A mid-sized lender looked at its dashboard and saw what most executives would call “fine.” Losses were on plan. Delinquency had ticked up but not alarmingly. Portfolio growth was steady enough to avoid tough conversations.
Then they cut the data by segment.
In one product, a single FICO band in a handful of geographies was suddenly responsible for an outsized share of new losses. Early-stage roll rates had quietly doubled. The portfolio-level averages had been telling a story that the segment-level data flatly contradicted.
That’s the real underwriting challenge right now. The biggest risk isn’t a visible blow-up. It’s slow, uneven erosion that hides inside metrics that still look “acceptable.”
On paper, many portfolios are holding up. Growth patterns have shifted. Loss profiles look different than they did two years ago. Regulators have a wide lens, not a narrow one aimed at your underwriting.
But under the surface, the mix has shifted. Customer behavior, acquisition channels, and product strategies have all evolved faster than most institutions’ policy cycles, data pipelines, and analytics routines. You’re being asked to defend returns and capital deployment in a world where small mispricings compound fast.
When underwriting stays “good enough” while conditions change around it, you end up taking risk you didn’t intend and leaving safe growth on the table.
Here are the three symptoms we see again and again.
1. Portfolio Averages Look Fine, but Dispersion Is Widening
Your overall loss and delinquency metrics stay within appetite. The board deck looks clean. But certain FICO bands, geographies, or products are quietly deteriorating and consuming more capital than anyone realized.
The problem with averages is that they’re designed to smooth things out. That’s exactly what you don’t want when risk is concentrating in specific pockets. By the time the average moves, the segment has already been bleeding for quarters.
2. Manual Overrides Are Becoming the Unofficial Policy
Your underwriters increasingly override scores and rules to “save” deals. Sometimes that instinct is right. But when no one is measuring whether those overrides actually perform, you’ve effectively replaced your model with a collection of individual judgment calls that nobody is tracking. The Federal Reserve has supervisory guidance on model risk management that addresses this issue.
Overrides aren’t the problem. Unmanaged overrides are. If you can’t tell me whether overridden accounts perform better or worse than model-only decisions, you have a blind spot sitting in the middle of your credit process.
3. Policy Refresh Cycles Lag the Environment
Cutoffs, pricing tiers, and exposure limits that worked in 2022–2023 haven’t been recalibrated to current performance cohorts and market dynamics. You’re effectively underwriting to a world that has already moved on.
This one is the most common and the least dramatic, which is exactly what makes it dangerous. Nobody sounds an alarm because the policies haven’t technically “failed.” They’ve just drifted out of alignment with the environment. And every month they stay static, the gap between your assumptions and reality widens a little more.
What This Means at the Executive Level
The impact of all three symptoms lands in the same place: diluted risk-adjusted returns, less reliable forecasts, and less confidence that your portfolio is aligned with today’s reality. You’re making capital allocation decisions based on a picture that’s slightly (or significantly) out of date.
The good news is that none of these problems require you to tear down and rebuild your models. They require better visibility into what’s actually happening at the segment level and a more intentional approach to tuning decisions based on current data.
We wrote a white paper that lays out five practical strategies for getting underwriting back on the front foot, each illustrated with real client outcomes and specific results. If the symptoms above sound familiar, it’s worth the read.
