Regulatory Compliance and Anti-Money Laundering Case Studies
Finance5 Case StudiesUpdated Sep 2026
This set of five cases follows practitioners who inherit a problem someone else created: a ledger that does not tie, a forecast that has stopped meaning anything, an alert queue that keeps refilling, a loan file that looks sound until someone reads it closely, and an audit report with more findings than anyone can fix at once. Each organization is fictional, but the decisions follow the ones working teams actually face. The focus stays on the financial records involved, the risk each gap creates, the control evidence that proves or disproves a fix, and the thresholds that tell a team when to stop investigating and act. Read them as a sequence of judgments rather than a checklist.
Case Study 1
Client trust account variance at a remittance firm
Quillfeather Remittance Ltd., a licensed money transmitter handling about 42,000 consumer transfers a month across eleven states
The situation
Quillfeather's operations group closes its settlement ledger each business day against the bank statements for its pooled customer trust account. In the third week of the month, the ledger showed a balance 38,450 dollars below the bank. The same kind of gap had appeared twice in the prior quarter, each time smaller and each time cleared by a manual journal posted near month end. Nobody had written down why those journals were correct. The reconciliation was prepared by one analyst, Dana, and signed by a supervisor who did not open the supporting detail. The pooled balance was roughly 9,200,000 dollars, so the variance was small in proportion but well above the 500 dollar daily tolerance.
The challenge
Marcus, the finance controller, had to determine whether the gap was timing, a posting error, a duplicate entry, or something more serious, such as a transfer that had left the account without authorization. Two systems fed the ledger, the payment processor export and the core transfer system, and their transaction identifiers did not share a format. Matching had been done by amount and date, which produced false matches and hid open items. He needed a credible explanation before the monthly certification to the chief financial officer.
Analysis
The first step is to establish what kind of record each source represents. The bank statement is external evidence and sets the reference point. The processor export is a system report that can be regenerated, and the core ledger is a book of account that may contain manual entries. A variance between an external record and an internal book is a different problem from a variance between two internal reports, because the first may mean cash has moved while the second usually points to a mapping error. The team assumed the second and spent four days searching for posting errors inside the ledger. When they finally pulled raw bank detail, they found two returned transfers that had been reversed in the processor but never reversed in the core system. The exposure was larger than the 38,450 dollars on the sheet, since every failed reversal overstated customer balances and understated the amount owed back to the bank.
Teams grounded in Regulatory Compliance and Anti-Money Laundering usually split a reconciliation into two questions: whether the books are right, and whether the controls would have caught the problem if they were wrong. Here the control evidence was weak. The sign-off existed, but nothing recorded which items the reviewer tested, and the manual journals carried no attached support. The decision thresholds were also poorly placed. A 500 dollar tolerance was reasonable for timing differences, yet nothing required escalation when the same unexplained item returned in later periods. Recurrence is the signal that matters. One unexplained item is a question, while three in a single quarter is evidence of a design weakness. The team agreed to escalate any unmatched item over 5,000 dollars, or any item unresolved after two business cycles, and to require reviewer-selected sampling with documented tick marks.
Actions taken
- Rebuild the matching key so every processor transaction carries the core system identifier, and reject any match made on amount and date alone without a second field agreeing.
- Trace the two returned transfers to their reversal records, confirm customer balances on both sides, and post correcting entries with support attached and approved by the controller.
- Replace the signed summary sheet with a reviewer checklist that lists each reconciling item, the evidence examined, and the tolerance applied, initialed item by item.
- Set escalation rules so that any unmatched item above 5,000 dollars, or any item that recurs across two cycles, reaches the chief financial officer within one business day.
Results
The team cleared the variance completely within the month. The two reversals had failed because a status code from an older integration was never mapped, so the fix went in at the interface rather than in the ledger. Over the following cycles the average number of unmatched items per close fell from 14 to 3, and no item stayed open past two cycles. The new checklist caught one manual journal that lacked support, which was reversed before certification. The controller noted that the work took longer than the old sign-off, but the gain was that the team could show an examiner exactly which items were tested, why each tolerance applied, and who approved each correction.
Questions for discussion
- Which of the two reconciliation questions, whether the books are right or whether the controls would have caught an error, would you examine first in a team with no prior history of variances, and why?
- How would you set the escalation threshold if the bank statement itself arrived late on some days, and what evidence would you require before accepting a timing explanation?
- What would convince you that the matching key change had actually solved the problem rather than just shifted the unmatched items into a different report?
Case Study 2
Cash shortfall threatens a revolving credit covenant
Tallow Ridge Freight Holdings, a regional trucking group with four terminals and about 1,300 employees
The situation
Tallow Ridge planned for operating cash of 4.1 million dollars over the twelve months under review. Fuel costs rose faster than the plan assumed, and a large retail customer began paying in roughly 58 days instead of its usual 35. Operating cash for the trailing period came in 22 percent under plan. The revolving credit agreement requires minimum liquidity of 2.5 million dollars, and on the current trajectory headroom was about 300,000 dollars. The forecast had been updated monthly, but nobody kept a record of which assumptions had changed between versions.
The challenge
Priya, the treasurer, had to present a revised forecast to the lender group and the board within a short window. She needed scenarios that were defensible rather than optimistic, tests of the assumptions most likely to be wrong, and a recommendation that avoided both overreaction and complacency. The board wanted to know whether covenant compliance was at risk and what the company could do before that became a question for the lenders.
Analysis
Start with the records behind the forecast. The plan model took its fuel price input from a single index with a one-month lag, and it assumed the company would recover 90 percent of fuel cost through surcharges. Billing data showed actual recovery at 71 percent, because the surcharge reset lagged a full billing cycle behind price changes. The collection assumption was also a single average across all customers, which hid the fact that three accounts made up most of the delay. Modeling those accounts individually changed the timing of the cash shortfall by about six weeks. Exposure was concentrated in the fourth month of the downside case, when projected liquidity fell below the covenant floor even before any new loss of volume.
Practitioners working in Regulatory Compliance and Anti-Money Laundering tend to treat a forecast miss as a control question before it becomes a forecasting question. In this case the version history was missing, so the board could not see which changes had moved the numbers, and the assumption owners had not signed off on their inputs. The team set decision thresholds in advance. A lender conversation would begin if projected headroom fell below 400,000 dollars, and a review of draws would begin below 200,000 dollars. Three scenarios were built: a base case with surcharge recovery restored and delayed accounts paying within fifteen days, a downside with the lag persisting, and a severe case that added an 8 percent volume decline. Ending liquidity was about 2.9 million, 2.5 million, and 2.2 million dollars respectively, so only the severe case breached.
- 1Restate fuel recoveryApply observed recovery of 71 percent, not the planned 90 percent.
- 2Separate delayed accountsModel the three largest late payers individually instead of using the average curve.
- 3Run three scenariosBase, downside, and severe cases, with liquidity tested every month.
- 4Set trigger pointsOpen lender talks below 400,000 dollars headroom; review draws below 200,000.
- 5Stage the responsePause discretionary capital, tighten collections, and hold the revolver in reserve.
Actions taken
- Rebuild the forecast with the fuel surcharge reset lag modeled explicitly, using billing cycle data from the prior year rather than the planned recovery rate.
- Model the three largest late-paying customers individually, with payment dates taken from receivable aging, and report the effect on monthly liquidity separately.
- Give each assumption a named owner who signs the input sheet, and store every forecast version with a note listing what changed and why.
- Open discussions with the lender group before headroom reaches the trigger, presenting the severe case and the staged response as the company's own plan.
Results
The revised forecast showed that covenant compliance was not breached in the base or downside cases, but the severe case would breach in the fourth month without action. The lenders accepted a short waiver conversation rather than a formal default process, largely because the company had raised the issue before headroom crossed the trigger. Capital spending was deferred for two quarters, and collections on the three large accounts improved after the company agreed to revised terms. Liquidity stabilized above the trigger, and the forecast file now carries a change log that the board reviews each month.
Questions for discussion
- Which assumption in this forecast would you have tested first, and what evidence would have made you distrust the single average collection curve?
- How would you justify the 400,000 dollar trigger to a lender who thinks it is too conservative, and what would change your view of the number?
- If the severe case had breached before the company approached its lenders, what would you have changed about the order of the response steps?
Case Study 3
Alert review on a merchant account with sudden volume
Corvane Payments Inc., a payments company serving small exporters with about 3,800 registered merchant accounts
The situation
A merchant account that had processed modest and steady volume for a long period began receiving inbound wires from eleven different originators in one week. Several amounts sat just under the internal review threshold of 9,000 dollars. The account's expected activity profile described a single domestic customer base and a monthly volume near 20,000 dollars. The alert was generated by the monitoring system, and the first-line analyst, Odell, had seen similar patterns that turned out to be seasonal. This time the originators were not previously linked to the merchant in any onboarding record.
The challenge
Odell had to decide whether the activity warranted a suspicious activity report, what documentation to collect, whether to restrict the account while the review was open, and how to document the reasoning either way. The compliance officer, Teresa, wanted a defensible file rather than a fast answer, and the internal procedure set a review window that would not wait for a complete investigation.
Analysis
The records that matter here are the wire detail, the onboarding file, the expected activity profile, and the disposition history of earlier alerts on the same account. The wire detail showed the eleven originators, their banks, and the purpose text, which was generic in nine of eleven cases. The onboarding file described a sole proprietor exporting machine parts, with no mention of the foreign buyers now sending funds. Exposure had two parts. The company faced possible loss if the inbound funds were later reversed, and it faced a regulatory exposure if it failed to escalate activity that its own procedures described as unusual. The account had been reviewed twice in the prior year without escalation, and the earlier notes did not explain why those alerts were closed, so the file was weak on the question an examiner would ask first.
Analysts trained through Regulatory Compliance and Anti-Money Laundering generally separate what the account's activity shows from what the file explains. Here the explanation was thin. The merchant's invoices could support some of the new buyers, but three of the originators matched no invoice at all. The threshold question was whether the unexplained portion crossed the internal escalation line, which the procedure set at any unexplained inbound amount above 15,000 dollars in a rolling thirty day period. It did. Restricting outbound transfers was a separate decision, and the team chose it on the basis that funds could still leave the account quickly once released. Restriction was documented as a temporary control with an explicit owner and a review point, not as a finding about the merchant's intent.
- Earlier periodSteady domestic volume near 20,000 dollars a month, with clean alert history.
- Start of the weekEleven inbound wires from unlinked originators arrive in succession.
- Second daySeveral amounts sit just under the 9,000 dollar internal review line.
- Fourth dayAnalyst opens review, restricts outbound transfers, and requests invoices.
- End of review windowUnexplained inflows documented; escalation decision recorded with reasoning.
Actions taken
- Restrict outbound transfers on the account while the review is open, record the owner and the review point, and notify the merchant only as permitted by the internal procedure.
- Collect invoices, shipping records, and originator bank details for each inbound wire, and mark which items are supported, partly supported, or unsupported in the file.
- Compare the unexplained inflow against the 15,000 dollar rolling threshold, and record the calculation and the reviewer who confirmed it in the case notes.
- Decide on escalation under the written procedure and document the reasoning whether the answer is to file or not, so that the decision can be reconstructed later.
Results
The unsupported inflows exceeded the escalation threshold, and the compliance officer approved a report to be filed within the review window. The merchant produced invoices for six of the eleven originators, but the remaining five could not be tied to any contract. The company exited the relationship after the review closed, with restrictions held until outbound funds were reconciled. The earlier alert closures were also reopened, and the team added a rule requiring a written rationale for any closure on an account with new counterparties. The file now shows the reasoning from start to finish, which was the main point of the exercise.
Questions for discussion
- What would you need to see in the earlier alert closures before you were willing to rely on them as precedent for closing a similar alert today?
- How would you weigh a partial explanation from invoices against the unsupported portion when the threshold is set on the unexplained amount?
- When restricting outbound transfers, what notice to the customer would you consider appropriate, and who should make that decision under a written procedure?
Case Study 4
Term loan request from a fastener manufacturer
Brackwell Fastener Works, a family-owned manufacturer of industrial fasteners with two plants and about 240 employees
The situation
Brackwell asked the commercial credit committee for a 6.5 million dollar seven-year term loan to fund a third press line and a warehouse extension. Revenue had risen steadily over three years to 36.8 million dollars, and reported EBITDA margin sat near 13 percent. The largest customer, an automotive supplier, accounted for 44 percent of revenue, and its supply agreement came up for renewal within the next twelve months. Existing debt was a 2.1 million dollar equipment line. The owner, Teodor, asked for fast approval because a press order carried a delivery deadline that the company could not meet without new capacity.
The challenge
Imani, the credit analyst, had to decide whether projected cash flow covered debt service under realistic downside conditions, whether the customer concentration was controlled or merely priced, and whether the financial statements were reliable enough to support the request. The credit policy set a minimum debt service coverage ratio of 1.25 times. Her recommendation had to state the conditions under which the committee would approve, reduce, or decline the request.
Analysis
The financial evidence came from reviewed rather than audited statements, three years of tax returns, the receivables aging, and the customer contract. The reported EBITDA included a one-time gain of about 0.9 million dollars from the sale of land, so the underlying figure was lower than the headline. On the reported basis, cash flow available for debt service covered projected annual payments on the new loan by about 1.4 times. On an adjusted basis that removed the gain and normalized working capital, coverage fell to about 1.1 times, which was below policy. The receivables aging also showed the largest customer paying in roughly 52 days against contract terms of 30, which tightened liquidity further. The collateral package included the new press at an orderly value, but the forced sale value was much lower and the presses had a useful life longer than the loan tenor.
Credit officers working in Regulatory Compliance and Anti-Money Laundering usually test concentration as a scenario rather than a ratio. Losing the largest customer would reduce revenue by roughly 16 million dollars, and coverage on the adjusted basis would fall to about 0.6 times. That result rules out the full request without a structure that handles the renewal risk. The committee's decision thresholds were set accordingly: coverage of at least 1.25 times on the adjusted base, coverage above 1.0 times in the customer-loss case, and a loan amount that can be serviced from the reduced cash flow. The recommended structure reduced the principal to 4.8 million dollars, tied first drawdown to a signed renewal or extension of the supply agreement, required a semiannual coverage test with a cash sweep if it fell below 1.25 times, and included a personal guarantee from the owner.
Actions taken
- Rebuild cash flow available for debt service on the adjusted basis, removing the land gain and normalizing receivables to the contract terms, and show both bases side by side in the memo.
- Run the largest-customer loss scenario and a milder case with a 15 percent price concession at renewal, recording coverage for each against the policy minimum.
- Reduce the principal to an amount the downside case can service, and tie first drawdown to a signed renewal or extension of the supply agreement.
- Set a semiannual coverage covenant at 1.25 times with a cash sweep if it is breached, and obtain a personal guarantee with a review of the owner's other obligations.
Results
The committee approved 4.8 million dollars, with first drawdown held until the supply agreement was extended for a further three years. The press line was funded in part from the company's own cash and in part from the approved loan, which slowed the delivery schedule slightly but kept the order. The first semiannual test passed comfortably, and the customer renewal closed on terms with a modest price concession that the base case had already absorbed. The committee noted that the original request would have passed the headline test and failed the adjusted one, which is why the adjusted basis now appears first in every memo.
Questions for discussion
- Which of the adjustments to reported EBITDA would you challenge most strongly with the owner, and what documents would you request to support each one?
- If the customer had offered a renewal without a price concession but with a shorter term, how would that change your coverage scenarios and the structure you recommended?
- Would you accept the personal guarantee as a substitute for a lower principal amount, and what would you need to know about the owner's other obligations to decide?
Case Study 5
Internal audit findings at a consumer lending company
Pembrook Lending Corp., a consumer and small commercial lender with a servicing book of about 58,000 loans
The situation
Internal audit completed a six-week review of onboarding, transaction monitoring, escrow reconciliation, and system access. The team recorded eleven findings, four of which the audit committee needed to see at the next meeting. The customer identity population contained 1,140 records whose documentation had passed its refresh date. Escrow balances were reconciled an average of nine business days after month end, and an unreconciled difference of 310,000 dollars had carried across two cycles. Seventeen analysts held payment release rights outside their assigned roles. The monitoring rule thresholds had no written rationale or calibration record.
The challenge
Sofia, the chief risk officer, had to rank the findings by materiality, assign accountable owners who would actually act, and define closure tests that the audit committee would accept as evidence rather than assurance. Some findings overlapped, since the access issue affected the escrow process, and at least two owners disputed the severity of their findings. Closing a finding on the strength of a status update was not acceptable to the committee.
Analysis
Ranking combined three factors: the amount of money or number of records exposed, the likelihood implied by the test failure rate, and the sensitivity of the area to regulatory scrutiny. The identity refresh gap was the largest by record count and carried the highest regulatory sensitivity. In the sample of 60 files, 13 had failed the refresh test, which projected to a population of roughly 250 records with similar issues. The escrow difference was smaller in dollars but was the most immediate exposure, since unreconciled balances could mask errors in customer funds. Access rights ranked lower on exposure, but they undermined the reliability of the other controls because the same analysts could both post and release. The monitoring calibration finding ranked fourth, since the rules were operating, but the company could not show why their thresholds had been chosen.
Audit and risk teams applying Regulatory Compliance and Anti-Money Laundering to remediation tend to insist that every closure rests on re-performance rather than assertion. The owners were named by first name and role, so the issue log showed who answered for each action. Closure tests were written before work began. For identity records, a new sample of 40 refreshed files would need to pass with no exceptions. For escrow, three consecutive months would need to reconcile within two business days, with reviewer evidence attached to each. For access, the role matrix would need to match the system, and the next quarterly recertification would need zero unexplained exceptions. For monitoring, a documented back test on the last full cycle of transactions would need compliance officer approval. A finding closed only when its test passed in two consecutive cycles, which prevented the team from declaring success after a single clean month.
- Rank findings Weigh money exposed, records affected, test failure rate, and regulatory sensitivity.
- Assign owners Name one accountable person by role for each finding in the issue log.
- Remediate controls Fix the root cause, not only the sampled records, and attach the change record.
- Test closure Re-perform against a fresh sample and pass in two consecutive cycles before closing.
Actions taken
- Assign each finding one accountable owner by name and role, record the owner in the issue log, and agree a completion window with the audit committee chair before work starts.
- Refresh identity documentation for the 1,140 records in order of loan balance, starting with the largest, and track completion weekly against the population file with exceptions listed.
- Document the monitoring rule thresholds with their rationale, historical alert volumes, and a back test on the last full cycle, then obtain compliance officer approval before any rule is changed.
- Remove payment release rights from the 17 analysts outside their role matrix, and run a quarterly access recertification that lists every remaining exception with its approver.
Results
Two findings closed after passing their tests in two consecutive cycles. The identity refresh reached 81 percent completion, with the remaining files scheduled in the order of balance. Escrow reconciliation moved to an average of three business days, and the 310,000 dollar difference was traced to a mapping error that had been carried forward and corrected. All out-of-role release rights were removed, and the first access recertification produced no unexplained exceptions. The monitoring documentation was approved, but the back test showed six rules with alert volumes too low to justify their thresholds, and those remain open with a revised calibration plan.
Questions for discussion
- How would you rank a finding involving a small number of records that carries a high regulatory sensitivity against a finding with a large dollar exposure but a low test failure rate?
- What evidence would convince you that an owner who disputes a finding's severity has a point, and how would you record that disagreement in the issue log?
- If the escrow reconciliation passed three months but the access recertification still showed exceptions, would you close the escrow finding, and how would you explain the dependency to the committee?
Learning Resources
Finance CASE STUDIES
China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.