Examples

To get you started we prepared a few examples of how to approach certain scenarios often used in trading.

For working strategies check out these repositories:

• https://github.com/jesse-ai/example-strategies
• https://github.com/nicolay-zlobin/jesse-indicators

Crossovers

Let's try to catch the moment the price moves above the bollinger middleband.

@property  
def long_cross(self):  
    return self.price > self.bb.middleband[-1] and self.candles[:, 2][-2] <= self.bb.middleband[-2]  
  
@property  
def bb(self):  
    return ta.bollinger_bands(self.candles, sequential=True)

As alternative you could use the crossed utility:

@property  
def long_cross(self):  
    return utils.crossed(self.candles[:, 2], self.bb.middleband, 'above')  
  
@property  
def bb(self):  
    return ta.bollinger_bands(self.candles, sequential=True)

Stoploss / Take Profit

Simple ATR:

def go_long(self):  
    take_profit = self.price + self.atr * 3  
    stop = self.price - self.atr * 2  
    qty = 10
    self.buy = qty, self.price  
    self.stop_loss = qty, stop  
    self.take_profit = qty, take_profit  
  
  
@property  
def atr(self):  
    return ta.atr(self.candles, period=22)

A trailing stop:

def update_position(self):  
    # update trailing_stop_loss only if in profit  
    if self.position.pnl > 0:  
        if self.is_long:  
            self.stop_loss = self.position.qty, self.price - self.atr * 2  
        else:  
            self.stop_loss = self.position.qty, self.price + self.atr * 2

You also can use Moving Averages to exit once they are hit.

def update_position(self):  
    if self.is_long and self.price <= self.exit_ema:  
       self.liquidate()  
       
@property  
def exit_ema(self):  
    return ta.ema(self.candles)

Special indicators you might want to try for your exits:

• safezonestop
• devstop
• kaufmanstop

Channels can also be good for stoploss or take profit:

• Keltner
• Donchian
• Bollinger Bands

Getting the size right

This example might help with the position size/qty. We use risk management. There might be situations where risk_to_qty returns a qty exceeding the available capital leading to an exception. The reason for this is a very close stop loss (often due to the usage of the ATR). That's not an error, but the expected behavior of the formula. We add a logic limiting the qty to a maximum percentage (in this case 25 %) of the capital for those cases.

  
def go_long(self):  
    stop = self.bb.lowerband
    qty = self.position_size(self.price, stop)  
    take_profit = self.bb.upperband
    self.buy = qty, self.price  
    self.stop_loss = qty, stop  
    self.take_profit = qty, take_profit  
  
@property  
def position_size(self, entry, stop):  
    # risk 10%
    risk_qty = utils.risk_to_qty(self.balance, 10, entry, stop, fee_rate=self.fee_rate)  
    # never risk more than 25% of the capital
    max_qty = utils.size_to_qty(0.25 * self.balance, entry, precision=6, fee_rate=self.fee_rate)  
    qty = min(risk_qty, max_qty) 
    return qty

Using the kelly criterion:

def kelly_qty(self, entry, stop):  
    if not self.metrics or self.metrics['total'] < 20:  
        win_rate = 0.46  
        avg_win_ratio = 1.1
        avg_loss_ratio = 0.5
    else:  
        win_rate = self.metrics['win_rate']  
        avg_win_ratio = self.metrics['avg_win_percentage'] / 100
        avg_loss_ratio = self.metrics['avg_loss_percentage'] / 100
    kc = utils.kelly_criterion(win_rate, avg_win_ratio, avg_loss_ratio) * 100  
    if not kc or kc <= 0:  
        raise ValueError("Bad Kelly criterion.")  
    risk_qty = utils.risk_to_qty(self.available_margin, kc, entry, stop, self.fee_rate)  
    # never risk more than 25%  
    max_qty = utils.size_to_qty(0.25 * self.available_margin, entry, precision=6, fee_rate=self.fee_rate)  
    qty = min(risk_qty, max_qty)  
    return qty

We need to check for a minimum of trades so we have a good win_rate and ratio_avg_win_loss to work with. In this case, we use 20 trades. For those first trades we use hardcoded values, we got from backtests. win_rate is the same as Percent Profitable / 100. ratio_avg_win_loss is called Ratio Avg Win / Avg Loss.