Forecasting for Credit Risks and Estimating Probability for Credit Defaulters (Math Problem Sample)

Forecasting for Credit risks and Estimating Probability for Credit Defaulters
Student’s Name:
Course Code:
Department:
Professor’s Name:
Date
Table 1.

Forecasting for Credit risks and Estimating Probability for Credit Defaulters

Step 1: Business Problem Understanding

What was the Business problem you are trying to solve?
In the contemporary business environment, consumers tend to largely use credit cards for their purchases as well as acquisition of goods and services. Recently credit card issuers such as banks, with the aim of maximizing their market share, over issued credit cards even to individuals who were unqualified for the product. Additionally, it was learnt that consumers of the service who had the potential of repaying overused their credit cards to an extent that it was difficult to recover the deficiencies. It is because of this that most card issuers faced card debt crisis which has affected the businesses financial performance. Besides the delinquency of credit card crisis is expected to dramatically increase in the comparing financial years if adequate risk prediction is not put in place. The current business problem of concern was also witnessed in Taiwan in the third quarter of 2006 when card-issuing banks over-issued credit cards to unqualified consumers while at the same time those with the potential of repaying exhausted their cards accumulating heavy credits (Yeh % Lien, 2009). The authors also observed that the problem caused a blow to consumer finance confidence and presented a significant challenge to both card-issuing banks and the consumers.
To solve the problem, the financial institutions of interest in Taiwan such as the card-issuers had to engage in robust risk production while leaving the crisis management at the downstream. The aim was to use consumer transaction and repayment records to alleviate the customer credit risk while also reducing the damages caused by uncertainty. While the case was experienced in Taiwan, the latter have emerged for most financial institutions across the globe and thus the need to upstream risk prediction while down streaming crisis management to alleviate such risks caused by credit debts. The present analysis has thus used statistical methods including the discriminant analysis, logistic regression and Bayes classifier as models for risk prediction with regards to the customer repayment data provided. The models were useful in forecasting credit risks and estimating the probability of default among consumers.

Step 2: Data