ChiahungTai/BackTester-BuyAndHold

This is Buy And Hold Strategy developed by BackTester Framework

Requirement

• python3 (>3.6)
• pandas (> v0.23.0)
• numpy (> v1.14)
• dateutl (>2.7.0)

Usage

1. Set parameters

   In Main.py:

   1. Prepare data for framework

      back_test_context = BackTesterContext(
          price_information_generator=PriceInformationGenerator(data_source).generate_price_information(
              start_date=datetime.strptime('1998/07/21', '%Y/%m/%d'),
              end_date=datetime.strptime('2016/12/30', '%Y/%m/%d')
          ),
          trading_status_calculator=TradingStatusCalculator(success_rate=100)
      )

      Data format ( in pandas.DataFrame ) required by framework is shown below:

      Column Name	Data Type
      date	datetime64[ns]
      contract	str/object
      delivery_month	str/object
      open_price	float64
      highest_price	float64
      lowest_price	float64
      close_price	float64

      You can add arbitrary columns if you need, but columns and their data type above are immutable when passing DataFrame to PriceInformationGenerator:

   2. Configure time interval and success rate of order

      back_test_context = BackTesterContext(
              price_information_generator=PriceInformationGenerator(data_source).generate_price_information(
                  start_date=datetime.strptime('1998/07/21', '%Y/%m/%d'),
                  end_date=datetime.strptime('2016/12/30', '%Y/%m/%d')
              ),
              trading_status_calculator=TradingStatusCalculator(success_rate=100)
      )

      1. BackTesterContext will iterate from start_date to end_date. During iterating progress, if any date in DataFrame you passed exist, all rows with same date will be read.

      2. When success_rate to BackTesterContext, you can set success rate when BackTesterContext filling orders. When you passing orders to BackTesterContext, if success_rate is less than 100, then every order will have chance to be fail traded.

   3. Configure initial capital

      asset_maintainer = AssetMaintainer(
          initial_capital=1000000, price_parameter=pd.read_csv(os.path.join('Source', 'price_parameter.csv'))
      )

   4. Configure your target contract

      buy_and_hold = BuyAndHold(
            asset_maintainer=asset_maintainer, back_test_context=back_test_context, target_contract='TX'
      )

2. Run Main.py

Note

1. When developing strategy, you can access

   • AssetMaintainer.current_position_status_maintainer.current_position_status (pandas.DataFrame)

     Column Name	Data Type
     contract	str/object
     delivery_month	str/object
     value	float64
     cost	float64
     close_price_of_prefious_trading_date	float64
     status	str/object

     • status always = 'holding'
     • Buying time of contracts with same contract and delivery_month can be specified through their sequence of index. Smaller means bought earlier. When selling, contracts with smaller index number will be sold first.

   • AssetMaintainer.current_capital_maintainer.current_capital (int)

   to get current_position_status and current_capital.

2. When sending order in strategy through BackTesterContext.send_order, you have to follow the format:

   • data_type: pandas.DataFrame

   • format:

     Column Name	Data Type
     contract	str/object
     delivery_month	str/object
     buy_position	int
     sell_position	int