A loan can look flawless on paper and still unravel fast. Finance professional Haleigh Truong opens her recent ProSight piece with exactly that experience: a borrower who met every traditional benchmark—repayment history, income trend, documentation—yet defaulted within a year. The lesson is increasingly familiar across banking: traditional credit indicators alone no longer capture how risk actually behaves.
Here are some practical takeaways from Truong’s article for banks navigating a credit environment where borrower stability can change rapidly:
Stable industries aren’t as stable as they used to be. Professions and sectors long considered low risk can change quickly when labor-market conditions, regulation, or policy shift. Truong points to healthcare as a clear example. According to Nurse.com’s 2024 Salary and Work-Life Report, 23% of nurses are considering leaving the profession, while nearly half report mental-health impacts. The National Council of State Boards of Nursing adds that more than 138,000 nurses have left the workforce since 2022. Those pressures reshape how lenders evaluate professions once viewed as exceptionally stable.
Logistics tells a similar story. After a pandemic-era surge, freight demand dropped sharply amid rising costs and excess capacity. Truong notes that even Standard Forwarding Freight, a 91-year-old carrier, paused operations in December 2025, creating immediate ripple effects driven by uncertainty.
Takeaway: Industry stability is a moment in time, not a permanent trait. Underwriting needs to reflect current labor and regulatory conditions, not historical reputation.
Documentation problems are early warning signs. Missing tax returns, inconsistent bank statements, or narratives that don’t align aren’t just clerical issues. Truong argues they often signal deeper organizational weakness. The same applies internally: inconsistent credit policy application or reliance on outdated models can create both financial and legal exposure.
Takeaway: Strong documentation standards and consistent internal processes aren’t bureaucracy—they’re protection in a heavily regulated environment.
Models lag reality—judgment closes the gap. Risk models are inherently backward-looking. Truong highlights food-service lending to show how assumptions can miss real-world resilience. During the pandemic, food trucks proved far more resilient than brick-and-mortar restaurants, yet many lenders treated them the same because policies grouped them under a single industry label.
She recommends more precise segmentation, such as using NAICS codes to differentiate subsectors with materially different cost structures and regulatory exposure.
Takeaway: Judgment informed by current market and regulatory conditions is a strategic asset—not a workaround.
Compliance risk is credit risk. State-level lending laws, licensing requirements, labor rules, and insurance costs can dramatically alter borrower viability. Truong stresses that two identical applicants in different states may present entirely different compliance exposures, and that ethical or compliance shortcuts often surface later as financial stress.
Bottom line: Algorithms are useful, but they can’t see regulatory pressure, workforce instability, or human behavior. Institutions that pair data with compliance-aware judgment will be better positioned to recognize emerging risk earlier.
