EC623/MF703 Financial Economics (Seminar in Corporate Finance) Data Exercise 1 Subhankar Nayak, 2022 Page 1 of 4 EC623/MF703 Financial Economics (Seminar in Corporate Finance) WINTER 2022 Data Exercise 1 Instructions and Notes: 1. The assignment submission is due back by 10.00 pm on Sunday, February 20th, 2022. 2. The submission needs to be in two forms: 1. The solution report. 2. The accompanying spreadsheets and/or outputs from statistical packages. 3. Submit all files in the dropbox folder titled “Data Exercise 1: Report, Data and Analysis” on MyLearningSpace. 4. The assignment exercise is worth 25 points 5. For Masters students: The assignment is designed as a group exercise; form groups of two students each. 6. For PhD students: The assignment is designed as a solo exercise. 7. Collaboration across groups (i.e., joint work with members of another group) is strictly discouraged. However, you are free to consult with me. 8. The assignment is designed as a do-it-yourself analytical exercise based on real world data. 9. Conduct comprehensive analysis and document the procedure as well as the results thoroughly. 10. Interpret all results. 11. A professional report is expected. 12. Grading will be based on (a) correctness and completeness of analysis, (b) creativity in analysis, (c) documentation and interpretation of results, and (d) quality of report. EC623/MF703 Financial Economics (Seminar in Corporate Finance) Data Exercise 1 Subhankar Nayak, 2022 Page 2 of 4 Asset Pricing Models and Corporate Valuation Objectives: (a) application and test of asset pricing models; (b) comparison of one-factor and multifactor models; (c) extraction of corporate valuation parameters; (d) linking asset pricing models to corporate valuation. Sample: the sample created in the Data Exercise 0 – all 30 component stocks constituting the Dow Jones Industrial Average (DJIA or Dow) index. Time period: 2003 through 2020; 18 years for each stock, monthly frequency. Sub-periods: pre-crisis era (2003-2006); financial crisis era (2007-2009); post-crisis Obama era (2010-2016); turbulent Trump era (2017-2020). Data Sources: (a) CRSP database (under WRDS system) for stock price returns; (b) French data library (on Kenneth R. French’s website) for (i) risk-free returns and (iii) returns on Fama-French five factors; and (c) COMPUSTAT database (under WRDS system) for dividends-per-share and earnings-per-share data. Asset pricing models: 1. Capital Asset Pricing Model, CAPM 2. Fama-French 3-factor model 3. Fama-French 5-factor model Equity valuation model: Gordon dividend discount constant growth model (DDM) PROCEDURAL DETAILS 1. Collection of returns and factors data: From CRSP database, collect the returns on all samples stocks on a monthly frequency for the 18-year sample period. From Kenneth R. French’s data library, collect the following on a monthly frequency for the 18-year sample period: a) excess return on the market, b) risk-free rate, c) returns on the size factor, d) returns on the book-to-market factor, e) returns on the profitability factor, and f) returns on the investment factor. 2. Implementation of asset pricing models: Implement the three alternate asset pricing models, i.e., conduct monthly time-series regressions of stock returns on: EC623/MF703 Financial Economics (Seminar in Corporate Finance) Data Exercise 1 Subhankar Nayak, 2022 Page 3 of 4 a) market return, b) Fama-French 3-factor returns, and c) Fama-French 5-factor returns. 3. Table 1: Report the summarized regression output of 1-factor CAPM for each individual stock – for the full period only present the results in a succinct way – all key regression parameters must be reported in a single row. 4. Table 2: Report the summarized regression output of 3-factor Fama-French for each individual stock – for the full period only present the results in a succinct way – all key regression parameters must be reported in a single row. 5. Table 3: Report the summarized regression output of 5-factor Fama-French for each individual stock – for the full period only present the results in a succinct way – all key regression parameters must be reported in a single row. 6. Table 4: Report the summarized regression results for the aggregate sample of all stocks (not individual stocks) and for all three models for the full period as well as the four sub-periods. 7. Comparison of asset pricing models from econometric validity perspective: Compare and discuss the validity of the 1-factor vs. 3-factor vs. 5-factor models. 8. Collecting dividends per share (DPS) and earnings per share (EPS): From COMPUSTAT database, for each stock collect: a) the first four quarterly dividends per share (on an ex-date basis), b) the last four quarterly dividends per share (on an ex-date basis), and c) the last four quarterly earnings per share (basic, including extraordinary items). Notes: o Not all stocks are dividend-paying. o For earnings, first four quarters encompass the start-year of the sample, and the last four quarters correspond to the last sample year. o For dividends, the calendar timing for the first and last quarters depend on corresponding dividend history of the stock. 9. Computing valuation parameters: Based on the data collected in the previous step, for each stock, compute: a) DPS–T = average of first four quarterly dividends per share, b) DPS0 = average of last four quarterly dividends per share, c) g = annualized growth rate based on DPS–T and DPS0, d) EPS0 = average of last four quarterly earnings per share, and e) b = earnings retention ratio computed based on DPS0 and EPS0. EC623/MF703 Financial Economics (Seminar in Corporate Finance) Data Exercise 1 Subhankar Nayak, 2022 Page 4 of 4 10. Table 5: Report the values of DPS–T, DPS0, g, EPS0 and b for each stock, and also the mean and median values of each for the overall sample. 11. Valuation of equity: Using the valuation parameters of Table 5 and the (three versions of) cost of equity computed alternately in Tables 1, 2 and 3, apply Gordon dividend discount constant growth model (DDM) to estimate (three) forecasted values of each stock at the end of 2020. Also collect the actual stock price as at the end of 2020. Note: It is possible that valuation parameters turn out to be so that DDM may or may not be applicable. Playing the role of either econometricians or financial managers, decide how to handle these applicability issues. 12. Table 6: For each stock, report: a) the 3 forecasted stock prices for 2020-end, b) the actual stock as of 2020-end, and c) the 3 percentage forecast errors. 13. Comparison of asset pricing models from equity valuation perspective: Based on the results of Table 6, compare and discuss the appropriateness of the 1-factor vs. 3-factor vs. 5-factor models towards equity valuation. ADDITIONAL NOTES A. Be creative. You are encouraged to conduct additional analysis (and report additional tables). If so, it will be suitably rewarded. B. Describe the steps and procedure. Be detailed enough so that a reader (namely, I) can replicate your work in entirety just by reading your report. C. It is important to understand, interpret and explain the findings in detail. D. Prepare a professional report. This is a very important criteria.
