Algorithmic Trading Strategy with Machine Learning and Python

What is Algorithmic Trading Strategy ?

Developing an Algorithmic trading strategy with Python is something that goes through a couple of phases, just like when you build machine learning models: you formulate a strategy and specify it in a form that you can test on your computer, you do some preliminary testing or back testing, you optimize your strategy and lastly, you evaluate the performance and robustness of your strategy.

Trading strategies are usually verified by back testing: you reconstruct, with historical data, trades that would have occurred in the past using the rules that are defined with the strategy that you have developed.

This way, you can get an idea of the effectiveness of your strategy, and you can use it as a starting point to optimize and improve your strategy before applying it to real markets.

This relies heavily on the underlying theory or belief that any strategy that has worked out well in the past will likely also work out well in the future.

And any strategy that has performed poorly in the past will probably also do badly in the future.

Let’s Build an Algorithmic Trading Strategy with Python and Machine Learning

This program uses the dual moving average crossover to determine when to buy and sell stocks.

import pandas as pd
import numpy as np
from datetime import datetime
import matplotlib.pyplot as plt"dark_background")

You can download the data set we need for this purpose from here:

To store the data

apple = pd.read_csv("AAPL.csv")

Let’s Visualize the data

plt.figure(figsize=(12, 5))
plt.plot(apple['Adj Close Price'], label='Apple')
plt.title('Apple Adj Close Price History')
plt.xlabel("May 27,2014 - May 25,2020 ")
plt.ylabel("Adj Close Price USD ($)")
plt.legend(loc="upper left")

Create a Simple moving average with a 30 day window

sma30 = pd.DataFrame()
sma30['Adj Close Price'] = apple['Adj Close Price'].rolling(window=30).mean()

To create a Simple moving average 100 day window

sma100 = pd.DataFrame()
sma100['Adj Close Price'] = apple['Adj Close Price'].rolling(window=100).mean()

Now let’s Visualize the new data

plt.plot(apple['Adj Close Price'], label='Apple')
plt.plot(sma30['Adj Close Price'], label='SMA30')
plt.plot(sma100['Adj Close Price'], label='SMA100')
plt.title("Apple Adj. Close Price History")
plt.xlabel('May 27,2014 - May 25,2020')
plt.ylabel('Adj. Close Price USD($)')
plt.legend(loc='upper left')

Now create a new Data Frame to store all the data

data = pd.DataFrame()
data['apple'] = apple['Adj Close Price']
data['SMA30'] = sma30['Adj Close Price']
data['SMA100'] = sma100['Adj Close Price']

Create a function to signal when to buy or sell stock

def buySell(data):
  sigPriceBuy = []
  sigPriceSell = []
  flag = -1
  for i in range(len(data)):
    if data ['SMA30'][i] > data['SMA100'][i]:
      if flag != 1:
        flag = 1
    elif data['SMA30'][i] < data['SMA100'][i]:
      if flag != 0:
        flag = 0
  return(sigPriceBuy, sigPriceSell)

To store the buy and sell  data into a variable

buySell = buySell(data)
data['Buy Signal Price'] = buySell[0]
data['Sell Signal Price'] = buySell[1]
# To show the data

Now let’s Visualize the data and strategy to buy and sell stock'classic')
plt.plot(data['apple'], label='Apple', alpha=0.35)
plt.plot(data['SMA30'], label='SMA30', alpha=0.35)
plt.plot(data['SMA100'],label='SMA100', alpha=0.35)
plt.scatter(data.index, data['Buy Signal Price'], label ='Buy', marker='^',color='green')
plt.scatter(data.index, data['Sell Signal Price'],label='Sell', marker='v', color='red')
plt.title('Apple Adj. Close Price History Buy and Sell Signals')
plt.xlabel("May 27,2014 - May 25,2020")
plt.ylabel("Adj Close Price USD($)")
plt.legend(loc='upper left')

Follow us on Instagram for all your Queries

Articles: 75

One comment

Leave a Reply