Testing Your Leadership's Read on Regulators Before You Spend

Start by writing down what you think you know

The first failure point is not being wrong about regulators. It is not knowing precisely what your leadership believes. Assumptions live in slide decks, board minutes, and offhand comments from the CRO. Before you validate anything, extract them.

Run a short exercise with the executives sponsoring the investment. Ask each to write, in one sentence each, the specific supervisory expectations driving the spend. You will get variance. That variance is the finding. If the CEO, CRO and General Counsel describe the regulator's position differently, you do not yet have a shared basis for a capital decision.

Good looks like a list of 10 to 15 concrete propositions. Not "the FCA expects strong governance," but "the FCA expects a named accountable executive under SMCR for our crypto onboarding controls, with quarterly attestation to the board." Testable claims. Falsifiable ones.

Separate the three types of assumption

Not all assumptions carry equal risk. Sort them into three categories.

Interpretive assumptions: what a rule or piece of guidance means in practice. These are the easiest to test through written channels.

Priority assumptions: how much the regulator cares about this issue relative to others. These are the ones most often wrong. Firms routinely over-invest in areas the supervisor considers settled and under-invest in areas quietly rising up the agenda.

Tolerance assumptions: what the regulator will accept as adequate. This is where money is wasted. Leadership often assumes a higher bar than the supervisor requires, then gold-plates the response.

Most validation effort should go into priority and tolerance assumptions. They drive the size of the cheque.

Choose your validation channels carefully

You have four practical routes. Each has a cost.

Direct supervisory dialogue

A structured conversation with your named supervisor, framed as horizon-scanning rather than confirmation-seeking. Be careful. Asking a supervisor to validate your plan invites them to raise the bar. Frame questions around understanding, not endorsement. "We are working through how firms are responding to X. Are there patterns you are seeing that concern you?" is very different from "Does our approach meet your expectations?"

Peer intelligence

Structured conversations with counterparts at three to five comparable firms. Not to copy, but to triangulate. If four peers heard something in their own supervisory meetings that your team did not, that is a signal. Trade body working groups are useful; informal CRO-to-CRO calls more so.

Ex-regulator advisers

Useful for interpretive questions, less useful for current priority questions. Someone who left the PRA eighteen months ago has stale information on what supervisors are focused on this quarter. Use them for context, not current intelligence.

Written supervisory correspondence

Underused. A carefully drafted letter seeking clarification on a specific point creates a written record and forces the supervisor to be precise. Reserve this for high-stakes interpretive questions where ambiguity would be expensive.

Test the assumptions that would change the decision

Do not validate everything. Validate the assumptions where being wrong would change the size, shape or timing of the investment. If a 30 percent reduction in scope is possible under one reading of supervisory expectations, that is the assumption to test first.

Build a small matrix: assumption, what changes if wrong, cost of testing, chosen channel. Sequence from cheapest to most expensive.

Watch for the confirmation trap

The most common failure is teams designing validation exercises that can only confirm the existing view. Signs you are in the trap: only briefing consultants who already know your position, only speaking to peers who made similar investments, only asking supervisors closed questions.

Good validation actively hunts for disconfirming evidence. Assign someone to argue the opposite case. Ask peers what they decided not to do and why.

Convert findings into a decision, not a report

End with a one-page reconciliation: original assumption, evidence gathered, revised assumption, implication for the investment. Take it to the sponsoring executive before it goes to the board. If the reconciliation changes the number by more than 15 percent in either direction, the board needs to see the working, not just the conclusion.

Your next step

Before your next investment committee, get the top three assumptions driving the spend written down in one sentence each by the three executives sponsoring it. If they do not match, you have your starting point.