Building and Evaluating an EWMAC Forecast from Interactive Brokers FX Data

# Modified introduction using forex data


## Modified introduction using forex data

This is the trading rule example shown in [the introduction](https://github.com/pst-group/pysystemtrade/blob/master/docs/introduction.md) but modified to use Interactive Brokers instead of CSV files as data source.

IB requires a minimum equity and a monthly subscription to provide historical data on future contracts.  This example was modified to use FX prices instead futures to make it runnable with free unfunded paper trading accounts.  Note that Rob [does not recommend trading FX spot data with IB due to their high fees](https://github.com/pst-group/pysystemtrade/issues/517#issuecomment-1010770678).

First, import the required packages and initialize ib_insync.  

```python
from sysbrokers.IB.ib_connection import connectionIB
from sysbrokers.IB.ib_Fx_prices_data import ibFxPricesData

from ib_insync import util
util.startLoop() #only required when running inside a notebook
```

Connecting to Interactive Brokers gateway...

```python
conn = connectionIB(111)
conn
```

See what fx instruments we have configured.  These are configured in `sysbrokers/IB/ib_config_spot_FX.csv`

```python
ibfxpricedata = ibFxPricesData(conn)
ibfxpricedata.get_list_of_fxcodes()
```

Now we select one instrument (`EURUSD`) and try to fetch historical data for it.

```python
ibfxpricedata.get_fx_prices('EURUSD')
```

Data can also be indexed as a python dict:

```python
ibfxpricedata['JPYUSD']
```

Create the trading rule

```python
import pandas as pd
from sysquant.estimators.vol import robust_vol_calc


def calc_ewmac_forecast(price, Lfast, Lslow=None):
    """
    Calculate the ewmac trading rule forecast, given a price and EWMA speeds Lfast, Lslow and vol_lookback

    """
    if Lslow is None:
        Lslow = 4 * Lfast

    ## We don't need to calculate the decay parameter, just use the span directly
    fast_ewma = price.ewm(span=Lfast).mean()
    slow_ewma = price.ewm(span=Lslow).mean()
    raw_ewmac = fast_ewma - slow_ewma

    vol = robust_vol_calc(price.diff())

    return raw_ewmac / vol
```

Run a forecast with the previous rule

```python
price=ibfxpricedata['EURUSD']
ewmac=calc_ewmac_forecast(price, 32, 128)
ewmac.tail(5)
```

```python
import matplotlib.pyplot as plt

plt.figure(figsize=(12,5))

ax1 = price.plot(color='blue', grid=True, label='Price')
ax2 = ewmac.plot(color='red', grid=True, secondary_y=True, label='Forecast')

h1, l1 = ax1.get_legend_handles_labels()
h2, l2 = ax2.get_legend_handles_labels()


plt.legend(h1+h2, l1+l2, loc=2)
plt.show()
```

The original introduction jumps directly to "Did we make any money?".  
I would like to see here the orders that were triggered by this forecast, but instead we jump directly into P&L.  Still, these are the P&L numbers for this forecast and data:

```python
from systems.accounts.account_forecast import pandl_for_instrument_forecast
account = pandl_for_instrument_forecast(forecast = ewmac, price = price)
account.percent.stats()
```

```python
account.curve().plot()
plt.show()
```

```python
conn.close_connection()
```
![notebook output](figures/p1_1.png)
![notebook output](figures/p1_2.png)