vaayaa Inclusive Education

Bank Reconciliation Operations Case Studies

Finance5 Case StudiesUpdated Sep 2026

Bank reconciliation looks like clerical work from the outside, but the people who run it know it is where the ledger either tells the truth or does not. The five cases here come from finance operations teams in mid-sized organizations, and each begins with a number that stopped agreeing with another number. What follows is how those teams traced the gap, what they decided, and what they changed afterward. Every organization and person named is fictional, and the figures are internal to each case. The aim is to show how transaction records, approval steps, and close deadlines interact when something goes wrong, and how a practitioner decides which fix will hold.

1

Case Study 1

A stubborn variance in a regional distributor's cash account

Ferris Lane Distributing, a regional wholesaler of building supplies with six warehouses and about 310 employees

The situation

Ferris Lane Distributing closes its books on a fixed calendar, with the operating account reconciled by the fifth business day after month end. In the third week of the close, Priya, the staff accountant who owns the cash reconciliation, found that the adjusted bank balance and the adjusted general ledger balance differed by 4,380 dollars. The month held 2,940 posted ledger entries and 1,116 bank transactions. Most reconciling items were routine, including outstanding checks, deposits in transit, and bank fees. The gap did not match any single item. Priya had cleared the prior month in full, so the difference had to come from something posted during this period. Marcus, the controller, wanted it explained before the close package reached the executive team.

The challenge

The first guesses, a missing deposit or a duplicated check, were tested on the first day and ruled out. The harder problem was that three records disagreed: the bank export, the ledger detail, and the scanned supporting documents in the approval system. A vendor credit had been posted to the ledger under the wrong period, and its supporting image sat in a different batch. Priya had two days before the close deadline, and treasury needed a clean balance for the cash forecast.

Analysis

Start with the three records and ask which one is the outlier. In this case the bank export was internally consistent, and its running balance tied to the statement total. The ledger detail was consistent with itself too, but one entry carried a posting date from the prior period and a document number that belonged to a vendor credit. That combination is easy to miss because the entry was valid on its own. The supporting image pointed the wrong way, which is why a reviewer who only checked attachments would have signed off. Ledger integrity problems of this type rarely show up as a broken total. They show up as a date and a reference that disagree, and the fastest way to see them is to sort unmatched items by posting date and compare document numbers against the source batch rather than against the description.

The approval trail explained why the entry got through. The vendor credit had been keyed by an accounts payable clerk and approved by a supervisor who saw only the batch total, not the line detail. The batch sat under the threshold that triggered second review, so no one outside the batch looked at it. Under deadline pressure, the reconciliation team had been accepting batch-level sign-offs for cash postings, which made the control thinner than the policy suggested. A practical test is to ask whether the approver could have spotted this specific error from what the approval screen showed. If the answer is no, the control exists on paper only. Close deadlines make this worse, because a late correction tends to be pushed through without rerunning the approval path.

Whether the posted credit can stay in the period
Does the entry match its original source document?
YesReverse and repost next periodKeep the audit trail, move the entry to the correct date, and document the reason.
NoReverse and investigate sourceHold the entry, pull the batch image, and confirm vendor and amount with payables.
Variance cleared, control added at line level approval

Actions taken

  1. Sort unmatched ledger items by posting date and flag any entry whose document number belongs to a different period or batch. Run this before comparing amounts, since date conflicts point to the source faster.
  2. Require line-level review for any cash posting above a set value, even when the batch total passes. The approver should see the vendor, the document reference, and the posting period on the same screen.
  3. Add a system check that rejects prior-period entries unless they carry a flag and a controller's initials. Test it against the previous twelve months before switching it on.
  4. Set a reconciliation calendar rule that cash items must be cleared two business days before the close deadline, with late corrections requiring a written exception. Track exceptions by owner so patterns become visible.

Results

The reversal and repost took one afternoon once the source image was pulled. The 4,380 dollar variance cleared, and the close went out on the fifth business day. The line-level review rule caught two similar postings in the following cycle, both from the same clerk and both corrected before close. The team also found that 37 batch approvals over the prior period had been signed without line detail, which led to a retrospective review of 212 entries. None were wrong, but the gap was documented and closed. Priya's note to the controller said the reconciliation now tied on the first pass in most months, which had not been true before.

Questions for discussion

  • Which of your current approval screens shows the vendor and the document reference together, and what would an approver need to see before signing?
  • How would you decide whether a prior-period entry belongs in the open period or should be corrected in its original period?
  • What evidence would convince you that a batch-level sign-off was a real control rather than a formality?
2

Case Study 2

Cash forecast drift at a veterinary supply distributor

Oakhurst Veterinary Supply, a distributor serving about 1,400 clinics across four states

The situation

Oakhurst Veterinary Supply runs a 13-week cash forecast that treasury rebuilds every week from reconciled bank balances, open receivables aging, and scheduled payables. Imani, the treasury analyst who maintains the model, saw in the sixth week of the quarter that actual cash sat 612,000 dollars below the plan curve. Collections from the largest clinic group had slipped. The plan assumed receivables would turn in 38 days, and the trailing figure had moved to 47. The reconciliation team also held 96 unapplied cash receipts, which inflated the aging report and made the collection picture hard to read.

The challenge

The forecast had to support a recommendation to the chief financial officer by the end of the week. Three assumptions were in question: collection timing, the treatment of unapplied cash, and whether a rebate payment had been counted twice. The reconciliation team could not say whether the 96 receipts belonged to invoices already closed, and until they did, the base case was unreliable. The choice on the table was to draw on the revolving line, delay a vendor payment run, or some mix of both.

Analysis

Teams grounded in Bank Reconciliation Operations usually begin by separating what the bank says happened from what the ledger says should happen, because a forecast built on an unreconciled base carries both errors forward. Here the unapplied receipts had been double counted: cash already received sat in the aging report as if it were still owed. Once those were separated, the collection shortfall shrank by about a third. That left a smaller but real timing gap, which was the part worth modeling. Each assumption needed a source record behind it. Collection timing came from invoice-level payment dates, rebate timing from the vendor agreement, and the cash base from bank statements tied to the ledger. An assumption without a record was marked as a judgment, with an owner and a review date.

The approval path mattered as much as the arithmetic. A forecast does not authorize a payment, but a worrying forecast can push a payment run through with less scrutiny than a budget change would get. The team agreed that any hold or draw would go to the controller with the reconciled base attached, and that scenarios would be labeled as planning tools, not approvals. The close calendar set the limits. The weekly forecast had to be ready before the Monday cash call, so the base reconciliation had to finish on Friday. Anything later was treated as a revised figure for the following week, with the change explained in writing.

Forecast refresh cycle under cash pressure
Weekly cash review
1Reconcile base
2Test assumptions
3Run scenarios
4Decide and log
  1. Reconcile base Tie bank balances and unapplied receipts before any projection is updated.
  2. Test assumptions Compare collection timing, days sales outstanding, and rebate timing against source records.
  3. Run scenarios Model base, slower collections, and a severe case with the revolver drawn in each.
  4. Decide and log Record the recommendation, owner, and review date so the next cycle can check it.

Actions taken

  1. Separate unapplied receipts from open receivables in the aging report so that cash already received is not counted twice as collections due.
  2. Rebuild the days sales outstanding assumption from invoice-level payment dates for the last six months, not from the monthly average the plan used.
  3. Hold the vendor payment run for one week only if the severe case breaches the minimum cash balance, and get the controller's written approval for the hold.
  4. Put the forecast change through the same approval path as a budget revision, with the reconciliation owner signing off on the base balance before the treasury team uses it.

Results

The team cleared 74 of the 96 unapplied receipts before the deadline. Most of the remaining 22 tied to a customer dispute that did not change cash timing. The base case moved by 318,000 dollars, mostly from reclassifying cash already received. The severe case still breached the minimum balance by a narrow margin, so the recommendation was a three-day hold on the lowest-priority vendor batch and a standby request for the revolver, with nothing drawn. The controller approved the hold in writing, and the remaining batches were released on their revised dates. The next weekly forecast tied to the bank balance on the first pass.

Questions for discussion

  • Which of your forecast inputs would you trust least if the bank reconciliation arrived three days late, and how would you label that uncertainty?
  • When unapplied cash and receivables disagree, what evidence would you accept before changing the collections assumption?
  • How would you choose between delaying payables and drawing on the credit line when both keep the balance above the minimum?
3

Case Study 3

Warning signs in a settlement account's outbound transfers

Halvard Merchant Services, a payments processor settling card receipts for about 2,300 small merchants

The situation

Halvard Merchant Services settles card receipts to its merchants, and its settlement account is reconciled every business day. Tomas, a reconciliation analyst, noticed in the seventh business day of the month that the account had 14 outbound transfers to a single beneficiary over nine days. Each transfer fell between 9,100 and 9,900 dollars, just under the internal review threshold of 10,000 dollars. The beneficiary had been added to the payment file about three weeks earlier. None of the transfers had a matching merchant ledger entry, and six had been parked in a suspense account across several periods. The total came to 131,400 dollars.

The challenge

Renata, the compliance officer, needed to decide whether this was a stuck reserve release, a legitimate sub-acquirer payout, or something requiring a suspicious activity review. The onboarding file for the beneficiary carried a signature from a manager who had since left the company. The next settlement batch was due in two days, and an automatic release would include the beneficiary. Tomas also had to explain the suspense balance in the reconciliation certification.

Analysis

The useful first step was to describe the pattern before deciding what it meant. Fourteen transfers, a narrow amount band, a single counterparty, and a new payee added shortly before activity began are each explainable on their own. Together they form a set of warning signs that calls for review rather than a conclusion. The ledger side was more telling. Suspense items that sit across several periods usually mean someone has been avoiding a decision, not that the bank is slow. Tomas traced each transfer to its bank reference and checked whether the merchant ledger carried any corresponding obligation. None did. That absence matters more than the amounts, because the outflows were not settling any recorded liability. Exposure here is the total outflow plus the cost of an unsupported suspense balance in the certified books.

Approval controls explain how the beneficiary got into the file. The payment system required two approvers for new payees, but the second approval had been recorded against the first approver's user ID under a delegated role that was never removed. That is a control that looks present in the log and is absent in practice. Escalation under the applicable rules does not wait for certainty. Once the pattern meets the firm's defined indicators, the compliance team decides the review path, including whether to hold the release and what the merchant may be told. Premature contact can compromise a review, so the reconciliation team should not raise it with the beneficiary. The close deadline adds pressure, but a suspense balance can be disclosed with a note while the review runs.

Where the settlement outflows sit on risk
Pattern concern, low to high
Escalate nowStrong pattern, weak paperwork: hold the next release and open a review.
Confirm then documentConcerning pattern, complete file: confirm purpose and document the closure.
Routine monitoringFew warning signs and weak paperwork: fix the file, keep watching the account.
Clear and filedLittle pattern concern and strong records; release on schedule and sample quarterly.
Documentation strength, low to high

Actions taken

  1. Hold the next automatic release to the beneficiary, and route the held items to compliance with the transaction references, amounts, and dates attached.
  2. Pull the onboarding file for the beneficiary and verify each approver's identity and authority at the time of approval, including any delegated roles still active.
  3. Review the suspense account for all items older than one period, and require a documented owner and a supporting document for each before the certification is signed.
  4. Test the two-approver rule against the full payee file to find any other approvals recorded under a single user ID, and remove delegated roles that have lapsed.

Results

Compliance opened a formal review and held the batch. The review found that the manager's signature was genuine, but his delegated access had never been revoked, and the 14 transfers had been initiated under that delegation. The matter was referred under the firm's reporting procedure, and the beneficiary account was frozen pending the outcome. Tomas certified the reconciliation with the suspense balance disclosed and a named owner for each item. The delegation audit found nine other lapsed roles, all removed the same day.

Questions for discussion

  • What threshold would you set for escalating a pattern like this, and how would you keep that threshold from being worked around by amounts just below it?
  • If the account holder offered a plausible business explanation, how much weight should that explanation carry against the documentary record?
  • Which staff should know about an open review, and what should they be told when a legitimate payment is delayed?
4

Case Study 4

Sizing a seasonal credit line for a bakery group

Marrow Creek Bakery Holdings, a family-owned group of three bakeries supplying grocery chains

The situation

Marrow Creek Bakery Holdings asked for a 1,250,000 dollar revolving line to fund seasonal flour and packaging purchases. Helena, a credit analyst at Tallow Ridge Commercial Bank, received 24 months of bank statements, the company's monthly reconciliations, and reported annual sales of about 8,400,000 dollars. The reconciliations were prepared on time, but three of them carried differences forward without a documented cause. Deposits in the most recent quarter ran 11 percent below the same quarter of the prior period, and the owner offered a personal guarantee.

The challenge

The credit committee would decide by the end of the month. The company's own forecast assumed stable renewal of a grocery contract that supplied 41 percent of revenue, and that contract was up for review within six months. Historical deposits supported the company's record, but forward-looking evidence was thin, and the carried-forward items raised a question about close discipline. The line had to be sized for a seasonal peak without assuming the best case for the largest customer.

Analysis

Start with cash rather than the income statement. Bank deposits are a record the company does not write, and they tell a different story from reported sales in two months where the gap widened. Helena rebuilt monthly cash flow from the statements and compared it with invoice totals. Most months tied within a narrow band. The two that did not were the same two holding the largest carried-forward differences, which pointed to a timing problem in receipts rather than missing sales. That makes the reconciliations themselves part of the evidence. Until those differences are cleared with documents, reported margins are a claim to test, not a basis for sizing the line.

Suitability and downside decide the size and the conditions. The seasonal need is real and predictable, which supports a revolving line rather than term debt. The downside case removed the largest contract, cut sales by 25 percent, and drew the line in full for six months. Under that case the full request would not have been covered by the borrowing base, while the reduced line would. Approval conditions followed from that test: a borrowing base tied to receivables from remaining customers, monthly reporting within fifteen days of month end, independent review of the next two reconciliations, and a covenant reset if concentration stayed high. Rejection would have been right if the carried-forward items could not be cleared.

Evidence strength behind the line request
Bank deposit consistency6/10
Deposits track reported sales in most months, with a wider gap in two.
Reconciliation discipline4/10
Three carried-forward differences point to weak close discipline.
Customer concentration3/10
One grocery contract supplies 41 percent of revenue and renews soon.
Seasonal cash need7/10
Peak purchases are predictable and sized close to the requested line.
Owner guarantee strength5/10
Personal recourse helps, but the asset values behind it are not current.

Scores set by the review team in this scenario, 10 = highest priority.

Actions taken

  1. Rebuild monthly cash flow from bank statements rather than the company's summary, and tie each month's deposits to invoice and sales records before using the figures.
  2. Require the three carried-forward reconciling items to be cleared with supporting documents, and have a third-party accountant review the next two reconciliations before any draw.
  3. Run a downside case where the largest contract lapses, sales fall 25 percent, and the line is drawn to its full amount for six months, then measure coverage against the borrowing base.
  4. Approve a reduced line of 800,000 dollars with a borrowing base, monthly reporting within fifteen days of month end, and a covenant reset if customer concentration exceeds the agreed level.

Results

Helena recommended a line of 800,000 dollars with conditions rather than the full request. The company accepted the independent review and cleared the carried-forward items before the next close. Under the downside case, the reduced line stayed covered through the seasonal peak even without the largest contract, though with little room to spare. Credit committee approved on those terms, and the first monthly report arrived inside the fifteen day window. The owner's guarantee was kept as recourse, but the committee did not rely on it to justify the amount.

Questions for discussion

  • Which single piece of evidence would most change your view of the line size, and what would you ask the company to produce to test it?
  • How should a lender weigh a personal guarantee when the asset values underlying it are several periods old?
  • At what level of customer concentration would you stop treating seasonal sizing as the main justification for the request?
5

Case Study 5

Internal audit findings and closure for clinic bank accounts

Stillwater Ambulatory Health, a network of 14 outpatient clinics with about 1,900 staff

The situation

Stillwater Ambulatory Health operates 14 outpatient clinics and reconciles 22 bank accounts, including operating, payroll, and a patient refund account. Internal audit reviewed reconciliation controls across all of them in the closing weeks of the review cycle. The team found 268,000 dollars in reconciling items older than 90 days, signoff evidence missing from 7 of the 22 reconciliations, and two periods of the patient refund account never reconciled at all. A bank feed mapping error was also routing some card settlements to the wrong general ledger account, and two users held both posting and approval rights on the same accounts.

The challenge

Omar, the audit manager, had to rank the five findings by materiality and assign accountable owners before the audit committee met. Management wanted every finding closed by the end of the quarter. The committee wanted evidence, not policy updates, behind each closure. Gwen, the revenue cycle director, and Blake, the controller, both had competing priorities, and the IT lead was working through a backlog that already ran several weeks long. Closure tests had to be specific enough that someone outside the team could repeat them.

Analysis

Ranking by materiality means combining factors that do not always agree. Dollar exposure is the obvious one: the aged items totaled 268,000 dollars, and the refund account had an unknown balance, which is itself a finding. Age matters because old reconciling items often hide errors already made. Account sensitivity matters too. A patient refund account and a payroll account carry obligations an operating account does not, even when the dollars look smaller. The feed mapping error sits in a different category because it creates silent misstatements that a reconciliation will not reveal on its own. Omar ranked it near the top despite its modest value, because it undermined the records the other findings rely on.

Approval and access findings are harder to close than they look, because a control can exist in a policy document and still fail in the system. The two users holding both posting and approval rights on the same accounts are the clearest case. Closure should require an access report showing the conflict removed, plus evidence that no entry was approved by its own preparer during the gap. Tests also need to run over time, since a signoff requirement that operated once has not been proven. Omar set re-performance of five reconciliations a month across three consecutive closes, and a sample of 60 transactions per account against source documents for the feed mapping. The committee closed a finding only after independent review of those results, not on the owner's report that the work was done.

Path from finding to evidenced closure
  1. 1Confirm exposureQuantify dollar value, aged items, and affected accounts for each finding.
  2. 2Assign one ownerName one person who can change the process, not just report on it.
  3. 3Fix root causeCorrect the feed mapping, access conflict, or signoff step that allowed the gap.
  4. 4Test operationRe-perform reconciliations over several closes and sample the evidence trail.
  5. 5Close with evidenceAudit committee closes a finding only after the test passes on independent review.

Actions taken

  1. Rank the five findings by a combined score of dollar exposure, age of items, and whether the account holds patient refunds or payroll, and publish the ranking with the reasoning.
  2. Assign Gwen to the aged outstanding items and the refund account, Blake to the signoff evidence and access conflicts, and the IT lead to the bank feed mapping, each with a named deadline.
  3. Define each closure test in advance: for signoff, re-perform five reconciliations per month for three consecutive closes; for feed mapping, match a sample of 60 transactions per account to source.
  4. Keep the refund account open to new periods until its two missing periods are rebuilt and independently checked, and report the residual exposure to the audit committee each month.

Results

Management corrected the feed mapping and removed the access conflicts within the first month. The aged reconciling items fell from 268,000 dollars to 41,000 dollars, most of the difference being bank adjustments already in transit. The signoff test failed once in the first cycle, because a new preparer had skipped the evidence attachment, which was the outcome the test was built to catch. The refund account took longer, since the missing periods had to be rebuilt by hand, and it closed three periods after the original target. Three findings closed on schedule, and two remained open at the end of the cycle, each with its test results attached.

Questions for discussion

  • How would you weigh a finding about aged items against one about access conflicts when the first has more dollars and the second has more potential to hide errors?
  • What would you accept as evidence that a signoff control operated, beyond the presence of a signature?
  • When a remediation owner asks to close a finding early on the basis of a policy update, what would you require before agreeing?

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.

Hourly Active: 37,008 Learners