Backtrader Community
    • Login
    • Search
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Groups
    • Search
    For code/output blocks: Use ``` (aka backtick or grave accent) in a single line before and after the block. See: http://commonmark.org/help/

    Execute on bid/ask

    General Code/Help
    3
    12
    5748
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • N
      Nigel last edited by

      First of all, many thanks to the developer and the contributors to this project!

      Hopefully somebody can help me with a problem I have and can't seem to be able to fix. I've got intra day bid/ask index data (sampled every 5 seconds for now), and I am trying to figure out how to properly execute on the bid or the offer at the next bid/ask quote coming in after order creation.

      Currently, I can sell the bid (market sell order) on the next quote, but the order to buy the offer is only executed at 23:59:55, at the end of the day (even when order creation was much earlier).

      Here is my data feed class:

      class BidAskCSV(btfeeds.GenericCSVData):
    linesoverride = True
    lines = ('bid', 'ask', 'datetime', 'close', 'open', 'high', 'low')
    params = (('datetime', 0), ('bid', 2), ('ask', 3), ('close', 3), ('open', 2), ('low', 2), ('high',3))

    def __init__(self):
        super(BidAskCSV, self).__init__()

      And some toy order creation/execution logic in a strategy class:

      
        # Check if we are in the market
        position = self.getposition(data=self.datas[0])
        rand = random.random()
        if rand < 0.2:
            if position:
                # SELL (with all possible default parameters)
                self.log('SELL CREATE {}'.format(self.datas[0].bid[0]))

                # Keep track of the created order to avoid a 2nd order
                self.order = self.sell(data=self.datas[0])
        elif rand > 0.8:
            if not position:
                # BUY (with all possible default parameters)
                self.log('BUY CREATE {}'.format(self.datas[0].ask[0]))

                # Keep track of the created order to avoid a 2nd order
                self.order = self.buy(data=self.datas[0], exectype=bt.Order.Close)

      The sell order gets executed on the bid of the next quote. It is a market order and the 'open' price is set to the bid-column of my data in the data feed class. Everything is fine so far....

      But, for buy orders I am trying to take the offer with a 'close' order (the 'close' price being set to the ask column). However, this order gets executed at 23:59:55 and not at the quote following the order generation. I would like it to be executed at the offer of the following quote that comes in.

      Should I be using a different order type, structure the CSV differently, is there an extra parameter to pass, etc? I much appreciate any help!

      Cheers,
      N

      1 Reply Last reply Reply Quote 1
      • B
        backtrader administrators last edited by

        Glad to see that linesoverride has found a use case.

        The problem here is the interpretation of the execution type Order.Close you are making. This is not going to give you the close price of a specific bar. It is meant to give you the close price which corresponds to the end of the session.

        Each broker/platform has a naming for it. Interactive Brokers, for example, calls it Market-On-Close or MOC.

        See the description: https://www.interactivebrokers.com/en/index.php?f=599

        Quoting directly from the BackBroker (the simulation broker in backtrader):

        • Close: meant for intraday in which the order is executed with the closing price of the last bar of the session

        Or directly at the documentation: Docs - Broker

        This order execution type is for intraday data and will be translated to the proper order type if you decide to connect to a live broker (in the aforementioned Interactive Brokers case it will be translated to MOC)

        If you wish to get the close price of the bar at which the order is generated (your ask price) you have to activate the cheat-on-close mode in the broker and then simply issue a Market order, which is the default.

        See here for example: https://community.backtrader.com/topic/15/convincing-strategy-to-buy-at-close-of-current-bar-in-backtest

        Or directly at the docs: https://www.backtrader.com/docu/broker.html?highlight=cheat

        N 1 Reply Last reply Reply Quote 1
        • N
          Nigel @backtrader last edited by

          @backtrader
          Many thanks for your reply! I experimented with the coc parameter:

          cerebro.broker.set_coc(True)

          This does execute a market buy order on the offer, but at the same time will also execute a market sell order on the offer price. However, the behaviour I am after is to execute a buy order on the offer and a sell order on the bid.

          An idea I've got is to just rework my data to represent the midpoint of the bid/ask data, use the cheat-on-close param and allow for the spread in the form of an extra commission charge.

          Cheers,
          N

          1 Reply Last reply Reply Quote 0
          • B
            backtrader administrators last edited by

            This really grants a small addition to the broker, because by activating cheat-on-close you get twice the closing price, and the 2nd price you want is the incoming opening price.

            Let's add the possibility to selectively disable coc for orders. Your sell action would now be as follows:

             self.order = self.sell(data=self.datas[0], coc=False)

            In regular code with coc deactivated, this plays no role. But if you activate it to get the closing price of the bar for your buy, the sell deactivates it to make sure the execution waits until the next bar.

            In the development branch now with this commit id: 959f6cf46623d750c73f75bde9457269b0d4aed2

            1 Reply Last reply Reply Quote 0
            • N
              Nigel last edited by

              This is great, thank you very much!

              1 Reply Last reply Reply Quote 0
              • R
                roch_02 last edited by roch_02

                @Nigel said in Execute on bid/ask:

                BidAskCSV

                Hello,

                I am also using Bid/ask data. Wanted to do a test where I buy and sell like @Nigel did.
                But it seems like the orders are not being executed because when I access the order.Exuecuted.price is always 0.

                from __future__ import (absolute_import, division, print_function,
                        unicode_literals)

import argparse

import backtrader as bt
import backtrader.feeds as btfeeds
#import backtrader.indicators as btind
import random


class BidAskCSV(btfeeds.GenericCSVData):
    linesoverride = True  # discard usual OHLC structure
    # datetime must be present and last
    lines = ('bid', 'ask', 'close', 'open','low', 'high', 'datetime')
    # datetime (always 1st) and then the desired order for
    params = (
        # (datetime, 0), # inherited from parent class
        ('bid', 1),  # default field pos 1
        ('ask', 2),  # default field pos 2
        ('close', 2), ('open', 1), ('low', 1), ('high',2)
    )
    
    def __init__(self):
        super(BidAskCSV, self).__init__()

                The strategy:

                class St(bt.Strategy):

    def log(self, txt, dt= None):
        """Logging function """
        dt= dt or self.datas[0].datetime[0]
        if isinstance(dt, float):
            dt = bt.num2date(dt)
        print('%s, %s' % (dt.isoformat(), txt))


    def __init__(self):
        self.order = None
        self.dataclose= self.datas[0].close
        #self.dataclose = self.datas[0].close[0]
            
            
    def notify_order(self, order):
       if order.status in [order.Submitted, order.Accepted]:
           # Buy/Sell order submitted/accepted to/by broker - Nothing to do
           return 

       # Check if an order has been completed
       # Attention: broker could reject order if not enougth cash
       if order.status in [order.Completed]:
           if order.isbuy():
               self.log(
                   'BUY EXECUTED, Price: %.2f, Cost: %.2f' %
                   (order.executed.price,
                    order.executed.value))

               self.buyprice = order.executed.price
               
           else:  # Sell
               self.log('SELL EXECUTED, Price: %.2f, Cost: %.2f' %
                        (order.executed.price,
                         order.executed.value))

           self.bar_executed = len(self)
       elif order.status in [order.Canceled, order.Margin, order.Rejected]:
           self.log('Order Canceled/Margin/Rejected')

       self.order = None
       
    def notify_trade(self, trade):
        if not trade.isclosed:
            return

        self.log('OPERATION PROFIT, GROSS %.2f, NET %.2f' %
                 (trade.pnl, trade.pnlcomm))
            

    def next(self):
        dtstr = self.datas[0].datetime.datetime().isoformat()
        txt = '%4d: %s - Bid %.4f - %.4f Ask, CLOSE %.4f ' % (
            (len(self), dtstr, self.datas[0].bid[0], self.datas[0].ask[0],
             self.datas[0].close[0]))

        print(txt)
        
        position = self.getposition(data=self.datas[0])
        rand = random.random()
        print('-------------')
        print ('Random:', rand)
        print('-------------')
        if rand < 0.2:
            if position:
                # SELL (with all possible default parameters)
                self.log('SELL CREATE '.format(self.data.bid[0]))

                # Keep track of the created order to avoid a 2nd order
                self.order = self.sell(data=self.datas[0])
                print ('Bid', self.data.bid[0])
        elif rand > 0.8:
            if not position:
                # BUY (with all possible default parameters)
                self.log('BUY CREATE {}'.format(self.data.ask[0]))

                # Keep track of the created order to avoid a 2nd order
                self.order = self.buy(data=self.datas[0])
                self.log('Executed Price:{}'.format(self.order.executed.price))
                print ('Executed Price:', self.order.executed.price)
                print ('ASK', self.data.ask[0])

                and finale code:

                def parse_args():
    parser = argparse.ArgumentParser(
        description='Bid/Ask Line Hierarchy',
        formatter_class=argparse.ArgumentDefaultsHelpFormatter,
    )

    parser.add_argument('--data', '-d', action='store',
                        required=False, default='bidask.csv',
                        help='data to add to the system')

    parser.add_argument('--dtformat', '-dt',
                        required=False, default='%m/%d/%Y %H:%M:%S',
                        help='Format of datetime in input')

    return parser.parse_args()


def runstrategy():
    args = parse_args()

    cerebro = bt.Cerebro()  # Create a cerebro
       
    data = BidAskCSV(dataname=args.data, dtformat=args.dtformat)
    cerebro.adddata(data)  # Add the 1st data to cerebro
    # Set our desired cash start
    cerebro.broker.setcash(100000.0)
    
    # Add the strategy to cerebro
    cerebro.addstrategy(St)    
    
        # Add a FixedSize sizer according to the stake
    cerebro.addsizer(bt.sizers.FixedSize, stake=10)
    
    # Print out the starting conditions
    print('Starting Portfolio Value: %.2f' % cerebro.broker.getvalue())

    # Run over everything
    cerebro.run()

    # Print out the final result
    print('Final Portfolio Value: %.2f' % cerebro.broker.getvalue())
  

if __name__ == '__main__':
    runstrategy()

                This is what I get as output:

                Starting Portfolio Value: 100000.00
   1: 2010-02-03T23:59:59.999989 - Bid 0.5346 - 0.5347 Ask, CLOSE 0.5347 
-------------
Random: 0.329840335116921
-------------
   2: 2010-02-03T23:59:59.999989 - Bid 0.5343 - 0.5347 Ask, CLOSE 0.5347 
-------------
Random: 0.680459956622627
-------------
   3: 2010-02-03T23:59:59.999989 - Bid 0.5543 - 0.5545 Ask, CLOSE 0.5545 
-------------
Random: 0.04838233726017549
-------------
   4: 2010-02-03T23:59:59.999989 - Bid 0.5342 - 0.5344 Ask, CLOSE 0.5344 
-------------
Random: 0.7977225266469058
-------------
   5: 2010-02-03T23:59:59.999989 - Bid 0.5245 - 0.5464 Ask, CLOSE 0.5464 
-------------
Random: 0.8661165246756014
-------------
2010-02-03T23:59:59.999989, BUY CREATE 0.5464
2010-02-03T23:59:59.999989, Executed Price:0.0
Executed Price: 0.0
ASK 0.5464
   6: 2010-02-03T23:59:59.999989 - Bid 0.5460 - 0.5470 Ask, CLOSE 0.5470 
-------------
Random: 0.7797905933237207
-------------
   7: 2010-02-03T23:59:59.999989 - Bid 0.5824 - 0.5826 Ask, CLOSE 0.5826 
-------------
Random: 0.585687043100518
-------------
   8: 2010-02-03T23:59:59.999989 - Bid 0.5371 - 0.5374 Ask, CLOSE 0.5374 
-------------
Random: 0.2171674683269379
-------------
   9: 2010-02-03T23:59:59.999989 - Bid 0.5793 - 0.5794 Ask, CLOSE 0.5794 
-------------
Random: 0.6576887524911412
-------------
  10: 2010-02-03T23:59:59.999989 - Bid 0.5684 - 0.5688 Ask, CLOSE 0.5688 
-------------
Random: 0.406599019686264
-------------
Final Portfolio Value: 100000.00

                I don't understand what I am doing wrong, and why the orders are not being executed. Can I get some help here? Thanks!

                1 Reply Last reply Reply Quote 0
                • B
                  backtrader administrators last edited by

                  @roch_02 said in Execute on bid/ask:

                              # BUY (with all possible default parameters)
            self.log('BUY CREATE {}'.format(self.data.ask[0]))

            # Keep track of the created order to avoid a 2nd order
            self.order = self.buy(data=self.datas[0])
            self.log('Executed Price:{}'.format(self.order.executed.price))
            print ('Executed Price:', self.order.executed.price)
            print ('ASK', self.data.ask[0])

                  The order has not been executed yet. Only issued, hence the 0.0 price.

                  Furthermore, see the resolution of this: Community - Getting executed twice on closing orders, because your data isn't obviously 1-day based

                  R 1 Reply Last reply Reply Quote 0
                  • R
                    roch_02 @backtrader last edited by roch_02

                    @backtrader

                    Thank you for the reply.

                    I changed the code and added: bt.TimeFrame.Ticks, compression, cerebro.broker.set_coc(True) and self.order = self.sell(data=self.datas[0], coc=False).

                    The code is here:

                    from __future__ import (absolute_import, division, print_function,
                        unicode_literals)

import argparse

import backtrader as bt
import backtrader.feeds as btfeeds
#import backtrader.indicators as btind
import random


class BidAskCSV(btfeeds.GenericCSVData):
    linesoverride = True  # discard usual OHLC structure
    # datetime must be present and last
    lines = ('bid', 'ask', 'close', 'open','low', 'high', 'datetime')
    # datetime (always 1st) and then the desired order for
    params = (
        # (datetime, 0), # inherited from parent class
        ('bid', 1),  # default field pos 1
        ('ask', 2),  # default field pos 2
        ('close', 2), ('open', 1), ('low', 1), ('high',2),
        ('timeframe', bt.TimeFrame.Ticks))
    
    def __init__(self):
        super(BidAskCSV, self).__init__()
    


class St(bt.Strategy):

    def log(self, txt, dt= None):
        """Logging function """
        dt= dt or self.datas[0].datetime[0]
        if isinstance(dt, float):
            dt = bt.num2date(dt)
        print('%s, %s' % (dt.isoformat(), txt))


    def __init__(self):
        self.order = None
        self.dataclose= self.datas[0].close
        #self.dataclose = self.datas[0].close[0]
            
            
    def notify_order(self, order):
       if order.status in [order.Submitted, order.Accepted]:
           # Buy/Sell order submitted/accepted to/by broker - Nothing to do
           return 

       # Check if an order has been completed
       # Attention: broker could reject order if not enougth cash
       if order.status in [order.Completed]:
           if order.isbuy():
               self.log(
                   'BUY EXECUTED, Price: %.4f, Cost: %.4f, bid: %.4f, ask: %.4f' %
                   (order.executed.price,
                    order.executed.value,
                    self.data.bid[0],
                    self.data.ask[0]))

               self.buyprice = order.executed.price
               
           else:  # Sell
               self.log('SELL EXECUTED, Price: %.4f, Cost: %.4f' %
                        (order.executed.price,
                         order.executed.value))

           self.bar_executed = len(self)
       elif order.status in [order.Canceled, order.Margin, order.Rejected]:
           self.log('Order Canceled/Margin/Rejected')

       self.order = None
       
    def notify_trade(self, trade):
        if not trade.isclosed:
            return

        self.log('OPERATION PROFIT, GROSS %.2f, NET %.2f' %
                 (trade.pnl, trade.pnlcomm))
            

    def next(self):
        dtstr = self.datas[0].datetime.datetime().isoformat()
        txt = '%4d: %s - Bid %.4f - %.4f Ask, CLOSE %.4f ' % (
            (len(self), dtstr, self.datas[0].bid[0], self.datas[0].ask[0],
             self.datas[0].close[0]))
        
        print (self.broker.getvalue())

        print(txt)
        
        position = self.getposition(data=self.datas[0])
        rand = random.random()
        print('-------------')
        print ('Random:', rand)
        print('-------------')          
        if rand < 0.2:
            if position:
                # SELL (with all possible default parameters)
                self.log('SELL CREATE '.format(self.data.bid[0]))

                # Keep track of the created order to avoid a 2nd order
                self.order = self.sell(data=self.datas[0], coc=False)
                print ('Bid', self.data.bid[0])
        elif rand > 0.5:
            if not position:
                # BUY (with all possible default parameters)
                self.log('BUY CREATE {}'.format(self.data.ask[0]))

                # Keep track of the created order to avoid a 2nd order
                self.order = self.buy(data=self.datas[0])
                #self.log('Executed Price:{}'.format(self.order.executed.price))
                #print ('Executed Price:', self.order.executed.price)
                print ('ASK', self.data.ask[0])
                
def parse_args():
    parser = argparse.ArgumentParser(
        description='Bid/Ask Line Hierarchy',
        formatter_class=argparse.ArgumentDefaultsHelpFormatter,
    )

    parser.add_argument('--data', '-d', action='store',
                        required=False, default='bidask.csv',
                        help='data to add to the system')

    parser.add_argument('--dtformat', '-dt',
                        required=False, default='%m/%d/%Y %H:%M:%S',
                        help='Format of datetime in input')
    
    parser.add_argument('--compression', required=False, default=2, type=int,
                        help='How much to compress the bars')

    return parser.parse_args()


def runstrategy():
    args = parse_args()

    cerebro = bt.Cerebro()  # Create a cerebro
           
    data = BidAskCSV(dataname=args.data, dtformat=args.dtformat)
    cerebro.adddata(data)  # Add the 1st data to cerebro
    # Set our desired cash start
    cerebro.broker.setcash(100000.0)
    cerebro.broker.set_coc(True)
    
    cerebro.resampledata(data,
                         timeframe=bt.TimeFrame.Ticks,
                         compression = args.compression)
    
    # Add the strategy to cerebro
    cerebro.addstrategy(St)    
    
        # Add a FixedSize sizer according to the stake
    cerebro.addsizer(bt.sizers.FixedSize, stake=10)
    
    # Print out the starting conditions
    print('Starting Portfolio Value: %.2f' % cerebro.broker.getvalue())

    # Run over everything
    cerebro.run()

    # Print out the final result
    print('Final Portfolio Value: %.2f' % cerebro.broker.getvalue())
  

if __name__ == '__main__':
    runstrategy()

                    For sell order it executes with the bid price from the SELL EXECUTED time but for the buy order it executes with the ask price from the BUY CREATE time and not with the ask price from the BUY EXECUTED time.

                    the output is here:

                    Starting Portfolio Value: 100000.00
100000.0
   2: 2010-02-03T16:53:51 - Bid 0.5343 - 0.5347 Ask, CLOSE 0.5347 
-------------
Random: 0.7585227506318774
-------------
2010-02-03T16:53:51, BUY CREATE 0.5347
ASK 0.5347
2010-02-03T16:53:52, BUY EXECUTED, Price: 0.5347, Cost: 5.3470, bid: 0.5543, ask: 0.5545
100000.198
   3: 2010-02-03T16:53:52 - Bid 0.5543 - 0.5545 Ask, CLOSE 0.5545 
-------------
Random: 0.7866456795035222
-------------
99999.997
   4: 2010-02-03T16:53:53 - Bid 0.5342 - 0.5344 Ask, CLOSE 0.5344 
-------------
Random: 0.5313830132332052
-------------
100000.11700000001
   5: 2010-02-03T16:53:54 - Bid 0.5245 - 0.5464 Ask, CLOSE 0.5464 
-------------
Random: 0.409841893004643
-------------
100000.123
   6: 2010-02-03T16:53:54 - Bid 0.5460 - 0.5470 Ask, CLOSE 0.5470 
-------------
Random: 0.009108455433875506
-------------
2010-02-03T16:53:54, SELL CREATE 
Bid 0.546
2010-02-03T16:53:56, SELL EXECUTED, Price: 0.5824, Cost: 5.3470
2010-02-03T16:53:56, OPERATION PROFIT, GROSS 0.48, NET 0.48
100000.477
   7: 2010-02-03T16:53:56 - Bid 0.5824 - 0.5826 Ask, CLOSE 0.5826 
-------------
Random: 0.4250130075545223
-------------
100000.477
   8: 2010-02-03T16:53:57 - Bid 0.5371 - 0.5374 Ask, CLOSE 0.5374 
-------------
Random: 0.9467974453210086
-------------
2010-02-03T16:53:57, BUY CREATE 0.5374
ASK 0.5374
2010-02-03T16:53:58, BUY EXECUTED, Price: 0.5374, Cost: 5.3740, bid: 0.5793, ask: 0.5794
100000.897
   9: 2010-02-03T16:53:58 - Bid 0.5793 - 0.5794 Ask, CLOSE 0.5794 
-------------
Random: 0.4769590069462516
-------------
100000.791
  10: 2010-02-03T16:53:59 - Bid 0.5684 - 0.5688 Ask, CLOSE 0.5688 
-------------
Random: 0.21570733606386583
-------------
Final Portfolio Value: 100000.79

                    How can I correct the ask price?

                    Another question, why is cost different than price? Is that the correct value?

                    (I am using the bidask.csv file from backtrader data)

                    B 1 Reply Last reply Reply Quote 0
                    • B
                      backtrader administrators @roch_02 last edited by

                      @roch_02 said in Execute on bid/ask:

                      the buy order it executes with the ask price from the BUY CREATE time and not with the ask price from the BUY EXECUTED time.

                      Because you have cheat-on-close active and you are effectively asking for that price.

                      @roch_02 said in Execute on bid/ask:

                      Another question, why is cost different than price?

                      The price is per unit of an asset. It cannot be the same as the total cost of the operation.

                      R 1 Reply Last reply Reply Quote 0
                      • R
                        roch_02 @backtrader last edited by

                        @backtrader

                        Thanks for the answer but I still did not get part of it.

                        So here is the issue:
                        Before I choose to put cerebro.broker.set_coc(True)the price for the Buy executed and Sell executed would always be from the bid from the execution time. (correct price for the sell order)

                        Once I set cerebro.broker.set_coc(True), the Buy executed and Sell executed would always be from the ask price from the created time (not from executed time).

                        Setting self.sell(data=self.datas[0], coc=False) would correct the problem for the sell order and the price used is now the bid from the execution time (correct price for sell order). However, for the buy order, the ask from the created time is the price used. My question is, if it is possible to use the ask price from the execution time for the buy order?

                        Thanks for your time in helping!

                        1 Reply Last reply Reply Quote 0
                        • B
                          backtrader administrators last edited by backtrader

                          Again: no. By setting cheat-on-close you get the price at creation time, unless the flag coc in buy/sell is set.

                          1 Reply Last reply Reply Quote 1
                          • R
                            roch_02 last edited by

                            Ok, thanks once again.

                            1 Reply Last reply Reply Quote 0
                            • 1 / 1
                            • First post
                              Last post
                            Copyright © 2016, 2017, 2018, 2019, 2020, 2021 NodeBB Forums | Contributors