Issues that should be paid attention to when backtrader conducts futures backtesting: Margin and other settings, splicing and rolling contracts

Click here to get the backtrader technical tutorial

==================

Many students asked how to use backtrader for futures backtesting. The following is an example. The difference from ordinary stock backtesting is the way commissions are set.

1 Commission settings

import datetime  # For datetime objects
import backtrader as bt
import backtrader.feeds as btfeeds
import backtrader.indicators as btind
import pandas as pd
import numpy as np
from datetime import datetime
class Par_SAR(bt.Strategy):
    params = (('period', 2), ('af', 0.02), ('afmax', 0.2))

    def __init__(self):
       pass
       



if __name__ == '__main__':
    # rb 15 min data 
    brf_min_bar = bt.feeds.GenericCSVData(
        dataname='E:/Quant/Backtrader/backtrader-master/China_Market/data/rb_15m.csv',
        timeframe=bt.TimeFrame.Minutes,
        #fromdate=datetime.datetime(2019, 5, 1),
        #todate=datetime.datetime(2020, 6, 1),
        nullvalue=0.0,
        dtformat=('%Y-%m-%d %H:%M:%S'),
        datetime=1,
        time=-1,
        high=4,
        low=5,
        open=3,
        close=6,
        volume=-1,
        # openinterest=-1
    )
    # Add the Data Feed to Cerebro
    cerebro.adddata(brf_min_bar)
    #cerebro.resampledata(brf_min_bar, timeframe=bt.TimeFrame.Days)

    # Add strategy
    cerebro.addstrategy(Par_SAR)
    # 设置初始资本为100,000
    cerebro.broker.setcash(1000000.0) 

    # 设置期货佣金模式
    cerebro.broker.setcommission(commission=0.0001, margin=2800.0, mult=10.0)

    cerebro.run()

The above commission settings actually use the built-in commission class CommInfoBase of backtrader, and the parameters that can be set in this class are as follows. Of course, most of the parameters can use the default values, and there is no need to set them. In the above case, only 3 parameters are set, and the others are default values. The leverage in the following parameters may also be related to futures. In addition, this class can also consider borrowing interest.

Params:

• commission (def: 0.0): base commission value in percentage or monetary units
• mult (def 1.0): multiplier applied to the asset for value/profit
• margin (def: None): amount of monetary units needed to open/hold an operation. It only applies if the final _stocklike attribute in the class is set to False
• automargin (def: False): Used by the method get_margin to automatically calculate the margin/guarantees needed with the following policy
  • Use param margin if param automargin evaluates to False
  • Use param mult and use mult * price if automargin < 0
  • Use param automargin and use automargin * price if automargin > 0

 

• commtype (def: None): Supported values are CommInfoBase.COMM_PERC (commission to be understood as %) and CommInfoBase.COMM_FIXED (commission to be understood as monetary units)
  The default value of None is a supported value to retain compatibility with the legacy CommissionInfo object. If commtype is set to None, then the following applies:
  • margin is None: Internal _commtype is set to COMM_PERC and _stocklike is set to True (Operating %-wise with Stocks)
  • margin is not None: _commtype set to COMM_FIXED and _stocklike set to False (Operating with fixed rount-trip commission with Futures)

If this param is set to something else than None, then it will be passed to the internal _commtype attribute and the same will be done with the param stocklike and the internal attribute _stocklike

• stocklike (def: False): Indicates if the instrument is Stock-like or Futures-like (see the commtype discussion above)
• percabs (def: False): when commtype is set to COMM_PERC, whether the parameter commission has to be understood as XX% or 0.XX
  If this param is True: 0.XX If this param is False: XX%
• interest (def: 0.0)
  If this is non-zero, this is the yearly interest charged for holding a short selling position. This is mostly meant for stock short-selling
  The formula: days * price * abs(size) * (interest / 365)
  It must be specified in absolute terms: 0.05 -> 5%
  Note
  the behavior can be changed by overriding the method: _get_credit_interest
• interest_long (def: False)
  Some products like ETFs get charged on interest for short and long positions. If ths is True and interest is non-zero the interest will be charged on both directions
• leverage (def: 1.0)
  Amount of leverage for the asset with regards to the needed cash

 

You can derive your own commission class from the commission base class to meet your own requirements. For example, if someone expects that the commission rate depends not only on the transaction amount, but also on the multiplier mult, you can define a new commission class as follows

import backtrader as bt

class CommInfo_Fut_Perc_Mult(bt.CommInfoBase):
    params = (
      ('stocklike', False),  # Futures
      ('commtype', bt.CommInfoBase.COMM_PERC),  # Apply % Commission
    # ('percabs', False),  # pass perc as xx% which is the default
    )

    def _getcommission(self, size, price, pseudoexec):
        return size * price * self.p.commission * self.p.mult

Then, use it like this

comminfo = CommInfo_Fut_Perc_Mult(
    commission=0.1,  # 0.1%
    mult=10,
    margin=2000  # Margin is needed for futures-like instruments
)

cerebro.addcommissioninfo(comminfo)

 

2 Stitched Rolling Contracts

In addition, backtrader also allows rolling splicing of futures contracts with different expiration dates to form continuous futures (The RollOver Data Feed), examples are as follows

import backtrader as bt

cerebro = bt.Cerebro()
data0 = bt.feeds.MyFeed(dataname='Expiry0')
data1 = bt.feeds.MyFeed(dataname='Expiry1')
...
dataN = bt.feeds.MyFeed(dataname='ExpiryN')

drollover = bt.feeds.RollOver(data0, data1, ..., dataN, dataname='MyRoll', **kwargs)
cerebro.adddata(drollover)

cerebro.run()

The RollOver class has two parameters that control how contracts are stitched together, checkdateand  checkcondition. Detailed description refer to  Data Feeds - Rollover - Backtrader