Hyperliquid LTC/BCH 4H Confirmed Swing-Structure Trend Following

Hypothesis

Derived by the Research Lead from 发明者量化-小小梦’s FMZ post《期货隔夜交易（趋势交易）的金石良言》: follow established rising or falling peaks and troughs, tolerate ordinary pullbacks, limit losses, and let profitable positions run. This operationalization trades LTC and BCH perpetuals independently in both directions on 4-HOUR bars, holding across UTC days without a daily flattening rule. The Research Lead chooses causal swing confirmation, structure-break entries, equity-relative sizing, intrabar protective stops, and delayed trailing protection. These executable rules are not supplied by the author, and the source’s performance assertions are not adopted.

The concentration map is empty, so instrument-level usage cannot be established. Hyperliquid is under its 20% target at 10.3%, and long-short trading addresses the largest directional imbalance: 80.8% long-only. Two independently traded instruments reduce reliance on one series. This tests directional swing structure rather than the failed LTC/BCH relative-return or PCA-reversion mechanisms. Four-hour bars respect Hyperliquid’s history limit. Options and cross-venue execution would add mechanisms absent from this essay, so they are not introduced merely to satisfy quotas.

negative_expectancy: decline optimization of this Research Lead operationalization of the FMZ essay. Across 58 closed positions, both instruments lose substantially before execution costs, and net per-trade return is -5.064% against the +0.15% venue floor. No measured improvement path justifies a six-parameter optimization. This verdict covers the specified LTC/BCH 4H confirmed-swing breakout with delayed trailing protection, not the source author's unspecified strategy or trend following generally.

Implementation

Trades LTC and BCH Hyperliquid perpetuals independently on 4H bars. Long entries require two rising confirmed highs and lows plus a close crossing above the latest swing high; shorts require falling highs and lows plus a close crossing below the latest swing low. Each entry risks up to 0.5% of settled equity, with 30% per-leg and 60% gross entry-notional caps. Intrabar stops start beyond the opposite confirmed swing with an ATR buffer. After a favorable excursion of 3R, a 3.5-ATR trailing stop ratchets protectively. No pyramiding, profit target or daily flattening.

Implements the Research Lead's operationalization with strictly causal pivots confirmed after pivot_width subsequent bars. Both legs have separate indicator, position, stop and cooldown state; no relative-value signal is introduced. BCH orders use only a bar matching the primary timestamp, preventing future secondary bars from influencing decisions. Bounded pivot buffers and incremental Wilder ATR keep per-bar computation bounded. Tunable names, baseline values and research domains are preserved without silent parameter clamps. Protective orders are placed after entry fills, and trailing levels apply prospectively. Verification remains the pipeline's responsibility. The required outbox file could not be written because this session's filesystem permissions are read-only; the complete bound JSON is returned here.

Verification Results

The sandbox reports -4.98% average net trade return and profit factor 0.379 across 58 closed positions. Full-history review should assess breakout follow-through and delayed trailing protection; these smoke-test losses do not establish a correctness defect.

Backtest Review

Sharpe

-1.93

Total return

-14.83%

Max drawdown

17.54%

Trades

58

Win rate

15.5%

Profit factor

0.37

Both bar inputs pass integrity audits; accounting reconciles and all 58 positions are closed.

Trade records show both directions and multi-day holdings, consistent with the intended independent trend trades.

Average net trade return is -5.064% versus the +0.15% HYPERLIQUID review floor; profit factor is 0.370 and total return is -14.834%.

Both instruments lose: BCH -$7,350 across 36 positions; LTC -$7,484 across 22. Aggregate price PnL is already -$14,503, so execution costs are not the primary cause.

All reported calendar years and volatility regimes lose. QA's economic concern is borne out.

No optimization sweep has been run; these results do not establish that every alternative swing-trend specification fails. Benchmark comparison is inapplicable because benchmark_meaningful is false.

Outcome Summary

This specific swing-breakout implementation needs a measured improvement path for its negative price expectancy before parameter optimization is justified.

Backtest review chose abandon because both instruments lost substantially before execution costs and average net trade return fell below the +0.15% venue floor. Optimization and subsequent analyst and risk reviews were not reached.

Trade LTC and BCH Hyperliquid perpetuals independently in both directions using confirmed 4-hour swing breakouts, protective stops, and delayed trailing protection.

The backtest returned -14.834% with a Sharpe ratio of -1.931 across 58 closed positions and a maximum drawdown of 17.540%. Average net trade return was -5.064%, and profit factor was 0.370.

Analysis

期货隔夜交易（趋势交易）的金石良言