Which stress testing approach fits your institution?

A question worth getting right

Stress testing has been part of the regulatory conversation long enough that most institutions have done some version of it. The question is whether what you've done would hold up today.

The bar keeps moving. Examiners who were satisfied with a high-level portfolio view a few years ago are now asking to see the assumptions behind it. Boards that once nodded at a stress testing line item in the audit report are starting to ask what the numbers actually mean. And institutions that were running basic scenario analyses are finding out — sometimes mid-exam — that basic isn't what the room was expecting.

None of that means you need to overhaul everything. But it does mean the question isn't just whether to stress test. It's whether the approach you're using actually fits your situation. Too lightweight and you're handing examiners a reason to dig deeper. Too complex and you've spent a meaningful chunk of your risk budget on something your team can't maintain or explain when it matters.

There are two broad approaches community banks typically land on. This lesson walks through both: what separates them, when each one applies, and what stress actually looks like when it hits a portfolio like yours.

The right approach starts with the right question

Most institutions don't struggle with stress testing because they used the wrong methodology. They struggle because they picked a solution built for someone else's problem — too lightweight to satisfy an examiner who wanted depth, or too elaborate for a team that just needed a clean, documented baseline they could defend without outside help.

Before evaluating any tool or approach, it helps to get honest about what's actually driving the need. Regulatory pressure looks different than routine monitoring. Responding to a specific examiner comment looks different than getting ahead of a CRE concentration you've been watching. A board that wants "something on credit risk" before the next meeting looks different from a credit team that needs account-level documentation for an upcoming exam.

The questions below are designed to cut through that. They take about two minutes and they'll give you a clear starting point — an honest read on which approach is better suited to your situation.

Same purpose. Very different approaches.

Both approaches are built around the same documented, scenario-based regulatory framework — baseline, adverse, severely adverse — and both produce a reliable, comprehensive, and examiner-ready analysis of how your loan portfolio would hold up when conditions deteriorate. Projected losses, earnings impact, capital adequacy. That's the shared goal.

What differs is everything underneath it. An automated, top-down stress test driven by publicly available Call Report data is fast and accessible, until an examiner starts pushing for borrower-level detail and assumptions built directly into the model. A loan-level analysis that digs into the nitty-gritty of individual credits & relationships is exactly right when you need it — and significant overkill when you don't.

The comparison below maps those differences across the dimensions that tend to matter most when institutions are making this call: data requirements, implementation effort, what it produces, and when each approach is the right fit.

The institutions that get this right aren't necessarily doing more. They're likely doing it earlier.

The pattern that holds across institutions that consistently move through exam cycles without significant credit risk findings isn't always a more sophisticated methodology, or a larger risk team... it's timing and discipline.

They proactively run stress analyses before they’ve being asked. They build the documentation before they’re under the gun. They identify the potential concentrations and significant exposures worth monitoring before an examiner does it for them. And when the exam conversation happens, they're presenting analysis they've been consistently maintaining — not analysis they assembled in response to a request.

That posture is available to any institution, regardless of size or internal capacity. It doesn't require the most complex solution on the market. It requires an approach that fits the current internal and external landscape, that the team can actually maintain on a consistent basis, and that produces output clear and concise enough to stand behind in a room where people are asking hard questions.

Choosing the right approach is where it all starts. The rest is execution.