EMA Cross Strategy (ema_cross) by Frogleim
By Frogleim
Performance Metrics
- Author: Frogleim
- Symbol: BINANCE:SOLUSDT.P
- Timeframe: 15 minutes
- Win Rate: 33.3%
- Profit Factor: 0.838
Description
A trend-following strategy built around a 50/200 EMA cross, filtered by RSI momentum, volume confirmation, and trend strength — designed to avoid low-conviction chop and only fire on trades with volume and momentum behind them.How it works:-Trend detection: Fast/slow EMA (default 50/200) establishes directional bias-Momentum filter: RSI must confirm overbought/oversold conditions in the direction of the trade-Volume confirmation: Entries require volume to exceed a multiple of its moving average, filtering out low-participation moves-Trend strength filter: A minimum EMA spread threshold avoids entries during flat/ranging conditions-Volatility-adaptive TP/SL: Take-profit and stop-loss percentages scale with a short/long ATR ratio, widening in high-volatility regimes and tightening in calm ones-Volatility-adaptive position sizing: Position size scales inversely with ATR%, so size shrinks automatically in choppy/volatile conditions-Risk controls: Daily trade cap, daily loss cap (% of equity), and an automatic pause after a configurable streak of consecutive take-profit wins (to avoid over-trading a hot streak into reversal)-Session filter: Optional block on entries during a low-liquidity UTC windowInputs are fully configurable — EMA/RSI/ATR lengths, volume and trend thresholds, TP/SL bounds, position sizing caps, leverage, and daily risk limits can all be adjusted to fit different symbols and timeframes.Notes:-Backtested with a 0.04% commission and 1-tick slippage assumption to reflect realistic futures execution costs-Designed for use with webhook-based automation (alerts fire structured JSON on entry/exit) but works standalone as a manual signal tool-This is a systematic strategy, not financial advice — past performance in the strategy tester does not guarantee future results. Test on paper/small size before live deployment.