How to entry trade on next candle?

[ghost]

What I observe from using backtesting.py is that long and short trades are executed on the same candle. Like in the given 'SMA-Crossover' example a short trade is closed and a long immediately starts.

Since I use freqtrade which does not support executing both signals on the same candle I was wondering how to implement a rule that tells backpy to entry a trade on the next candle after closing the previous trade?

[kernc]

IOW, you:

self.position.close()
self.buy(...)

This places two orders to be executed when the next candle opens.

If you want to skip a candle, simply consider it all as a state machine.

[ghost]

I must admit that I dont understand your hint. Ive extracted four (2 short, 2 long) trades. It seems that the position.close() and buy() orders are executed on the very same candle.

Size  EntryBar  ExitBar 
-3      20           35 
 3      35           50
-3      50           62
 3      62           64

My code is stated below. So what can I do for skipping buy/sell orders after closing the previous one?

  if not self.position:
            if crossover(self.sma1, self.sma2):
                self.buy()
            elif crossover(self.sma2, self.sma1):
                self.sell()
  elif self.position:
            if self.position.is_long:
                if crossover(self.sma2, self.sma1):
                    self.position.close()
                    self.sell()
            elif self.position.is_short:
                if crossover(self.sma1, self.sma2):
                    self.position.close()
                    self.buy()`

Thank you for your help!

[kernc]

class S(Strategy):
    def init(self):
        self.state_variable = None
    
    def next(self):
        # State checking needs to come before
        # state setting to be executed on next `next`.
        # On the next next, the order will be placed,
        # to be executed on next**2 `next`.
        if self.state_variable:
            if self.state_variable == 'should_buy':
                self.buy()
            elif self.state_variable == 'should_sell':
                self.sell()
            self.state_variable = None
    
        # Set state:
        if position.is_long and crossover(...):
            self.position.close()
            self.state_variable = 'should_sell'
        ...

An even better alternative might be to precompute your crossovers (with required lags and derivatives) in vectorized form in init().

[ghost]

class macross(Strategy):
    fast = 3
    slow = 12
    
    def init(self):

        self.mafast = self.I(SMA, self.data.Close, self.fast)
        self.maslow = self.I(SMA, self.data.Close, self.slow)
        self.state_variable = None

    def next(self):

        if self.state_variable:
            if self.state_variable == 'should_buy':
                self.buy()
            elif self.state_variable == 'should_sell':
                self.sell()
            self.state_variable = None

            if self.position.is_long and crossover(self.maslow, self.mafast):
                self.condition.close()
                self.state_variable = 'should_sell'

            elif self.position.is_short and crossover(self.mafast, self.maslow):
                self.condition.close()
                self.state_variable = 'should_buy'

I get empty backtest results. No trades were made.

def next(self):

    if self.state_variable:
         if self.state_variable == 'should_buy':
             self.buy()
         elif self.state_variable == 'should_sell':
             self.sell()
         self.state_variable = None

     if not self.position:
         if crossover(self.mafast, self.maslow):
             self.state_variable = 'should_buy'
         elif crossover(self.maslow, self.mafast):
             self.state_variable == 'should_sell'

     elif self.position:
         if self.position.is_long and crossover(self.maslow, self.mafast):
             self.condition.close()
             self.state_variable = 'should_sell'

         elif self.position.is_short and crossover(self.mafast, self.maslow):
             self.condition.close()
             self.state_variable = 'should_buy'

I get error: macross object has no attribute 'condition'.

Dont get my head around it. Can you help a third time?

[kernc]

Apologies, the correct code is obviously self.position.close().

[ghost]

Your solution worked out. This topic can be closed.

Many thanks to you!