Enterprise and Financial Risk Management Case Studies
Finance5 Case StudiesUpdated Sep 2026
Financial risk management tends to be learned through the moment a number refuses to agree with another number. The cases below follow one practitioner's work across five fictional engagements, each starting from a different kind of trouble: a ledger that will not close, a forecast that has drifted, a flagged transaction, a credit request, and an audit with more findings than expected. Each case records the evidence that was available, the thresholds that governed the decision, and what changed afterward. The emphasis throughout is on how records, exposure, controls, and thresholds interact in practice.
Case Study 1
A Cash Variance That Moved Between Three Ledgers
Harrow Lane Distributors, a regional wholesale food distributor with four warehouses and a centralized accounts payable team of nine
The situation
At Harrow Lane Distributors, the monthly bank reconciliation for the operating account had closed with an unexplained difference of 38,600 dollars. The general ledger showed one balance, the bank statement showed another, and the accounts payable subledger sat between them with its own total. The controller, Dana, had asked for the close to be finished before the audit committee packet went out in the third week. Four warehouses posted payments through the same disbursement file, and two of them had changed their remittance batching process in the prior month without telling accounting. The team had rerun the bank import twice and found nothing wrong with the file format.
The challenge
The difference was not large relative to monthly cash activity, which ran above 12,000,000 dollars, but it was unexplained and sat in a balance that the auditors would test. Dana needed to know whether the gap was timing, a duplicate posting, a missing entry, or a reconciling item that had been carried forward without support. Writing a plug entry to clear it was not an option Dana would accept, and the close could not wait for a full rebuild of the subledger.
Analysis
Teams grounded in Enterprise and Financial Risk Management usually start by separating the three records rather than adjusting any of them. The bank statement is outside evidence, the general ledger shows what the company recorded, and the subledger shows the detail behind the payables balance. Here the bank-to-subledger gap traced to 214 checks that cleared in a different period from the one in which they were recorded. The ledger-to-subledger gap was a separate set of manual entries posted at month end with no payment reference. Only after isolating these two populations did the exposure become visible, because the unsupported entries, not the timing items, carried the real risk.
Control evidence determined how far the team could go without escalating. The manual journal entries had preparer initials but no approver, which broke the segregation the company's own policy required. The policy set a threshold of 5,000 dollars for any manual entry to a cash or payables account requiring a second signature, and 22 of the unsupported entries fell below that line individually but exceeded it in aggregate by the same posting user on the same day. That pattern pointed toward splitting rather than honest error, and the team treated it as a control failure. Escalation therefore turned on the pattern, not the gap size.
- 1Bank statement balanceExternal evidence used as the anchor for every comparison.
- 2Subledger detailCheck payments mapped to the batch file and the period they cleared.
- 3Ledger journal reviewIsolate manual entries posted at month end without a payment reference.
- 4Unsupported populationGroup entries by preparer and day to expose split postings.
- 5Approval and escalationEscalate any unapproved pattern regardless of the single-entry amount.
Actions taken
- Match every open reconciling item to a bank line or a payment batch identifier, and mark anything without a match as unsupported rather than timing.
- Require a second approver for all manual journal entries touching cash or payables, with the policy threshold applied to same-day totals per user, not just single entries.
- Move remittance batching changes through a written change request that names accounting as a reviewer before the next file run at each warehouse.
- Run a monthly exception report that lists unmatched items by age and preparer, and send it to the controller before the close calendar opens.
Results
The unexplained difference fell to 1,900 dollars after the 22 unsupported entries were reversed and rebooked with proper support. The remaining amount was traced to a deposit in transit. The split posting pattern stopped after the approval rule took effect and the next two closes finished in the first week rather than the third. The audit committee received a memo with the root cause and listed the change request as a condition of sign-off. Two warehouses said the new batching form added about ten minutes per file, a cost the team accepted for a testable trail.
Questions for discussion
- Which of the three record pairs would you reconcile first, and how would a late bank statement change that order?
- How would you set an approval threshold for manual entries when the business has a legitimate reason to post many small adjustments on the same day?
- What evidence would convince you that a reconciling item was a timing difference rather than a carried-forward error?
Case Study 2
Cash Flow Drifting Below Plan in a Seasonal Business
Brightwater Outdoor Supply, a mid-sized seller of camping and marine equipment with wholesale and online channels
The situation
Brightwater's finance group had built its annual plan around a strong second half, with inventory purchases timed to a spring sales peak that the planning team expected to exceed the prior run-rate by roughly a fifth. Eight weeks into the year, booked revenue was tracking about 9 percent below the plan line, while supplier invoices had arrived on schedule. The variance was not evenly spread. Wholesale orders had slipped in two regions, and online conversion had softened after a pricing change that marketing had approved without a finance review. The treasurer, Anil, needed a forecast he could defend to the lender before the covenant test at the end of the quarter.
The challenge
The question was not simply whether revenue would recover. Anil had to decide how much of the shortfall was timing and how much was structural, because the response differed sharply. A timing gap might justify deferring a purchase order and drawing on the revolving line for a few weeks. A structural gap would mean resizing the inventory commitment, which carried a cancellation fee and a supplier relationship the company valued. The covenant allowed a minimum liquidity balance that the base case already brushed against.
Analysis
The rebuild had to start at the driver level. Brightwater's plan had been built from a single growth rate applied to last year's monthly revenue, which hid the mix between channels with very different collection timing. The team split revenue into wholesale, direct online, and marketplace sales, and tied each to its own order pipeline and days-to-cash profile. Inventory purchases were broken into committed and uncommitted orders, because only the uncommitted portion could be moved without penalty. Once the record was rebuilt, the liquidity gap in the sixth week stood out as the real exposure, not the annual revenue shortfall that had dominated the earlier discussion.
Scenarios were built around three assumptions the team could test against actual data: quote-to-order conversion, the recovery in online conversion after the pricing change was reversed for a subset of products, and the collection period for marketplace receivables. Each scenario carried a decision threshold. If liquidity fell below the internal floor of 750,000 dollars, the company would defer uncommitted purchase orders. If it fell below the covenant level, Anil would open discussions with the lender before the test date. The team also marked which assumptions had evidence behind them. Only the quote conversion rate had a clean history, so the downside case leaned on it most.
- Eighth weekBooked revenue sits about 9 percent under plan while invoices arrive on schedule.
- Ninth weekChannel-level rebuild shows a liquidity low below the internal floor.
- Tenth weekUncommitted orders deferred in two regions while the pricing test begins on slow lines.
- Eleventh weekConditional waiver agreed with monthly liquidity reporting as the trigger.
- Quarter endCash position matches the revised forecast and the waiver holds.
Actions taken
- Rebuild the cash forecast weekly by channel with separate collection curves, and reconcile each week's actuals to the prior forecast before changing any assumption.
- Defer uncommitted purchase orders in the two weakest regions until the liquidity floor is restored, and document the supplier notice for each deferral.
- Reverse the pricing change on the slowest product lines for a four week test, and measure conversion against a held-out group before expanding it.
- Give the lender a written range with the downside case and the specific trigger that would prompt a waiver request, so the conversation begins before the test.
Results
The base case showed the sixth-week low about 300,000 dollars under the floor. Deferring 1,200,000 dollars of uncommitted purchase orders restored the floor, although two suppliers charged cancellation fees the team had underpriced. The pricing test recovered about half of the lost online conversion within its window, which supported a partial rollout rather than a full reversal. Anil's lender meeting took place in the eleventh week, and the company agreed to a conditional waiver with monthly reporting. Revenue for the year still landed below plan, but the cash position at the test date matched the revised forecast.
Questions for discussion
- Which assumption in the downside case would you test first, and what data would you need before trusting it?
- When does deferring discretionary purchases protect the business, and when does it damage supplier relationships that matter later?
- How would you decide whether to disclose a forecast miss to a lender before a covenant test is at risk?
Case Study 3
Structured Payments That Tripped the Monitoring Rules
Calloway Harbor Payments, a licensed payments processor that settles transfers for about 300 small merchants
The situation
Calloway Harbor Payments processed transfers for roughly 300 small merchants and kept a rules engine that flagged patterns for review. In the third week of the quarter, the engine raised a cluster of alerts on one merchant account, a building supply retailer that had been a client for four years. Over nine days, the account received 46 deposits of just under 9,000 dollars each, then sent most of the funds to a set of personal accounts held by three individuals who had no listed connection to the business. The relationship manager, Priya, believed the pattern was seasonal restocking, but the documentation in the file had not been refreshed since the original onboarding.
The challenge
The compliance officer, Mateo, had to decide whether the alerts warranted a formal report, and on what evidence. The business had a plausible story, the transfers were within the account's stated volume range, and the merchant had cooperated with routine questions in the past. But the split of deposits under a reporting threshold and the rapid onward movement were the kind of indicators the monitoring policy required to be escalated, and the firm could not afford a decision that rested on the relationship manager's confidence alone.
Analysis
Exposure and suspicion need separate handling, and the first step here was to keep them apart. Exposure was measurable: about 414,000 dollars moved through the account in nine days, and the firm carried the loss on funds it had settled. Suspicion is a judgment built from the pattern. Mateo laid the transactions against the account's documented expected activity, which the file described as monthly volume of roughly 120,000 dollars. The nine-day total was well above that baseline, and no recipient appeared in the vendor master. The documentation gap mattered as much as the transactions, because without a refreshed profile, the firm could not show what normal looked like.
Controls and escalation rules set the path. The monitoring policy required a documented review within two business days of any alert cluster and a decision on whether to file a report within a further set period. It also set a threshold: any pattern of deposits below the internal reporting threshold, repeated more than 10 times in 30 days, required a senior review. Mateo's review confirmed the rule had fired correctly, but also that the alert had sat in a queue for four days because the rules engine routed merchant alerts to a general pool. The control evidence showed the detection worked and the escalation did not.
Actions taken
- Place a hold on outbound transfers from the account pending senior review, using the standard hold procedure that does not disclose the indicators to the client.
- Route all merchant-level alerts to the senior compliance queue with a two business day review deadline, so the escalation path no longer depends on which pool an alert lands in.
- Require a refreshed expected-activity profile for any merchant whose file is older than twelve months, signed by the merchant and the relationship manager.
- Decide on the suspicious activity report within the policy window, and record the reasoning, including the explanations offered and why they were or were not accepted.
Results
The merchant supplied invoices for part of the restocking, covering roughly 60 percent of the outbound value, but could not document the recipient accounts. Mateo filed the report within the policy window and recorded the reasoning in full. The account was restricted and later closed at the merchant's request. The routing change cut average alert-to-review time from four days to under two in the following quarter, and the refreshed profile process reached 85 percent of older files before the quarter ended. A later examiner review found the file complete.
Questions for discussion
- What would you need to see from the merchant before you were willing to release the hold, and who should make that call?
- How would you tune a rule that catches structured deposits without flooding the queue with legitimate seasonal patterns?
- When a control detects a problem but the escalation fails, which fix should come first, and how would you prove it worked?
Case Study 4
A Growth Loan Request Built on Optimistic Projections
Tidewater Ceramics, a family-owned manufacturer of commercial tile with two plants and about 140 employees
The situation
Tidewater Ceramics asked the commercial lending group for a term loan of 4,200,000 dollars to fund a third kiln line and a new distribution contract. The request came with a three-year model that projected revenue growth of about 35 percent in the first year, driven by the distribution agreement. The company's historical statements showed a steady but slower trajectory, with gross margin in a narrow band and a customer base where the top five accounts supplied roughly 58 percent of revenue. The credit analyst, Lucia, had the audited statements for two years, management accounts for the current year to date, and the signed distribution term sheet, which carried a conditional volume commitment.
The challenge
Lucia had to decide whether the projections deserved any weight, what downside the loan could survive, and what structure would protect the bank if the distribution volumes arrived late or not at all. The committee expected a recommendation with explicit conditions, not a simple approve or decline. The requested amount was about 40 percent of the company's current tangible net worth, and the proposed repayment schedule assumed the new line would be earning its full rated output in its second quarter of operation.
Analysis
A projection is a claim, and the evidence behind it has to be tested apart from the model's arithmetic. Lucia rebuilt the forecast from the bottom up. The distribution contract's volume commitment was conditional on the partner's own sales targets, and the term sheet set no minimum. She converted the contract into a range of volumes, applied the historical margin on that mix, and compared the result with management's figure. The management case required the partner to take nearly three times the volume it had taken from its existing supplier in the first year. That gap between projected and supportable revenue became the central credit issue.
Downside scenarios and decision thresholds turned the discussion into testable conditions. Lucia ran three cases: a base case, a stress case in which the largest existing customer reduced orders by a fifth, and a severe case combining both with a six-month delay in kiln commissioning. The bank's policy required debt service coverage of at least 1.25 times in the base case and at least 1.0 times in the stress case. The base case cleared the first test narrowly. The severe case fell below 1.0 in the first year, so approval would need more equity or tranched funding. Suitability also mattered, because the owners had personally guaranteed existing facilities and this loan would deepen that exposure.
- Test projections Rebuild revenue from contract volume ranges rather than management's single figure.
- Run downside cases Apply base, stress, and severe scenarios against coverage thresholds.
- Set conditions Translate each failing case into a structure, covenant, or equity requirement.
- Monitor and revisit Track verified volumes against the tranche triggers and the covenant reports.
Actions taken
- Require a signed distribution agreement with a minimum annual volume and a termination right if the partner falls below it for two consecutive quarters.
- Structure the facility in two tranches, with the second released only after the kiln line reaches agreed output and the first six months of distribution revenue are verified.
- Raise the equity contribution so the stress case meets the coverage floor without counting on distribution revenue in year one.
- Obtain monthly management accounts and a customer concentration report as covenants, with a cure period and a right to reprice if the top five accounts exceed an agreed share of revenue.
Results
The committee approved a reduced facility of 3,100,000 dollars in two tranches, the second conditional on verified distribution volumes. The company agreed to contribute 900,000 dollars of equity before the first draw, which lifted the stress case above the coverage floor. The partner signed with a lower minimum volume than management had projected. Eight months after closing, the first tranche was drawn and the second stayed undrawn at about 60 percent of plan volume. The bank's exposure stayed within the approved limit, and covenant reporting flagged the shortfall early.
Questions for discussion
- Which assumption would you most want independently verified before the committee met, and how would you verify it?
- Is a tranched facility a real protection or mainly a way to defer a decision that should be made now?
- How should personal guarantees factor into suitability when the owners' wealth is concentrated in the business?
Case Study 5
Ranking Control Gaps After an Internal Audit Review
Pinecrest Health Partners, a multi-site outpatient group whose shared services team handles billing and payables for six clinics
The situation
The internal audit team at Pinecrest Health Partners completed a review of the shared services finance function and issued fourteen findings. The review covered billing adjustments, vendor master changes, payroll journal approvals, and month-end accrual reviews. Several findings were small, such as an out-of-date signature card and a missing reviewer sign-off on a template. Others were more serious. Billing adjustments above a set amount could be posted by the same person who approved them, and vendor bank details could be changed without a callback to the vendor. Management had a remediation deadline in the second quarter, and the audit committee wanted a ranked plan before its next meeting.
The challenge
The finance director, Gareth, had to decide which findings to fix first, who should own each one, and what evidence would show that a finding had truly closed. Fourteen findings with uneven severity could easily become a list in which everything looked urgent, and the teams involved had limited capacity. The audit report also described some issues in terms that were technically accurate but did not say how much money or how many records were exposed. Gareth needed the exposure figures before he could rank anything credibly.
Analysis
Ranking by exposure and control weight keeps the plan honest. Gareth's group measured each finding against two numbers: the value of transactions the gap could affect, and whether one person could complete the risky action without a second check. The billing adjustment gap touched about 2,300,000 dollars of adjustments in the period, with the same user approving and posting in roughly a third of cases. The vendor bank change gap had a smaller total but a higher loss per event, since one fraudulent change could redirect an entire payment run. The signature card issue had no transaction exposure at all and went to the bottom of the list.
Closure tests were where remediation most often failed quietly. For each finding, Gareth required a named owner with authority over the process, a written fix, and a closure test an outsider could run. The billing gap test was a sample of 60 adjustments, with no adjustment in the sample showing the same user for both steps. The vendor bank change test required a callback record for every change in the following quarter, matched to the verification log. Management had proposed closing the signature card finding on the basis of a new form. Gareth rejected that, since a new form did not show the old card had been retired.
Actions taken
- Remove the ability for any user to both approve and post billing adjustments above the policy threshold, and add a system rule that blocks posting until a second user confirms.
- Require a callback to a verified number on file for every vendor bank change, with the verification log retained and reviewed by someone outside payables each month.
- Assign each finding a named owner with process authority, a target date within the second quarter, and a closure test written before the work starts.
- Close the low-severity findings only after an independent tester confirms the sample evidence, and report any reopened item to the audit committee with its original rank.
Results
By the end of the second quarter, nine of the fourteen findings had closed on evidence. The billing sample of 60 adjustments showed no same-user approval and posting, two months after the system rule went live. The vendor log covered 47 bank detail changes, each with a callback record. The signature card finding reopened once, because a retired card was still on file at one clinic bank. Three findings carried into the next period with revised dates. The audit committee accepted the ranked plan and asked that closure tests for the remaining items be drafted first.
Questions for discussion
- How would you decide whether a low-severity finding deserves a full closure test or can close on a management attestation?
- What would a closure test look like for a month-end accrual review, where the risk is judgment rather than transaction value?
- When an owner resists a closure test because the fix is already in place, how would you resolve the disagreement without weakening the standard?
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.