Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

Assignment 2:

Part 1: Historical Data and Stock Betas (10pts)

a) Collect 10 years historical price/return data for stocks with following tickers: "RACE", "JPM","XOM", “AMZN“ and S&P 500 Index. Collect Risk-Free Interest Rate data from FRED. (you can choose 3Month US Treasury Rate)

b) Estimate the beta (in Sharpe’s Model) by running OLS with weekly returns.

c) First use the complete 10-year sample to report the estimated betas, and standard error of beta. What is the t-stat? (For each of 4 stocks)

d) Then run OLS over moving windows.

- Use 3-year window (start with 2012-Jan to 2015-Jan) window and then repeat the regression by moving the 3-year (~160-week) window by 1 week. (use for-loop)

- Collect estimated betas and then plot the beta values to see the changes over time for each stock.

e) Repeat a-d with Python in Jupyter Notebooks.

Part 2: Historical Data and Fama-French Factors (5 pts)

- Briefly explain what the 3-Factor Model is. Refer to K. French’s website, internet searches and textbooks.

- Collect 20 years of monthly historical price/return data for stocks with following tickers: “CAT", “XOM", “IBM", “WMT” and S&P 500 Index, and F-F factor returns data

- Estimate the Fama-French 3 Factor model coefficients for the given stocks over 2 different samples. (2002-Sep to 2011-Sep) and (2012-Sep to 2021-Sep) periods, separately.

- What do the estimated coefficients tell you about the selected stocks’ factor exposures? Are the estimated coefficients statistically significant? Do they look useful in explaining the stock returns in any way? Explain

Do the coefficients change significantly between the 2 different 10-year periods? Show the differences with bar charts

Part 3: Common issues in back-testing practices(5 pts)

Explain the following terms in the context of financial portfolio back-test practices. How do they occur? Briefly describe in a paragraph. Provide examples if possible.

(1) Survival Bias; (2) Look-ahead Bias; (3) Over-fitting; (4) Data snooping

Part 4: Machine Learning and Back-testing Discussion

Read the attached articles on the uses of back-testing and machine learning in financial market strategy and asset pricing research. Related questions will follow