Skip to content
All hypotheses

Hypotheses

CEZLSMA port (geraked, MIT): Chandelier Exit flip + Heikin-Ashi close vs ZLSMA(50), 50% trailing stop, 1.5x grid adds, long-short on LTCUSDT.BINANCE 15m

A line-for-line port of geraked's MetaTrader 5 expert CEZLSMA.mq5 v1.6 (MIT licence, github.com/geraked/metatrader5, commit d3eb29c). The original ran on AUDUSD M15. Credit for the strategy goes to the author.…

Hypothesis

A line-for-line port of geraked's MetaTrader 5 expert CEZLSMA.mq5 v1.6 (MIT licence, github.com/geraked/metatrader5, commit d3eb29c). The original ran on AUDUSD M15. Credit for the strategy goes to the author. Indicators are read once per new bar, from the last CLOSED bar [1] and the bar before it [2]. (1) Heikin-Ashi close HA_C = (O+H+L+C)/4 of the regular candle. (2) Chandelier Exit with ATR period 1 and multiplier 0.75: longStop = highest(1) - 0.75*ATR(1), shortStop = lowest(1) + 0.75*ATR(1), each ratcheted……Show moreShow less

A line-for-line port of geraked's MetaTrader 5 expert CEZLSMA.mq5 v1.6 (MIT licence, github.com/geraked/metatrader5, commit d3eb29c). The original ran on AUDUSD M15. Credit for the strategy goes to the author. Indicators are read once per new bar, from the last CLOSED bar [1] and the bar before it [2]. (1) Heikin-Ashi close HA_C = (O+H+L+C)/4 of the regular candle. (2) Chandelier Exit with ATR period 1 and multiplier 0.75: longStop = highest(1) - 0.75*ATR(1), shortStop = lowest(1) + 0.75*ATR(1), each ratcheted the usual way. Direction flips to long when close > the prior shortStop and to short when close < the prior longStop. The buy buffer CE_B holds the longStop value ONLY on the bar where direction flips to long (0 otherwise); CE_S does the same for the sell side. The developer should mirror geraked's Indicators/ChandelierExit.mq5 exactly, including its useClose setting. (3) ZLSMA(50): lsma = linreg(src, 50, 0), lsma2 = linreg(lsma, 50, 0), ZLSMA = 2*lsma - lsma2. The EA passes a second argument 'true', which in geraked's ZLSMA.mq5 selects the Heikin-Ashi close as the source. Mirror that file. Venue translation: the FX source maps to BINANCE USD-M perps, because the strategy is long-short and uses margin (MarginLimit 300%). LTCUSDT was chosen because it is a liquid, low-share instrument in our corpus with multi-year 15m history on Binance (15m is not viable on Hyperliquid, and OKX 15m depth is unverified). BINANCE is over quota, and we accept that because of the 15m history requirement. Point-based distances are translated relative to price: SLDev = 650 points on AUDUSD (about 0.0065 at ~0.68) becomes 0.95% of price. Settings that are off in the source stay off: News=false, SpreadLimit=-1, EquityDrawdownLimit=0, MultipleOpenPos=false, Reverse=false, GridTrailingStopLevel=0. Martingale on loss is off by default. Fees: Binance USD-M taker 0.05% per side on every market order, including each grid add. Caution: faithfulness carries the source's risks. Our distilled lessons L167 (faithful ports of FX/zero-fee scripts die) and L171 (grid/inventory designs die) apply directly. This is pre-registered as a faithful replication for the backtest to judge, not a claim of edge.

Analysis

CEZLSMA

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