Added roll costs in backtest

README.md

@@ -171,6 +171,8 @@ In finmarketpy/examples you will find several examples, including some simple tr
 
 # finmarketpy log
 
+* 23 Jul 2021
+ * Added roll costs in backtest
 * 30 May 2021
  * Added S3 Jupyter notebook for use with findatapy
 * 21 May 2021

finmarketpy/backtest/backtestengine.py

@@ -152,7 +152,9 @@ def calculate_trading_PnL(self, br, asset_a_df, signal_df, contract_value_df, ru
  signal_df = signal_df.mask(non_trading_days) # fill asset holidays with NaN signals
  signal_df = signal_df.fillna(method='ffill') # fill these down
 
+ # Transaction costs and roll costs
  tc = br.spot_tc_bp
+ rc = br.spot_rc_bp
 
  signal_cols = signal_df.columns.values
  asset_df_cols = asset_df.columns.values
@@ -245,7 +247,7 @@ def calculate_trading_PnL(self, br, asset_a_df, signal_df, contract_value_df, ru
  = self.calculate_exposures(portfolio_signal)
 
  # Calculate final portfolio returns with the amended portfolio leverage (by default just 1s)
- portfolio = calculations.calculate_signal_returns_with_tc_matrix(portfolio_leverage_df, portfolio, tc=tc)
+ portfolio = calculations.calculate_signal_returns_with_tc_matrix(portfolio_leverage_df, portfolio, tc=tc, rc=rc)
 
  # Assign all the property variables
  # Trim them if we have asked for a different plot start/finish
@@ -284,7 +286,7 @@ def calculate_trading_PnL(self, br, asset_a_df, signal_df, contract_value_df, ru
 
  # P&L components of individual assets after all the portfolio level risk signals and position limits have been applied
  self._components_pnl = self._filter_by_plot_start_finish_date(calculations.calculate_signal_returns_with_tc_matrix(portfolio_signal_before_weighting,
- returns_df, tc=tc), br)
+ returns_df, tc=tc, rc=rc), br)
  self._components_pnl.columns = pnl_cols
 
  # TODO FIX very slow - hence only calculate on demand
@@ -1997,6 +1999,7 @@ def optimize_portfolio_weights(self, returns_df, signal_df, signal_pnl_cols, br=
  br = self._br
 
  tc = br.spot_tc_bp
+ rc = br.spot_rc_bp
 
  individual_leverage_df = None
 
@@ -2015,7 +2018,7 @@ def optimize_portfolio_weights(self, returns_df, signal_df, signal_pnl_cols, br=
 
  individual_leverage_df = leverage_df # Contains leverage of individual signal (before portfolio vol target)
 
- signal_pnl = self._calculations.calculate_signal_returns_with_tc_matrix(signal_df, returns_df, tc=tc)
+ signal_pnl = self._calculations.calculate_signal_returns_with_tc_matrix(signal_df, returns_df, tc=tc, rc=rc)
  signal_pnl.columns = signal_pnl_cols
 
  adjusted_weights_matrix = None

finmarketpy/backtest/backtestrequest.py

@@ -31,7 +31,7 @@ def __init__(self):
 
  self.__signal_name = None
 
- # output parameters for backtest (should we add returns statistics on legends, write CSVs with returns etc.)
+ # Output parameters for backtest (should we add returns statistics on legends, write CSVs with returns etc.)
  self.__plot_start = None
  self.__plot_finish = None
  self.__calc_stats = True
@@ -46,7 +46,7 @@ def __init__(self):
 
  self.__portfolio_weight_construction = None
 
- # default parameters for portfolio level vol adjustment
+ # Default parameters for portfolio level vol adjustment
  self.__portfolio_vol_adjust = False
  self.__portfolio_vol_period_shift = 0
  self.__portfolio_vol_rebalance_freq = None
@@ -57,7 +57,7 @@ def __init__(self):
  self.__portfolio_vol_periods = 20
  self.__portfolio_vol_obs_in_year = 252
 
- # default parameters for signal level vol adjustment
+ # Default parameters for signal level vol adjustment
  self.__signal_vol_adjust = False
  self.__signal_vol_period_shift = 0
  self.__signal_vol_rebalance_freq = None
@@ -68,12 +68,12 @@ def __init__(self):
  self.__signal_vol_periods = 20
  self.__signal_vol_obs_in_year = 252
 
- # portfolio notional size
+ # Portfolio notional size
  self.__portfolio_notional_size = None
  self.__portfolio_combination = None
  self.__portfolio_combination_weights = None
 
- # parameters for maximum position limits (expressed as whole portfolio)
+ # Parameters for maximum position limits (expressed as whole portfolio)
  self.__max_net_exposure = None
  self.__max_abs_exposure = None
 
@@ -82,18 +82,21 @@ def __init__(self):
  self.__position_clip_resample_type = 'mean'
  self.__position_clip_period_shift = 0
 
- # take profit and stop loss parameters
+ # Take profit and stop loss parameters
  self.__take_profit = None
  self.__stop_loss = None
 
- # should we delay the signal?
+ # Should we delay the signal?
  self.__signal_delay = 0
 
- # annualization factor for return stats (and should we resample data first before calculating it?)
+ # Annualization factor for return stats (and should we resample data first before calculating it?)
  self.__ann_factor = 252
  self.__resample_ann_factor = None
 
- # how do we create a cumulative index of strategy returns
+
+ self.__spot_rc_bp = None
+
+ # How do we create a cumulative index of strategy returns
  # either multiplicative starting a 100
  # or additive starting at 0
  self.__cum_index = 'mult' # 'mult' or 'add'
@@ -366,6 +369,25 @@ def spot_tc_bp(self, spot_tc_bp):
  else:
  self.__spot_tc_bp = float(spot_tc_bp) / (2.0 * 100.0 * 100.0)
 
+ @property
+ def spot_rc_bp(self):
+ return self.__spot_rc_bp
+
+ @spot_rc_bp.setter
+ def spot_rc_bp(self, spot_rc_bp):
+ if isinstance(spot_rc_bp, dict):
+ spot_rc_bp = spot_rc_bp.copy()
+
+ for k in spot_rc_bp.keys():
+ spot_rc_bp[k] = float(spot_rc_bp[k]) / (100.0 * 100.0)
+
+ self.__spot_rc_bp = spot_rc_bp
+
+ elif isinstance(spot_rc_bp, DataFrame):
+ self.__spot_rc_bp = spot_rc_bp # assume that DataFrame is in the percentage form (bid to mid)
+ else:
+ self.__spot_rc_bp = float(spot_rc_bp) / (100.0 * 100.0)
+
  #### FOR FUTURE USE ###
 
  @property