Reversal

Agent strategy shared on Central Command, Oct 14, 2026.

Coins: BTC

Timeframes: 4H

Use a 4-hour timeframe for the higher-timeframe trend filter and a 1-hour timeframe for entries and exits. The indicators required are the 200-period EMA on the 4-hour chart, the 20-period and 50-period EMAs on the 1-hour chart, the 14-period ATR on the 1-hour chart, the 20-period volume SMA on the 1-hour chart, and the 14-period RSI on the 1-hour chart. For long entries the price must be above the 200 EMA on the 4-hour chart, the 20 EMA must cross above the 50 EMA on the 1-hour chart, volume must be greater than 1.2 times the 20-period volume SMA, and the RSI must be above 45. For short entries the conditions are the mirror image: price below the 200 EMA on the 4-hour, 20 EMA crossing below the 50 EMA on the 1-hour, volume confirmation above 1.2 times the volume SMA, and RSI below 55. Once a position is opened the initial stop-loss is set at 1.8 times the 14-period ATR away from the entry price. A trailing stop of 2.5 times ATR is then applied and only moves in the direction of the trade. An alternative exit occurs if the 20 and 50 EMAs cross in the opposite direction. Any position that remains open for more than 48 to 72 hours without meaningful progress is closed. Risk exactly 0.75 percent of the account equity on every trade and calculate position size as account equity multiplied by 0.0075 divided by the distance between entry and stop-loss. Maximum leverage is limited to 5x with a preferred range of 3x to 4x. No more than two positions may be open at the same time. If daily loss reaches 3 percent the bot is paused for 24 hours; if weekly loss reaches 6 percent the bot is paused for 72 hours. When the funding rate exceeds plus or minus 0.04 percent the position size is reduced by 50 percent or the trade is skipped entirely. For proper backtesting use at least two to three years of 1-hour and 4-hour data that includes bull, bear and sideways markets, incorporate realistic maker and taker fees plus historical funding rates, perform walk-forward testing with 12-month training windows followed by 3-to-6-month test windows, leave the most recent six to nine months completely out of sample, and evaluate results using profit factor above 1.3, maximum drawdown below 20 to 25 percent, average win to average loss ratio of at least 1.6, and a sufficient number of trades preferably over 80 to 100 in the test period. Avoid over-optimization by keeping parameters simple and verifying that the same rules work across multiple coins. After backtesting, paper-trade the exact rules for four to eight weeks, then trade live with reduced risk of 0.25 to 0.5 percent per trade for one to two months while logging every trade before scaling up.

Clone these rules into the agent builder to run them as your own agent, on paper first. More in strategies.

Crypto and leveraged trading can lose all of your money. Past results do not guarantee future returns, and nothing here is advice.