Correlating Stock Returns Using Python

In this tutorial I'll walk you through a simple methodology to correlate various stocks against each other. We'll grab the prices of the selected stocks using python, drop them into a clean dataframe, run a correlation, and visualize our results.

(1) Import libraries, select our list of stocks to correlate

 import numpy as np
 import pandas as pd
 #used to grab the stock prices, with yahoo
 import pandas_datareader as web
 from datetime import datetime
 #to visualize the results
 import matplotlib.pyplot as plt
 import seaborn

 #select start date for correlation window as well as list of tickers
 start = datetime(2017, 1, 1)
 symbols_list = ['AAPL', 'F', 'TWTR', 'FB', 'AAL', 'AMZN', 'GOOGL', 'GE']

(2) Pull stock prices, push into clean dataframe

#array to store prices
symbols=[]

#pull price using iex for each symbol in list defined above
for ticker in symbols_list: 
    r = web.DataReader(ticker, 'yahoo', start)
    # add a symbol column
    r['Symbol'] = ticker 
    symbols.append(r)

# concatenate into df
df = pd.concat(symbols)
df = df.reset_index()
df = df[['Date', 'Close', 'Symbol']]
df.head()

We're now left with a table in this format:

3) However, we want our symbols represented as columns so we'll have to pivot the dataframe:

 df_pivot = df.pivot('date','symbol','close').reset_index()
 df_pivot.head()

(4) Next, we can run the correlation. Using the Pandas 'corr' function to compute the Pearson correlation coeffecient between each pair of equities

 corr_df = df_pivot.corr(method='pearson')
 #reset symbol as index (rather than 0-X)
 corr_df.head().reset_index()
 del corr_df.index.name
 corr_df.head(10)

5) Finally, we can plot a heatmap of the correlations (with Seaborn and Matplotlib) to better visualize the results:

 #take the bottom triangle since it repeats itself
 mask = np.zeros_like(corr_df)
 mask[np.triu_indices_from(mask)] = True
 #generate plot
 seaborn.heatmap(corr_df, cmap='RdYlGn', vmax=1.0, vmin=-1.0 , mask = mask, linewidths=2.5)
 plt.yticks(rotation=0) 
 plt.xticks(rotation=90) 
 plt.show()

Here we can see that, as expected, the tech companies are generally pretty highly correlated (as indicated by dark green -- AMZN/GOOGL, FB/AAPL, etc). Conversely, Ford (F) and General Electric (GE) are either not correlated or negatively correlated with the rest of the group.