I am currently running a backtest of a wide range of stocks.
At any given time, the system I am backtesting holds multiple positions concurrently. I would like to track the maximum profit of each position from the point of it is taken to the "current" bar. Why? The maximum profit of each position is a key exit parameter.
To do this I created a list called "self.position_pnl_list". Here is how I integrated it into my code:
class BTStrategy(bt.Strategy):
def __init__(self):
pass
def start(self):
self.bought_bars={d:0 for d in self.getdatanames()}
self.sold_bars={d:0 for d in self.getdatanames()}
self.size = None
self.initial_risk = None
self.position_pnl_list = []
def next(self):
for i, d in enumerate(self.getdatanames()):
pos = self.getposition(data=self.getdatabyname(d))
comminfo = self.broker.getcommissioninfo(data=self.getdatabyname(d))
position_pnl = comminfo.profitandloss(pos.size, pos.price, self.getdatabyname(d).close[0])
risk_parity_size = int((self.broker.getvalue() * 0.01)/self.getdatabyname(d).atr[0])
self.trailing_stop = self.getdatabyname(d).trailing_stop
self.position_pnl_list.append(position_pnl)
When I close each position, I delete the list values just to make sure that does not mix old position profit metrics with new ones: as follows:
if pos.size:
if len(self.position_pnl_list) > 0:
maximum_position_profit = max(self.position_pnl_list)
if maximum_position_profit >= 1000:
self.close(data=self.getdatabyname(d), size=self.size)
del self.position_pnl_list[:]
Questions:
Is this the correct approach?
Is there a better way or an inbuilt Backtrader I could use to accomplish my goals?
