Your combined ratio is still under 100, but renewals are closing at lower rates every quarter. The cushion that hid pricing errors for the past two years is thinning, and nobody on the team can say which accounts carry premium that no longer matches their risk.
Pricing accuracy decides whether a policy earns an underwriting profit or a loss. When the premium collected falls short of the exposure insured, the gap does not show up at binding. It shows up later, when claims develop against a premium that was never enough.
This article explains where pricing errors enter the policy lifecycle, how a softening market exposes them, and which five checkpoints a pricing or underwriting lead can review first.
What Does Pricing Accuracy Decide?
Pricing accuracy is the match between the premium charged and the true risk insured. The combined ratio scores that match: losses plus expenses, divided by earned premium, with anything under 100% marking an underwriting profit.
Because the ratio sits on earned premium, an error made at binding stays in the book until the policy expires and renews. A wrong price is not a one-time event. It is a recurring cost.
The Margin Cushion Is Shrinking
US P&C insurers posted an estimated net underwriting gain of $31.7 billion in the first half of 2026, with a combined ratio of 92.7 against 96.5 at midyear 2025, according to Verisk and the American Property Casualty Insurance Association (APCIA), as reported by Beinsure on September 10, 2026.
Premium growth moved the other way. Net written premiums rose 2.1% in the first half of 2026, compared with 5.2% a year earlier. APCIA said growth fell below general inflation and below increases in building material and labor costs.
Softening Rates Expose Small Errors
WTW reported that rates for large and complex property risks in its portfolio fell 14.5% in the second quarter of 2026, after an 8.4% decline a year earlier, as covered by Beinsure on October 5, 2026.
When rate is rising, a mispriced account rides on the increase. When rate is falling, the same account loses that cover.
Saurabh Khemka, president of Verisk Underwriting Solutions, cautioned that a strong half-year does not shrink the exposures behind it. He pointed to a more segmented property market as competition sharpens.
What Is Premium Leakage?
Premium leakage is the revenue an insurer loses when it collects less premium than the exposure warrants, through misclassified risk, incomplete data, fraud, or processing errors.
The term matters because it names the mechanism behind a pricing miss. A miss is no longer a vague "bad year." It is a specific failure at a specific point in the policy lifecycle.
How Large Is the Problem?
ReSource Pro's February 2026 research report states that most sources suggest P&C insurers lose more than $30 billion a year to leakage, excluding fraud. The report adds that industrywide figures vary.
The report also says the damage spreads past revenue. Systemic leakage distorts portfolio performance, weakens underwriting discipline, and erodes the data used to set future rates.
Where Does Leakage Show Up?
Leakage is not one department's problem. The same report maps it across the full policy lifecycle.
Lifecycle area | Typical leak |
|---|---|
Underwriting | Underpriced accounts, incomplete data, wrong class codes |
Distribution | Discounts applied without verifying eligibility |
Policy administration | Mid-term exposure changes never reported or corrected |
Premium audit | Findings that reach renewal underwriting late or not at all |
Claims | A mismatch between the risk priced and the actual cost of claims |
Why Does It Stay Hidden?
Leakage rarely announces itself. Smaller accounts are more prone to misclassification, in part because renewals roll forward without a fresh look. Middle-market accounts get more thorough underwriting and more frequent audits.
The report adds that renewal processes bypassing proper review perpetuate existing errors and compound leakage over time.
Where Does Pricing Go Wrong? The Rate Accuracy Chain
Pricing is a chain of five links. A weak link at any point passes its error to every link after it, and the combined ratio only reports the damage at the end.
We call this the Rate Accuracy Chain. Each link has a named mechanism, a typical failure, and an owner.
Link | What happens | Typical failure | Owner |
|---|---|---|---|
1. Capture | Exposure data enters at submission or application | Missing or misstated data: payroll, headcount, vehicles, locations | Underwriting, agents |
2. Classify | The risk is placed in a class code, territory, or segment | Wrong class code or misclassified exposure | Underwriting |
3. Rate | The rating engine applies filed rates, factors, and discounts | Discounts applied without an eligibility check | Pricing, underwriting |
4. Renew | The policy rolls forward with updated exposure | Errors carried forward when renewal skips review | Underwriting, operations |
5. Feed back | Audit and claims findings return to pricing | Findings stay in audit or claims files | Audit, claims, actuarial |
Links 1 and 2: Capture and Classify
Everything downstream trusts the first two links. In the report's commercial lines example, premium is payroll multiplied by a class-code rate, so an understated payroll or a wrong code changes the premium directly.
Knowledge gaps among less seasoned staff also distort the inputs, because the right questions about operations and employee duties never get asked. The input is wrong before the rating engine ever runs.
Links 3 and 4: Rate and Renew
Rating is where an error becomes a number. The engine applies discounts exactly as the inputs instruct it. If an unverified discount enters, the engine charges the lower premium and records it as correct.
Renewal is where the error repeats. Skip the review, and the same class code and the same exposure roll into another term.
Link 5: Feed Back
Premium audits check whether the premium reflects actual risk, and annual audits sharpen the data that underwriting and pricing rely on. But audits only help as far as they reach.
In one illustration from the report, a carrier audits only policies above $8,000 in annual premium. That leaves a substantial share of the book unexamined.
Claims findings face the same gap. The report says carriers should not treat claims leakage as separate from pricing, because claim outcomes bear directly on underwriting and price. When audit and claims data sit in separate systems, findings struggle to reach the people who set the rate.
How Do You Check Pricing Accuracy Before the Combined Ratio Shows It?
The combined ratio is a lagging score. By the time it moves, the policies that caused the move have already renewed.
Start With Leading Indicators
ReSource Pro closes its report with diagnostic questions for carriers. Four of them translate directly into checks a pricing lead can run this quarter.
Compare audit adjustments. Does the carrier bill additional premium or issue refunds, and in what proportion by line?
Trace how exposure data enters. Is it ingested and validated, or re-keyed by hand?
Review renewals for accuracy on a sample, not only on exceptions.
Confirm that trained staff and a defined process handle each leak once it is found.
Track the combined ratio by accident year as well as calendar year. A ratio creeping toward 100% across accident years signals rate inadequacy or rising expenses before a reported loss.
Segment Before You Average
Industry averages hide opposite stories. Verisk and APCIA reported that property pricing softened across much of the market in the first half of 2026 while casualty stayed under greater pressure. APCIA's Robert Gordon said bodily injury and commercial liability loss trends continued to deteriorate over the same period.
Check each line against its own trend. A single blended ratio hides a casualty problem behind a property cushion.
What Can a Pricing Team Do in the Next 90 Days?
A 90-day plan builds a leak list, not a finished fix. Corrected premium earns over the policy term, so the effect reaches the combined ratio across the next renewal cycle.
Weeks 1 to 3: Sample and Measure
Pull recent renewals from two lines. Compare the submission data against audit and claims findings, and record each mismatch against one of the five links.
Weeks 4 to 8: Trace Root Causes
Group the mismatches by link. If most sit at Capture, the intake is the problem. If most sit at Feed back, the handoff from audit and claims to pricing is the problem.
Weeks 9 to 13: Close the Loop
Assign an owner to each link and set a recurring review. Route audit and claims findings into underwriting before the next renewal cycle begins.
How InsOps Helps
InsOps builds an insurance-trained AI that assists underwriting teams in moving exposure data accurately from source systems into Guidewire. LiLa, our insurance-trained LLM, runs inside your own environment, so PII and PHI never leave controlled infrastructure. A person reviews and validates every mapping before it is finalized.
Our Integration Gateway provides pre-built connectors for Guidewire PolicyCenter, ClaimCenter, BillingCenter, UnderwritingCenter, PricingCenter, and Quoting Services. Policy, claims, and billing data flows directly into your rating and underwriting workflows without custom engineering, and validation stays with a person before anything deploys.
InsOps migrates legacy data into Guidewire and keeps it flowing in real time.
InsOps is building toward AI-assisted pricing recommendations inside LiLa, with a person reviewing every recommendation before it is used. If you are evaluating how to find pricing errors before they reach the combined ratio without adding manual reconciliation, contact us to talk through what this looks like for your operation.
Frequently Asked Questions
What is premium leakage?
Premium leakage is the gap between the premium an insurer should have collected and the premium it did collect. It starts at product design, underwriting, or policy issuance and continues through audit and claims.
What does a combined ratio above 100% mean?
It means losses and expenses exceeded the premium earned, so the underwriting operation lost money before any investment income. A ratio below 100% marks an underwriting profit.
How do insurers determine the premium for a policy?
Insurers estimate the expected cost of losses and expenses for a risk, using exposure data such as payroll, vehicles, location, or building characteristics. They then apply filed rates, class codes, and factors. A wrong input produces a wrong premium, which is why capture and classification carry so much weight.
What factors increase the combined ratio?
Anything that raises the loss ratio or the expense ratio: higher claim frequency or severity, catastrophe losses, claims inflation, and premium that fell short of the exposure at binding. Pricing errors act on the loss ratio, because the same claims land against less premium.
Where in the policy lifecycle do pricing errors enter?
At five links: capture, classify, rate, renew, and feed back. The earlier a leak starts, the more of the policy lifecycle it touches, because later links inherit the mistake.
How can AI assist with pricing accuracy?
An insurance-trained AI assists by mapping and validating the exposure data that moves between source systems and rating or underwriting workflows, with a person signing off before anything goes live. InsOps is building toward AI-assisted pricing recommendations inside LiLa. Contact us to talk through what this looks like for your operation.
How long does it take to see the effect of pricing corrections?
Building the first leak list takes about one quarter. The combined ratio responds more slowly, since each correction earns in as its policy renews. The common pitfall is judging results too early and dropping the review before a renewal cycle completes.