Data Visualization

For this data visualization project we use top 5 bank stock price data.

import pandas as pd
import yfinance as yf
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
from mywebstyle import plot_style
plot_style('#f4f4f4')

Data

start = pd.to_datetime('2020-01-01')
end = pd.to_datetime('today')
df = yf.download(['WFC','BAC','JPM','C','GS'], start=start, end=end)
df.index = df.index.date
df.tail()

[                       0%                       ][*******************   40%                       ]  2 of 5 completed[**********************60%****                   ]  3 of 5 completed[**********************80%*************          ]  4 of 5 completed[*********************100%***********************]  5 of 5 completed

Price	Close					High					...	Open					Volume
Ticker	BAC	C	GS	JPM	WFC	BAC	C	GS	JPM	WFC	...	BAC	C	GS	JPM	WFC	BAC	C	GS	JPM	WFC
2025-01-03	44.810001	71.000000	580.130005	242.029999	71.309998	44.849998	71.089996	582.140015	243.621784	71.419998	...	44.750000	70.879997	581.000000	242.636865	70.349998	23455700	11342900	1422600	9491100	9152800
2025-01-06	45.400002	72.739998	583.390015	240.850006	72.029999	46.150002	73.989998	593.650024	245.690002	73.250000	...	45.160000	72.230003	584.979980	243.699997	71.930000	30518500	19199700	2257000	9917800	15565600
2025-01-07	46.080002	73.680000	580.119995	243.169998	71.589996	46.650002	74.290001	588.280029	245.259995	73.129997	...	46.279999	73.900002	587.109985	242.139999	72.970001	41111200	18309400	2021900	8753400	13522000
2025-01-08	46.209999	73.260002	580.020020	243.130005	71.570000	46.259998	73.470001	581.200012	244.250000	72.080002	...	45.900002	73.449997	581.039978	242.750000	71.510002	40246000	13742800	1513400	8675300	16544700
2025-01-10	45.110001	71.400002	560.000000	239.869995	69.959999	46.049999	72.459999	576.000000	243.809998	70.910004	...	46.000000	72.459999	576.250000	243.089996	70.900002	39048804	10557376	2241713	7745890	11267550

5 rows × 25 columns

Now we compute the maximum closing prices of all these 5 banks during this time period

df.xs(key='Close', axis=1, level='Price').max()

Ticker
BAC     47.505741
C       73.680000
GS     605.570007
JPM    249.003983
WFC     77.349998
dtype: float64

Now we compute the returns for each of the stock

# Retrieve the 'Close' prices for each ticker directly
close_prices = df.xs(key='Close', axis=1, level=0)

# Calculate the daily percentage change (returns) for all tickers
returns = close_prices.pct_change()
returns.index = pd.to_datetime(returns.index)

# Display the first few rows of returns
returns.head()

Ticker	BAC	C	GS	JPM	WFC
2020-01-02	NaN	NaN	NaN	NaN	NaN
2020-01-03	-0.020763	-0.018835	-0.011693	-0.013196	-0.006140
2020-01-06	-0.001433	-0.003137	0.010234	-0.000795	-0.005990
2020-01-07	-0.006600	-0.008685	0.006583	-0.017001	-0.008286
2020-01-08	0.010110	0.007618	0.009640	0.007801	0.003038

Visualization

Let’s create a pairplot of the returns

sns.pairplot(returns[1:])

Now to check the maximum and minimum return and on what dates that happened

print('Minimum Return')
print(' ')
print(returns.idxmin())
print(' ')
print('Maximum Return')
print(' ')
print(returns.idxmax())

Minimum Return
 
Ticker
BAC   2020-03-16
C     2020-03-16
GS    2020-03-16
JPM   2020-03-16
WFC   2020-03-12
dtype: datetime64[ns]
 
Maximum Return
 
Ticker
BAC   2020-03-13
C     2020-03-13
GS    2020-03-13
JPM   2020-03-13
WFC   2020-03-24
dtype: datetime64[ns]

To find which bank is more risky we can simply check the standard deviations of the returns of each bank

returns.std()

Ticker
BAC    0.022482
C      0.024632
GS     0.020837
JPM    0.020450
WFC    0.024600
dtype: float64

It seems like CITI na groop has the maximum value in the standared deviations.

Now let’s check the distribution of Wells Fargo’s return in 2023

return_2023_wf = returns.loc['2023-01-01':'2023-12-31','WFC']
sns.displot(return_2023_wf, color='blue', bins=80, kde=True)

Next, we create the timeseries plot of the closing prices

df.xs(key='Close', axis=1, level='Price').plot(figsize=(9,5))

import plotly
import cufflinks as cf
cf.go_offline()
df.xs(key='Close', axis=1, level='Price').iplot()

Trend

Next we plot moving average for Wells Fargo

wfc = df['Close']['WFC']
start_date = pd.to_datetime('2023-10-10').date()
end_date = pd.to_datetime('2024-10-20').date()
wfc.loc[start_date:end_date].rolling(window=30).mean().plot(
    figsize=(9,5),label='30 Day Avg'
    )
wfc.loc[start_date:end_date].plot(
    label='WFC Close',figsize=(9,5)
    )
plt.legend()

Price Movement Correlation

sns.heatmap(df.xs(key='Close', axis=1, level='Price').corr(), annot=True)

Techniqal Analysis Plots

Candle PLot

wfc = df[['Open', 'High', 'Low', 'Close']].xs('WFC', level=1, axis=1)

# Define start and end dates
start_date = pd.to_datetime('2023-10-10').date()
end_date = pd.to_datetime('2024-10-20').date()

wfc_filtered = wfc.loc[start_date:end_date]
wfc_filtered.iplot(kind='candle')

Bollinger Plot

wfc_filtered['Close'].ta_plot(study='boll')

Share on