FIN 6123 Investment Management and Analysis
Group Project Guideline
For this project, you are going to form. groups with four to six students in each group. You can form. groups on your own before midnight October 20th. After that, students without groups will be randomly assigned into new groups. You can choose one from the following two options to work on for the project.
Option I: Portfolio optimization and performance evaluation
If you choose this option, all the data work can be done in Excel. That being said, you are more than welcome to use other software if you want as long as the outputs are readable.
1. Pick at least 10 stocks that you want to invest in. Briefly discuss the reasons for picking these stocks. Download monthly returns of these stocks from January 2019 to December 2024.
2. Calculate the mean and standard deviation of the returns of each stock during January 2019 and December 2021. Estimate the covariance matrix for each stock during the same sample period.
3. Also using data from January 2019 and December 2021, graph the efficient frontier of the stocks, find the weight of each stock in the optimal risky portfolio, and graph the capital allocation line for the optimal risky portfolio.
4. Calculate monthly returns of the optimal risky portfolio from January 2022 to December 2024 (using returns from January 2022 to December 2024 and portfolio weights estimated in Step 3).
5. Pick a mutual fund, ideally one that primarily invests in similar stocks to those in your portfolio. Download monthly returns ofthe mutual fund from January 2022 to December 2024. Compare the following performance evaluation metrics of your optimal risky portfolio and the mutual fund during the sample period:
a. Average return
b. Standard deviation
c. Sharpe ratio
6. Estimate CAPM regressions foryour optimal risky portfolio and the mutual fund. Discuss the alpha and market beta of your portfolio and compare them with those of the mutual fund.
7. Estimate the Fama-French three-factor regression for your optimal risky portfolio. Discuss the alpha and factor betas ofthe portfolio.
8. Use regression to examine the return reaction of your optimal risky portfolio to changes in one or more state variables of your choice, controlling for market excess return. Discuss your findings.
9. Estimate the conditional CAPM regression for your optimal risky portfolio using state variables of your choice. Discuss your findings.
Option II: Factor investing and performance evaluation
If you choose this option, the data work will be hard to do with Excel. You might need other programming tools (Python, R, SAS, Stata, etc.).
1. Design a long-short portfolio investment strategy based on a factor (can be either a firm characteristic such as P/E ratio or a market pattern such as past returns). You are free to design the details of the strategy, such as stock selection criteria and rebalancing frequency. Briefly discuss why you believe this portfolio might generate a positive CAPM alpha.
2. Download necessary return data and calculate the historical monthly returns of the long-short portfolio for an appropriate sample period of your choice.
3. Estimate the CAPM regression for the portfolio. Discuss the alpha and market beta of the portfolio.
4. Estimate the Fama-French three-factor regression for the portfolio. Discuss the alpha and factor betas ofthe portfolio.
5. Use regression to examine the return reaction of your portfolio to changes in one or more state variables of your choice, controlling for market excess return. Discuss your findings.
6. Estimate the conditional CAPM regression for the portfolio using state variables of your choice. Discuss your findings.
Deliverables:
1. The Excel file or the code you used to create the portfolios and estimate the models. You can refer to the Excel files for Lecture 4 and Lecture 6 and the sample code posted on Blackboard for roughly what the file should look like.
2. A group-written report including at least the following components.
For Option I:
a. The reasons for picking these stocks
b. Graphs and statistics mentioned in the instructions
c. Description of the portfolio optimization process
d. Comparison of the performance of your portfolio and the mutual fund
e. Discussion of the regression outputs and analysis of the risk premium of the portfolio
f. An overall discussion of the performance of the investment strategy. Do you recommend this strategy? Why or why not?
For Option II:
a. Your hypothesis on why the factor might generate a positive CAPM alpha
b. Description of the construction of the long-short portfolio
c. Discussion of the regression outputs and analysis of the risk premium of the portfolio
d. An overall discussion of the performance of the investment strategy. Do you recommend this strategy? Why or why not?
3. Group presentation:
a. The presentations will be conducted in class. A total of 20 minutes will be allocated to each group, including 15 minutes for the presentation and 5 minutes for a Q&A session.
b. Each group member has to participate in the presentation. Presentation performance will be graded for each individual student to satisfy the requirement of an AACSB Learning Objective. Any students who do not participate in the presentation will receive a presentation grade of zero.
c. The presentations might be recorded for grading purposes
For the Monday session, all deliverables except the presentation are due at 10pm on November 9th. The presentations will be delivered on November 10th and November 17th.
For the Wednesday session, all deliverables except the presentation are due at 10pm on November 11th. The presentations will be delivered on November 12th and November 19th.
For the part-time session, all deliverables except the presentation are due at 10pm on November 21st. The presentations will be delivered on November 22nd.
Some useful tips:
1. If you choose Chinese stocks for the project, you should use risk factors and state variables of the Chinese market. If you choose U.S. stocks for the project, you should use data on the U.S. market. The file “Group project data.xlsx” on Blackboard contains some useful data for the project. You can also download the data from your own sources. The Excel file contains state variables for the U.S. but not for China. If you want to use Chinese stocks for the project and cannot find the equivalent of these variables for the Chinese market, it is ok to use other commonly used measures for macroeconomy conditions such as GDP growth rate, unemployment rate, etc.
2. You may choose stocks from other markets if you want but it might be hard to obtain necessary data for smaller markets. Choosing both Chinese stocks and U.S. stocks is not recommended as it will make things overcomplicated. Choosing both U.S. stocks and ADRs of Chinese stocks is fine.
3. Your investment strategy does not have to create a significant alpha. The regressions do not have to generate significant beta coefficients. You will not be penalized for having insignificant alphas or other regression coefficients.
4. If you choose to work on Option I, under certain circumstances you might get negative portfolio variance, which is an inherent problem of the Markowitz model potentially caused by negative sample covariances or negative asset weights. You can get around the problem by imposing short- sale constraints on the portfolio (setting a constraint that the asset weights must be non-negative).
Grading:
Code/Excel spreadsheet and group-written report 10%
Presentation 10%