Predictive models of expenditure and over-indebtedness for assessing the affordability of new consumer credit applications

Polarization or “Squeezed Middle” in the Great Recession?: A Comparative European Analysis of the Distribution of Economic Stress

“…In addition, our indicator of economic stress allows a role for objective factors relating to debt burdens and financial obligation while also permitting a role for subjective elements in relation to variable ability to cope with such demands and obligations and differential reference points against which financial pressures are evaluated. This is in line with findings in the literature on the measurement of debt showing that neither purely subjective or objective approaches prove to be adequate (Ferreira, 2000, Finlay, 2006and Betti et al 2007. As we will set out in more detail subsequently, the deprivation items we employ are more likely to capture longer term economic circumstances while the economic stress items focus on recent experiences.…”

“…The models employed for measuring consumer over-indebtedness include objective and subjective versions (Ferreira, 2000;Finlay, 2006;Betti et al, 2007). The former is based on the notion of unsustainable spending behaviour (consumption/income ratio) or unsustainable level of debt (debt/asset ratio) or inability to service debt (debt payment/income ratio).…”

Dynamic affordability assessment: predicting an applicant’s ability to repay over the life of the loan

“…The other characteristics are assumed to be constant, as they typically remain relatively time-invariant. Sex is included since, "after discussion with industry experts", Finlay (2006) came to the conclusion that it may be allowed in affordability models, although its use is debatable. If T denotes the number of months in the training …”

“…Although it was originally formulated at the individual level, most models are developed at the household level, since household surveys are the main source of panel data on consumption. The model built by Finlay (2006) to estimate the expenditure-to-income ratio is also at the household level. Nevertheless, a loan application (including affordability) is usually assessed at the individual level (unless it is a joint application).…”

“…Finlay (2006) proposed a linear regression model to estimate the expenditure-toincome ratio for such purposes and a logistic regression model to estimate probability of overindebtedness, both based on application data and credit reports. However, in the conclusions to his paper, the dynamic nature of income and expenditure is mentioned as a possible argument against the use of those static models.…”