Abstract
This paper illustrates the power of Genetic Programming (GP) with two simple examples: a housing price prediction based on only one variable (house size) and a mortgage default model based on the usual underwriting criteria. The general GP approach is described and the results are compared to the results obtained from regressions analysis. The superiority of GP results appears to be convincing. Less convincing is the Darwinian evolutionary metaphor that underpins the whole concept.