Principles of Finance

Erwan Morellec Assignment 4 Instructions Assignments should be done in groups no larger than 3 students. You should remain with the same group through the entire course. Submit on Moodle only one copy of solutions per group. For each assignment you can get a maximum of 100 points. All assignments turned in late will not be graded (zero points). Due date The due date is indicated on Moodle. For this assignment, you should hand in two les: First, a separate le (Word, PDF) where you record all your results (tables, graphs, etc.) and interpretations. Second, you should also hand in your Excel le with your calculations. The Excel le we only need if we do not understand something you did. The instructions are very precise on what you should do, what results to report, and what questions to answer. Part I (20 points) Today is May 24, 2006, and you have just started your new job with a nancial planning rm. In addition to studying for your exams, you have been asked to review a portion of a clients stock portfolio to determine the risk/return proles of 12 stocks in the portfolio. Unfortunately, your small rm cannot aord the expensive databases that would provide all this information with a few simple keystrokes, but thats why they have you. Specically, you have been asked to determine the monthly average returns and standard deviations for the 12 stocks for the past ve years. The stocks are: 3M Company (MMM), Adobe Systems Inc. (ADBE), American Express Company (AXP), BlackRock, Inc. (BLK), The Coca-Cola Company (KO), Johnson & Johnson (JNJ), Kellogg Company (K), Pzer Inc. (PFE), Ryder System, Inc. (R), Tyson Foods, Inc. (TSN), The Walt Disney Company (DIS), and Whirlpool Corp. (WHR). 1

 Collect price information for each stock from Yahoo! Finance (http://finance. yahoo.com) as follows: Enter the stock symbol. On the page for that stock, click Historical Prices on the left side of the page. Enter the start date as May 24, 2001 and the end date as May 1, 2006 to cover the ve-year period. Make sure you click monthly next to the date. Download the data to a spreadsheet. Delete all columns except the date and the adjusted close (the rst and last columns). Keep the Excel le open and go back to the Yahoo! Finance Web page. When you return to the prices page, enter the next stock symbol and hit Get Prices again. Do not change the dates or frequency, but make sure you have the same dates for all the stocks you will download. Again, click Download to Spreadsheet and then open the le. Copy the last column, Adj. Close, paste it into the Excel le and change Adj. Close to the stock symbol (so that you know which column belongs to which stock). Make sure that the rst and last prices are in the same rows as the rst stock. Repeat these steps for the remaining ten stocks, pasting each adjusted closing price right next to the other stocks, again making sure that the correct prices on the correct dates all appear on the same rows. Convert these prices to monthly returns as the percentage change in the monthly prices. Hint: Create a separate worksheet within the Excel le. Compute the mean monthly returns and standard deviations for the monthly returns of each of the stocks. Convert the monthly statistics to annual statistics for easier interpretation. Add a column in your Excel worksheet with the average returns across stocks for each month. These are the monthly returns to an equally weighted portfolio of the 12 stocks. Compute the mean and standard deviation of monthly returns for the equally weighted portfolio. Double check that the average return on this equally weighted portfolio is equal to the average return of all of the individual stocks. Again, convert these monthly statistics to annual statistics. Report the following results: Report the annual mean return and annualized standard deviation of the 12 stocks and of the equally weighted portfolio in a table in the separate le. Create one plot with the annual standard deviation on the x -axis and the annual return on the y -axis and represent the 12 stocks and the portfolio in this plot. Copy the plot into the separate le. How does the portfolio standard deviation compare with those of individual rms? Comment.

Part II (40 points) Now you want to rebalance your portfolio with the optimal weights that will provide the best risk and return combination for the 12-stock portfolio. Use the Solver function in Excel to perform this analysis (the time-consuming alternative is to nd the optimum weights by trial and error). Begin with the equally weighted portfolio analyzed above. Establish the portfolio returns for the stocks in the portfolio using a formula that depends on the portfolio weights. Initially, these weights will all equal 1/12. You would like to allow the portfolio weights to vary, so you will need to list the weights for each stock in separate cells and establish another cell that sums the weights of the stocks. The portfolio returns for each month MUST reference these weights for Excel solver to be of any use. Compute the values for the monthly mean return and standard deviation of the portfolio. Convert these values to annual numbers. Compute the ecient frontier when short sales are NOT allowed. To activate the Solver function in Excel, click the Tools tab, select Add-Ins..., check Solver-Addin in the pop-up dialog box, and then click OK. To set the solver parameter: Set the target cell as the cell of interest, making it the cell that computes the (annual) portfolio standard deviation. Minimize this value. Establish the By Changing Cells by holding the control key and clicking in each of the 12 cells containing the weights of each stock (all in all you select 12 cells). Add constraints by clicking on the add button next to the Subject to the constraints box. One set of constraints will be that the weight of each stock is greater than or equal to zero. A second constraint is that the weights will sum to one. Compute the portfolio with the lowest standard deviation for a given expected return. Start by nding this portfolio with an expected return of 5%. To do this, add a constraint that the (annual) portfolio return equals 0.05. If the parameters are set correctly, you should get a solution when you click Solve. If there is an error, you will need to double-check the parameters, especially the constraints. Record the resulting standard deviation for the optimally weighted portfolio with a return of 0.05 in a separate cell on the spreadsheet. Repeat the previous step to solve for the portfolio with the lowest standard deviation for several dierent choices of expected returns: 0.07, 0.1, 0.15, 0.2, 0.3, 0.4. Record these values. Plot the ecient frontier with the constraint of no short sales. To do this, use a plot with portfolio standard deviation on the x-axis and the return on the y -axis.

 Redo your analysis to allow for short sales by removing the constraint that each portfolio weight is greater than or equal to zero. Use again the solver to calculate the (annual) portfolio standard deviation when the annual portfolio returns are set to 0.05, 0.07, 0.1, 0.15, 0.2, 0.3, 0.4. Plot the unconstrained ecient frontier into the same plot as the constrained ecient frontier. Report the following results: Report the values that you used to produce the plots of the constrained and unconstrained ecient frontier in a table in the separate le. Copy the plot into the separate le. What do you observe in the plot? Comment. Part III (40 points) Finally, you would like to use the CAPM to compute expected returns for all 12 stocks in the portfolio. Specically, you should estimate betas for each stock using ve years of monthly data and an expected return using the historical risk premium of 4.5% and a risk free rate of 6%. You should compute the betas using the S&P500 as market index. Get the returns for the S&P 500 from Yahoo! Finance (symbol ^GSPC). Again, use May 24, 2001 as the start date and May1, 2006 as the end date to obtain the prices and remember to click monthly. Download those prices and add the adjusted closing price to your spreadsheet. Then create monthly returns for the S&P 500 following the procedure you used for individual stocks. For each stock, estimate the following regression: Rit = i + i RS &P 500,t +
it

where Rit is the monthly return of stock i at time t, RS &P 500,t is the return of the S&P 500 at time t, and it is a normally distributed error term. To do those regression, you can use any program you like (Excel, Stata, Matlab, Eviews etc.). Report the following results: Report in one table the following statistics for each stock: the estimate of the constant, the estimate of the slope coecient (beta), the t-statistic of the slope coecient, the adjusted R-squared, the number of observations you used for the estimation, and the Jensens Alpha (assume a monthly interest rate of 0.4% to do the Jensens Alpha computation). Finally, compute and report in the same table the expected return (on an annual basis) for each stock using the beta estimates and the other information provided in this exercise (Hint: Also have a look at your class notes). Insert this table into the separate le. How do you interpret the slope coecient (beta)? 4

 What does the adjusted R-squared mean? What does Jensens Alpha measure? Are the estimations of expected returns for BlackRock, Inc. (BLK) and Kellogg Company (K) reliable? Why? Why not? Contrast this to the estimate of the expected return for American Express Company (AXP).