So, Thomas, Seow, and Mues (2014) developed a logistic regression model for estimating the profitability of consumers’ use of credit cards using a transactor/revolver scorecard and explained how such a scorecard can be used as part of a risk assessment system. Building a scorecard that enables credit card issuers to classify an applicant as either good (low risk) or bad (high risk) has been the traditional approach to deciding whether or not to extend credit. According to the authors, credit card companies could better inform their decisions by also distinguishing between transactors and revolvers, or between those who pay off their balances to avoid interest charges and those who do not. The authors argued that while transactors are inherently good, they are also less profitable than revolvers and by scoring applicants along both good/bad and transactor/revolver dimensions can help credit card issuers make pricing decisions and more accurately estimate profits.
The authors collected three years’ of data from a major Hong Kong institution consisting of 1577 defaulted accounts and 4731 non-defaulted accounts. The independent variables were predetermined from prior research and included occupation, education type, citizenship, residential type, employment status, annual income, months associated with the bank, and age. Weight of evidence was examined to assess the relative default risk of each variable. Stepwise logistic regression was used to obtain the coefficients of three models including the standard scorecard, the transactor/revolver scorecard, and the standard scorecard restricted to revolvers. Scores from the latter two models were then used to develop a fourth model to assess the probability that an applicant is a revolver and is likely to be a good risk.
Based on the fourth model, the authors were then able to use a linear function to compare the traditional credit card profitability model with their proposed model that takes into account the transactor/revolver score. The results showed that the proposed model was significantly more accurate in that it does not overestimate profits generated by transactors. Further, the authors argued that by generating a transactor/revolver score at the time acceptance/rejection decisions are made credit card companies would be able to optimize profits and more accurately screen applicants.
Anderson, Sweeney, Williams, Camm, and Martin (2012) stated there are numerous linear programming applications used in finance, such as in capital budgeting, make-or-buy decisions, asset allocation, portfolio selection, and financial planning and in risk assessment. The research done by So, Thomas, Seow, and Mues (2014) demonstrates how logistic regression and linear models can help credit card companies make better informed applicant acceptance/rejection decisions and more accurately minimize risks and maximize profits.
A long time CPA turned online professor, Kathleen’s life was transformed upon completion of her dissertation An Investigation of the Factors that Influence Faculty and Student Acceptance of Mobile Learning in Online Higher Education.
Her statistical analyses was called ”pioneering” by her committee chair Dr. Marlyn K. Littman and brought Kathleen full circle back to her number-crunching roots inspiring her to earn a second master’s in Business Intelligence.
Kathleen plans to continue her studies of contemporary issues related to teaching, learning, and technology and loves to help undergrad and grad students achieve their academic and professional goals. As a lifelong learner she also plans on continuing her quest to understand the problems posed by mobile and micro learning formats and find innovative ways of helping people maximize the benefits these emerging technologies afford.