Reading a Self-Funded Dashboard: PEPM, Budget Performance, and Paid vs. Incurred

Every month, a self-funded employer receives a claims experience report. Buried in it are the handful of numbers that say whether the plan year is on track: PMPM, loss ratio, the gap between paid and incurred claims, and an estimate for claims that have happened but not yet reached the ledger. Reading those numbers well is a skill, and it pays: the plans that get surprised in the fourth quarter are usually the ones that misread the second. This guide walks through each metric with worked numbers, using one example plan throughout: 400 employees, 900 covered members.

PMPM and Member Months

Raw claim dollars mislead the moment enrollment moves. A plan that added 60 members will spend more this March than last March even if nothing about its cost profile changed. PMPM, per member per month, is the normalization that makes months and years comparable.

A member month is one member covered for one month. Our example plan covers 900 members, so a full month contributes 900 member months and the first half of the year contributes 5,400. If the plan incurred $585,000 of medical and pharmacy claims in a month, that is $650 PMPM. Six months at $3.51 million is the same $650 PMPM, and now the number can be compared to last year, to the budget, and to benchmarks, regardless of how enrollment drifted in between.

Watch the denominator. Members means employees plus dependents; dividing by employees instead produces a number roughly twice as large (2.25 times in our example plan), and mixing the two in one exhibit is one of the most common reporting errors in the industry. The distinctions are covered in detail in PEPM, PMPM, and PEPY: the three metrics every benefits consultant should track.

Loss Ratio and What Good Looks Like

Loss ratio is claims divided by premium. In a fully insured plan the premium is real, and federal MLR rules require large-group carriers to spend at least 85 cents of every premium dollar on claims and quality improvement. In a self-funded plan there is no premium, so the denominator becomes the budget: the premium-equivalent rates the plan set at renewal to cover expected claims plus expenses.

Our example plan set its premium equivalents at $820 PMPM: $700 budgeted for claims plus $120 for administration and stop-loss. Incurring $650 of claims puts the loss ratio at 79%, which is favorable. Some reports show a second ratio, claims against the claims budget alone: $650 against $700 here, or 93%. The two differ by the full fixed-cost load, so check which basis an exhibit is using before judging the number. On the loss-ratio basis, anything above roughly 85% here means claims are eating into the money set aside for expenses, and above 100% they have consumed the entire budget; how urgent that is depends on duration and cause.

Calibrate "good" to the environment. Surveys published in the summer of 2026 put 2027 medical trend between 9 and 11 percent before plan changes, with PwC at 9.0 percent for the group market and Segal at 9.9 percent for PPO plans, the highest readings in nearly twenty years. Against that backdrop, a plan holding PMPM flat year over year is not merely on budget, it is outperforming the market by a wide margin.

Why Paid Is Not Incurred

A claim is incurred when the care happens and paid when the check clears, and the gap between the two dates is where dashboards go wrong. An inpatient stay in December may not be processed until February. On a paid basis it looks like a February problem in a new plan year; on an incurred basis it belongs to December, where it actually happened.

Paid-basis reporting distorts in two predictable ways. A growing plan looks artificially healthy, because enrollment shows up in the denominator immediately while its claims arrive months later. And a plan that terminates or changes funding mid-stream discovers runout: months of incurred-but-unpaid claims that still belong to the old plan. Serious analysis is done on incurred claims organized by the month of service, restated each cycle as payments complete.

There is a seasonal wrinkle on top of the lag. Most plans run calendar-year deductibles, so members shoulder more of the cost early in the year and the plan's share grows as deductibles are met. A clean incurred exhibit will show claims climbing from the first quarter to the fourth in a perfectly normal year. Comparing January to the prior December and declaring victory is the classic seasonal misread; compare each month to the same month a year earlier, or to a budget that carries the same shape.

IBNR and Completion Factors

If incurred claims are the true measure, and recent claims have not all been paid, then recent months must be estimated. IBNR, incurred but not reported, is that estimate: the liability for claims that have happened but have not yet surfaced. Its sibling, claims reported but not yet paid, is a separate reserve; financial statements usually lump both under the IBNR label. The standard tool is the completion factor: the share of a month's ultimate claims that has been paid so far. Administrators derive it from their book of business; a plan with a long paid history can fit its own, and industry curves stand in when it cannot.

Suppose the most recent month shows $390,000 paid and history says a month is about 60% complete at this point. Estimated incurred is $390,000 / 0.60 = $650,000, so the month's true cost is about $650,000 and the $260,000 gap is the reserve the plan's financials should carry. Without the adjustment, the freshest month always looks like the best month of the year, and every plan "improves" right before it gets expensive.

At any point in time the unpaid liability for a typical medical plan tends to run on the order of one to two months of average claims. That reserve is real money the plan owes; a year-end financial view that ignores it overstates how well the year went.

High-Cost Claimants Move Everything

The metrics above assume claims arrive as a smooth stream. They do not. Sun Life's 2026 high-cost claims report found $1 million-plus stop-loss claims rose 46% in frequency from 2022 to 2026. Segal's SHAPE analysis of 2019–2023 claims found members with claims above $100,000 are fewer than 1% of claimants yet account for roughly 30% of plan spend. A dashboard that does not track large claimants explicitly is hiding the single biggest driver of variance.

Run the arithmetic on our example plan. One $800,000 claimant adds $74 PMPM across the plan's 10,800 annual member months: an 11% addition to the $650 baseline from a single person. With specific stop-loss at a $200,000 deductible, the plan retains the first $200,000 (about $18.50 PMPM) and the carrier reimburses $600,000. That is why experience should be shown both gross and net of stop-loss reimbursements: a terrible-looking month that is fully reimbursed is not a budget problem, and a mediocre month concentrated in one claimant tells a different story than broad utilization growth. How the coverage itself works is the subject of our stop-loss guide.

A Monthly Reading Routine

Five checks, in order, every time a carrier report or TPA experience package lands:

Twenty minutes on these five checks each month is worth more than most year-end analysis, because a mid-year finding still leaves time to act: revise the budget, brief finance, or open the renewal conversation early.