Stock Analysis

In this project we will focus in exploratory data analysis of stock prices.Keep in mind, this project is just meant to practice visualizations and pandas skills, it is not meant to be a robust financial analysis or be taken as financial advice.

We will focus on some US bank stocks and see how they progressed thoughtout the financial crisis of 2016. Then do a quick comparison with last 10 years data.

Import your important libraries:

import numpy as np
import pandas as pd 
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline
import datetime as dt
import yfinance as yf

Import plotly and cufflinks for interactive graphs

cufflinks connects pandas dataframe with plotly library and helps to run the visualizations directly

!pip install plotly
!pip install cufflinks

import plotly 
import cufflinks as cf
cf.go_offline()

sns.set_style('whitegrid')

Data

We need to get data using yahoo finance.We will get stock information for the following banks:

• Bank of America
• CitiGroup
• GoldmanSachs
• JPMorgan Chase
• Morgan Stanley
• Wells Fargo

Pull th stock data from Jan 1st 2006 to Jan 1st 2016 for each of these banks. Set each bank to be a separate dataframe, with variable name for that bank being its ticker symbol. This will involve few steps:**

1. Use datetime to set start and end datetime objects.
2. Figure out ticker symbol for each bank.
3. Figure out how to use yfinance to grab info on the stock

start = dt.date(year = 2006, month = 1 , day = 1)
end = dt.date(2016,1,1)

#pull the stock information from yfinanc
#Process: create a df(BAC)->pull data using yf library

BAC = yf.download('BAC', start, end)          #Bank of America
C = yf.download('C', start, end)              #Citigroup
GS = yf.download('GS', start, end)            #Goldmansachs
JPM = yf.download('JPM', start, end)          #JPMorgan Chase
MS = yf.download('MS', start, end)            #MorganStanley
WFC = yf.download('WFC', start, end)          #WellsFargo

[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed

WFC

	Open	High	Low	Close	Adj Close	Volume
Date
2006-01-03	31.600000	31.975000	31.195000	31.900000	18.979551	11016400
2006-01-04	31.799999	31.820000	31.365000	31.530001	18.759415	10870000
2006-01-05	31.500000	31.555000	31.309999	31.495001	18.738600	10158000
2006-01-06	31.580000	31.775000	31.385000	31.680000	18.848658	8403800
2006-01-09	31.674999	31.825001	31.555000	31.674999	18.845688	5619600
...	...	...	...	...	...	...
2015-12-24	54.970001	55.090000	54.709999	54.820000	42.477665	4999400
2015-12-28	54.549999	54.779999	54.169998	54.680000	42.369175	8288800
2015-12-29	55.110001	55.349998	54.990002	55.290001	42.841831	7894900
2015-12-30	55.270000	55.310001	54.790001	54.889999	42.531898	8016900
2015-12-31	54.509998	54.950001	54.220001	54.360001	42.121220	10929800

2517 rows × 6 columns

Create a list of ticker symbols(as strings) in alphabetical order. Call this list: tickers.

tickers = 'BAC C GS JPM MS WFC'.split()
tickers

['BAC', 'C', 'GS', 'JPM', 'MS', 'WFC']

Use pd.concat to concatenate the bank dataframes together to a single dataframe called bank_stocks. Set the key argument equal to the tickers list.

bank_stocks = pd.concat([BAC, C, GS, JPM, MS, WFC], axis = 1,keys=tickers)
bank_stocks

	BAC						C				...	MS				WFC
	Open	High	Low	Close	Adj Close	Volume	Open	High	Low	Close	...	Low	Close	Adj Close	Volume	Open	High	Low	Close	Adj Close	Volume
Date
2006-01-03	46.919998	47.180000	46.150002	47.080002	31.544905	16296700	490.000000	493.799988	481.100006	492.899994	...	56.740002	58.310001	32.661312	5377000	31.600000	31.975000	31.195000	31.900000	18.979551	11016400
2006-01-04	47.000000	47.240002	46.450001	46.580002	31.209898	17757900	488.600006	491.000000	483.500000	483.799988	...	58.349998	58.349998	32.683727	7977800	31.799999	31.820000	31.365000	31.530001	18.759415	10870000
2006-01-05	46.580002	46.830002	46.320000	46.639999	31.250097	14970700	484.399994	487.799988	484.000000	486.200012	...	58.020000	58.509998	32.773338	5778000	31.500000	31.555000	31.309999	31.495001	18.738600	10158000
2006-01-06	46.799999	46.910000	46.349998	46.570000	31.203197	12599800	488.799988	489.000000	482.000000	486.200012	...	58.049999	58.570000	32.806950	6889800	31.580000	31.775000	31.385000	31.680000	18.848658	8403800
2006-01-09	46.720001	46.970001	46.360001	46.599998	31.223280	15619400	486.000000	487.399994	483.000000	483.899994	...	58.619999	59.189999	33.154228	4144500	31.674999	31.825001	31.555000	31.674999	18.845688	5619600
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2015-12-24	17.320000	17.379999	17.219999	17.270000	14.271372	29369400	52.480000	52.970001	52.450001	52.709999	...	32.439999	32.480000	25.353054	2798200	54.970001	55.090000	54.709999	54.820000	42.477665	4999400
2015-12-28	17.219999	17.230000	16.980000	17.129999	14.155680	41777500	52.570000	52.570000	51.959999	52.380001	...	31.950001	32.169998	25.111073	5420300	54.549999	54.779999	54.169998	54.680000	42.369175	8288800
2015-12-29	17.250000	17.350000	17.160000	17.280001	14.279634	45670400	52.759998	53.220001	52.740002	52.980000	...	32.330002	32.549999	25.407690	6388200	55.110001	55.349998	54.990002	55.290001	42.841831	7894900
2015-12-30	17.200001	17.240000	17.040001	17.049999	14.089570	35066400	52.840000	52.939999	52.250000	52.299999	...	32.200001	32.230000	25.157906	5057200	55.270000	55.310001	54.790001	54.889999	42.531898	8016900
2015-12-31	17.010000	17.070000	16.830000	16.830000	13.907768	47153000	52.070000	52.389999	51.750000	51.750000	...	31.770000	31.809999	24.830067	8154300	54.509998	54.950001	54.220001	54.360001	42.121220	10929800

2517 rows × 36 columns

Set the column name levels 'Bank Tickers' and 'Stock info':

bank_stocks.columns.names =['Bank Tickers', 'Stock Info']
bank_stocks.columns.names

FrozenList(['Bank Tickers', 'Stock Info'])

Check the head of the bank_stocks dataframe

bank_stocks.head()

Bank Tickers	BAC						C				...	MS				WFC
Stock Info	Open	High	Low	Close	Adj Close	Volume	Open	High	Low	Close	...	Low	Close	Adj Close	Volume	Open	High	Low	Close	Adj Close	Volume
Date
2006-01-03	46.919998	47.180000	46.150002	47.080002	31.544905	16296700	490.000000	493.799988	481.100006	492.899994	...	56.740002	58.310001	32.661312	5377000	31.600000	31.975000	31.195000	31.900000	18.979551	11016400
2006-01-04	47.000000	47.240002	46.450001	46.580002	31.209898	17757900	488.600006	491.000000	483.500000	483.799988	...	58.349998	58.349998	32.683727	7977800	31.799999	31.820000	31.365000	31.530001	18.759415	10870000
2006-01-05	46.580002	46.830002	46.320000	46.639999	31.250097	14970700	484.399994	487.799988	484.000000	486.200012	...	58.020000	58.509998	32.773338	5778000	31.500000	31.555000	31.309999	31.495001	18.738600	10158000
2006-01-06	46.799999	46.910000	46.349998	46.570000	31.203197	12599800	488.799988	489.000000	482.000000	486.200012	...	58.049999	58.570000	32.806950	6889800	31.580000	31.775000	31.385000	31.680000	18.848658	8403800
2006-01-09	46.720001	46.970001	46.360001	46.599998	31.223280	15619400	486.000000	487.399994	483.000000	483.899994	...	58.619999	59.189999	33.154228	4144500	31.674999	31.825001	31.555000	31.674999	18.845688	5619600

5 rows × 36 columns

Exploratory Data Analysis

Let's explore the data a bit! What is the max Close price for each bank's stock throughout the time period

# use cross section for multiindex data frames

bank_stocks.xs('Close',axis=1,level=1) # Returns the dataframe with Close columns

Bank Tickers	BAC	C	GS	JPM	MS	WFC
Date
2006-01-03	47.080002	492.899994	128.869995	40.189999	58.310001	31.900000
2006-01-04	46.580002	483.799988	127.089996	39.619999	58.349998	31.530001
2006-01-05	46.639999	486.200012	127.040001	39.740002	58.509998	31.495001
2006-01-06	46.570000	486.200012	128.839996	40.020000	58.570000	31.680000
2006-01-09	46.599998	483.899994	130.389999	40.669998	59.189999	31.674999
...	...	...	...	...	...	...
2015-12-24	17.270000	52.709999	182.470001	66.599998	32.480000	54.820000
2015-12-28	17.129999	52.380001	181.619995	66.379997	32.169998	54.680000
2015-12-29	17.280001	52.980000	183.529999	67.070000	32.549999	55.290001
2015-12-30	17.049999	52.299999	182.009995	66.589996	32.230000	54.889999
2015-12-31	16.830000	51.750000	180.229996	66.029999	31.809999	54.360001

2517 rows × 6 columns

bank_stocks.xs('Close',axis=1,level=1).max() #Returns the max value of each column

Bank Tickers
BAC     54.900002
C      564.099976
GS     247.919998
JPM     70.080002
MS      89.300003
WFC     58.520000
dtype: float64

Create a new empty DataFrame called returns. The dataframe will contain the returns for each bank's stock. returns are typically defined by

$$ r_t = \frac{P_t - P_{t-1}}{ P_t-1} = \frac{P_t}{P_{t-1}}-1 $$

returns = pd.DataFrame()
print(returns)

Empty DataFrame
Columns: []
Index: []

Use pandas pct_change() method on the Close column to create a column representing this return value. Create a for loop that goes and for each Bank Stock Ticker creates this returns column and sets it as a column in the returns DataFrame.

for i in tickers: 
    returns[i + ' Returns'] = bank_stocks.xs('Close', axis = 1, level = 1)[i].pct_change() 
    

returns

	BAC Returns	C Returns	GS Returns	JPM Returns	MS Returns	WFC Returns
Date
2006-01-03	NaN	NaN	NaN	NaN	NaN	NaN
2006-01-04	-0.010620	-0.018462	-0.013812	-0.014183	0.000686	-0.011599
2006-01-05	0.001288	0.004961	-0.000393	0.003029	0.002742	-0.001110
2006-01-06	-0.001501	0.000000	0.014169	0.007046	0.001025	0.005874
2006-01-09	0.000644	-0.004731	0.012030	0.016242	0.010586	-0.000158
...	...	...	...	...	...	...
2015-12-24	-0.004037	0.001520	-0.002624	-0.001948	-0.003681	-0.003997
2015-12-28	-0.008107	-0.006261	-0.004658	-0.003303	-0.009544	-0.002554
2015-12-29	0.008757	0.011455	0.010516	0.010395	0.011812	0.011156
2015-12-30	-0.013310	-0.012835	-0.008282	-0.007157	-0.009831	-0.007235
2015-12-31	-0.012903	-0.010516	-0.009780	-0.008410	-0.013031	-0.009656

2517 rows × 6 columns

Observations

• What pct_change() does here is it computes the fractional change from the immidiately previous row by default. It is useful for comparing the fraction of change in a time series of elements. Remeber here it is a fraction if yoou want to change it to percentage u want to multiply it with 100
• So we have got the columns for each of the stock tickers for each trading day
• For the first day the return is obiviously Nan as there was no prior day element to calculate it
• From the second row we can see the positive and negative returns , negatives are the days where it went down and positives are the days when it went up

Create a pair plot using seaborn of the returns dataframe. Do any stocks stands out ? Why or why not?

sns.pairplot(returns)

<seaborn.axisgrid.PairGrid at 0x3392183a0>

Observations

• Diagonal elements are univariant data i.e, just for that particular column
• Off diagonal elements are bivariant data
• What pairplot does here is it creates a group of scatter plots for each pair of the numerical data in the data frame,For exit will create a 'BAC' vs 'C' scatterplot, 'BAC' vs 'GS' scatter plot and so on.
• So we have all of these tickers/banks on X-axis and the same banks as well as on the Y-axis.
• So if you compare 'WFC Returns' to the 'WFC Returns' it is a univariant data so therefore we have a Histogram.
• If you compare 'WFC Returns' to the other banks we have the scatter plots to analyse the data.
• Its very important to check the scales for it , almost everything are in decimals on both X and Y axis.
• Sometimes based on the data you can get different scales for one of the columns and that graph will look very different.
• Lets look at the 'MS Returns' graph is going from -0.25 to 0.75 and there is only one outlier point that is there at 0.75 which makes the slope is much higher but if you ignore the outlier the slope will be much equal as the distance from zero on both sides is almost equal for rest of the datapoints.
• Same with the 'Citi Group' except few(countable) points most of them lies in the range of -0.25 to 0.25.
• So if you analyse this for 'BAC Returns' compared to other banks, there is a very strong correlation between these banks or the returns of these banks as the X-axis goes up Y-axis goes up as well and that pattern is prevalent for all combinations of the all pairs of the banks here.

returns.corr()

	BAC Returns	C Returns	GS Returns	JPM Returns	MS Returns	WFC Returns
BAC Returns	1.000000	0.802752	0.685271	0.815410	0.643840	0.828337
C Returns	0.802752	1.000000	0.668989	0.739045	0.634352	0.717386
GS Returns	0.685271	0.668989	1.000000	0.738671	0.805155	0.662847
JPM Returns	0.815410	0.739045	0.738671	1.000000	0.646711	0.826276
MS Returns	0.643840	0.634352	0.805155	0.646711	1.000000	0.598681
WFC Returns	0.828337	0.717386	0.662847	0.826276	0.598681	1.000000

Observations

• We can see the correlation in the matrix format.
• We can use these later to create a heat map.
• 'BAC Returns have strong relationship with the 'City Bank', 'JPMorgan' and 'WellsFargo' compared to the 'GoldmanSachs' and 'Morgan Stanley' banks.

Using this returns DataFrame, figure out on what dates each bank stock had the best and worst single day returns. Did anything significant happen on those days?

#Best single day returns
returns.max()  #Returns the max value of each column 

BAC Returns    0.352691
C Returns      0.578249
GS Returns     0.264678
JPM Returns    0.250967
MS Returns     0.869835
WFC Returns    0.327645
dtype: float64

#Best single day returns
returns.idxmax() #Returns the index of the max value

BAC Returns   2009-04-09
C Returns     2008-11-24
GS Returns    2008-11-24
JPM Returns   2009-01-21
MS Returns    2008-10-13
WFC Returns   2008-07-16
dtype: datetime64[ns]

#Worst single day returns
returns.idxmin() #Returns the index of the min value

BAC Returns   2009-01-20
C Returns     2009-02-27
GS Returns    2009-01-20
JPM Returns   2009-01-20
MS Returns    2008-10-09
WFC Returns   2009-01-20
dtype: datetime64[ns]

Did anything significant happen on those days?

Best days

• If you notice the JPMorgan returns the best performance day(2009-01-21) was exactly after the worst performance day(2009-01-20) and the reason for that was it received a 700 billion dollar bailout from the federal reserve.

Worst days

• If you look carefully four of the banks Bank of America, Goldman Sachs, JPMorgan and Wells Fargo had the worst return on the same day.

  What happened exactly?

• On the particular day 2009-01-20, The Icelands banks collapsed.

• InShort, About Iceland banks, During 2000-2007 the Icelandic banks were privatised and saw quiet a bit of Boom years so much so that the country's entire economy and GDP became dependent on the performance of the banks, and when the economy collapsed theirs banks and government collapsed . That might have created a widespread of panic in the market which might be one of the reasons for the decline on the particular day [ 1 ][2 . Finacncial crisis 2009].

• For the City Bank,On 27th February 2009,The US government has reached a deal to take a stake of 30 to 40 percent , which raised concerns among the Investors regarding the nationalisation of the banks and thats why it has the biggest decline of its stock worldwide on that particular day.

More Details

Take a look at the standard deviation of the returns, which stock would you classify as the riskiest over the entire period? Which would u classify as the riskiest for the year 2015?

#For the entire 10 years period (More riskier higher the std.deviation)
returns.std()

BAC Returns    0.036647
C Returns      0.038672
GS Returns     0.025390
JPM Returns    0.027667
MS Returns     0.037819
WFC Returns    0.030238
dtype: float64

Observations

• There isnt much to differentiate between these stocks. Statistically, you can say Citi bank is the most riskiest but its only by the bearest of margin really insignificant. Practically they all had similiar risk.

#For year 2015

returns.loc['2015-01-01':'2015-12-31'].std()

BAC Returns    0.016163
C Returns      0.015289
GS Returns     0.014046
JPM Returns    0.014017
MS Returns     0.016249
WFC Returns    0.012591
dtype: float64

Obseravtion

• There isnt much difference among these stocks they all posses similiar risk.

Create a histplot using seaborn of the 2015 returns for the Morgan Stanley

sns.histplot(returns.loc['2015-01-01':'2015-12-31']['MS Returns'], color ='g', bins =50, kde = True)

<Axes: xlabel='MS Returns', ylabel='Count'>

Observations

• If you see the data most of the data is in between -0.02 and 0.02. If you imagine this in percentage by multiplying the numbers with 100,so most of the daily return within plus or minus 2% of the previous day number.

Create a histplot using seaborn of the 2008 returns for the CitiGroup

sns.histplot(returns.loc['2008-01-01':'2008-12-31']['C Returns'], color = 'r', bins = 50, kde= True)

<Axes: xlabel='C Returns', ylabel='Count'>

Observations

• If you notice the Kde graph there is a long tail on the right side that is because of some point 0.6 . Basically it says that they have a day where there is very high return approx. 6 percent but its only few days maybe only one or two days.

Create a line plot showing Close price for each bank for the entire index of time.

bank_stocks.xs('Close', axis = 1 ,level = 1).iplot(kind = 'line')

Moving Averages

Lets analyze the moving average for these stocks in the year 2008

BAC.loc['2008-01-01':'2008-12-31'].rolling(window=30).std().head(30)

	Open	High	Low	Close	Adj Close	Volume
Date
2008-01-02	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-03	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-04	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-07	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-08	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-09	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-10	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-11	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-14	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-15	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-16	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-17	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-18	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-22	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-23	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-24	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-25	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-28	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-29	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-30	NaN	NaN	NaN	NaN	NaN	NaN
2008-01-31	NaN	NaN	NaN	NaN	NaN	NaN
2008-02-01	NaN	NaN	NaN	NaN	NaN	NaN
2008-02-04	NaN	NaN	NaN	NaN	NaN	NaN
2008-02-05	NaN	NaN	NaN	NaN	NaN	NaN
2008-02-06	NaN	NaN	NaN	NaN	NaN	NaN
2008-02-07	NaN	NaN	NaN	NaN	NaN	NaN
2008-02-08	NaN	NaN	NaN	NaN	NaN	NaN
2008-02-11	NaN	NaN	NaN	NaN	NaN	NaN
2008-02-12	NaN	NaN	NaN	NaN	NaN	NaN
2008-02-13	2.490761	2.092302	2.555124	2.313858	1.698958	2.104923e+07

Plot the rolling 30 day average against the Close Price for Bank Of America's stock for the year 2008

BAC.loc['2008-01-01':'2008-12-31']['Close'].rolling(window=30).mean().plot(figsize = (10,4), label = '30 Day Average of Close Price')
BAC.loc['2008-01-01':'2008-12-31']['Close'].plot(label = 'Close Price')
plt.legend()

Observations

• If you see the difference between rolling average and actual closing stock prices, the real stock prices can have sharp declines and inclines and what the 30 day average does is it ignores those sharp edges and smoothens the data based on average of the last 30 days at each point. This is basically used to analyse the stock data so the sudden falls can be ignored and its not the part of the stock analysis unless it is sustained over a longer period of time.
• If u see the data between 2008-05 to 2008-07 the stock price started to fall but the average price was still high and because the fall sustained over a period of time the avg value came down as well. The avg values has much more roundish edge and the actual closing price has a sharp edge and the same thing happens between the period 2008-07 to 2008-11 as well.

Create a heatmap of the correlation between the stocks Close price.

bank_stocks.xs('Close', axis=1, level=1).corr()

Bank Tickers	BAC	C	GS	JPM	MS	WFC
Bank Tickers
BAC	1.000000	0.971516	0.550898	0.103874	0.944218	0.008542
C	0.971516	1.000000	0.434123	0.003515	0.933609	-0.068536
GS	0.550898	0.434123	1.000000	0.685286	0.683792	0.499897
JPM	0.103874	0.003515	0.685286	1.000000	0.250427	0.940269
MS	0.944218	0.933609	0.683792	0.250427	1.000000	0.131835
WFC	0.008542	-0.068536	0.499897	0.940269	0.131835	1.000000

sns.heatmap(bank_stocks.xs('Close', axis=1, level=1).corr(),cmap = 'crest', annot = True)

Observations:

• If you see correlation between bank of america to bank of america is same .
• Correlation basically says if the stock price for the bank of america/any bank increases what is the movement of the stock price for the other banks. So it basically denotes the linear relationship between those banks.
• +ve relationship means if price of one bank increases, price of the other bank increases as well, -ve relation will show movement on the negative direction compared to the stock.
• Correlation always varies between -1 to 1.
• +ve 1 could mean perfect correlation which means both the stock goes up or down at the same time.
• -ve 1 means everytime the stock goes up the other stock goes down.
• Correlation of zero(0) means that they dont have correlation among themselves. The movement of one stock has no relationship with the movement of the other stock.
• These are used by the hedgefunds and other guys who take lot of risk to hedge their risk based on the movement of the stock price. So if the are betting on BAC to go up it has a -ve correlation with some other stock and they can make some position in the other stock as well. So just in case if BAC goes down the other stock might go up there by they might not be losing as much money.
• You can clearly see the BAC has a strong correlationship with the City Bank and Morgan Stanley Banks where as a very weak correlationship with WellsFargo. Even with JPMorgan it is 0.1 which is a weak correlation and close to zero .

Use seaborn's clustermap to cluster the correlations together:

sns.clustermap(bank_stocks.xs('Close', axis=1, level=1).corr(),figsize=(6,6),cmap = 'crest',annot= True)

Observations

• Basically what clustermap does is that it creates a cluster based on the closed data.
• If you see on the both X and the Y axis, the map BAC lies between MS and C whereas in the heatmap its had very strong relationship with the City bank and Morgan Stanley and Wells Fargo and JPMorgan were close to eachother.So in the cluster map WFC and JPM has put together.

USING CUFFLINKS

Use .plot(kind='candle') to create a candle plot of Bank of americas stock from Jans 1st 2015 to Jan 1st 2016.

BAC.loc['2015-01-01':'2016-01-01'].iplot(kind='candle')

Use .ta_plot(study='sma') to create a Simple Moving Average plot of Morgan Stanley for the year 2015.

MS.loc['2015-01-01':'2016-01-01']['Close'].ta_plot(study='sma',period=[14,21,55])

Observations

• Simple moving average smoothens the edges of the stock.
• period=[14,21,55] -> This gives the simple moving averages of 14th 21 and 55th day.

MS.loc['2015-01-01':'2016-01-01']['Close'].ta_plot(study='rsi')

Use .ta_plot(study='boll') to create a Bollinger Band plot for Bank of Americs for the year 2015.

BAC.loc['2015-01-01':'2015-12-31']['Close'].ta_plot(study='boll')

Observations

• What the bollinger does is it creates a simple moving average and has a upper line and a lower line aswell.

Just update the start and end dates, and return the codes to analyze the last 10 years performance.

Findings

• The maximum closing price for each bank's stock throughout the time period was calculated using the max() function.
• The returns for each bank's stock were calculated using the pct_change() function and stored in a new dataframe called returns.
• A pairplot was created using seaborn to visualize the returns dataframe. The plot showed that there was a strong correlation between the returns of the banks.
• The best and worst single day returns for each bank were identified using the idxmax() and idxmin() functions. The worst single day returns for four of the banks occurred on the same day, which was the day that the Icelandic banks collapsed.
• The standard deviation of the returns was calculated to determine which stocks were the riskiest over the entire period and in the year 2015. The results showed that there was not much difference in the riskiness of the stocks.
• Histograms were created using seaborn to visualize the returns for Morgan Stanley in 2015 and for CitiGroup in 2008. The histograms showed that most of the daily returns were within plus or minus 2% of the previous day's number.
• A line plot was created to show the close price for each bank for the entire index of time.
• Moving averages were calculated for the stocks in the year 2008 using the rolling() function.
• A line plot was created to show the rolling 30 day average against the close price for Bank of America's stock for the year 2008.
• A heatmap was created to show the correlation between the stocks' close price using the corr() function and seaborn's heatmap() function. The heatmap showed that Bank of America had a strong correlation with CitiGroup and Morgan Stanley, while it had a weak correlation with Wells Fargo.
• A clustermap was created to cluster the correlations together using seaborn's clustermap() function. The clustermap showed that Bank of America's stock was clustered with Morgan Stanley and CitiGroup, while Wells Fargo's stock was clustered with JPMorgan Chase.

Reports

• The maximum closing price for each bank's stock throughout the time period was analyzed. Bank of America had the highest maximum closing price, while CitiGroup had the lowest.
• The returns for each bank's stock were analyzed using a pairplot. The plot showed that there was a strong correlation between the returns of the banks.
• The best and worst single day returns for each bank were identified. The worst single day returns for four of the banks occurred on the same day, which was the day that the Icelandic banks collapsed.
• The standard deviation of the returns was calculated to determine which stocks were the riskiest over the entire period and in the year 2015. The results showed that there was not much difference in the riskiness of the stocks.
• Histograms were created to visualize the returns for Morgan Stanley in 2015 and for CitiGroup in 2008. The histograms showed that most of the daily returns were within plus or minus 2% of the previous day's number.
• A line plot was created to show the close price for each bank for the entire index of time. The plot showed that the close prices for the banks were generally trending upwards over the time period.
• Moving averages were calculated for the stocks in the year 2008. The moving averages were used to smooth out the noise in the data and to identify trends.
• A line plot was created to show the rolling 30 day average against the close price for Bank of America's stock for the year 2008. The plot showed that the rolling 30 day average followed the overall trend of the close price, but with less volatility.
• A heatmap was created to show the correlation between the stocks' close price. The heatmap showed that Bank of America had a strong correlation with CitiGroup and Morgan Stanley, while it had a weak correlation with Wells Fargo.
• A clustermap was created to cluster the correlations together. The clustermap showed that Bank of America's stock was clustered with Morgan Stanley and CitiGroup, while Wells Fargo's stock was clustered with JPMorgan Chase.
• Overall, the analysis showed that there was a strong correlation between the returns of the banks, and that Bank of America had a strong correlation with CitiGroup and Morgan Stanley, while it had a weak correlation with Wells Fargo. The histograms and moving averages were used to smooth out the noise in the data and to identify trends. The clustermap was used to group the stocks based on their correlations.

Conclusions

This project demonstrates my skills as a data analyst by utilizing Python and various libraries pandas,numpy,matplotlib,seaborn and plotly to analyze stock data. Through exploratory data analysis, we identified patterns, trends, and correlations in the stock prices of major US banks over a 10-year period. By visualizing the data using interactive graphs and statistical analysis, we gained insights into the risk and return profiles of these stocks. This project showcases my ability to handle data, perform statistical analysis, and communicate findings effectively.

Author - Thanveer Ahmed Shaik

This project is part of my portfolio, showcasing the Python skills essential for data analyst roles. If you have any questions, feedback, or would like to collaborate, feel free to get in touch!

• LinkedIn: Connect with me professionally
• Mail: shaikthanveerahmed123@gmail.com