Herron Topic 3 - Practice Blank

FINA 6333 for Spring 2024

Author

Richard Herron

1 Announcements

2 10-Minute Recap

3 Practice

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import pandas_datareader as pdr
import seaborn as sns
import statsmodels.formula.api as smf
import yfinance as yf
%precision 4
pd.options.display.float_format = '{:.4f}'.format
%config InlineBackend.figure_format = 'retina'

3.1 Plot the security market line (SML) for a variety of asset classes

Use the past three years of daily data for the following exhange traded funds (ETFs):

  1. SPY (SPDR—Standard and Poor’s Depository Receipts—ETF for the S&P 500 index)
  2. BIL (SPDR ETF for 1-3 month Treasury bills)
  3. GLD (SPDR ETF for gold)
  4. JNK (SPDR ETF for high-yield debt)
  5. MDY (SPDR ETF for S&P 400 mid-cap index)
  6. SLY (SPDR ETF for S&P 600 small-cap index)
  7. SPBO (SPDR ETF for corporate bonds)
  8. SPMB (SPDR ETF for mortgage-backed securities)
  9. SPTL (SPDR ETF for long-term Treasury bonds)

3.2 Plot the SML for the Dow Jones Industrial Average (DJIA) stocks

Use the past three years of daily returns data for the stocks listed on the DJIA Wikipedia page. Compare the DJIA SML to the asset class SML above.

3.3 Plot the SML for the five portfolios formed on beta

Download data for portfolios formed on \(\beta\) (Portfolios_Formed_on_BETA) from Ken French. For the value-weighted portfolios, plot realized returns versus \(\beta\). These data should elements [2] and [6], respectively.

3.4 Estimate the CAPM \(\beta\)s on several levered and inverse exchange traded funds (ETFs)

Try the following ETFs:

  1. SPY
  2. UPRO
  3. SPXU

Can you determine what these products do from the data alone? Estimate \(\beta\)s and plot cumulative returns. You may want to pick short periods of time with large market swings.

3.5 Explore the size factor

3.5.1 Estimate \(\alpha\)s for the ten portfolios formed on size

Academics started researching size-based portfolios in the early 1980s, so you may want to focus on the pre-1980 sample.

3.5.2 Are the returns on these ten portfolios formed on size concentrated in a specific month?

3.5.3 Compare the size factor to the market factor

You may want to consider mean excess returns by decade.

3.6 Repeat the exercises above with the value factor

3.7 Repeat the exercises above with the momentum factor

You may find it helpful to consider the worst months and years for the momentum factor.

3.8 Plot the coefficient estimates from a rolling Fama-French three-factor model for Berkshire Hathaway

Use a three-year window with daily returns. How has Buffett’s \(\alpha\) and \(\beta\)s changed over the past four decades?

3.9 Use the three-, four-, and five-factor models to determine how the ARKK Innovation ETF generates returns