How to use val_price? about vectorbt HOT 8 CLOSED

Close is used for calculating return metrics such as sharpe, val_price for using target value and percentage, price is order price, so setting price to close * 100 should work.

from vectorbt.

@polakowo Thank you! but I can not get right result.

import numpy as np
import pandas as pd
import vectorbt as vbt

close = pd.Series([1, 2, 3, 4, 5], index=pd.date_range('2020-01-01', periods=5))
price = close * 100
val_price = close * 100
size = [1, np.nan, np.nan, np.nan, -1]
portfolio = vbt.Portfolio.from_orders(close, size,
                                      price=price,
                                      val_price=val_price, close_first=True)
#
print(portfolio.positions().records)
#
print(portfolio.value())

portfolio.trades().plot()

   id  col  size  entry_idx  entry_price  entry_fees  exit_idx  exit_price  \
0   0    0   1.0          0        100.0         0.0         4       500.0   

   exit_fees    pnl  return  direction  status  
0        0.0  400.0     4.0          0       1  
2020-01-01      1.0
2020-01-02      2.0
2020-01-03      3.0
2020-01-04      4.0
2020-01-05    500.0
Freq: D, dtype: float64

the entry_price should be 1, maybe need a new column entry_value.
the figure is wrong. entry_maskers not on the close price line.

from vectorbt.

But your order price is 100 * close, so why do you expect the entry price to be 1? Entry price in your example is close[0] * 100, which is 100, and not 1. Your exit price is close[-1] * 100, which is 500, thus a P&L of 400. Value at the beginning is 1 instead of 100 because it uses your close (last price at each bar) instead of your order price to calculate various time series such as value and returns for consistency. Can you elaborate more on what you want to achieve?

Edit: I don't think there is currently a way to backtest futures, haven't checked that.

from vectorbt.

I think right output is this

   id  col  size  entry_idx  entry_price  entry_fees  exit_idx  exit_price  exit_fees    pnl  return  direction  status  entry_value  exit_value
0   0    0   1.0          0          1.0         0.0         4         5.0       0.0   400.0     4.0          0       1  100.0        500.0


2020-01-01    100.0
2020-01-02    200.0
2020-01-03    300.0
2020-01-04    400.0
2020-01-05    500.0
Freq: D, dtype: float64

price is quote price.
value is asset value.

sorry I am not good at english. have to use code to explain it.

from vectorbt.

There is no easy way to implement your example without workarounds such as applying a factor of 100 directly on close:

import numpy as np
import pandas as pd
import vectorbt as vbt

close = pd.Series([100, 200, 300, 400, 500], index=pd.date_range('2020-01-01', periods=5))
size = [1, np.nan, np.nan, np.nan, -1]
portfolio = vbt.Portfolio.from_orders(close, size, close_first=True)
#
print(portfolio.positions().records)
#
print(portfolio.value())

   id  col  size  entry_idx  entry_price  entry_fees  exit_idx  exit_price  \
0   0    0   1.0          0        100.0         0.0         4       500.0   

   exit_fees    pnl  return  direction  status  
0        0.0  400.0     4.0          0       1  
2020-01-01    100.0
2020-01-02    200.0
2020-01-03    300.0
2020-01-04    400.0
2020-01-05    500.0
Freq: D, dtype: float64

As said, futures trading is no part of vectorbt as of now.

Do you have a similar example of how you do it in let's say backtrader?

from vectorbt.

thank you all the same

Commissions: Stocks vs Futures

https://www.backtrader.com/docu/commission-schemes/commission-schemes/

mult (default: 1.0) For future like instruments this determines the multiplicator to apply to profit and loss calculations.

cerebro.broker.setcommission(commission=2.0, margin=2000.0, mult=10.0)

https://github.com/mementum/backtrader/blob/master/samples/commission-schemes/commission-schemes.py

Futures and Spot Compensation

https://www.backtrader.com/docu/order-creation-execution/futurespot/future-vs-spot/
https://github.com/mementum/backtrader/blob/master/samples/future-spot/future-spot.py

https://github.com/mementum/backtrader/search?p=1&q=mult

from vectorbt.

I will mark this as an enhancement, as it means adding support for future-like instruments. But I can't give any estimates on when or whether it can be implemented - a lot of changes have to be done to the backbone to make it possible. For example, you may want to buy 10 units with a multiplier of 100 and another 20 units with a multiplier of 50, so you need to track those multipliers at each timestamp to generate the value - this adds complexity to and further reduces the performance of the solution.

from vectorbt.

value = price * multiplier
margin = value* 0.2

vbt.Portfolio.from_orders(
    close,
    val_price,
    value,
    margin,
    
)

maybe prepare some data by user before call from_orders is an easy way

from vectorbt.