Abstract: Prior research shows that underwriter affiliation inflates analyst recommendations but overlooks the issuer-specific information affiliation provides. Using LLMs and 97,425 post-IPO reports, I compare broker-specific recommendations with rating-blind narrative scores within each report. Affiliation shifts recommendations more than narratives, creating a larger wedge, especially near the IPO day and among lead underwriters. In joint regressions, market reactions follow narratives rather than recommendations, especially for affiliated reports, even after controlling for quantitative signals. When narratives deteriorate, affiliated analysts downgrade less often, particularly after positive offer-price revisions with relatively weak prospectus revisions. Affiliation thus protects formal ratings while preserving value-relevant information.
Presentations: AAA Global Doctoral Student Consortium, AI in Finance Conference, Edinburgh Financial Technology Conference, ENTFIN Conference, Financial Reporting and Business Communication Conference, Frontiers in Finance Conference, Greater China Area Finance Conference, International Risk Management Conference, Next-Gen AI and Autonomous Finance Workshop, RCF-ECGI Corporate Finance and Governance Conference, University of Pittsburgh (Katz), Vietnam International Conference in Finance
Presentations: CMU-Pitt-PSU Finance Conference, FMA, Research Symposium on Finance and Economics, University of Pittsburgh (Katz)
Presentations (*Presented by Co-Authors): AEA CSQIEP Mentoring Conference*, Australasian Finance and Banking Conference*, University of Auckland (Business School)*, University of California, Irvine (Economics)*, University of California, Irvine (Paul Merage)*, University of Pittsburgh (Katz)
Presentations (*Presented by Co-Authors): University of Auckland (Business School)*, University of Pittsburgh (Katz)