FINS5530 Financial Institution Management 2014 S1 Individual Project

Assignment Requirements

FINS5530 Financial Institution Management

2014 S1 Individual Project

Submission deadline: 9pm, 1/6/2014 (Sunday)

Submission method: Submit your 3-page report via the assignment submission link under

“Exams and Assessments” section on Moodle course website. The link will be created soon.

Assume that you work as a credit-risk analyst at a newly-opened U.S. commercial bank whose

target clients are mainly U.S. publicly-listed firms. Your supervisor would like you to develop

a credit scoring model for credit risk evaluation. The essentials of developing a linear credit

scoring model is to determine the set of factors causing or related to defaults and decide on

the relative importance of these factors. So the typical procedure is to: 1) generate a list of

potential determinants of credit risk, based on economic theories, common sense and business

experience; 2) calibrate these potential factors in determining defaults with historical data

representative of your target clients.

As a starting point, you try to discover the potential determinants of credit rating by S&P to

facilitate the building of your own credit scoring model. S&P ratings about U.S.-listed firms could

be requested from S&P, and firm characteristics potentially relevant for credit risk could be

downloaded from Compustat – a data vendor affiliated with S&P. Assuming that you decided to

sample the data for two years only, 1990 and 2010, which were assembled in the provided excel

file. Specifically, the file contains the following variables:

[1] S&P Rating: original rating for each firm-year provided by S&P.

[2] S&P rating, numerical value: the transformed numerical value of the rating.

For the description of ratings and the transformation of the original rating to numerical values,

please refer to the sheet “SP_Rating_Details”.

[3] Ln(S&P rating, numerical value) :log transformed value of the numerical rating values

[4] Other firm characteristics as defined in Altman credit scoring model. You can refer to the

textbook for their definition. The only exception is that book value rather than the market value

of the equity is used to calculate Capital Ratio (Book equity/Long term debt).

Please prepare a three-page report integrating the following tasks:

[1] Run a regression of Ln(S&P rating, numerical value)on the five firm characteristics,

separately for 1990 data and 2010 data. If you don’t know how to run a multiple regression,

please refer to the guidance in the attached PDF file:https://faculty.fuqua.duke.edu/~pecklund/

ExcelReview/Use%20Excel%202007%20Regression.pdf

[2] Please compare the two credit scoring models as suggested by 1990 data and 2010 data with

the one estimated by Professor Altman and listed in the textbook. Please also compare the two

credit scoring models estimated by you. Please discuss the possible reasons why they differ and

the lessons for your credit scoring model development.

[3] The list of these five firm characteristics is surely not exhaustive in terms of potential credit

risk determinants. Could you think of other factors relevant for credit risk and explain why they

[4] Now take the 2010 data, calculate the predicted rating value (Ln(S&P rating, numerical value)

) based on your regression coefficients, i.e., Predicted value of Ln(S&P rating, numerical value)

= intercept + coefficient of X * X (there are five Xs here). Then generate the difference between

actual and predicted values of Ln(S&P rating, numerical value). Now assuming S&P ratings

provide a correct description of firms’ credit quality, a positive value of difference means the

model underestimates the actual credit quality of the firm, and the opposite holds for a negative

Please sort the observations into 5 (roughly equally-sized) groupsby the difference between

actual and predicted ratings, and report the average values for each group for the following

variables: a) the difference, b) the numerical rating score (not the log one), c) five firm

Based on the average values of the above variables for those groups, you should be able to see

the performance of the model across firms with different characteristics, i.e., the model could

explain the credit quality of specific types of firms very well (with very small difference btw

predicted and actual ratings), but very badly for some types of firms (with very big difference

btw predicted and actual ratings).

What could you learn from this exercise to help you search for potential factors relevant for

[5] Finally a note: I list these tasks one by one, but you should integrate all these tasks

coherently in your 3-page report. Your overall mission is to develop a credit scoring model –

determining the list of factors relevant for credit risk, assigning the relative importance, and

being aware of issues in this modeling procedure.

 

Order Now

http://zelessaywritings.com/order/

 

You can place an order similar to this with us. You are assured of an authentic custom paper delivered within the given deadline besides our 24/7 customer support all through.

 

[order_calculator]