!pip install statsmodels

!pip install seaborn matplotlib

!pip install pandas

!pip install yfinance==0.2.28

!pip install numpy

!pip install PyPortfolioOpt

import pandas as pd import yfinance as yf import numpy as np from scipy.optimize import minimize import scipy.stats as stats # Load Fama French factors data with error handling try: fama_french_data = pd.read_csv('fama_french_factors.csv', parse_dates=True, index_col=0) / 100 # Convert percentages to decimals except pd.errors.ParserError as e: print(f"Error reading CSV file: {e}") fama_french_data = pd.DataFrame() # Create an empty DataFrame as a placeholder # Manually set the risk-free rate (you can replace this with actual data) risk_free_rate = 0.02 # Replace with the appropriate risk-free rate (in decimal form) # Define the stock symbols and dates stock_symbols = ['XOM', 'FSLR', 'JNJ', 'AMZN'] start_date = '2018-09-06' end_date = '2023-09-06' # Download historical stock price data data = yf.download(stock_symbols, start=start_date, end=end_date) # Calculate daily returns returns = data['Adj Close'].pct_change() returns.dropna(inplace=True) df=pd.DataFrame(returns) # Calculate daily returns returns = data['Adj Close'].pct_change() returns.dropna(inplace=True) # Define the objective function for Sharpe ratio maximization def sharpe_ratio(weights, returns, risk_free_rate): portfolio_returns = np.dot(returns, weights) mean_return = np.mean(portfolio_returns) portfolio_stddev = np.std(portfolio_returns) sharpe = (mean_return - risk_free_rate) / portfolio_stddev return -sharpe # Minimize the negative Sharpe ratio # Set initial guess for portfolio weights initial_weights = np.array([0.25, 0.25, 0.25, 0.25]) # Set optimization constraints (sum of weights equals 1) constraints = ({'type': 'eq', 'fun': lambda weights: np.sum(weights) - 1}) # Set optimization bounds (each weight is between 0 and 1) bounds = tuple((0, 1) for asset in range(len(stock_symbols))) # Run the optimization to find the maximum Sharpe portfolio optimized_weights = minimize(sharpe_ratio, initial_weights, args=(returns, risk_free_rate), method='SLSQP', bounds=bounds, constraints=constraints) optimized_weights = optimized_weights.x # Calculate the daily returns of the maximum Sharpe portfolio portfolio_returns_max_sharpe = np.dot(returns, optimized_weights) # Create a DataFrame with the portfolio returns portfolio_returns_df = pd.DataFrame({'Portfolio Returns': portfolio_returns_max_sharpe}, index=returns.index) # Merge the portfolio returns with the Fama French factors (if available) if not fama_french_data.empty: merged_data = pd.concat([portfolio_returns_df, fama_french_data], axis=1, join='inner') else: merged_data = portfolio_returns_df # You can further analyze and work with the merged data as needed # Calculate statistics like mean, standard deviation, skewness, kurtosis, etc. mean_return = np.mean(merged_data['Portfolio Returns']) volatility = np.std(merged_data['Portfolio Returns']) skewness = stats.skew(merged_data['Portfolio Returns']) kurtosis = stats.kurtosis(merged_data['Portfolio Returns']) # Print the portfolio weights, statistics, and the merged data print("Optimal Portfolio Weights:", optimized_weights) print("Mean Return:", mean_return) print("Volatility (Standard Deviation):", volatility) print("Skewness:", skewness) print("Kurtosis:", kurtosis) print(merged_data.head())

import seaborn as sns import matplotlib.pyplot as plt # Calculate the correlation matrix between portfolio and factor returns correlation_matrix = merged_data.corr() # Plot the correlation matrix as a heatmap plt.figure(figsize=(10, 8)) sns.heatmap(correlation_matrix, annot=True, cmap='coolwarm', fmt=".2f") plt.title("Correlation Between Portfolio and Factor Returns") plt.show()

import statsmodels.api as sm for factor in fama_french_data.columns: X = sm.add_constant(merged_data[factor]) Y = merged_data['Portfolio Returns'] model = sm.OLS(Y, X).fit() beta = model.params[factor] alpha = model.params['const'] print(f"Factor: {factor}") print(f"Beta: {beta}") print(f"Alpha: {alpha}") print(model.summary())

# Define the independent variables (multiple factors) X = merged_data[fama_french_data.columns] X = sm.add_constant(X) # Add the intercept term Y = merged_data['Portfolio Returns'] # Fit the multi-factor regression model model = sm.OLS(Y, X).fit() # Obtain the regression results factor_betas = model.params[fama_french_data.columns] # Factor betas alpha = model.params['const'] # Intercept (alpha) print("Multi-Factor Regression Results:") print("Factor Betas:") print(factor_betas) print("Alpha (Intercept):", alpha) print(model.summary())