time-analysis-finance-2

September 24, 2023

[1]: import pandas as pd

import numpy as np
import matplotlib.pyplot as plt
import yfinance as yf

1 Data Cleaning
[2]: datasets = ['AAPL']

for dataset in datasets :

Ticker = yf.Ticker(dataset)
data = Ticker.history(start='2023-01-01' , end='2023-09-23')
filename = f"{dataset}_data.csv"
data.to_csv(filename)
print(f"Download data for {dataset} and saved as {filename}")

Download data for AAPL and saved as AAPL_data.csv

[3]: Ticker = 'AAPL'

start_date = '2019-01-01'
end_date = '2023-09-23'

[4]: stock_data = yf.download(Ticker, start=start_date, end=end_date)

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

[5]: stock_data

[5]: Open High Low Close Adj Close \

Date
2019-01-02 38.722500 39.712502 38.557499 39.480000 37.943260
2019-01-03 35.994999 36.430000 35.500000 35.547501 34.163837
2019-01-04 36.132500 37.137501 35.950001 37.064999 35.622250
2019-01-07 37.174999 37.207500 36.474998 36.982498 35.542976
2019-01-08 37.389999 37.955002 37.130001 37.687500 36.220524
… … … … … …

1
 2023-09-18 176.479996 179.380005 176.169998 177.970001 177.970001
2023-09-19 177.520004 179.630005 177.130005 179.070007 179.070007
2023-09-20 179.259995 179.699997 175.399994 175.490005 175.490005
2023-09-21 174.550003 176.300003 173.860001 173.929993 173.929993
2023-09-22 174.669998 177.078995 174.054993 174.789993 174.789993

Volume
Date
2019-01-02 148158800
2019-01-03 365248800
2019-01-04 234428400
2019-01-07 219111200
2019-01-08 164101200
… …
2023-09-18 67257600
2023-09-19 51826900
2023-09-20 58436200
2023-09-21 63047900
2023-09-22 56682928

[1190 rows x 6 columns]

[12]: # Data Cleaning

# Remove duplicate row if any

stock_data = stock_data.drop_duplicates()

[13]: stock_data

[13]: Open High Low Close Adj Close \

Date
2019-01-02 38.722500 39.712502 38.557499 39.480000 37.943249
2019-01-03 35.994999 36.430000 35.500000 35.547501 34.163818
2019-01-04 36.132500 37.137501 35.950001 37.064999 35.622253
2019-01-07 37.174999 37.207500 36.474998 36.982498 35.542973
2019-01-08 37.389999 37.955002 37.130001 37.687500 36.220531
… … … … … …
2023-09-18 176.479996 179.380005 176.169998 177.970001 177.970001
2023-09-19 177.520004 179.630005 177.130005 179.070007 179.070007
2023-09-20 179.259995 179.699997 175.399994 175.490005 175.490005
2023-09-21 174.550003 176.300003 173.860001 173.929993 173.929993
2023-09-22 174.669998 177.078995 174.054993 174.789993 174.789993

Volume
Date
2019-01-02 148158800
2019-01-03 365248800

2
 2019-01-04 234428400
2019-01-07 219111200
2019-01-08 164101200
… …
2023-09-18 67257600
2023-09-19 51826900
2023-09-20 58436200
2023-09-21 63047900
2023-09-22 55110610

[1190 rows x 6 columns]

[14]: #handling Missing Value

#Forward fill missing value in case of gaps in data
stock_data['Close'].fillna(method='ffill', inplace=True)

[15]: stock_data

[15]: Open High Low Close Adj Close \

Date
2019-01-02 38.722500 39.712502 38.557499 39.480000 37.943249
2019-01-03 35.994999 36.430000 35.500000 35.547501 34.163818
2019-01-04 36.132500 37.137501 35.950001 37.064999 35.622253
2019-01-07 37.174999 37.207500 36.474998 36.982498 35.542973
2019-01-08 37.389999 37.955002 37.130001 37.687500 36.220531
… … … … … …
2023-09-18 176.479996 179.380005 176.169998 177.970001 177.970001
2023-09-19 177.520004 179.630005 177.130005 179.070007 179.070007
2023-09-20 179.259995 179.699997 175.399994 175.490005 175.490005
2023-09-21 174.550003 176.300003 173.860001 173.929993 173.929993
2023-09-22 174.669998 177.078995 174.054993 174.789993 174.789993

Volume
Date
2019-01-02 148158800
2019-01-03 365248800
2019-01-04 234428400
2019-01-07 219111200
2019-01-08 164101200
… …
2023-09-18 67257600
2023-09-19 51826900
2023-09-20 58436200
2023-09-21 63047900
2023-09-22 55110610

[1190 rows x 6 columns]

3
[16]: #calculate Daily Returns
stock_data['Daily Return'] = stock_data['Close'].pct_change() * 100

[17]: #Calculate log Return

stock_data['Log_Return'] = (stock_data['Close']/ stock_data['Close'].shift(1)).
↪apply(lambda x: None if pd.isnull(x) else (100*(np.log(x))))

[18]: stock_data

[18]: Open High Low Close Adj Close \

Date
2019-01-02 38.722500 39.712502 38.557499 39.480000 37.943249
2019-01-03 35.994999 36.430000 35.500000 35.547501 34.163818
2019-01-04 36.132500 37.137501 35.950001 37.064999 35.622253
2019-01-07 37.174999 37.207500 36.474998 36.982498 35.542973
2019-01-08 37.389999 37.955002 37.130001 37.687500 36.220531
… … … … … …
2023-09-18 176.479996 179.380005 176.169998 177.970001 177.970001
2023-09-19 177.520004 179.630005 177.130005 179.070007 179.070007
2023-09-20 179.259995 179.699997 175.399994 175.490005 175.490005
2023-09-21 174.550003 176.300003 173.860001 173.929993 173.929993
2023-09-22 174.669998 177.078995 174.054993 174.789993 174.789993

Volume Daily Return Log_Return

Date
2019-01-02 148158800 NaN NaN
2019-01-03 365248800 -9.960737 -10.492436
2019-01-04 234428400 4.268930 4.180324
2019-01-07 219111200 -0.222583 -0.222831
2019-01-08 164101200 1.906312 1.888370
… … … …
2023-09-18 67257600 1.691336 1.677192
2023-09-19 51826900 0.618085 0.616183
2023-09-20 58436200 -1.999219 -2.019474
2023-09-21 63047900 -0.888947 -0.892922
2023-09-22 55110610 0.494452 0.493234

[1190 rows x 8 columns]

[19]: stock_data.dropna(subset=['Daily Return', 'Log_Return'], inplace=True)

[20]: stock_data['Cumulative Return'] = (1+ stock_data['Daily Return'] / 100).

↪cumprod() - 1

[21]: print(stock_data.head())

Open High Low Close Adj Close Volume \

4
 Date
2019-01-03 35.994999 36.430000 35.500000 35.547501 34.163818 365248800
2019-01-04 36.132500 37.137501 35.950001 37.064999 35.622253 234428400
2019-01-07 37.174999 37.207500 36.474998 36.982498 35.542973 219111200
2019-01-08 37.389999 37.955002 37.130001 37.687500 36.220531 164101200
2019-01-09 37.822498 38.632500 37.407501 38.327499 36.835617 180396400

Daily Return Log_Return Cumulative Return

Date
2019-01-03 -9.960737 -10.492436 -0.099607
2019-01-04 4.268930 4.180324 -0.061170
2019-01-07 -0.222583 -0.222831 -0.063260
2019-01-08 1.906312 1.888370 -0.045403
2019-01-09 1.698174 1.683916 -0.029192

[22]: plt.figure(figsize=(12, 6))

plt.subplot(2, 1, 1 )
plt.plot(stock_data['Close'], label='Close Price', color='Blue')
plt.title('Stock Prices Analysis')
plt.ylabel('Price USD')
plt.legend()

[22]: <matplotlib.legend.Legend at 0x7f37e9ccf2e0>

[23]: #Return Analysis

plt.figure(figsize=(12, 6))
plt.subplot(2, 1, 2)
plt.plot(stock_data['Daily Return'], label='Daily Return', color='green')
plt.title('Daily Return Analysis')
plt.ylabel('Daily Return')
plt.legend()
plt.tight_layout()

5
[24]: fig ,ax = plt.subplots(figsize=(12, 4))
stock_data['Daily Return'].plot(ax=ax);

2 Time Series Decompositions

[19]: from statsmodels.tsa.seasonal import seasonal_decompose

[20]: stock_prices = stock_data['Adj Close']

[21]: result= seasonal_decompose(stock_prices, model='addictive', period=252)

[22]: plt.figure(figsize=(12, 6))

plt.subplot(411)
plt.plot(stock_prices, label='Original Stock Prices')
plt.legend()

plt.subplot(412)
plt.plot(result.trend, label='Trend Component')
plt.legend()

6
 plt.subplot(413)
plt.plot(result.seasonal, label='Seasonal Component')
plt.legend()

plt.subplot(414)
plt.plot(result.resid, label='Residual component')
plt.legend()

plt.show()
plt.tight_layout()

<Figure size 640x480 with 0 Axes>

3 Volatility Garch
[23]: !pip install arch

Collecting arch
Downloading
arch-6.1.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (916 kB)
�������������������������������������� 916.4/916.4 kB
25.1 MB/s eta 0:00:00
Requirement already satisfied: numpy>=1.19 in
/usr/local/lib/python3.10/dist-packages (from arch) (1.23.5)
Requirement already satisfied: scipy>=1.5 in /usr/local/lib/python3.10/dist-
packages (from arch) (1.11.2)
Requirement already satisfied: pandas>=1.1 in /usr/local/lib/python3.10/dist-

7
 packages (from arch) (1.5.3)
Requirement already satisfied: statsmodels>=0.12 in
/usr/local/lib/python3.10/dist-packages (from arch) (0.14.0)
Requirement already satisfied: python-dateutil>=2.8.1 in
/usr/local/lib/python3.10/dist-packages (from pandas>=1.1->arch) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-
packages (from pandas>=1.1->arch) (2023.3.post1)
Requirement already satisfied: patsy>=0.5.2 in /usr/local/lib/python3.10/dist-
packages (from statsmodels>=0.12->arch) (0.5.3)
Requirement already satisfied: packaging>=21.3 in
/usr/local/lib/python3.10/dist-packages (from statsmodels>=0.12->arch) (23.1)
Requirement already satisfied: six in /usr/local/lib/python3.10/dist-packages
(from patsy>=0.5.2->statsmodels>=0.12->arch) (1.16.0)
Installing collected packages: arch
Successfully installed arch-6.1.0

[24]: from arch import arch_model

[25]: model = arch_model(stock_data['Daily Return'], vol='Garch', p=1, q=1)

results = model.fit()

Iteration: 1, Func. Count: 6, Neg. LLF: 7094.188659020681

Iteration: 2, Func. Count: 16, Neg. LLF: 180592995930.79388
Iteration: 3, Func. Count: 24, Neg. LLF: 2800.9432418003344
Iteration: 4, Func. Count: 31, Neg. LLF: 2784.186029104576
Iteration: 5, Func. Count: 38, Neg. LLF: 2419.5217245842377
Iteration: 6, Func. Count: 44, Neg. LLF: 2419.0228627426386
Iteration: 7, Func. Count: 49, Neg. LLF: 2419.0220910105472
Iteration: 8, Func. Count: 54, Neg. LLF: 2419.0220837354263
Iteration: 9, Func. Count: 58, Neg. LLF: 2419.0220837353813
Optimization terminated successfully (Exit mode 0)
Current function value: 2419.0220837354263
Iterations: 9
Function evaluations: 58
Gradient evaluations: 9

[26]: conditional_volatility = results.conditional_volatility

[27]: plt.figure(figsize=(12, 6))

#Plot Stock Price

plt.subplot(2, 1, 2)
plt.plot(stock_prices, label='Adjusted Close Prices', color='Blue')
plt.title('Historical Stock Prices')
plt.ylabel('Prices')
plt.legend()

8
 #Plot Conditional Volatility
plt.subplot(2, 1, 2)
plt.plot(conditional_volatility, label='Conditional Volatility (GARCH)',␣
↪color='green')

plt.title('Conditional Volatility (GARCH)')

plt.ylabel('Volatility')
plt.legend()

plt.tight_layout()
plt.show()

4 Statistical Descriptions
[28]: mean_return = stock_data['Daily Return'].mean()
median_return = stock_data['Daily Return'].median()
std_deviation = stock_data['Daily Return'].std()
skewness = stock_data['Daily Return'].skew()
kurtosis = stock_data['Daily Return'].kurtosis()

[29]: mean_return

[29]: 0.14673047261990926

[30]: plt.subplot(2, 1, 2)
plt.axhline(mean_return, color='green', linestyle='--', label=f'Mean Return␣
↪({mean_return:.4f})')

plt.axhline(median_return, color='orange', linestyle='--',␣

↪label=f'Median_return ({median_return:.4f})')

plt.axhline(std_deviation, color='red', linestyle='--',␣

↪label=f'Standard_Deviation ({std_deviation:.4f})')

plt.title('Statistical Decsriptions of Daily Returns')

plt.ylabel('Value')
plt.legend()

9
 plt.tight_layout()
plt.show()

5 Correlations / cointegrations
[31]: correlation_matrix = stock_data.corr()

[32]: correlation_matrix

[32]: Open High Low Close Adj Close \

Open 1.000000 0.999633 0.999555 0.999061 0.999003
High 0.999633 1.000000 0.999471 0.999552 0.999477
Low 0.999555 0.999471 1.000000 0.999574 0.999546
Close 0.999061 0.999552 0.999574 1.000000 0.999939
Adj Close 0.999003 0.999477 0.999546 0.999939 1.000000
Volume -0.427227 -0.419438 -0.437098 -0.428742 -0.431130
Daily Return -0.046858 -0.034352 -0.032134 -0.015809 -0.016420
Log_Return -0.044253 -0.031982 -0.029379 -0.013254 -0.013865
Cumulative Return 0.999061 0.999552 0.999574 1.000000 0.999939

Volume Daily Return Log_Return Cumulative Return

Open -0.427227 -0.046858 -0.044253 0.999061
High -0.419438 -0.034352 -0.031982 0.999552
Low -0.437098 -0.032134 -0.029379 0.999574
Close -0.428742 -0.015809 -0.013254 1.000000
Adj Close -0.431130 -0.016420 -0.013865 0.999939
Volume 1.000000 -0.034130 -0.048782 -0.428742
Daily Return -0.034130 1.000000 0.999622 -0.015809
Log_Return -0.048782 0.999622 1.000000 -0.013254

10
 Cumulative Return -0.428742 -0.015809 -0.013254 1.000000

[33]: from statsmodels.tsa.stattools import coint

[35]: price_series_1 = np.random.randn(100)

price_series_2 = 0.5 * price_series_1 + np.random.randn(100)

[36]: cointegration_test = coint(price_series_1, price_series_2)

[37]: cointegration_test

[37]: (-8.10242004427434,
1.6696221220460294e-11,
array([-4.01048603, -3.39854434, -3.08756793]))

[40]: plt.figure(figsize=(12, 6))

plt.plot(price_series_1, label='Price Series 1', linestyle='--', marker='o',␣
↪markersize=3)

plt.plot(price_series_2, label='Prices Series 2', linestyle='--', marker='o',␣

↪markersize=3)

plt.title('Price Series')
plt.xlabel('time')
plt.ylabel('Price')
plt.legend()
plt.grid(True)
plt.show()

11
 6 Econometrics Model
[63]: import statsmodels.api as sm

[64]: Stock_0 = '^GSPC'

start = '2019-01-01'
end = '2023-09-22'

[65]: market_returns = yf.download(Stock_0, start, end)

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

[66]: market_returns

[66]: Open High Low Close Adj Close \

Date
2019-01-02 2476.959961 2519.489990 2467.469971 2510.030029 2510.030029
2019-01-03 2491.919922 2493.139893 2443.959961 2447.889893 2447.889893
2019-01-04 2474.330078 2538.070068 2474.330078 2531.939941 2531.939941
2019-01-07 2535.610107 2566.159912 2524.560059 2549.689941 2549.689941
2019-01-08 2568.110107 2579.820068 2547.560059 2574.409912 2574.409912
… … … … … …
2023-09-15 4497.979980 4497.979980 4447.209961 4450.319824 4450.319824
2023-09-18 4445.129883 4466.359863 4442.109863 4453.529785 4453.529785
2023-09-19 4445.410156 4449.850098 4416.609863 4443.950195 4443.950195
2023-09-20 4452.810059 4461.029785 4401.379883 4402.200195 4402.200195
2023-09-21 4374.359863 4375.700195 4329.169922 4330.000000 4330.000000

Volume
Date
2019-01-02 3733160000
2019-01-03 3858830000
2019-01-04 4234140000
2019-01-07 4133120000
2019-01-08 4120060000
… …
2023-09-15 6932230000
2023-09-18 3161230000
2023-09-19 3614880000
2023-09-20 3308450000
2023-09-21 3662340000

[1189 rows x 6 columns]

[67]: market_data = market_returns['Adj Close'].pct_change().dropna()

Ticker_data = stock_data['Adj Close'].pct_change().dropna()

12
[68]: X = sm.add_constant(market_data)

[69]: Ticker_data = Ticker_data.reindex(X.index)

[70]: model = sm.OLS(Ticker_data, X).fit()

[71]: plt.scatter(market_data, Ticker_data, label='Data')

[71]: <matplotlib.collections.PathCollection at 0x7b527f8a0e20>

[79]: prices = stock_data['Adj Close']

[80]: model = sm.tsa.ARIMA(prices, order=(1, 1, 1)).fit()

/usr/local/lib/python3.10/dist-packages/statsmodels/tsa/base/tsa_model.py:473:
ValueWarning: A date index has been provided, but it has no associated frequency
information and so will be ignored when e.g. forecasting.
self._init_dates(dates, freq)
/usr/local/lib/python3.10/dist-packages/statsmodels/tsa/base/tsa_model.py:473:
ValueWarning: A date index has been provided, but it has no associated frequency
information and so will be ignored when e.g. forecasting.
self._init_dates(dates, freq)

13
 /usr/local/lib/python3.10/dist-packages/statsmodels/tsa/base/tsa_model.py:473:
ValueWarning: A date index has been provided, but it has no associated frequency
information and so will be ignored when e.g. forecasting.
self._init_dates(dates, freq)

[81]: plt.figure(figsize=(12,6))
plt.plot(prices, label='Original Price', color='blue')

[81]: [<matplotlib.lines.Line2D at 0x7b527ce0a530>]

[82]: lags = 5

[84]: plt.plot(model.fittedvalues, label='Fitted Value', color='green')

[84]: [<matplotlib.lines.Line2D at 0x7b527cec2470>]

14
[101]: forecast = model.forecast(steps=5)
plt.plot(range(len(prices), len(prices) + 5), forecast, label='Forecasts',␣
↪color='orange')

plt.xlabel('Time')
plt.ylabel('Stock Price')
plt.title('Model ARIMA')
plt.xticks(rotation = 25)
plt.legend()
plt.show()

/usr/local/lib/python3.10/dist-packages/statsmodels/tsa/base/tsa_model.py:836:
ValueWarning: No supported index is available. Prediction results will be given
with an integer index beginning at `start`.
return get_prediction_index(
/usr/local/lib/python3.10/dist-packages/statsmodels/tsa/base/tsa_model.py:836:
FutureWarning: No supported index is available. In the next version, calling
this method in a model without a supported index will result in an exception.
return get_prediction_index(

15
 7 Strategies Based on Simple Moving Averages
[11]: data = pd.DataFrame(stock_data)

[14]: data

[14]: Open High Low Close Adj Close \

Date
2019-01-02 38.722500 39.712502 38.557499 39.480000 37.943260
2019-01-03 35.994999 36.430000 35.500000 35.547501 34.163837
2019-01-04 36.132500 37.137501 35.950001 37.064999 35.622250
2019-01-07 37.174999 37.207500 36.474998 36.982498 35.542976
2019-01-08 37.389999 37.955002 37.130001 37.687500 36.220524
… … … … … …
2023-09-18 176.479996 179.380005 176.169998 177.970001 177.970001
2023-09-19 177.520004 179.630005 177.130005 179.070007 179.070007
2023-09-20 179.259995 179.699997 175.399994 175.490005 175.490005
2023-09-21 174.550003 176.300003 173.860001 173.929993 173.929993

16
 2023-09-22 174.669998 177.078995 174.054993 174.789993 174.789993

Volume
Date
2019-01-02 148158800
2019-01-03 365248800
2019-01-04 234428400
2019-01-07 219111200
2019-01-08 164101200
… …
2023-09-18 67257600
2023-09-19 51826900
2023-09-20 58436200
2023-09-21 63047900
2023-09-22 56682928

[1190 rows x 6 columns]

[15]: stock_data['SMA1'] = stock_data['Adj Close'].rolling(42).mean()

stock_data['SMA2'] = stock_data['Adj Close'].rolling(252).mean()

[16]: stock_data.tail()

[16]: Open High Low Close Adj Close \

Date
2023-09-18 176.479996 179.380005 176.169998 177.970001 177.970001
2023-09-19 177.520004 179.630005 177.130005 179.070007 179.070007
2023-09-20 179.259995 179.699997 175.399994 175.490005 175.490005
2023-09-21 174.550003 176.300003 173.860001 173.929993 173.929993
2023-09-22 174.669998 177.078995 174.054993 174.789993 174.789993

Volume SMA1 SMA2

Date
2023-09-18 67257600 183.063979 160.650982
2023-09-19 51826900 182.735418 160.767088
2023-09-20 58436200 182.349915 160.854076
2023-09-21 63047900 181.908008 160.925327
2023-09-22 56682928 181.465892 161.012536

[19]: stock_data['SMA1'].plot(title='APPL | 42 days SMA2', figsize=(10,6));

stock_data['SMA2'].plot (title='APPL | 252 days SMA2', figsize=(10,6))

[19]: <Axes: title={'center': 'APPL | 252 days SMA2'}, xlabel='Date'>

17
[20]: stock_data['Position'] = np.where(stock_data['SMA1']> stock_data['SMA2'], 1, -1)

[21]: stock_data.dropna(inplace=True)

[23]: stock_data['Position'].plot(ylim=[-1.1, 1.1], title='Market', figsize=(10,6));

18
[25]: stock_data['Return'] = np.log(stock_data['Adj Close'] / stock_data['Adj Close'].
↪shift(1))

[26]: stock_data['Return'].hist(bins=35, figsize=(10, 6));

[27]: stock_data['Strategy'] = stock_data['Position'].shift(1) * stock_data['Return']

[29]: stock_data[['Strategy', 'Return']].sum()

[29]: Strategy 0.418708

Return 0.892280
dtype: float64

[31]: stock_data[['Strategy', 'Return']].sum().apply(np.exp)

[31]: Strategy 1.519997

Return 2.440689
dtype: float64

[32]: stock_data[['Strategy', 'Return']].cumsum().apply(np.exp).plot(figsize=(10, 6));

19
[33]: stock_data[['Strategy', 'Return']].mean() * 252

[33]: Strategy 0.112489

Return 0.239717
dtype: float64

[34]: np.exp(stock_data[['Strategy', 'Return']].mean() * 252) - 1

[34]: Strategy 0.11906

Return 0.27089
dtype: float64

[35]: stock_data[['Strategy','Return']].std() * 252 ** 0.5

[35]: Strategy 0.345133

Return 0.344875
dtype: float64

[36]: (stock_data[['Strategy','Return']].apply(np.exp) - 1).std() * 252 ** 0.5

[36]: Strategy 0.344744

Return 0.345029
dtype: float64

[37]: stock_data['cumret'] = stock_data['Strategy'].cumsum().apply(np.exp)

20
[38]: stock_data['cummax'] = stock_data['cumret'].cummax()

[39]: stock_data[['cumret', 'cummax']].dropna().plot(figsize=(10, 6));

[40]: drawdown = stock_data['cummax'] - stock_data['cumret']

[41]: drawdown.max()

[41]: 1.1597999478192935

8 Predictions Stock
[47]: X= stock_data['Close']

[49]: y = X + np.random.standard_normal(len(X))

[50]: reg = np.polyfit(X, y, deg=1)

[51]: reg

[51]: array([ 1.00090991, -0.1240214 ])

[52]: plt.figure(figsize=(10, 6))

plt.plot(X, y, 'bo', label='data')

21
 plt.plot(X, np.polyval(reg, X), 'r', lw=2.5, label='Linear Regression')
plt.legend(loc=0)

[52]: <matplotlib.legend.Legend at 0x7b528c7d7dc0>

[53]: lags = 5

[54]: cols = []
for lag in range(1, lags + 1):
col =f'lag_{lag}'
stock_data[col] = stock_data['Adj Close'].shift(lag)
cols.append(col)
stock_data.dropna(inplace=True)

[55]: reg = np.linalg.lstsq(stock_data[cols], stock_data['Adj Close'], rcond=None)[0]

[56]: reg

[56]: array([0.95667901, 0.01357397, 0.00697459, 0.00724528, 0.01613076])

[57]: stock_data['Predictions'] = np.dot(stock_data[cols], reg)

[58]: stock_data[['Predictions', 'Adj Close']].plot(figsize=(10, 6));

22
[59]: stock_data[['Predictions', 'Adj Close']].loc['2023-1-1':].plot(figsize=(10, 6));

23