Beta estimation

# Initialize empty dictionaries to save results
beta,alpha = dict(), dict()

# Make a nice subplot
fig, axes = plt.subplots(1,3, dpi=150, figsize=(15,8))
axes = axes.flatten()

# Loop on every stock daily return
for idx, stock in enumerate(stocks_daily_return.columns):

# Ignoring the date and S&P500 Columns
    if stock != "Date" and stock != "SP500":

        # scatter plot between stock and the S&P500
        stocks_daily_return.plot(kind = "scatter", x = "SP500", y = stock, ax=axes[idx-1])

        # Fit a line (regression using polyfit of degree 1)
        b_, a_ = np.polyfit(stocks_daily_return[stock], stocks_daily_return["SP500"], 1)

        regression_line = b_ * stocks_daily_return["SP500"] + a_
        axes[idx-1].plot(stocks_daily_return["SP500"], regression_line, "-", color = "r")

        # save the regression coeeficient for the current stock
        beta[stock] = b_
        alpha[stock] = a_

plt.suptitle("Beta estimation: regression between SP500 and individual stock daily performance", size=20)