Multi-EMA and RSI Trend Following Strategy

Author: ChaoZhang, Date: 2024-03-01 13:26:24



This article mainly analyzes the quantitative trading strategy developed by Ravikant_sharma based on multiple exponential moving averages (EMA) and relative strength index (RSI). The strategy identifies price trends and determines entry and exit points by crossing EMAs with different cycles and values of RSI.

Strategy Principle

Indicator Calculation

The strategy uses 5 EMAs with different periods, including 9-day, 21-day, 51-day, 100-day and 200-day lines. Only the first 4 EMAs are plotted in the code. The RSI parameter is set to 14.

Entry Conditions

One of the following conditions must be met before buying:

  1. The 9-day EMA crosses above the 21-day EMA
  2. The 9-day EMA crosses above the 51-day EMA
  3. The 51-day EMA crosses below the 100-day EMA

At the same time, RSI must be greater than 65, indicating a strong uptrend.

Exit Conditions

One of the following conditions must be met before closing the position:

  1. 9-day EMA crosses below 51-day EMA, indicating trend reversal
  2. The closing price exceeds 125% of the entry price, reaching the profit target
  3. RSI drops below 40, signaling reversal
  4. The closing price falls below 98% of the entry price, stop loss triggered

Advantage Analysis

It is a typical trend following strategy with the following strengths:

  1. Using EMA crossovers to determine trend direction for effective trend tracking
  2. Combining EMAs of different periods identifies more reliable trend signals
  3. RSI filter avoids false signals in range-bound markets
  4. Take profit and stop loss settings lock in profits and control risks

Risks and Solutions

There are still some risks:

  1. Uncertain signals may occur frequently in range-bound markets, causing over-trading. EMA periods and RSI filter conditions can be adjusted.
  2. EMA crossover signals may lag during sharp reversals, unable to exit in time. Other indicators can be added to determine long/short signal strength.
  3. Improper profit target and stop loss settings lead to premature stop loss or failure to lock in profits in time. Parameters should be optimized according to different products and market environments.

Optimization Directions

The strategy can be further optimized in the following ways:

  1. Parameter optimization for different products
  2. Adding other technical indicators to build multifactor models
  3. Incorporating machine learning algorithms to judge signal quality
  4. Combining sentiment analysis to avoid emotional pitfalls
  5. Testing different take profit/stop loss strategies to find optimum


In conclusion, this is an overall reliable and easy-to-implement trend following strategy. With EMA crossover for trend direction and RSI filter for false signals, good backtest results provide a solid foundation for further parameter and model optimization to obtain steady profits. However, traders should still be cautious of sharp reversals and improper parameters that pose risks.

start: 2024-01-30 00:00:00
end: 2024-02-29 00:00:00
period: 3h
basePeriod: 15m
exchanges: [{"eid":"Futures_Binance","currency":"BTC_USDT"}]

// This source code is subject to the terms of the Mozilla Public License 2.0 at
// © Ravikant_sharma


strategy('new', overlay=true)

start = timestamp(1990, 1, 1, 0, 0)
end = timestamp(2043, 12, 12, 23, 59)
ema0 = ta.ema(close, 9)
ema1 = ta.ema(close, 21)
ema2 = ta.ema(close, 51)
ema3 = ta.ema(close, 100)
ema4 = ta.ema(close, 200)

plot(ema0, '9',, 0))
plot(ema1, '21',, 0))
plot(ema2, '51',, 0))
plot(ema3, '200',, 0))   

//plot(ema4, '100',, 0)) 

//LongEntry = (  ta.crossover(ema0,ema3)  or  ta.crossover(ema0,ema2) or  ta.crossunder(ema2,ema3) ) // ta.crossover(ema0,ema1) //
if ta.crossover(ema0,ema1) 
    if rsi2>65
if ta.crossover(ema1,ema2)
    if rsi2>65
LongExit =  ta.crossunder(ema0,ema2) or close >(strategy.position_avg_price*1.25) or rsi2 <40 or close < (strategy.position_avg_price*0.98)

if time >= start and time <= end 
    if(LongEntry and rsi2>60)
        strategy.entry('Long', strategy.long, 1)