Credit Risk

Question: If you were going to estimate the probability of a company failing in the next period, what variables do you think would be useful?
Firm Characteristics Employed in Previous Studies
The characteristics are taken from both accounting and market data • Accounting-based variables
For example: NITA = net income / book value of total assets
For example: TLTA = total liabilities / total assets
• Market-based variables
RELSIZE = Log [Firm mkt cap. / S&P500 mkt cap.]
SIGMA = standard deviation firm’s stock return
• Combined market- and accounting-based variables
CASHMTA = cash and short-term assets / market value of assets We create these variables in SAS using the data in the dataset hazard1.
Excel File
The descriptions of the variables in the dataset hazard are in an excel file (‘L2_Spreadsheets for Hazard Model Variables’)
The variables are listed in the order in which they appear in SAS.
• Open the excel file – BB_Hazard1 Variable Descriptions • Scroll through the variables.
• Ensure that you understand the description of each variable • Now pause this recording to look at the excel file
A short video is available to accompany the excel file
Creating Financial Ratios
• Credit risk models often employ financial ratios
➢ See page 114 of Shumway, Table 1 with further discussion on page 115. ➢ See page 2907 of Campbell et al., Table II.
• The SAS dataset ‘Hazard1’ was prepared with all values given either at the beginning of the quarter or over the previous quarter (e.g. SIGMA = st. dev returns over the previous quarter).
• The names of the existing variables in the SAS dataset are taken from Compustat/CRSP (the source of the data).
• The names of the variables to be employed to create the required variables for our models are as follows (see definitions in excel file):
Instructions for Opening SAS and Locating the Dataset Open SAS as follows:
• From the windows icon select “S”
• SAS 9.4
Create a library in SAS called ‘CREDIT’ and point that to your folder. Here’s how: Below is the screen when you open SAS.


Close the dialogue box. Note that you have various windows:
• Log tells you what is happening and highlights and problems. You should look at this after running each piece of SAS code.
• Editor is for typing SAS commands.
• Explorer is the filing system.
• Output shows the results
Highlight ‘Libraries’ then File→New to arrive at the below screen


Type CREDIT into the Name
Browse and find the folder on your own drive which contains Hazard1 Click ‘OK’.
Double click on Libraries
You should now see the dataset ‘Hazard1’ in the SAS library ‘CREDIT’. There is also a library called ‘Work’.
If you fail to specify the SAS folder (e.g. CREDIT) the file created will go into the folder ‘Work’ by default.
Any files in the ‘Work’ folder are NOT saved when exiting SAS.
Examining the SAS dataset Hazard1
Open the dataset by clicking on it.
Scroll down to the bottom and across to the right.
How many rows of data are there? (Answer = 49,412)
The dataset contains variables for firms for several years.
Each variable name is across the top of the dataset.
Each firm has a unique PERMNO.
For example, PERMNO 10032 relates to PLEXUS CORP.
Period = 1 (min) for 1992 Q2 and Period = 55 (max) for 2012 Q4.
Right click on PERMNO and go to ‘Column attributes’.
“Name” is the name that SAS recognises as the variable and which must therefore be used in all SAS code.
“Label” is the variable name that you see, which can be different. The PERMNO variable type is “Numeric”. Please ensure that you can see this. Numeric and character variables are treated differently in SAS code.


Study Cred Tutor

4.6 (24k+)

Purchase the answer to view it



Click one of our contacts below to chat on WhatsApp

× How can I help you?