Project: Analyzing Stocks

In this project, we will focus on exploratory data analysis of stock prices for following companies from 2006 to 2015.

• Amazon
• Microsoft
• Bank of America
• Citibank

#Imports
import numpy as np
import pandas as pd
import pandas_datareader.data as web
import datetime as dt

#visualization libraries
import matplotlib.pyplot as plt
import seaborn as sns

#Start and end period
start = dt.datetime(2006,1,1)
end = dt.datetime(2016,1,1)

#Tickers of the companies
tickers = ['AMZN','MSFT','BAC','C']

#Fetch data from yahoo stocks using DataReader
amazon = web.DataReader('AMZN',data_source='yahoo',start=start,end=end)
microsoft = web.DataReader('MSFT',data_source='yahoo',start=start,end=end)
bank_of_america = web.DataReader('BAC',data_source='yahoo',start=start,end=end)
citigroup = web.DataReader('C',data_source='yahoo',start=start,end=end)

#Concating all the data frames of different companies into one large data frame. 
stock_data = pd.concat([amazon,microsoft,bank_of_america,citigroup],axis=1,keys=tickers)

#Adding column names
stock_data.columns.names = ['Tickers','Stock Info']

#Glimpse of data
stock_data.head()

Tickers	AMZN						MSFT						BAC						C
Stock Info	High	Low	Open	Close	Volume	Adj Close	High	Low	Open	Close	Volume	Adj Close	High	Low	Open	Close	Volume	Adj Close	High	Low	Open	Close	Volume	Adj Close
Date
2006-01-03	47.849998	46.250000	47.470001	47.580002	7582200	47.580002	27.000000	26.10	26.250000	26.840000	79973000.0	19.602528	47.180000	46.150002	46.919998	47.080002	16296700.0	35.298687	493.799988	481.100006	490.000000	492.899994	1537600.0	440.882477
2006-01-04	47.730000	46.689999	47.490002	47.250000	7440900	47.250000	27.080000	26.77	26.770000	26.969999	57975600.0	19.697485	47.240002	46.450001	47.000000	46.580002	17757900.0	34.923801	491.000000	483.500000	488.600006	483.799988	1870900.0	432.742950
2006-01-05	48.200001	47.110001	47.160000	47.650002	5417200	47.650002	27.129999	26.91	26.959999	26.990000	48245500.0	19.712091	46.830002	46.320000	46.580002	46.639999	14970700.0	34.968796	487.799988	484.000000	484.399994	486.200012	1143100.0	434.889679
2006-01-06	48.580002	47.320000	47.970001	47.869999	6152900	47.869999	27.000000	26.49	26.889999	26.910000	100963000.0	19.653666	46.910000	46.349998	46.799999	46.570000	12599800.0	34.916302	489.000000	482.000000	488.799988	486.200012	1370200.0	434.889679
2006-01-09	47.099998	46.400002	46.549999	47.080002	8943100	47.080002	27.070000	26.76	26.930000	26.860001	55625000.0	19.617136	46.970001	46.360001	46.720001	46.599998	15619400.0	34.938789	487.399994	483.000000	486.000000	483.899994	1680700.0	432.832489

#Max close price for each company's stock
stock_data.xs(key='Close',level='Stock Info',axis=1).max()

    Tickers
    AMZN    693.969971
    MSFT     56.549999
    BAC      54.900002
    C       564.099976
    dtype: float64

#Return of each company
returns = pd.DataFrame()
for ticker in tickers:
    returns[ticker] = stock_data[ticker]['Close'].pct_change()

returns.head()

	AMZN	MSFT	BAC	C
Date
2006-01-03	NaN	NaN	NaN	NaN
2006-01-04	-0.006936	0.004843	-0.010620	-0.018462
2006-01-05	0.008466	0.000742	0.001288	0.004961
2006-01-06	0.004617	-0.002964	-0.001501	0.000000
2006-01-09	-0.016503	-0.001858	0.000644	-0.004731

sns.pairplot(returns)

    <seaborn.axisgrid.PairGrid at 0x1877dbb83c8>

#Worst return date for each company
returns.idxmin()

    AMZN   2006-07-26
    MSFT   2009-01-22
    BAC    2009-01-20
    C      2009-02-27
    dtype: datetime64[ns]

#Best return date for each company
returns.idxmax()

    AMZN   2007-04-25
    MSFT   2008-10-13
    BAC    2009-04-09
    C      2008-11-24
    dtype: datetime64[ns]

#Standard deviation of returns
returns.std()

    AMZN    0.026638
    MSFT    0.017764
    BAC     0.036647
    C       0.038672
    dtype: float64

#Standard deviation of returns for year 2015
returns.loc['2015-01-01':'2015-12-31'].std()

    AMZN    0.021147
    MSFT    0.017801
    BAC     0.016163
    C       0.015289
    dtype: float64

#Lets create distribution plot of the stock returns of Amazon for year 2015
sns.distplot(returns.loc['2015-01-01':'2015-12-31']['AMZN'],bins=100)

    <matplotlib.axes._subplots.AxesSubplot at 0x1877f3e4588>

#Lets create distribution plot of the stock returns of Citigroup for year 2008
sns.distplot(returns.loc['2008-01-01':'2008-12-31']['C'],bins=100)

    <matplotlib.axes._subplots.AxesSubplot at 0x187009ce518>

#Close price of each company throughout the entire time period of 2006-2015
stock_data.xs(key='Close',axis=1,level='Stock Info').plot(figsize=(10,5))
plt.legend()

    <matplotlib.legend.Legend at 0x1877fdf1278>

#30 day moving average of Amazon for year 2015
stock_data.loc['2015-1-1':'2015-12-31','AMZN']['Close'].rolling(30).mean().plot(figsize=(10,5),label='MA30')
stock_data.loc['2015-1-1':'2015-12-31','AMZN']['Close'].plot(label='Close')
plt.tight_layout()
plt.legend()

    <matplotlib.legend.Legend at 0x1870483e048>

#30 day moving average of Citigroup for year 2008
stock_data.loc['2008-1-1':'2008-12-31','C']['Close'].rolling(30).mean().plot(figsize=(10,5),label='MA30')
stock_data.loc['2008-1-1':'2008-12-31','C']['Close'].plot(label='Close')
plt.tight_layout()
plt.legend()

    <matplotlib.legend.Legend at 0x18704839898>

#Heatmap for the relation between close prices of different stocks
sns.heatmap(stock_data.xs(key='Close',axis=1,level='Stock Info').corr(),cmap='magma_r',annot=True)

    <matplotlib.axes._subplots.AxesSubplot at 0x18704ae9a58>