Alex P. Günsberg

Finding. Most rejections by Finnish consumer lenders are protective — the applicants who slip through and get the loan default at much higher rates (+55.6pp at 18 months, p = 0.006).

Why it matters. The standard algorithmic-fairness remedy — stripping local repayment signals — would force lenders to fund bad loans and ration viable ones. My results overturn it.

How. IV from weekday processing delays among already-offered applicants; Finnish marketplace data linked to credit-registry defaults.

Finding. Documents four facts about consumer behavior in near-zero-search-cost online loan marketplaces: (1) substantial within-applicant dispersion in offered terms makes search valuable; (2) marketplace nudges mitigate choice complexity; (3) applicants search in ways consistent with their creditworthiness; (4) dynamic adverse selection constrains search as lenders penalize repeat applicants.

Why it matters. Surfaces the trade-off in fintech lending between informational gains from search and the reputational and cognitive costs that limit it — relevant for marketplace design and for understanding adverse selection in low-search-cost environments.

How. Administrative data on 730,000 applications, 750,000 offers, and 200,000 individuals from a Finnish online loan marketplace, merged with credit-registry records.