CAPM_Returns.py

# Importing libraries

import streamlit as st
import pandas as pd
import yfinance as yf
import pandas_datareader.data as web
import datetime
import capm_functions

st.set_page_config(page_title = "CAPM",
    page_icon = "chart_with_upwards_trend",
    layout = 'wide') 

st.title("Capital Asset Pricing Model")

# Getting input from user

col1, col2 = st.columns([1,1])
with col1:
    stock_list = st.multiselect("Choose 4 stocks", ('TSLA', 'AAPL', 'NFLX', 'MSFT', 'MGM', 'AMZN', 'NVDA', 'GOOGL'), ['TSLA', 'AAPL', 'AMZN', 'GOOGL'])
with col2:
    year = st.number_input("Number of years",1,10)

# Downloading data for SP500
try:
    end = datetime.date.today()
    start = datetime.date(datetime.date.today().year-year, datetime.date.today().month, datetime.date.today().day)
    SP500 = web.DataReader(['sp500'], 'fred', start, end)
    print(SP500.tail())

    stocks_df = pd.DataFrame()

    for stock in stock_list:
        data = yf.download(stock, period = f'{year}y')
        stocks_df[f'{stock}'] =  data['Close']

    stocks_df.reset_index(inplace = True)
    SP500.reset_index(inplace = True)
    SP500.columns = ['Date', 'sp500']
    stocks_df['Date'] = stocks_df['Date'].astype('datetime64[ns]')
    stocks_df['Date'] = stocks_df['Date'].apply(lambda x:str(x)[:10])
    stocks_df['Date'] = pd.to_datetime(stocks_df['Date'])
    stocks_df = pd.merge(stocks_df, SP500, on = 'Date', how = 'inner')
    print(stocks_df) 

    col1, col2 = st.columns([1,1])
    with col1:
        st.markdown("### Dataframe head")
        st.dataframe(stocks_df.head(), use_container_width = True)
    with col2:
        st.markdown("### Dataframe tail")
        st.dataframe(stocks_df.tail(), use_container_width = True)

    col1, col2 = st.columns([1,1])
    with col1:
        st.markdown("### Price of all the Stocks")
        st.plotly_chart(capm_functions.interactive_plot(stocks_df))
    with col2:
        st.markdown("### Price of all the Stocks (After Normalizing)")
        st.plotly_chart(capm_functions.interactive_plot(capm_functions.normalize(stocks_df)))

    stocks_daily_return = capm_functions.daily_return(stocks_df)
    print(stocks_daily_return.head())

    beta  = {}
    alpha = {}

    for i in stocks_daily_return.columns:
        if i !='Date' and i != 'sp500':
            b, a = capm_functions.calculate_beta(stocks_daily_return, i)

            beta[i] = b
            alpha[i] = a
    print(beta, alpha)

    beta_df = pd.DataFrame(columns = ['Stocks', 'Beta Value'])
    beta_df['Stock'] = beta.keys()
    beta_df['Beta Value'] = [str(round(i,2)) for i in beta.values()]

    with col1:
        st.markdown('### Calculated Beta Value')
        st.dataframe(beta_df, use_container_width = True)

    rf = 0
    rm = stocks_daily_return['sp500'].mean()*252

    return_df = pd.DataFrame()
    return_value = []
    for stock, value in beta.items():
        return_value.append(str(round(rf+(value*(rf-rm)),2)))
    return_df['Stock'] = stock_list

    return_df['Return Value'] = return_value

    with col2:
        st.markdown('### Calculated Return using CAPM')
        st.dataframe(return_df, use_container_width = True) 

except:
    st.write("Please select valid inputs")