To identify the primary characteristics of the targeted and neglected consumers, we
conduct principal components analysis, which constructs uncorrelated linear combinations of
the individual attributes. Table 6 shows that four factors explain 57.43% of the variance for
targeted consumers. The most important factor is whether the consumers had serious
payment difficulties and credit damage, followed by some attributes indicating credit use
(loan balances), financial resources (income and homeownership), and credit history (auto