TRADLEWARE-Gaussian Channel + StochRSI ETH — Strategy by cs_lev
By cs_lev
Performance Metrics
- Author: cs_lev
- Symbol: BINANCE:ETHUSDT
- Timeframe: 1 day
- Win Rate: 44.1%
- Profit Factor: 5.331
Description
Gaussian Channel + Stochastic RSI ETHThis strategy combines a fast Gaussian Channel with a Stochastic RSI filter and a 200-day SMA bull-market gate, aimed at catching trend continuation while sitting out confirmed downtrends.How it worksThe Gaussian Channel is a smoothed price envelope built with an IIR (infinite impulse response) filter — a mathematically elegant alternative to a simple moving average. It applies a bell-curve weighting across recent bars, producing smooth, low-lag output. The channel is formed by adding and subtracting a filtered measure of true range (volatility) around the central filter line.The channel turns green when the filter is rising (uptrend) and red when it is falling (downtrend). A separate 200-day simple moving average acts as a bull/bear regime switch: the strategy only trades when price is above it.EntryA long position is opened when all five conditions are true simultaneously: The channel is green (filter rising — uptrend confirmed) Price closes above the upper band (breakout above the channel; an optional buffer above the band can require more room, but testing found this counterproductive — see Parameters) Stochastic RSI %K is either above 80 (strong momentum confirming the breakout) or below 25 (oversold dip within the uptrend) Price is above the 200-day SMA (bull regime — can be disabled) The signal bar itself closes above its own open — a bullish candle (can be disabled)The bullish-candle check filters out breakout bars that clear the upper band intrabar but still close weak — a common precursor to an immediate whipsaw exit on the next bar.The 200-day SMA gate exists specifically to block breakout entries that fire during bear-market bounces — dead-cat rallies that look like trend resumption on the channel and oscillator alone but occur underneath a still-falling long-term average.ExitThe position is closed when either: Price closes back below the upper band (breakout has failed or the trend is cooling), or The channel reverses from green to red (trend direction has flipped)An optional stop-loss (on by default) is placed at the lower band and trails as the channel moves, providing a floor on losses if price drops sharply through both the upper and lower bands in the same move. The regime gate only blocks new entries — it does not force an exit on its own if price falls back below the 200-SMA mid-trade.Parameters Poles: 4 (filter smoothness — higher = smoother but more lag) Sampling Period: 89 (faster channel than the baseline version, reacts sooner to trend changes) True Range Multiplier: 1.5 (controls channel width) Stochastic RSI overbought threshold: 80 Stochastic RSI oversold threshold: 25 (a parameter sweep found a stable plateau from 22-28; 25 sits at its center rather than its single best value) 200-SMA regime gate: on by default, can be disabled; length is adjustable Bullish entry candle requirement: on by default, can be disabled Entry breakout buffer: 0% (off) by default; tested at multiple levels above 0% and found to reduce returns at every level, so left disabled Stop-loss at lower band: on by default, can be disabled Start/End date range inputs let you restrict the backtest window without editing codeCosts modelled0.1% commission per side, 3 ticks slippage, fills at next bar's open.Intended assets and timeframeDaily bars. Designed and validated on ETH/USDT. Likely applicable to other trending crypto assets; not validated on equities.Known limitationsUnderperforms in choppy or ranging markets — the upper band breakout condition generates whipsaws when price oscillates without directional conviction. The regime gate is a trade-off: it blocks bear-bounce false starts, but it also means the strategy can miss the first leg of a genuine new uptrend until price reclaims the 200-day SMA. The filter requires several hundred bars of history to fully converge; results on very short histories may differ from the validated backtest. The strategy trades infrequently (around 28 trades on the validated window), so treat any single backtest run as a small sample rather than a statistically strong result.CreditThe Gaussian Channel filter is from the open-source "Gaussian Channel (DW)" indicator by DonovanWall. This script reuses that filter and adds the Stochastic RSI entry filter, the 200-day SMA regime gate, exit rules, stop-loss, and full strategy order management on top of it.