Credit Risk Leaders Flag Valuation Uncertainties at New York Meeting
Senior credit risk professionals meeting in New York in September 2026 flagged severe valuation uncertainties and limited look-through capabilities across private credit markets, according to a Risk.net Credit Risk Leaders’ Network report sponsored by KPMG.
The closed-door gathering exposed widespread vulnerabilities in how institutional lenders monitor growing exposures. Participants warned that overlapping investments create hidden concentrations that are nearly impossible to detect before broader market shifts occur.
More frequent asset pricing alone offers little protection without independent oversight of the underlying collateral, the report found. Financial institutions currently lack sufficient visibility across complex operational chains linking borrowers, investment funds, sponsors, and insurers.
As artificial intelligence embeds itself deeper into credit risk workflows for document analysis and ongoing portfolio tracking, banks are colliding with persistent data quality issues and internal governance roadblocks.
“Independent challenge and experienced credit professionals remain essential as AI takes on more analytical tasks,” the report said.
Proponents argue that granular borrower data could allow institutions to catch credit deterioration far earlier than traditional metrics permit. Yet current automated deployment remains severely limited by a lack of baseline visibility.
Existing risk monitoring practices continue to struggle with capturing the true scale of interconnected liabilities spanning the private debt sector. Comprehensive stress testing demands visibility across multiple overlapping layers of sponsorship and insurance coverage.
Lending institutions and asset managers are actively testing whether automated oversight tools can effectively substitute for traditional underwriting judgment during periods of market stress. The findings indicate that while machine learning successfully streamlines document review, seasoned human expertise remains strictly necessary to challenge baseline assumptions.