General and Property Insurance Operations Case Studies
Finance5 Case StudiesUpdated Sep 2026
Finance teams in general and property insurance spend a surprising share of their time asking whether two sources of truth agree. Premium ledgers, claim reserves, agency statements and reinsurance recoverables run on different calendars and through different systems, so a gap in one place often points to a weakness somewhere else. The five cases follow fictional firms through moments when records diverged, plans slipped, or a transaction raised questions. Each one names the records, controls and thresholds that mattered. The judgment calls matter since the same variance can be harmless in one quarter and serious in the next.
Case Study 1
A premium ledger that refused to balance at quarter close
Ostrander Mutual Property Insurance, a regional mutual writing homeowners and small commercial property across four states, with about 61,000 policies in force.
The situation
Ostrander Mutual closes its books monthly, but the quarter-end close is where the premium ledger gets its hardest test. In the third week after quarter end, the controller's team found that written premium in the policy system did not match cash receipts in the general ledger or the agency bordereaux from its twelve producers. The gap came to 318,600 dollars. That is small against a quarterly written base in the tens of millions, but large enough to draw an auditor's question. Two sub-ledgers had been reconciled by different people using different cutoff rules, and a batch of late endorsements was dated by effective date in one system and by transaction date in the other.
The challenge
The immediate question was whether the 318,600 dollars was a timing problem, a data problem, or a control failure that could repeat next quarter. Each explanation pointed to a different fix, and the regulatory filing deadline was close. The team also had to decide how much could be cleared through suspense without controller sign-off on each item, since the old practice had let small items age for months.
Analysis
Teams grounded in General and Property Insurance Operations tend to split a variance into its parts before chasing the total. The team at Ostrander divided the 318,600 dollars into three buckets: endorsements dated differently across systems, bordereaux received after cutoff, and cash receipts posted to the wrong period. The first two were timing differences that reversed within the close window, and each could be traced through system logs. The third was a posting error with real exposure, because those receipts had supported a reserve release that now looked premature. Under the written policy, any unmatched item above 15,000 dollars required controller sign-off, and the posting error was the only item that crossed it.
The control fix started with evidence, not with a new report. The team asked for three things from each sub-ledger owner: the tie-out workpaper, the list of open reconciling items with ages, and a signed statement of the cutoff rule used. Tie-out workpapers that lacked a preparer and reviewer sign-off were treated as missing evidence, even when the arithmetic was right. Once the cutoff rule was written into one shared standard, the controller added a monthly test: items dated differently across systems must match or appear on an exception log with a named owner.
Actions taken
- Rebuild the cutoff rule as one written standard that all three sub-ledgers reference, with the effective date and transaction date defined separately.
- Require a preparer and reviewer sign-off on every tie-out workpaper, and reject any workpaper that lacks either, regardless of whether the arithmetic agrees.
- Add a monthly exception log for items dated differently across systems, with an owner and a target clearing date for each entry.
- Review the reserve release that depended on the misposted receipts, and document whether the release stands, reverses, or is restated before the filing.
Results
By the end of the quarter the variance was fully explained. Of the 318,600 dollars, 301,200 dollars were timing differences that reversed within the close window, and the remaining 17,400 dollars were the posting error, corrected in the period it belonged to. The reserve release was reversed in part, and the restated figure went to the audit committee with a short note on cause. The monthly exception log cleared all but two items in the following cycle. Two owners had used different cutoffs for several quarters without anyone noticing.
Questions for discussion
- Which of the three buckets would you have tested first, and what would have changed if the total had been below the escalation threshold?
- How would you decide whether a reviewer's sign-off on a tie-out workpaper is enough evidence when the arithmetic is correct but the cutoff rule was never written down?
- What evidence would convince an auditor that a suspense balance is being cleared rather than carried forward?
Case Study 2
Flat new business and rising repair costs break the plan
Keel Harbor Commercial Property Group, a mid-sized carrier writing commercial property, inland marine and contractor equipment, with a finance function of about 40 people.
The situation
Keel Harbor built its annual plan on two assumptions: new business would grow by a steady share each quarter, and repair costs on commercial property claims would track the prior cycle. By the second quarter, submissions ran below plan in three regions, and average severity on water and fire claims had risen enough to lift the loss ratio above threshold. The rolling forecast still showed the full-year combined ratio within tolerance, because reserve assumptions had been carried forward unchanged. Cash was not yet under strain, but the surplus buffer the board wanted to keep above a stated minimum was narrowing each month.
The challenge
The finance lead had to decide what to change before the next board meeting: re-forecast the year with updated severity and volume, or hold the plan and explain the variance. The board had asked for a defensible view rather than a rosy one, and the reinsurance renewal discussion depended on the surplus projection. Any scenario needed assumptions the team could justify from its own records.
Analysis
The first step was to stop treating the forecast as one number. The team rebuilt it as three scenarios, each tied to a specific driver found in the claims records. The base case kept submissions at the level of the last two quarters and assumed severity would settle at the current run rate. The pressure case added a fall in regional submissions and a severity trend drawn from the last four quarters of closed water claims. The severe case layered in a catastrophe-year loss load and tested whether the surplus minimum held. Each scenario was linked to a source report so each assumption had a named owner.
The response was built around triggers rather than a single cut. Management agreed that if the severe case showed surplus below the minimum in any quarter, the team would reduce new commercial property writings in the affected regions and raise the rate filing on water damage. A second trigger tied to the loss ratio would authorize a review of the reinsurance renewal budget. The team did not recommend reacting to the base case alone, since a single quarter of higher severity had not yet shown a durable trend. Each trigger had a named decision owner and a date by which the data would be reviewed.
- Test assumptions Compare each driver with the latest claims and submission records before rerunning.
- Rerun scenarios Run base, pressure and severe cases against the surplus minimum and loss ratio limit.
- Check triggers Confirm which thresholds were crossed and which named owner must act.
- Commit response Approve the agreed action and record the review date for the next cycle.
Actions taken
- Rebuild the reserve assumptions from the current claims run rate and test them against the last four quarters, instead of carrying forward the original figures without change.
- Write each scenario as a short table of drivers, values and sources, so the board can see which numbers are observed and which are judgment.
- Set explicit triggers for writing restrictions, rate filings and reinsurance review, each tied to a surplus or loss ratio level the board has approved in advance.
- Require two consecutive quarters below a trigger before a permanent pricing change, so that a single difficult quarter does not reset the rate plan.
Results
Over the following two quarters the team ran the trigger framework. The severe case did not fully materialize, but surplus came close to the minimum in the third quarter, enough to trigger a writing restriction in two regions. The rate filing on water damage was approved, and the reinsurance review produced a revised renewal budget about 6 percent higher. The board received a forecast that showed the range of outcomes and the action tied to each, rather than a single revised number. Two of the four original assumptions were revised, and the other two were confirmed with evidence.
Questions for discussion
- Which assumption in the rolling forecast would you test first, and what evidence would make you change it before the next board meeting?
- How would you set the trigger level for a writing restriction so that it responds to a real trend rather than one difficult quarter?
- When the base case and severe case disagree on whether the surplus minimum holds, how should the finance lead present that disagreement to the board?
Case Study 3
Cash premium pattern and mismatched addresses on one account
Vantage Lakeshore Surplus Lines, a non-admitted carrier placing hard-to-place commercial property risks through a small panel of wholesale brokers.
The situation
A wholesale broker submitted a commercial property account for coverage on a portfolio of six warehouses and light manufacturing sites. The premium was 412,000 dollars for the first year, and the broker proposed payment in cash-equivalent instruments through a third-party premium finance company that was not on the carrier's approved list. Within the first month, underwriting noticed that the named insured had changed twice, that three of the six insured addresses matched a single commercial mail drop, and that the broker had asked to bind before the loss runs were received. None of these facts alone was unusual. Together they formed a pattern the compliance unit had seen before in unrelated files.
The challenge
The underwriter wanted to quote with conditions, but the premium finance arrangement and the address pattern raised questions the underwriting file alone could not answer. The carrier had to decide whether the warning signs required a hold on binding, a referral to the financial crimes officer, a report under applicable anti-fraud or suspicious activity procedures, or simply tighter documentation. Each path carried different exposure, and a wrong choice in either direction would be hard to undo.
Analysis
The team began by sorting each indicator into what was documented and what was only inferred. The premium amount and the proposed payment route were on paper, and the premium finance company's identity could be checked against the approved list. The address overlap was an inference from a shared mail drop, which might be innocent for a small broker operating several entities. The request to bind early was a timing fact that depended on the broker's explanation. Keeping observed facts separate from inferences prevented the file from turning a suspicion into a conclusion, and it gave the compliance officer a clear record of what had been verified.
Escalation depended on the rules that applied to the carrier and the broker, and the team did not try to settle those rules from memory. The compliance unit reviewed the state requirements for surplus lines filings, the carrier's own anti-money laundering procedures for premium payments, and the internal threshold that required referral for any payment method outside the approved list. The third item was clear: a non-approved premium finance company triggered mandatory referral regardless of the other facts. The reporting question was less clear, so the officer asked outside counsel to confirm which obligations applied before any filing decision was made.
Actions taken
- Hold binding until the premium finance company is either approved or replaced with an approved lender, and document the reason in the underwriting file.
- Obtain independent verification of the named insured, the ownership of each address, and the mail drop arrangement, using records outside the broker's submission.
- Refer the file to the financial crimes officer with a written chronology of each indicator, its source, and whether it was verified or inferred.
- Record the outside counsel view on which filing obligations apply, and set a decision date so the account is not left in an open hold indefinitely.
Results
The account was not bound on the original terms. The broker replaced the premium finance company with an approved lender, and the officer confirmed the referral was closed. Two of the three addresses were verified as separate operating sites, while the third could not be tied to a legitimate occupant, so the carrier excluded that location. The reporting decision was made within the window counsel advised. The broker's other submissions were reviewed, and two were placed under the same referral step.
Questions for discussion
- Which of the indicators would you have weighted most heavily on the first day, and how would your answer change if the broker had supplied verified ownership records?
- What would make you comfortable binding with a conditional hold rather than a full referral, and who should have authority to make that call?
- How should a team record an inference in a compliance file so that it does not later read as a verified fact?
Case Study 4
Whether to back a managing agency's credit line request
Coldwater Gate Managing General Agency, a program manager binding contractor liability and equipment coverage for four appointed carriers.
The situation
Coldwater Gate asked Harlan Point Insurance Company, its largest capacity provider, for a 4,500,000 dollar revolving credit line to fund the timing gap between premium collected from contractors and remittance to carriers. The request came after volume grew faster than working capital over the last several quarters. Coldwater Gate reported a combined ratio below 97 percent on its bound book, but its premium receivables had aged past sixty days on about 18 percent of balances. The agency proposed to pledge its commission stream and a portion of its trust account as collateral, and it wanted a decision within the month so it could renew two large contractor programs.
The challenge
The credit committee had to decide whether the line was a sound extension of capacity to a growing agency or a way to cover premium that had already been collected and spent. The agency's trust account was the central question, because commingling risk would make any collateral worth less in a downside scenario. Approval conditions had to be specific enough that the committee could say no to a later draw without a dispute.
Analysis
The financial evidence was stronger on the loss side than on the cash side. Loss ratios on the contractor liability book had held within a narrow band across several quarters, and the agency's audited statements showed no related-party premiums. The cash position told a different story. Premium receivables over sixty days, the liquidity headroom under the existing facility, and the timing of remittances to carriers all pointed to a business that could be solvent on paper and still short of cash in a slow month. The committee therefore weighted liquidity more heavily than profitability when it scored the request.
The downside case assumed that one of the four carriers paused remittance acceptance for a month and that receivables aged further by fifteen points. Under that combination, the line would need to be drawn close to its full limit in the slowest month, and the trust account would be short of the amount owed to carriers unless the commission stream was pledged first. The committee set the approval threshold at a minimum liquidity coverage of 1.25 times peak draw in the downside case. The agency met that only if the line was reduced to 3,200,000 dollars and the collateral included a first claim on the commission stream.
Actions taken
- Approve a reduced revolving line of 3,200,000 dollars, with each draw permitted only against eligible receivables aged under sixty days, reported weekly.
- Take a first claim on the commission stream and a defined portion of the trust account, with a monthly certificate of balances from the agency's administrator.
- Set a 1.25 times liquidity coverage test on peak draw in the downside case, checked each month, with automatic suspension of new draws if the test fails.
- Require an independent review of trust account reconciliations before the first draw, and a standing right to inspect remittance records on fifteen days notice.
Results
The committee approved the reduced line with these conditions and declined the original request. Draws stayed well under the limit, and receivables over sixty days fell from 18 percent of balances to 12 percent by the end of the quarter. The trust reconciliation review found a timing gap of 41,000 dollars that the agency cleared before the second draw. The agency renewed both contractor programs, though at a lower commission share than it had proposed. Two conditions were tightened after a carrier reduced its appointments.
Questions for discussion
- Which of the four tests would you have scored differently, and what evidence would have moved that score?
- How much weight should a credit committee give to a strong loss ratio when receivables are aging and cash is tight?
- What conditions would you attach to a draw if the agency pushed back on the trust account reporting requirement?
Case Study 5
Five control gaps found in a commercial claims payment review
Greywater Mutual Assurance, a regional mutual writing personal lines and small commercial property, settling about 9,000 claims each quarter.
The situation
Greywater's internal audit team reviewed commercial property claim payments over the trailing two quarters, sampling 240 payments totaling 3,870,000 dollars. The review was scheduled as routine, but three sample items had payments that exceeded the authority limits of the adjusters who approved them, and two items had settlement amounts that did not match the final estimate in the claim system. Early results suggested the issues were not isolated. Reserve release entries had been posted without the supervisor review the procedures called for, and the payment approval matrix in the claims system had not been updated after a reorganization of the regional teams.
The challenge
The audit committee wanted the findings ranked by materiality rather than by how alarming they sounded, and it wanted each finding tied to a named owner and a closure test that someone outside the claims team could verify. Management had already started fixing the approval matrix, which raised a question about whether the fix would be treated as remediated before the old transactions had been tested.
Analysis
Ranking started with the money and the exposure, not the audit label. The reserve release entries were the most material, because the 287,000 dollars released without supervisor review had reduced reported reserves on open claims at the period end. Payments above limit totaled 186,500 dollars across three files, each later supported by a senior approval that was missing from the system at payment. The settlement mismatches were smaller in total but pointed to a data integrity problem that could affect any report built on the claim system. The approval matrix was the root cause of the breaches, so it was ranked by what it enabled rather than its size.
Closure tests were written so that a reviewer outside the claims team could confirm the result from evidence alone. For the reserve releases, the test required the supervisor review record for every release above the threshold of 25,000 dollars and a recalculated reserve for each affected claim. For the approval matrix, the test was a re-performance: the team pulled every payment above the old limits, checked the approval timestamp against the payment timestamp, and sampled the new matrix against the current delegation letters. A finding was not closed on management's word that the fix was in place. It closed when the test produced zero exceptions on a fresh sample.
- 1Rank by materialityOrder findings by dollar exposure, control role, and chance of recurrence.
- 2Assign named ownersEach finding gets one accountable executive and one operational owner in writing.
- 3Define closure testsEach test re-performs the control on a fresh sample with an exception threshold.
- 4Verify before closingClosure requires zero exceptions and a signed result, not a management statement.
Actions taken
- Rank the five findings by dollar exposure, control role, and likelihood of recurrence, and record the basis for each ranking in the audit file.
- Assign each finding to one accountable executive and one operational owner, and confirm both names in writing before the remediation plan is accepted.
- Write each closure test as a re-performance on a fresh sample, stating the population, the sample size, the exception threshold, and who signs the result.
- Hold the approval matrix finding open until transactions from before and after the update have both been tested, so that the fix is not mistaken for a cure.
Results
The reserve release finding closed first, after the team recalculated 61 affected claims and restated two reserves. The approval matrix finding stayed open longer. The first re-performance found four payments above limit approved after the update, traced to a delegation letter that had not been loaded. After the correction, a fresh sample of 90 payments produced no exceptions. The settlement mismatch came from a template default and was fixed, with a test added to the monthly close checklist.
Questions for discussion
- How would you rank a finding whose dollar exposure is small but whose root cause affects every other control in the claims system?
- What evidence would you accept as proof that a delegation letter has been loaded correctly, and who should be allowed to confirm it?
- When a management team says a fix is complete before any transactions have been tested, what is the minimum you would require before reporting the finding as closed?
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.