Linear discriminate analysis is a statistical method that is used to classify observations into predefined groups. This is done by estimating a discriminate function using historical data. In the context of finance, this historical data often pertains to borrowers who have either performed well or defaulted. Once a good discriminate function has been estimated, it can be used to assign any new borrower to a specific predefined group (such as 'performing' or 'defaulting') based on the score produced by the function. This score is essentially a measure of how similar the new borrower's characteristics are to those of the borrowers in the predefined groups. Therefore, the statement in choice D accurately describes the application of linear discriminate analysis in finance.

Choice A is incorrect. Linear discriminate analysis, as the name suggests, produces a linear scoring function, not a non-linear one. The scoring function is used to classify observations into predefined groups based on their characteristics.

Choice B is incorrect. While it's true that variables can be selected from a wide range of sources in linear discriminate analysis, they are typically chosen based on their statistical significance and relevance to the classification task at hand. It's not accurate to say that there's little regard for their statistical significance.

Choice C is incorrect. The score produced by the function in linear discriminate analysis does not necessarily have to lie between 0 and 1. The score can take any value depending on the [context].