Connor McDonald Jesse MACD_EMA — LTC Spot 1H, Frozen Source Defaults

Hypothesis

Faithful reproduction of Connor McDonald's MACD_EMA, credited as MIT-licensed in the submission. Use LTCUSDT.BINANCE_SPOT with 1-hour OHLCV bars, long-only, no leverage. Preserve defaults ema=100, fastperiod=12, slowperiod=26 and signalperiod=9. The executable rules test indicator states, despite comments describing crossings. Preserve all-balance sizing and the absence of protective stops, profit targets, trailing stops and holding deadlines. Performance is unverified. Source: [pinned MACD_EMA code](https://github.com/jesse-ai/example-strategies/blob/7c91e0a37bf62165790120d730442e4f6eb00364/MACD_EMA/__init__.py).

The explicit request for a faithful source reproduction takes precedence over introducing a new mechanism or risk overlay. Spot is appropriate because the source never shorts or requests leverage, and its 11.0% portfolio share is below the 15% target. LTC avoids the heavily tested BTC series and the named pending ports. The supplied corpus_concentration is empty, so LTC's recent experimental share and novelty cannot be established. This remains a generic indicator strategy in an overrepresented single-symbol, long-only bucket; selecting LTC does not establish an edge.

negative_expectancy: decline optimization of this frozen LTC spot reproduction. Executed trades lose materially after costs, including before reported liquidation, and the contract permits no parameter search. This verdict concerns this LTC adaptation, not the source strategy on other markets. Independently review post-liquidation trading and inconsistent position metrics.

Implementation

Long-only LTC spot strategy on hourly bars. Enters when close exceeds EMA(100) and MACD(12,26,9) exceeds its signal; exits when both inequalities reverse. Uses the entire balance less the source fee reserve, with no protective exits.

Addresses Layer 1.5 feedback by explicitly setting leverage to 1.0 in both config and parameters and preserving the exact frozen research plan. Existing indicator, entry, exit and sizing implementations are unchanged; only the source header is clarified to avoid a byte-identical code submission. FactoryStrategy market-order helpers provide the platform's bar-level execution equivalent. Performance remains unverified. The required outbox file could not be written because this session has read-only filesystem permissions; the complete task-bound JSON is returned here.

Verification Results

Compare each entry notional with the contemporaneous USDT balance and reconcile commissions, modeled impact, and realized PnL against ending cash before trusting the bankruptcy and exposure diagnostics.

The sandbox reports average position exposure of 326% and bankruptcy despite this sizing expression requesting at most 99.7% of the stable balance while flat. These diagnostics require account-ledger reconciliation; the supplied metrics alone do not establish excess sizing in this strategy.

The sandbox reports -0.626% per closed trade over 331 positions and PF 0.593. Full-history evaluation should examine whether this fixed MACD/EMA conjunction clears the approximately 0.20% spot round-trip commission. Sandbox losses do not establish a coding defect.

Backtest Review

Sharpe

-2.12

Total return

-100.00%

Max drawdown

100.00%

Trades

1206

Win rate

20.1%

Profit factor

0.76

The supplied code preserves long-only MACD/EMA state conjunctions and frozen defaults; all 1,206 recorded positions are long.

QA's economic concern is borne out: average net trade return is -0.3934% versus the 0.25% spot review floor, with PF 0.7635.

Losses precede reported liquidation: 116 trades closed before that date have aggregate recorded PnL less commissions of approximately -115,459 USDT and PF 0.411; this calculation excludes any separately charged impact.

Trades continue after reported liquidation, while average position size is reported as 405.8% despite peak entry leverage below 1. These reporting/accounting inconsistencies require separate verification; their mechanism is unverified.

Buy-and-hold ends down 78.79%, versus the reported strategy loss of 100%. No tunable parameters are authorized.

Iteration History

Verification failed (Layer 1.5 — hypothesis/config consistency) [class=hypothesis_mismatch]: - optimization_plan fixes leverage at 1.0

Outcome Summary

This LTC adaptation showed negative expectancy after costs; its accounting inconsistencies need independent verification, and its outcome does not establish performance on other markets.

Backtest review voted abandon because executed trades lost materially after costs, including before reported liquidation, and frozen parameters permitted no search. Optimization and subsequent review or promotion stages were not reached.

Reproduce Connor McDonald's frozen MACD/EMA defaults on LTC spot hourly bars, using long-only all-balance sizing without leverage or protective exits.

The backtest reported -100% return, Sharpe -2.1184, and 1,206 closed positions, with average net trade return -0.3934% and profit factor 0.7635. Post-liquidation trading and inconsistent position-size metrics require separate verification.

Analysis

MACD_EMA