For the variable
AGE, we find a monotonically and significantly increasing value
of business credit information, which is consistent with H2b. As explained
earlier, one possible explanation for this result is that there
are more cross-sectional and time-series data available (i.e., a higher
number of business partners and a higher number of repeated
interactions with the same business partner) for older firms than
for younger firms, allowing the older ones to establish a more reliable
payment history. For SALES, our proxy for firm size, we find an
improvement of default prediction accuracy in all terciles but no
significant differences between terciles. Thus, there is no evidence
supporting H2c. This is most likely due to the fact that we explicitly
focus on SMEs, which makes our sample deliberately homogeneous
in firm size. The positive effect of FIRMS_PER_EMPLOYEE
might be explained with the learning effect associated with soft
information production as suggested by H3b. However, using the
aggregate business credit information does not allow for a detailed
examination of the effects due to hard or soft information. We will
further analyze this issue in Section 4.4.