Saturday, 2 May 2020

Baker and Wurgler, 2004 – Appearing and Disappearing dividends

Baker and Wurgler, 2004 – Appearing and Disappearing dividends
For this assignment, you will be extending the analysis performed by Baker and Wurgler (2004) and replicated in class to the post article sample period. Specifically, in class we reproduced results for Tables 1 of the original paper using the original sample. In your homework assignment, you will reproduce Table 1 for the full sample (1963-2019) and (2001-2019).
Step 1 – Get the data, apply the filters, and create the main variables.
Download Compustat data and all variables that are listed in the Appendix to FF2001 paper. Add prcc_c and exchg (Use the Compustat translation file on Blackboard to determine what variable names you need to include in your Compustat download or follow the code that we developed in class).
You will be re-creating most of the variables originally defined by Fama and French.
Create firm characteristics / explanatory variables used for logistic regressions to create propensity to pay. Follow definitions provided on pages273-274 of the paper. We created most of these variables in class.
In addition, you will need NYP and dividend premium. You can get NYP from the dataset that I posted on the Blackboard. The NYP variable is created for the full sample through 2019. Follow definition of dividend premium provided on page 277.
Apply filters, exclude utilities (SIC codes 4900-4949) and financial firms (SIC codes 6000-6999), exclude firms not publicly traded in US, remove firms with missing variables as detailed in the appendix.
Step 2 – Plot and examine the trends in PTP and DP
Re-create figures 1 and 2 of the BW paper. (We created some panels of figure 1). Since our data seems slightly different that BW data, please recreate the figures for the full sample 1963-2019.
Copy the results in word
Additionally print the annual values of PTP and DP into a table and compare to BW values. BW report PTP values from 1978-2001 in their Table 2.
In their companion BW paper they also report annual values of dividend premium. I am copying the table from that paper (and including the link to the paper) here. Please discuss how your values of PTP and DP differ from theirs in the early part of the sample (1963-2000). By comparing the numbers can you pinpoint why our numbers are different?
 Step 3 – Perform and report regression results of Table 1
Report regression results for the full sample period (1963-2019) and for the period post BW paper (2001-2019). Use both PTP definitions. For the full sample period use univariate and bivariate model. For the post-BW period use only univariate model because Nixon dummy will be zero and will play no role.
Step 4 – Write a brief explanation of your findings. Does the predictive power of dividend premium strengthen? Does it weaken or stay the same? Why do you think that happens?
In a word document copy the figures 1 & 2 that you produced for the extended (full sample, 1963-2019). Additionally, copy your regression results into a table resembling Table 1 from the paper. Write a few sentences discussing the results.

No comments:

Post a Comment

Note: only a member of this blog may post a comment.