Every claim your program pays comes with a second stream of cost: the attorney who defends it, the adjuster who works the file, the claims system the adjuster logs into. That stream is loss adjustment expense, or LAE. If a claim is open at year-end, some of the cost of finishing it is still ahead of you, and that unpaid cost is a liability in the same sense the unpaid loss is.
For a self-insured employer, a captive, or a public entity pool, it is the part of the reserve most likely to be scoped out by accident. The report says “loss and ALAE” or “loss only” on page two, nobody in finance reads that line as a decision, and a real cost goes unrecorded or gets counted twice. This explainer covers what the two kinds of LAE are, how each is estimated, and the scope questions that determine whether your number is complete.
Two kinds of adjustment expense
The actuarial standard that governs reserve estimates, ASOP No. 43, defines claim adjustment expense as the costs of administering, determining coverage for, settling, or defending claims, even if the claim is ultimately found to be invalid. A claim you defend to a zero payment still cost money.
Actuaries have traditionally divided that cost in two. The Casualty Actuarial Society’s standard reserving text, Jacqueline Friedland’s Estimating Unpaid Claims Using Basic Techniques, describes the split this way:
- Allocated loss adjustment expense (ALAE) is cost that can be assigned to a specific claim. Friedland’s examples are legal fees, expert witnesses, police reports, engineering reports, and independent adjusters assigned to a particular claim.
- Unallocated loss adjustment expense (ULAE) is cost that cannot be assigned to one claim: the expense of operating the claims department, including rent, technology, salaries, and management and administrative expenses.
The practical difference is data. An allocated expense carries the accident date of the claim that generated it, so it can be put in a loss triangle and developed like loss. An unallocated expense has no accident date. It is a calendar year overhead number, and it takes a different kind of method.
DCC and A&O: the labels on the annual statement
On an NAIC annual statement you will not see ALAE and ULAE. You will see DCC and A&O. Friedland notes that the NAIC adopted these two categories effective January 1, 1998 for U.S. insurers reporting on Schedule P.
The NAIC’s accounting statement on unpaid losses, SSAP No. 55, defines them by function:
- Defense and cost containment (DCC) covers defense, litigation, and medical cost containment expenses, whether internal or external. The listed items include surveillance, litigation management, attorney fees incurred under a duty to defend, and the cost of engaging experts.
- Adjusting and other (A&O) covers the remaining loss adjustment expense. The listed items include fees and expenses of adjusters and settling agents and attorney fees incurred in determining coverage.
The two pairs of labels do not map one to one. ALAE versus ULAE asks whether the cost can be tied to a claim. DCC versus A&O asks what the cost was for. Friedland’s example is the adjuster: A&O includes all claim adjuster costs, whether they come from in-house staff, which look like overhead, or from an outside independent adjuster billing to a specific file, which looks like ALAE.
Friedland advises actuaries to investigate which claim expenses are included in the data and how the terms are defined. The same applies to report readers. If your TPA’s loss run has an “expense” column, find out what is in it before assuming it equals ALAE.
How ALAE is estimated
Because allocated expense has an accident date, Friedland says all of the standard development techniques can be used on it. In practice there are three common treatments.
Combined with loss. Many actuaries simply add paid ALAE to paid loss and develop the total with the chain ladder and Bornhuetter-Ferguson methods. Friedland attributes this to data limitations: many insurers record only ALAE payments and set no separate case reserve for expense. The caution in the text is that for some lines the ALAE pattern differs significantly from the loss pattern. On third-party liability claims, defense and expert costs can be incurred well before any payment to the claimant, and some costs lag the settlement.
ALAE developed on its own. Paid ALAE is put in its own triangle and developed with its own loss development factors. Where separate ALAE case reserves exist, a reported ALAE triangle can be used as well.
The ratio of paid ALAE to paid loss. The actuary builds a triangle of ratios, paid ALAE divided by paid loss at each age, develops the ratio to ultimate, and multiplies the ultimate ratio by the separately estimated ultimate loss. Friedland lists the trade-offs. The ratio recognizes the relationship between expense and loss, and its development factors tend to be less leveraged than those on ALAE dollars. On the other side, any error in ultimate loss flows straight into ultimate ALAE, and the method struggles where large amounts of ALAE are spent on claims that close with no loss payment. It also assumes a stable relationship between ALAE and loss, which a change in defense strategy can break.
In Friedland’s worked example for an auto property damage book, paid ALAE developed more slowly than paid loss: at 12 months the implied share paid was 46.8 percent for ALAE against 63.1 percent for loss. That is one illustrative book, not a benchmark. But if expense lags loss in your program and the report uses a loss pattern for the combined total, the tail may be short.
The Wire has covered what a shift in defense cost does to a self-insured program in practice, including Kemper’s California auto defense cost development and the WCRI finding on California workers compensation litigation expense.
How ULAE is estimated
ULAE has no accident year, so the methods work from the relationship between what the claims operation costs in a calendar year and the volume of work it did. Friedland’s chapter on the subject, which relies heavily on a 2003 CAS Forum paper by Robert Conger and Alejandra Nolibos, sorts the techniques into dollar-based and count-based families.
The classical paid-to-paid method
The classical technique compares calendar year paid ULAE to calendar year paid claims, selects a ratio, and applies it to the unpaid claim estimate. The well-known feature is the 50/50 assumption: half of the unallocated work on a claim is done when it is opened and half when it is closed. So the method applies:
- 50 percent of the ULAE ratio to case reserves, because those claims are already open and half of their handling cost is behind you, and
- 100 percent of the ULAE ratio to IBNR, because for claims not yet reported all of the work is still to come.
The method uses data every program has. Friedland also sets out its problems. Paying a claim and closing a claim are not the same event on lines such as workers compensation, where payments continue for years. The ratio is distorted when claim volume is growing or shrinking, because paid ULAE responds quickly to new volume and paid claims do not. The text concludes that the classical technique may not be appropriate for long-tail lines, periods of changing inflation, rapid changes in volume, or operations where 50/50 does not describe the workflow.
There is a further subtlety. The IBNR in most reports is broad IBNR: it includes development on claims that are already open and already set up. Friedland notes that in theory the 100 percent factor belongs only on pure IBNR, with 50 percent applied to case reserves plus the development on known claims, and shows that making this correction in the text’s example lowered the indicated unpaid ULAE by about 24 percent.
The Kittel refinement
John Kittel’s refinement addresses the growth problem. His observation was that a claims department also opens claims, and paid losses miss the work done on claims opened during the year that are still open at year-end. The fix is to change the denominator. Instead of dividing paid ULAE by paid claims, divide it by the average of calendar year paid claims and calendar year incurred claims, where incurred means paid plus the change in total claim liabilities, case reserves and IBNR together. The 50/50 application to case reserves and IBNR is unchanged. For a growing program, incurred exceeds paid, the denominator is larger, and the selected ratio is lower.
The Conger and Nolibos generalized approach
Conger and Nolibos extended Kittel’s idea by dropping the fixed 50/50 split. Their approach assigns weights to three stages of a claim’s life: U1 for the share of ULAE spent opening claims, U2 for the share spent maintaining them, and U3 for the share spent closing them, with the three summing to 100 percent. Each calendar year’s ULAE is compared to a weighted claims basis:
Claims basis = (U1 x ultimate cost of claims reported in the period)
- (U2 x claims paid in the period) + (U3 x ultimate cost of claims closed in the period)
The authors’ preferred estimate of unpaid ULAE multiplies the selected ratio of ULAE to that basis by ultimate claims less the claims basis to date. Like Bornhuetter-Ferguson, it does not let the amount already paid drive the answer directly.
The weights can reflect how your claims are actually handled. Friedland notes that a litigation-heavy liability book concentrates effort near settlement, while workers compensation claims carry more front-end cost. Conger and Nolibos derived their weights from interviews with claims personnel and tested the sensitivity of the result.
Count-based methods
Dollar-based methods assume a $1,000 claim takes ten times the handling effort of a $100 claim. Count-based methods assume the opposite: the same kind of transaction costs the same amount whatever the claim size. The approach associated with Wendy Johnson projects the number of claims that will be newly reported and the number that will be open in each future year, weights those counts by the relative effort of each activity, and prices them using the historical relationship between ULAE spent and weighted counts. A claim that stays open for several years is counted in each of them.
Friedland identifies data as one of the most significant challenges: accurate and consistent claim count and transaction information is often not available. A self-insured program whose TPA bills per claim may already have those counts.
The TPA contract decides whether you have a ULAE liability
Most self-insureds pay a third-party administrator instead of running a claims department, and the fee looks like a period expense. Whether there is a liability behind it depends on what the fee already paid has bought.
Friedland makes the connection directly: ULAE liabilities have a “market value” in the fees a TPA would require to take over management of the book of claims, and many self-insurers use such market values to determine unpaid ULAE for financial reporting. That makes it a contract-reading exercise:
- Life-of-claim or “cradle to grave” fee, paid up front. If the fee paid when a claim was reported obligates the TPA to handle it to closure at no further charge, the future handling cost of that claim has been prepaid. What may remain is the handling cost of claims not yet reported, your own internal claims oversight cost, and the risk that the contract ends before the claims do.
- Annual fee per open claim, or a flat annual fee. Every claim open at year-end will generate fees in each future year until it closes. That is unpaid ULAE in the plain sense, and the count-based logic above fits it naturally.
- Termination and run-off terms. If either party can end the contract with claims still open, someone will charge to finish them.
These are contract patterns, not accounting conclusions. The accounting depends on your framework. For entities on NAIC statutory accounting, SSAP No. 55 is explicit: the liability for unpaid loss adjustment expense is to be established regardless of any payments made to third-party administrators, with an exception for capitated payments under managed care contracts. For GAAP reporting, see ASC 450 vs. ASC 944, which covers how adjustment expense is treated differently for a captive and for a parent that retains risk directly. Public entities and pools reporting under GASB Statement No. 10 have their own rules on which adjustment expenses belong in the claims liability and should confirm the treatment with their auditor.
Does ALAE erode the retention?
For a program with excess insurance or reinsurance above a retention, the treatment of defense cost under the excess contract changes the retained reserve. Friedland lists three possible treatments of ALAE in excess of loss reinsurance:
- included with the claim amount in determining excess coverage, which the text calls the most common treatment,
- not included in the coverage, and
- included on a pro rata basis, with the excess share of the loss determining the excess share of ALAE.
Take an illustrative claim with $800,000 of loss and $200,000 of defense cost under a $500,000 retention.
| ALAE treatment | Retained by the program | Paid by the excess carrier |
|---|---|---|
| Included with loss (erodes the retention) | $500,000 | $500,000 |
| Pro rata (excess pays 300/800, or 37.5%, of ALAE) | $625,000 | $375,000 |
| Not covered by the excess contract | $700,000 | $300,000 |
Same claim, three different retained amounts. If ALAE erodes the retention, the actuary should cap loss and ALAE combined at the retention, and a combined triangle is the natural data. If ALAE sits outside, loss is capped and ALAE is unlimited, so the two usually need to be developed separately and the retained ALAE on large claims has no ceiling. The mechanics of retentions are covered in Self-Insurance, Captives, Large Deductibles, and SIRs, and the gross and net layers in Fronting, Reinsurance, and Why Your Captive’s Net IBNR Can Exceed Gross.
A worked ULAE illustration
The figures below are hypothetical and chosen for round arithmetic. A group captive runs its own claims unit. Its recent history:
| Calendar year | Paid ULAE | Paid loss and ALAE | Ratio |
|---|---|---|---|
| 2023 | $380,000 | $4,000,000 | 9.5% |
| 2024 | $420,000 | $4,200,000 | 10.0% |
| 2025 | $450,000 | $4,400,000 | 10.2% |
At year-end 2025 the reserve study shows case reserves of $6,000,000 and broad IBNR of $4,000,000. The actuary selects a 10 percent ratio.
Classical method. (10% x 50% x $6,000,000) + (10% x 100% x $4,000,000) = $300,000 + $400,000 = $700,000.
Classical method with the pure IBNR correction. Suppose $800,000 of the IBNR is for unreported claims and $3,200,000 is development on known claims. (10% x 50% x $9,200,000) + (10% x 100% x $800,000) = $460,000 + $80,000 = $540,000.
Kittel refinement. The captive is growing. In 2025 total claim liabilities rose by $1,200,000, so incurred claims were $5,600,000 and the average of paid and incurred was $5,000,000. The 2025 ratio becomes $450,000 / $5,000,000 = 9.0 percent. Applied on the classical basis: (9% x 50% x $6,000,000) + (9% x 100% x $4,000,000) = $630,000.
Count-based cross-check. Suppose the actuary projects that claims already incurred will require 340 open-claim-years of handling in 2026, 190 in 2027, 100 in 2028, 50 in 2029, and 60 in total after that: 740 claim-years. At an assumed all-in cost of $900 per open claim per year, unpaid ULAE is 740 x $900 = $666,000, before any allowance for inflation in that cost.
Four defensible calculations give a range of $540,000 to $700,000 on $10,000,000 of unpaid loss and ALAE. The spread is the reason to ask which method was used and why it fits the way your claims are handled.
What ASOP 43 requires
ASOP No. 43, Property/Casualty Unpaid Claim Estimates, was adopted by the Actuarial Standards Board in June 2007 and updated for deviation language effective May 1, 2011. Its scope section says it applies to estimates for all classes of entities, and names self-insureds alongside insurance companies, reinsurers, and governmental entities.
Three points bear on adjustment expense:
- The purpose section states that any reference to “unpaid claims” in the standard includes the associated unpaid claim adjustment expense, unless explicitly stated otherwise.
- In identifying the scope of the estimate, the actuary should identify the specific types of unpaid claim adjustment expenses covered, with coverage dispute costs, defense costs, and adjusting costs given as examples.
- The standard names no required method and leaves the selection of methods or models to the actuary. In comments on the second exposure draft, a commentator asked for specific guidance on prepaid expenses and TPAs; the reviewers declined, calling prepayment to TPAs a specific situation too detailed for the standard’s general guidance.
The standard makes expense scope something the actuary identifies and leaves the TPA question to judgment and the facts of your contract. The scope statement is one of the first things to check when you evaluate an actuary’s report.
Questions to ask your actuary
- Is the estimate loss only, loss and ALAE, or loss and all adjustment expense? Where in the report is that stated?
- What does the “expense” field in our loss run contain, and has its definition changed during the experience period?
- Was ALAE developed with loss or separately? If combined, what evidence shows the two develop on a similar pattern?
- Does ALAE erode our retention under each policy year’s excess terms, and does the capping in the analysis match?
- Is there a ULAE provision? If not, is that because the TPA contract prepays handling to closure, or because it was outside the engagement scope?
- Which ULAE method was used, what ratio or cost per claim was selected, and what would a second method indicate?
- If our TPA contract ended tomorrow, what would it cost to have the open claims handled to closure, and how does that compare with the booked amount?
Sources
- Estimating Unpaid Claims Using Basic Techniques, Jacqueline Friedland, Casualty Actuarial Society, Version 3, July 30, 2010 (Chapters 1, 3, 16, and 17)
- Estimating ULAE Liabilities: Rediscovering and Expanding Kittel’s Approach, Robert F. Conger and Alejandra Nolibos, Casualty Actuarial Society Forum, Fall 2003
- Actuarial Standard of Practice No. 43, Property/Casualty Unpaid Claim Estimates, Actuarial Standards Board, adopted June 2007, updated May 2011
- Statement of Statutory Accounting Principles No. 55, Unpaid Claims, Losses and Loss Adjustment Expenses, National Association of Insurance Commissioners
Frequently asked questions
What is the difference between ALAE and ULAE?
ALAE is loss adjustment expense that can be assigned to a specific claim, such as defense attorney fees, expert witnesses, and independent adjusters assigned to that claim. ULAE is the general cost of operating the claims function, such as salaries, rent, and technology, which cannot be tied to one claim.
How is a ULAE reserve calculated?
The classical method selects a ratio of calendar year paid ULAE to calendar year paid claims, then applies 50 percent of that ratio to case reserves and 100 percent to IBNR. Refinements described in Casualty Actuarial Society literature change the ratio's denominator, replace the 50/50 split with program-specific weights, or use claim counts and a cost per claim handled.
Does a self-insured that uses a TPA need a ULAE reserve?
It depends on the contract. If the fee already paid obligates the TPA to handle existing claims to closure, little future handling cost remains; if the TPA bills annually per open claim, or the contract can end with claims still open, the cost of handling those claims in future years is still unpaid.
Does ALAE erode a self-insured retention?
Only if the excess policy says so. Defense costs can be included with the loss in measuring the retention, excluded from the excess coverage entirely, or shared pro rata, and each treatment produces a different retained reserve for the same claims.