Interview Quiz: Risk, Compliance and Audit

Question 1

Losses are lower than the model predicted. Is that good news? A panel for a credit risk seat puts a portfolio in front of you instead of a file. For the second year running, losses in one lending segment have come in well under what the bank's model expected, and the business is asking for more appetite in that segment on the strength of it. The question comes from the head of the portfolio team, and it is not rhetorical.

Accepted Answer

B. Not yet. Lower losses are a fact; what they mean is a question — a kind cycle, loans too young to have failed, or a model wrong in our favour. I'd start with what we have been lending to lately.

Answer

A. Yes. The model was calibrated conservatively, the portfolio is outperforming it, and the business has earned more appetite. I'd raise the segment limit and recalibrate the model to recent experience.

Answer

C. It's good news with caveats. I'd benchmark the segment against industry loss rates and peer banks first, and let that decide how much more appetite it has earned.

Answer

D. It's bad news: a model that misses in either direction is a broken model, and no basis for more appetite. I'd have it withdrawn and rebuilt before any decision is made on that segment, whatever that costs the business in the meantime.

Question 2

The business wants to launch on Friday and you are not comfortable. What do you do? A product team has a new lending product ready for Friday, and the marketing is booked. In your review you found that the way the product's cost is shown to customers could leave some of them believing it is cheaper than it is. The product owner says every competitor shows it the same way, and that the wording is legal. It is Wednesday.

Accepted Answer

A. Say what is wrong and what would fix it — show the cost the way a customer would compare it — and offer them Friday with the fixed wording. Escalate only if they refuse.

Answer

B. Stop the launch. Until the disclosure is rewritten and reviewed again, the product does not go out, whatever marketing has booked and whatever the competitors do.

Answer

C. Escalate it to the risk committee with both views written up, and let the committee decide whether Friday stands. The call is above my pay grade and the record should show it was made properly.

Answer

D. If it is legal and the market does it, sign it off, and record the concern on the file so the bank's position is documented if anyone asks later.

Question 3

This customer's transactions do not match its stated business. Where do you start? An alert has come up on a business customer that describes itself as a small wholesaler of building supplies. Over three months, most of the money coming into its account has arrived in round amounts from a handful of individuals, and most of it has left again within days for a company overseas. There are almost no payments to anyone who looks like a supplier. You have the account history and the customer's onboarding file.

Accepted Answer

D. Start with the business it says it is. A supplies wholesaler buys stock, pays suppliers, is paid on a rhythm. Read the account for that shape and write down each place it does not fit.

Answer

A. Report it now. The pattern matches a known typology, round amounts in and the money leaving the country within days, and the cost of being late is higher for the bank than the cost of being wrong.

Answer

B. Contact the customer, ask them to explain the payments, and request invoices for the overseas transfers. Their answer is the fastest way to know whether this is ordinary trade.

Answer

C. Screen the individuals and the overseas company against sanctions and watch lists. If nothing matches, clear the alert and note the check that was done.

Question 4

The business disagrees with your finding. What now? You have audited how a lending team watches its borrowers after the money is out. In a sample of files, annual reviews were signed off without the financial information the bank's own standard requires. The head of the team says the finding is unfair: the borrowers are long-standing, none of them has defaulted, and the information was “known” even if it was not on the file. They want the finding taken out before the report goes to the audit committee.

Accepted Answer

C. Go back to the evidence: the standard asks for the information the files do not have. What are they seeing that the sample missed? If they can show it, the finding changes; if not, it stands with their view.

Answer

A. Take it out, or downgrade it. None of the borrowers defaulted, so there is no loss, and the relationship with the business matters for the audits still to come.

Answer

B. Keep it exactly as written and send the report. The team's disagreement goes in management's response; that is what the response is for, and a finding that moves under pressure is worth nothing.

Answer

D. Keep the rating, soften the wording so it reads less critically of the team, and agree an action date they can live with. The rating is what the committee tracks.

Question 5

How would you tell whether a control is real or decorative? A lending operations team shows you its control for making sure the security on every new loan is registered before the money goes out. There is a policy, a checklist with a box for it, and a monthly sign-off by the team leader. The last internal review rated the control effective. You have a day with the team.

Accepted Answer

A. Take a few recent loans, one of them rushed, and trace each to the register, not the checklist. Then ask what happens when the team leader is away, and what the person ticking the box looks at.

Answer

B. Check the design. The policy is current, the checklist covers the risk, the sign-off is evidenced every month. If those hold and the last review was clean, the control is real.

Answer

C. Test the sign-offs over the past year for completeness. A control that is performed and evidenced every month, without a gap, is a control that is working.

Answer

D. Interview the team about their workload. Controls fail when people are stretched, so the real risk here is capacity rather than the design of the check itself.

Interview Quiz: Risk, Compliance and Audit

// FOUR BLOCKS, FOUR QUIZZES