450 likes | 468 Views
This research investigates the impact of demographics on creditworthiness, analyzing data from diverse household groups. Findings from Equifax data show correlations between income levels, race, education, and credit scores. The study delves into the relationship between demographic factors and credit performance, providing insight into the predictive power of credit scoring systems. By examining household finances and credit scores over time, the research aims to uncover patterns and disparities in credit outcomes based on demographic characteristics. The analysis utilizes various statistical models like OLS regression, Probit, and Logit to assess the probability of credit denial or approval based on demographic attributes.
E N D
Do Demographics Predict Creditworthiness? Presented by Kelli Jones ECON 616 April 2, 2003
Introduction • What is a credit score ? • Measure of relative creditworthiness / credit performance • Based on items from credit history such as bankruptcies, delinquent payments, revolving credit balances
Introduction • How is a credit scoring system built? • It is determined how effective each risk characteristic is in predicting credit performance • Each element is given a weight depending on that effectiveness • The combination of each element and weight results in the best predictor of credit performance • Generally, the higher the score, the better your credit
Introduction • How are credit scores used? • Credit applications • Mortgage loan applications • Insurance underwriting and/or pricing for personal auto and homeowners policies
Purpose of Research • To test whether certain demographic groups have a tendency to have worse credit (i.e. lower credit scores)
Avery, Bostic, Calem, Canner(1996, 2000) • Data obtained from Equifax on 3.4 million individuals making up 2.5 million households • income: • 33% of households in lowest income range have low credit scores, compared to 23% of households overall and 17% of households in the highest income range • As median family income ↑, median credit score ↑
Race: • as the %age of minority households ↑, median credit score ↓ • Education: • As the %age of high school graduates ↑, median credit score ↑ • Location: • No statistically significant relationship shown between credit scores and urban/suburban/rural classification • Age: • As the median age ↑, median credit score ↑
Kennickell, Starr-McCluer, Surette(2000) • Comparison of family finances from data obtained from 1995 and 1998 Survey of Consumer Finances • 1998 survey samples 4,309 households • Income: • As income ↑, the # of payments 60+ days past due ↓ • Age: • As age ↑, the # of payments 60+ days past due ↓
Fair, Isaac(1997) • Develops and markets credit scoring systems • Provided research paper in response to concerns that the use of credit scores results in unfair treatment to low-to-moderate-income (LMI) and high-minority area (HMA) populations
Income: • At a given credit score, the level of risk is the same regardless of income • Race: • Distribution of credit scores differs between HMA and non-HMA populations • For HMAs, 25.3% have scores < 620 compared to 13.8 % for non-HMA’s • At any given score, the odds (ratio of good to bad accounts) are lower for HMA’s; however, this difference seemed to be significant only at lower scores
Database • 1998 Survey of Consumer Finances • Complete sample is 21,525 observations • Reduced sample used for my analysis of those who have applied for credit in the last 5 years consists of 13,664 observations
Creditworthiness / credit score: • Y = 1 if credit denied or approved for lower amount based on credit history • Y = 0 if approved for full amount or denied for reasons other than credit history • Location: • No urban/suburban/rural classification • 9 categories describing area of country (e.g. New England, Midatlantic) • Not available in 2001 public dataset
Education: • 4 dummy variables to capture years of education • High school diploma • 1 – 3 years college • 4 years college • Graduate school • Having less than high school diploma is base case • Race: • 3 dummy variables • Black • Hispanic • Asian / Native American / Hawaiian / other • White is base case
Income: • Continuous variable • Age: • Continuous variable
Model • Yi = α+ βXi + εi • E(Yi) = Pi = P(Y = 1) = P( bad credit) = αhat + βhat Xi
Model • Zi = α+ βXi + εi • Zihat= αhat+ βhatXi = F-1(Pihat ) • Pihat = F(Zihat) where F is the normal distribution • Probability modeled is Y = 1
Model • Zi = α+ βXi + εi • Zihat= αhat+ βhatXi = ln (Pihat / (1 - Pihat )) • Pihat = exp(Zihat) / (1 + exp(Zihat) ) • Probability modeled is Y = 1
Comparison of Phat • ECON 616 Comparison.xls
Enhancements • Update data to 2001 SCF • Look at multivariate results • Analyze goodness of fit of models