!pip install yfinance

import yfinance as yf import pandas as pd import numpy as np # Define ETF symbols #Luxury portfolio #Amundi Index Solutions - Amundi S&P Global Luxury luxury_etf_symbol1 = 'GLUX.DE' # Global X Millennials Consumer ETF luxury_etf_symbol2= 'MILN' #Retail portfolio #VanEck Retail ETF retail_etf_symbol1 = 'RTH' #SPDR S&P Retail ETF retail_etf_symbol2 = 'XRT' # Fetch ETF data from Yahoo Finance luxury_etf1 = yf.Ticker(luxury_etf_symbol1) luxury_etf2 = yf.Ticker(luxury_etf_symbol2) retail_etf1 = yf.Ticker(retail_etf_symbol1) retail_etf2 = yf.Ticker(retail_etf_symbol2) # Download historical data start_date = '2018-11-01' end_date = '2023-11-01' luxury_data1 = luxury_etf1.history(start=start_date, end=end_date) luxury_data2 = luxury_etf2.history(start=start_date, end=end_date) retail_data1 = retail_etf1.history(start=start_date, end=end_date) retail_data2 = retail_etf2.history(start=start_date, end=end_date) # Inspect the available columns print(luxury_data1.columns) print(retail_data1.columns) #Cleaning data luxury_data1.index = luxury_data1.index.date luxury_data1 = luxury_data1.drop(columns=['High', 'Low', 'Stock Splits', 'Capital Gains', 'Dividends']) print(luxury_data1) luxury_data2.index = luxury_data2.index.date luxury_data2 = luxury_data2.drop(columns=['High', 'Low', 'Stock Splits', 'Capital Gains', 'Dividends']) print(luxury_data2) retail_data1.index = retail_data1.index.date retail_data1 = retail_data1.drop(columns=['High', 'Low', 'Stock Splits', 'Capital Gains', 'Dividends']) print(retail_data1) retail_data2.index = retail_data2.index.date retail_data2 = retail_data2.drop(columns=['High', 'Low', 'Stock Splits', 'Capital Gains', 'Dividends']) print(retail_data2)

# Extract adjusted closing prices (you can choose other columns if needed) luxury_prices1 = luxury_data1['Close'] luxury_prices2 = luxury_data2['Close'] retail_prices1 = retail_data1['Close'] retail_prices2 = retail_data2['Close'] # Create a DataFrame with ETF prices etf_data = pd.DataFrame({ luxury_etf_symbol1: luxury_prices1, luxury_etf_symbol2: luxury_prices2, retail_etf_symbol1: retail_prices1, retail_etf_symbol2: retail_prices2, }) # Calculate daily returns etf_returns = etf_data.pct_change().dropna() print(etf_returns) # Print the mean print("mean : ", etf_returns.mean()) print("*"*40) # Print the standard deviation print("Std. dev : ", etf_returns.std()) print("*"*40) # Print the skewness print("skew : ", etf_returns.skew()) print("*"*40) # Print the kurtosis print("kurt : ", etf_returns.kurtosis()) print("*"*40)

# Equal weights for the portfolios luxury_weight = 0.5 retail_weight = 0.5 # Calculate the portfolio returns portfolio_returns_luxury =(etf_returns[luxury_etf_symbol1] * luxury_weight + etf_returns[luxury_etf_symbol2] * luxury_weight) portfolio_returns_retail=(etf_returns[retail_etf_symbol1] * retail_weight + etf_returns[retail_etf_symbol2] * retail_weight) print("Daily Luxury Portfolio return: ", portfolio_returns_luxury) print("Daily Retail Portfolio return: ", portfolio_returns_retail) # Calculate cumulative returns cumulative_returns_luxury = (1 + portfolio_returns_luxury).cumprod() cumulative_returns_retail = (1 + portfolio_returns_retail).cumprod() print("Cumulative Luxury Portfolio return: ", cumulative_returns_luxury) print("Cumulative Retail Portfolio return: ", cumulative_returns_retail)

print(portfolio_returns_luxury) print(portfolio_returns_retail)

# Plot the cumulative returns import matplotlib.pyplot as plt cumulative_returns_luxury.plot(label='Luxury Portfolio', color='blue') cumulative_returns_retail.plot(label='Retail Portfolio', color='orange') plt.title('Portfolio Cumulative Returns') plt.legend() plt.xlabel('Date') plt.ylabel('Cumulative Returns') plt.show()

import numpy as np # Risk-free rate (10-year Treasury yield) is 4.566% as of 8th Nov 2023 risk_free_rate = 0.04566 # Calculate the annualized return of the portfolio cumulative_returns_luxury = portfolio_returns_luxury[-1] n1 = len(portfolio_returns_luxury) # Number of trading days in the data annualized_return_luxury = (cumulative_returns_luxury ** (1/n1) - 1) # Calculate the annualized return of the portfolio cumulative_returns_retail = portfolio_returns_retail[-1] n2 = len(portfolio_returns_retail) # Number of trading days in the data annualized_return_retail = (cumulative_returns_retail ** (1/n2) - 1) # Calculate the annualized volatility (standard deviation) of portfolio returns portfolio_returns_luxury = portfolio_returns_luxury.dropna() portfolio_returns_retail = portfolio_returns_retail.dropna() annualized_volatility_luxury = portfolio_returns_luxury.std() * np.sqrt(252) # Assuming 252 trading days in a year annualized_volatility_retail = portfolio_returns_retail.std() * np.sqrt(252) # Assuming 252 trading days in a year # Calculate the excess return excess_return_luxury = annualized_return_luxury - risk_free_rate excess_return_retail = annualized_return_retail - risk_free_rate # Calculate the Sharpe ratio sharpe_ratio_luxury = excess_return_luxury / annualized_volatility_luxury sharpe_ratio_retail = excess_return_retail / annualized_volatility_retail print(f"Portfolio Luxury Annualized Return: {annualized_return_luxury:.2%}") print(f"Portfolio Annualized Luxury Volatility: {annualized_volatility_luxury:.2%}") print(f"Sharpe Ratio Luxury: {sharpe_ratio_luxury:.2f}") print(f"Portfolio Retail Annualized Return: {annualized_return_retail:.2%}") print(f"Portfolio Annualized Retail Volatility: {annualized_volatility_retail:.2%}") print(f"Sharpe Ratio Retail: {sharpe_ratio_retail:.2f}")

!pip install pyfolio import pyfolio as pf

#input the global inflation rate--using the headline consumer price index !pip install --upgrade pip import pandas as pd !pip install openpyxl inflation_rate=pd.read_excel('Inflation-Rate.xlsx')

# Calculate the world's monthly inflation rate by summing up all country's inflation rates inflation_rate.dropna() selected_columns = inflation_rate.iloc[:, 591:-4] # Calculate the average for each time period # Replace non-numeric values (e.g., string values) with NaN selected_columns = selected_columns.apply(pd.to_numeric, errors='coerce') selected_columns=pd.DataFrame(selected_columns) average_inflation = selected_columns.mean(axis=0) # Assuming average_inflation is your Series average_inflation_df = average_inflation.reset_index() # Rename the columns for better clarity average_inflation_df.columns = ['Datetime', 'Average Inflation Rate'] average_inflation_df['Datetime'] = pd.to_datetime(average_inflation_df['Datetime'], format='%Y%m') average_inflation_df.index=average_inflation_df['Datetime'] # Display the updated DataFrame print(average_inflation_df)

#Downloading unemployment rate unemployment_data=pd.read_excel('unemployment.xlsx') # Reading the Excel file into a pandas DataFrame unemployment_data = pd.read_excel('unemployment.xlsx', index_col=0) # Assuming 'Country' is in the first column # Droping rows with NaN values unemployment_data = unemployment_data.dropna() # Converting the column names to datetime objects unemployment_data.columns = pd.to_datetime(unemployment_data.columns, errors='coerce') # Extracting columns from May-2011 to Sep-2023, including May-2011 and excluding the last column start_month = pd.to_datetime('2018-11-01') end_month = pd.to_datetime('2023-09-01') # Filtering columns based on date range selected_columns2 = unemployment_data.columns[(unemployment_data.columns >= start_month) & (unemployment_data.columns <= end_month)] # Selecting columns within the specified date range, including May-2011 to Sep-2023 selected_data = unemployment_data[selected_columns2] selected_data=selected_data.apply(pd.to_numeric, errors='coerce') # Calculating the average for each month average_unemployment_per_month = selected_data.mean() # Creating a new DataFrame with the average unemployment rate for each month average_unemployment_df = pd.DataFrame(average_unemployment_per_month) average_unemployment_df.columns = ['Average Unemployment Rate'] # Assuming your DataFrame is named 'average_unemployment_df' average_unemployment_df['Datetime'] = average_unemployment_df.index # Displaying the new DataFrame print(average_unemployment_df)

#Import csv file with monthly long-term interest rate data interest_rate_data = pd.read_csv('global interest rate.csv') df = interest_rate_data #Data cleaning process #Drop unnecessary columns df = df.drop(columns=['INDICATOR', 'SUBJECT', 'MEASURE', 'FREQUENCY', 'Flag Codes']) #Convert column 'TIME' to datetime df['TIME'] = pd.to_datetime(df['TIME'], format='%Y-%m') #Set 'Date' as index for both dataframes df.set_index('TIME', inplace=True) #Create pivot table df = df.rename(columns={'LOCATION': 'Country'}) df_pivoted = df.pivot_table(index=df.index, columns='Country', values='Value') #Dropping countries with incomplete data df_cleaned = df_pivoted.dropna(axis=1) #Dropping individual EA19 data sets and only using relevant countries df_final = df_cleaned.drop(columns=['BRA', 'BEL', 'DEU', 'ESP', 'FRA', 'HUN', 'ISR', 'PRT', 'ZAF']) # Assuming df_final is your DataFrame # Convert the index to datetime if it's not already df_final.index = pd.to_datetime(df_final.index) # Specify the date range to keep start_date = '2000-02-01' end_date = '2018-10-01' # Use boolean indexing to keep rows outside the specified date range df_final = df_final[(df_final.index < start_date) | (df_final.index > end_date)] #Add column of annual average rate df_final['Average'] = df_final.mean(axis=1) print("Cleaned data of interest rates globally:") print(df_final) #Visualise import matplotlib.pyplot as plt for country in df_final.columns: plt.plot(df_final.index, df_final[country]) plt.plot(df_final.index, df_final['Average'], label='Average') plt.xlabel('Time') plt.ylabel('Monthly long-term interest rates (%)') plt.title('Global monthly long-term interest rates') plt.show()

#Import csv file with monthly long-term interest rate data cci_data = pd.read_csv('CCI.csv') df_cci = cci_data #Data cleaning process #Drop unnecessary columns df_cci = df_cci.drop(columns=['INDICATOR', 'SUBJECT', 'MEASURE', 'FREQUENCY', 'Flag Codes']) #Convert column 'TIME' to datetime df_cci['TIME'] = pd.to_datetime(df_cci['TIME'], format='%Y-%m') #Set 'Date' as index for both dataframes df_cci.set_index('TIME', inplace=True) #Create pivot table df_cci = df.rename(columns={'LOCATION': 'Country'}) df_cci = df_cci.pivot_table(index=df.index, columns='Country', values='Value') #Dropping countries with incomplete data df_cci = df_cci.dropna(axis=1) #Dropping individual EA19 data sets and only using relevant countries df_cci = df_cci.drop(columns=['BRA', 'BEL', 'DEU', 'ESP', 'FRA', 'HUN', 'ISR', 'PRT', 'ZAF']) # Convert the index to datetime df_cci.index = pd.to_datetime(df_cci.index) # Specify the date range to keep start_date = '2000-02-01' end_date = '2018-10-01' # Use boolean indexing to keep rows outside the specified date range df_cci = df_cci[(df_cci.index < start_date) | (df_cci.index > end_date)] #Add column of annual average rate df_cci['Average cci'] = df_cci.mean(axis=1) print("Cleaned data of CCI globally:") print(df_cci) #Visualise import matplotlib.pyplot as plt for country in df_cci.columns: plt.plot(df_cci.index, df_cci[country]) plt.plot(df_cci.index, df_cci['Average cci'], label='Average cci') plt.xlabel('Time') plt.ylabel('Monthly CCI') plt.title('Global monthly CCI') plt.show()

#Import csv file with monthly CSI data csi_data = pd.read_csv('CSI.csv') df_csi = csi_data #Data cleaning process #Drop unnecessary columns df_csi = df_csi.drop(columns=['INDICATOR', 'SUBJECT', 'MEASURE', 'FREQUENCY', 'Flag Codes']) #Convert column 'TIME' to datetime df_csi['TIME'] = pd.to_datetime(df_csi['TIME'], format='%Y-%m') #Set 'Date' as index for both dataframes df_csi.set_index('TIME', inplace=True) #Create pivot table df_csi = df.rename(columns={'LOCATION': 'Country'}) df_csi = df_csi.pivot_table(index=df.index, columns='Country', values='Value') #Dropping countries with incomplete data df_csi = df_csi.dropna(axis=1) #Dropping individual EA19 data sets and only using relevant countries df_csi = df_csi.drop(columns=['BRA', 'BEL', 'DEU', 'ESP', 'FRA', 'HUN', 'ISR', 'PRT', 'ZAF']) # Convert the index to datetime df_csi.index = pd.to_datetime(df_csi.index) # Specify the date range to keep start_date = '2000-02-01' end_date = '2018-10-01' # Use boolean indexing to keep rows outside the specified date range df_csi = df_csi[(df_csi.index < start_date) | (df_csi.index > end_date)] #Add column of monthly average rate df_csi['Average csi'] = df_csi.mean(axis=1) print("Cleaned data of CSI globally:") print(df_csi) #Visualise import matplotlib.pyplot as plt for country in df_csi.columns: plt.plot(df_csi.index, df_csi[country]) plt.plot(df_csi.index, df_csi['Average csi'], label='Average csi') plt.xlabel('Time') plt.ylabel('Monthly CSI') plt.title('Global monthly CSI') plt.show()

us_consumption_data=pd.read_excel('US_consumption.xlsx', index_col=0) # Droping rows with NaN values us_consumption_data = us_consumption_data.dropna() # Converting the column names to datetime objects us_consumption_data.columns = pd.to_datetime(us_consumption_data.columns, errors='coerce') us_consumption_data.index=pd.to_datetime(us_consumption_data.index) us_consumption_data['us consumption']=us_consumption_data['NaT'] print(us_consumption_data)

!pip install statsmodels import statsmodels.api as sm

portfolio_returns_luxury=pd.DataFrame(portfolio_returns_luxury) portfolio_returns_luxury.columns = ['portfolio returns luxury'] # Print the updated DataFrame print(portfolio_returns_luxury) # Print the updated DataFrame portfolio_returns_luxury=pd.DataFrame(portfolio_returns_luxury) portfolio_returns_luxury.columns = ['portfolio returns luxury'] # Print the updated DataFrame print(portfolio_returns_luxury) portfolio_returns_retail=pd.DataFrame(portfolio_returns_retail) portfolio_returns_retail.columns = ['portfolio returns retail'] # Print the updated DataFrame print(portfolio_returns_retail)

#get the average monthly return of the portfolio portfolio_returns_luxury.index = pd.to_datetime(portfolio_returns_luxury.index) portfolio_returns_luxury=pd.DataFrame(portfolio_returns_luxury) # Change the datetime index to the first day of each month portfolio_returns_luxury.index = portfolio_returns_luxury.index.to_period('M').to_timestamp('M') overall_luxury_average_monthly_return = portfolio_returns_luxury.resample('M').mean() print(overall_luxury_average_monthly_return) #get the average monthly return of the portfolio portfolio_returns_retail.index = pd.to_datetime(portfolio_returns_retail.index) portfolio_returns_retail=pd.DataFrame(portfolio_returns_retail) # Change the datetime index to the first day of each month portfolio_returns_retail.index = portfolio_returns_retail.index.to_period('M').to_timestamp('M') overall_retail_average_monthly_return = portfolio_returns_retail.resample('M').mean() print(overall_retail_average_monthly_return)

# Assuming your DataFrame is named 'overall_luxury_average_monthly_return' overall_luxury_average_monthly_return.index = pd.to_datetime(overall_luxury_average_monthly_return.index) # Create new columns 'Year' and 'Month' with the year and month extracted from the index overall_luxury_average_monthly_return['Year'] = overall_luxury_average_monthly_return.index.year overall_luxury_average_monthly_return['Month'] = overall_luxury_average_monthly_return.index.month # Merge 'Year' and 'Month' into a single column 'YearMonth' overall_luxury_average_monthly_return['YearMonth'] = overall_luxury_average_monthly_return['Year'].astype(str) + '-' + overall_luxury_average_monthly_return['Month'].astype(str) # Drop 'Year' and 'Month' columns if you don't need them anymore overall_luxury_average_monthly_return = overall_luxury_average_monthly_return.drop(['Year', 'Month'], axis=1) overall_luxury_average_monthly_return['Datetime'] = pd.to_datetime(overall_luxury_average_monthly_return['YearMonth'], format='%Y-%m') # Print the updated DataFrame overall_luxury_average_monthly_return.index=overall_luxury_average_monthly_return['Datetime'] print(overall_luxury_average_monthly_return)

# Assuming your DataFrame is named 'overall_retail_average_monthly_return' overall_retail_average_monthly_return.index = pd.to_datetime(overall_retail_average_monthly_return.index) # Create new columns 'Year' and 'Month' with the year and month extracted from the index overall_retail_average_monthly_return['Year'] = overall_retail_average_monthly_return.index.year overall_retail_average_monthly_return['Month'] = overall_retail_average_monthly_return.index.month # Merge 'Year' and 'Month' into a single column 'YearMonth' overall_retail_average_monthly_return['YearMonth'] = overall_retail_average_monthly_return['Year'].astype(str) + '-' + overall_retail_average_monthly_return['Month'].astype(str) # Drop 'Year' and 'Month' columns if you don't need them anymore overall_retail_average_monthly_return = overall_retail_average_monthly_return.drop(['Year', 'Month'], axis=1) overall_retail_average_monthly_return['Datetime'] = pd.to_datetime(overall_retail_average_monthly_return['YearMonth'], format='%Y-%m') # Print the updated DataFrame overall_retail_average_monthly_return.index=overall_retail_average_monthly_return['Datetime'] print(overall_retail_average_monthly_return)

data=pd.concat([average_inflation_df, average_unemployment_df,overall_luxury_average_monthly_return, overall_retail_average_monthly_return, df_final,df_cci,df_csi,us_consumption_data], axis=1) columns_to_delete = ['Datetime', 'YearMonth', 'AUS', 'CAN','CHE', 'EA19', 'GBR', 'JPN', 'NOR','NZL','SWE','USA'] data = data.drop(columns_to_delete, axis=1) data['log Inflation Rate']=np.log(data['Average Inflation Rate']) data['log cci']=np.log(data['Average cci']) data['log csi']=np.log(data['Average csi']) data.dropna(inplace=True) print(data)

# Define your independent variable (X) and dependent variable (y) X1 = sm.add_constant(data[['Average', 'Average Unemployment Rate','log Inflation Rate', 'Average csi','us consumption']]) y = data['portfolio returns luxury'] # Fit the regression model model = sm.OLS(y, X1).fit() # Print the regression results print(model.summary())

# Define your independent variable (X) and dependent variable (y) del average inflation rate X1 = sm.add_constant(data[['Average Unemployment Rate','Average','us consumption']]) y = data['portfolio returns luxury'] # Fit the regression model model = sm.OLS(y, X1).fit() # Print the regression results print(model.summary())

# Define your independent variable (X) and dependent variable (y) X2 = sm.add_constant(data[['Average','Average Unemployment Rate','us consumption']]) y = data['portfolio returns retail'] # Fit the regression model model = sm.OLS(y, X2).fit() # Print the regression results print(model.summary())

# Define your independent variable (X) and dependent variable (y) del average inflation rate X2 = sm.add_constant(data[['Average Unemployment Rate', 'Average','us consumption']]) y = data['portfolio returns retail'] # Fit the regression model model = sm.OLS(y, X2).fit() # Print the regression results print(model.summary())