We study how competition between asymmetrically informed banks, one specialized and one non-specialized, affects loan prices. Both banks possess “general” signals regarding the borrower’s quality, which they use to screen loans. The specialized bank also has access to a “specialized” signal on which it bases its loan pricing. This private-information-based pricing makes the specialized bank bid more aggressively, mitigating the informational rent effect that gives...

Recent technological change in lending converts previously subjective assessments into structured, easily accessible data. We study this transformation in a credit market competition model that distinguishes between information span (breadth) and signal precision (quality). Borrower quality depends on multidimensional fundamentals, assessed through hard or soft signals. Two banks observe private hard signals, but only the specialized bank receives a soft signal. Expanding the span of...

While large language models (LLMs) perform well on well-defined tasks, effective
prompts are challenging when tasks depend on users’ soft traits and latent prefer-
ences. We formalize this friction by introducing preference uncertainty—capturing
soft information—into a cheap talk framework (Crawford and Sobel, 1982) and model soft
information communication with AI as the investor’s optimal stopping problem with
Brownian information flow. We propose a novel empirical methodology to test theoret-
ical predictions...