Intangibles and credit decisions: results from an experiment