or introduce optimization to rescue the reproduction.

verification_loop: Verification failed (Layer 2 — synthetic scenarios): Parameters used: ['leverage', 'kline_length', 'min_notional', 'atr_day_length', 'max_risk_ratio', 'pyramid_step_n', 'stop_distance_n', '_optimization_plan', 'unit_risk_fraction', 'contract_multiplier', 'donchian_channel_stop_profit', 'donchian_channel_open_position'] Check that __init__ sets all attributes from self.parameters.get(). - steady_uptrend: Setup error: RuntimeError: parameters.leverage must equal config.leverage - steady_downtrend: Setup error: RuntimeError: parameters.leverage must equal config.leverage - flat_ranging: Setup error: RuntimeError: parameters.leverage must equal config.leverage - volatility_spike: Setup error: RuntimeError: parameters.leverage must equal config.leverage - zero_volume: Setup error: RuntimeError: parameters.leverage must equal config.leverage - price_gap: Setup error: RuntimeError: parameters.leverage must equal config.leverage

Implementation

Faithful translation of limin's TqSdk Turtle on LTCUSD.HYPERLIQUID: trade-tick-driven 20-day Donchian breakout entries (long/short), frozen-N unit sizing (1% equity / N, N = simple-MA ATR(20) including the developing candle), 0.5N pyramiding gated by margin ratio <= 0.5, 2N stop from last adjustment price, and 10-completed-day channel exit. Daily candles are rebuilt from trade ticks.

Iteration 3 fixes Layer 1.5 (optimization_plan fixes leverage at 2.0): parameters now include 'leverage': 2.0 (float, type-matched) mirroring config.leverage=2.0; the code reads it, asserts it equals config.leverage, and uses it for the target margin-capacity check. Every other fixed plan value is present with the plan's exact type. No signal, entry, exit or sizing logic changed. Daily bars are reporting-only; execution requires historical trade ticks, and missing tick days fail loudly rather than falling back to closes.

Verification Results

Verification failed (Layer 2 — synthetic scenarios): Parameters used: ['leverage', 'kline_length', 'min_notional', 'atr_day_length', 'max_risk_ratio', 'pyramid_step_n', 'stop_distance_n', '_optimization_plan', 'unit_risk_fraction', 'contract_multiplier', 'donchian_channel_stop_profit', 'donchian_channel_open_position'] Check that __init__ sets all attributes from self.parameters.get(). - steady_uptrend: Setup error: RuntimeError: parameters.leverage must equal config.leverage - steady_downtrend: Setup error: RuntimeError: parameters.leverage must equal config.leverage - flat_ranging: Setup error: RuntimeError: parameters.leverage must equal config.leverage - volatility_spike: Setup error: RuntimeError: parameters.leverage must equal config.leverage - zero_volume: Setup error: RuntimeError: parameters.leverage must equal config.leverage - price_gap: Setup error: RuntimeError: parameters.leverage must equal config.leverage

Iteration History

Verification failed (Layer 1.5 — hypothesis/config consistency) [class=hypothesis_mismatch]: - optimization_plan fixes leverage at 2.0

Outcome Summary

Do not add hard constructor assertions that tie parameters to config values the verification harness may set differently; read leverage directly from self.config.leverage so the strategy can start in the synthetic and sandbox environments.

After 3 iterations the run was abandoned in the verification loop at Layer 2 (synthetic scenarios). All six scenarios failed at setup with 'RuntimeError: parameters.leverage must equal config.leverage', because the strategy's own __init__ guard raised an error whenever the leverage parameter did not match the harness config.

A faithful translation of limin's TqSdk Turtle demo to LTCUSD.HYPERLIQUID: tick-driven entries on a 20-day Donchian breakout, exits on a 10-day channel, and pyramiding sized by an N (20-day ATR) and unit that stay frozen while a position is open, with max_risk_ratio=0.5 and leverage fixed at 2.0.

No performance data exists: the backtest report is empty and there is no optimization report, so there are no return, Sharpe or trade-count figures.

Abandon Reason

Analysis

Turtle

Backtest and paper results are hypothetical. Trading involves risk of loss.