For this exercise, we will try to see what determines the managers’ incentives to provide

guidance. In order to do so, you will have to merge data from IBES and Compustat. Here

are the steps to follow:

• Import ibes.rds. This contains information on quarterly management guidance from

2013 to 2020. The unique identifier for this database is ticker. You can find the

column details on WRDS.

• Since the unique firm identifier in Compustat is gvkey, you will need to use gvkey.rds

to link gvkey and ticker.

• You will have to download quarterly data from Compustat. Before downloading the

data, read through the problem set and think about which variables you want to

download.

Panel data from voluntary disclosure

We want to see whether managers provide guidance or not for every quarter. However, if the

managers do not guide, there would be no observations in the data. In order to take care

of this issue, we will create a new panel dataframe using expand.grid. Here is an example

below:

expand.grid(

x = 1:5,

y = c('a','b')

)

• Create a new dataframe using expand.grid. The columns should be firm identifiers

(either ticker or gvkey) and year-quarters.

1

• Get only the EPS estimates from the guidance data and merge them to the newly

created database above. You should have a column like guidance that is populated

with 0s or 1s depending on whether the manager provided guidance or not.

• Lastly, merge the firm fundamentals from Compustat that you think have an effect on

managers’ incentives to provide guidance.

Analysis

• Run a logit regression, where the dependent variable represents whether the manager

discloses or not and the independent variables are a number of firm fundamentals that

you think will be important for the managers’ decision to provide guidance.

• Run a linear regression using the same regression model as above. Do your results

change (in terms of the direction of the coefficients)?

• Describe your results in words and/or using plots.