Price Changes and Financial Returns

import pandas as pd
import matplotlib.pyplot as plt
pd.options.display.float_format = '{:.4f}'.format
plt.style.use("seaborn")

close = pd.read_csv("close.csv", index_col = "Date", parse_dates = ["Date"])
close

Drop the NaN value without changing the original data frame

msft = close.MSFT.dropna().to_frame().copy()

Change a column label MSFT to Price:

msft.rename(columns = {"MSFT":"Price"}, inplace = True)

Last day of price:

msft["P_lag1"] = msft.shift(periods = 1)

Compare between two days of price:

msft["P_diff"] = msft.Price.sub(msft.P_lag1) # Alternative 1
msft["P_diff2"] = msft.Price.diff(periods = 1)  # Alternative 2

#print msft table
msft

Absolute price changes are not meaningful yet, this is how we can calculate relative changes

#percentage change method

msft["Returns"] = msft.Price.pct_change(periods = 1) * 100 # Alternative 2
msft

The output of percentage change:

save the file to csv format:

msft.drop(columns = ["P_lag1", "P_diff", "P_diff2"], inplace = True)
msft.to_csv("msft.csv")