Fundamentals

Case Reserve Strengthening: What It Looks Like in Your Triangle and What It Does to Your Reserves

When adjusters start setting higher case reserves, your reported triangle inflates, your development factors shift, and the chain ladder overstates the ultimate. Here is how to spot it, what it costs, and what to require from your actuary.

By Sam · April 2026 · 13 min read

You hired a new claims director. Or your TPA completed a case adequacy audit. Or a regulator flagged that initial reserves were consistently low. Whatever the trigger, your claims team started setting higher case reserves on new claims and, in some cases, re-evaluating open claims upward.

From an operational standpoint, this is good news. Stronger case reserves mean the filed amounts are closer to what claims will actually cost. But from a reserving standpoint, the improvement creates a distortion. The reported loss development triangle now contains a break: the most recent diagonals reflect the new, higher standard, while older diagonals reflect the old, lower standard. The age-to-age factors that bridge them are distorted, not because claims are developing more, but because the starting point moved up.

If the actuary does not recognize the shift, the chain ladder reads the inflated factors as genuine development and projects them forward. The result: an overstated ultimate for every accident year still developing through the affected maturities. Your reserve goes up not because your losses are worse, but because the model mistook an accounting change for a claims trend.

This article explains what case reserve strengthening looks like in the triangle, how it distorts the reported chain ladder, what the standard correction is, and what to ask your actuary when you suspect (or know) that case adequacy shifted during the experience period.

What case reserve strengthening actually is

Case reserves (also called case outstanding) are the adjuster’s estimate of the remaining cost to settle an individual open claim. When a new claim is reported, the adjuster evaluates the facts and posts an initial case reserve. As the claim develops (new medical bills, a deposition, a settlement demand), the adjuster revises the reserve up or down.

Case reserve strengthening occurs when the organization changes its reserving standard so that adjusters post higher initial reserves, or re-evaluate existing open claims upward, or both. The trigger is almost always one of a handful of events:

A new claims director establishes a higher adequacy standard.
The TPA completes an internal adequacy audit and raises reserves on a block of open claims.
A regulatory examination finds reserves are inadequate and requires correction.
The program transitions to a new TPA that uses different reserving guidelines.

The key distinction is that the underlying claims did not change. A $100,000 bodily injury claim is still a $100,000 claim. What changed is the number posted in the triangle at early maturities. That number is now higher, not because the claim is more expensive, but because the adjuster recognized more of the ultimate cost earlier.

How strengthening distorts the triangle

To see the distortion, consider a simplified reported triangle for a self-insured workers’ compensation program. Suppose the program has stable exposure and claim volume, and case reserves were set at a consistent adequacy level through accident years 2018 to 2022. Then, starting with the evaluation as of December 2023, case reserves were strengthened.

Before strengthening, a typical accident year might show $2.0 million in reported incurred losses at 12 months. After strengthening, the same exposure and claim volume produces $2.6 million at 12 months, because initial case reserves are 30% higher.

Now look at the age-to-age factors. For the pre-change years, the 12-to-24-month reported factor might be 1.35, meaning reported losses grew by 35% between the first and second evaluation. For the post-change year, the same factor drops to 1.15, because the claims started at a higher base but developed to roughly the same 24-month level. The lower factor does not mean the most recent accident year is developing better. It means the starting point was inflated by the adequacy change.

The distortion works in the other direction too. Look down the column of 12-to-24-month factors across accident years. If the actuary selects a weighted average and the most recent (lower) factor pulls the average down, the selected factor understates future development for all years. Alternatively, if the actuary excludes the most recent year and selects from the pre-change history, the selected factor overstates development for the post-change years, because those years already have more of the ultimate recognized at 12 months.

Neither selection is correct without an adjustment. The triangle contains two regimes (pre-change and post-change), and the development pattern from one regime does not apply to the other.

A numerical example

Consider five accident years, all with the same true ultimate of $4.0 million, evaluated at 12 and 24 months.

Before strengthening (accident years 2019 to 2022):

Reported at 12 months: $2.0M
Reported at 24 months: $2.7M
12-to-24 factor: 1.350

After strengthening (accident year 2023):

Reported at 12 months: $2.6M (same claims, higher case reserves)
Reported at 24 months: $2.9M (earlier recognition means less development remains)
12-to-24 factor: 1.115

The all-year average of the 12-to-24 factor across these five years: (1.350 + 1.350 + 1.350 + 1.350 + 1.115) / 5 = 1.303.

If the actuary applies the all-year average (1.303) to accident year 2023’s reported amount of $2.6M, the projected 24-month incurred is $3.39M. That is 18% higher than the actual $2.9M. The error compounds at every subsequent maturity, inflating the projected ultimate for 2023.

Now suppose the actuary recognizes the break and uses the pre-change average (1.350) for accident years 2019 to 2022 and the post-change factor (1.115) for 2023. The projections are accurate for both regimes.

But this approach has limits. It requires at least two post-change evaluations to establish a reliable post-change pattern. If the adequacy shift happened at the most recent evaluation, there is only one data point in the new regime, not enough to select a factor. This is where the Berquist-Sherman case adjustment becomes necessary.

The fix: Berquist-Sherman case adjustment

The standard correction for case reserve strengthening is the Berquist-Sherman case adjustment. The idea is to restate the entire case outstanding triangle to a consistent adequacy level (typically the most recent level), so that the development factors measure genuine claim development rather than the artifact of changing case practices (Friedland, p. 283).

The actuary works backward from the latest diagonal. Using a selected severity trend (which approximates annual claim cost inflation), the actuary deflates the latest case reserves back through the triangle to estimate what historical case reserves would have been if the current adequacy standard had been in place all along.

In the example above, restating the pre-change years’ 12-month reported amounts from $2.0M up toward $2.6M (adjusted for trend) produces a triangle where all accident years start from a comparable base. The restated development factors are consistent across years, and the chain ladder can operate without the distortion.

The severity trend is the critical assumption. A higher trend restates older diagonals more aggressively. A lower trend restates less. The projected ultimate is directly sensitive to this single parameter.

Alternatively, the actuary can sidestep the reported triangle entirely and rely more heavily on the paid triangle, which is unaffected by case adequacy changes. Or the actuary can use Bornhuetter-Ferguson for the affected accident years, which anchors part of the estimate to an expected claim ratio rather than the distorted development pattern. Neither alternative eliminates the problem, but both reduce the leverage of the distorted factors on the final estimate.

Why strengthening usually overstates the ultimate

The direction of the error is important and often counterintuitive. Case reserve strengthening is good claims practice: recognizing more of the ultimate cost earlier. But the reported chain ladder, applied without adjustment, interprets the inflated early-maturity amounts as evidence of higher ultimate losses rather than earlier recognition of the same losses. The result is an overstated ultimate.

This means the reserve in your financial statements may be too high after strengthening, not too low. For a CFO, that sounds conservative and therefore safe. But a materially overstated reserve has real costs: higher collateral requirements on letters of credit, larger actuarial central estimates that drive up excess insurance pricing, inflated self-insured retention funding, and balance sheet distortions that can affect borrowing capacity.

Overstatement is also fragile. If next year’s actuary recognizes the distortion and corrects for it, the reserve drops, producing favorable development that management may misinterpret as improved loss experience rather than a correction of a prior overstatement. This cycle of overshoot and correction erodes confidence in the reserve process.

Detecting strengthening before the actuary does

Not every case reserve shift is announced. Sometimes the change is gradual: a new claims supervisor raises reserves incrementally on their caseload, or a TPA rolls out updated reserving guidelines without flagging it to the policyholder. The actuary may not know the shift happened until the triangle shows it, and by then, the distorted factors have already affected the projection.

This is why buyers who monitor a few key metrics between annual reviews can identify a case adequacy shift before it distorts the reserve estimate. The diagnostic indicators include:

Average case outstanding at early maturities. If the average case reserve at 12 months increased materially between consecutive accident years without a corresponding increase in claim frequency or severity, case adequacy likely shifted.
Paid-to-incurred ratios. A declining ratio at early maturities (less paid relative to incurred) suggests case reserves are being set higher while payments have not caught up.
Reported development factors. A declining trend in the most recent diagonals of the reported triangle, without a corresponding change in the paid triangle, is the classic case strengthening signal.
Closed-without-payment rates. If these are stable, the change is in reserving practice, not in claim outcomes.

For a broader set of leading indicators that predict reserve distortions before they show up in the actuarial report, see Five Leading Indicators of Adverse Reserve Development.

The self-insured amplification

Case reserve strengthening hits self-insured programs harder than carriers, for the same reason every triangle distortion hits harder with thin data: there are fewer claims to average over.

A carrier with 5,000 claims per accident year sees the adequacy shift as a smooth change in aggregate development factors. The distortion is real but moderate, and the Berquist-Sherman adjustment operates on stable averages.

A self-insured program with 150 claims per accident year has no such cushion. If the new adequacy standard increases reserves on 30 claims by an average of $15,000 each, that is $450,000 of additional case outstanding in a year where total reported incurred might be $3 million. The 12-to-24-month factor for that year shifts materially, and the chain ladder projects the shift forward through every subsequent maturity.

The Berquist-Sherman adjustment for thin-data programs is also less reliable, because the severity trend must be estimated from a small number of claims. A handful of large claims that happen to coincide with the adequacy change can distort the trend estimate, and the restatement inherits that distortion.

For self-insured buyers, this means case reserve strengthening is one of the highest-impact operational changes in the reserving process. It happens frequently (TPA changes and claims leadership transitions are routine in the self-insured market), it distorts the most commonly used method (chain ladder), and the standard correction depends on a judgment-heavy assumption (the severity trend) that is hard to validate with limited data.

When broad IBNR is a large share of the total reserve, the stakes are higher still. The IBNER component (development on known claims) is exactly what case strengthening distorts. An inflated development pattern overstates IBNER, which overstates broad IBNR, which overstates the total reserve.

What a buyer should ask their actuary

These questions test whether case reserve strengthening was identified, quantified, and addressed.

1. Did case reserve adequacy change during the experience period? The actuary should be able to describe the direction and approximate magnitude of any shift, supported by data: average initial case reserves by accident year, paid-to-incurred ratios at consistent maturities, or information from claims management. If the actuary does not know, the diagnostic step was skipped.

2. How did the change affect the reported development factors? The actuary should show the age-to-age factors for the reported triangle with the affected diagonals highlighted, and explain whether the factors in those diagonals are higher or lower than the pre-change history.

3. What adjustment did you make? The options are: Berquist-Sherman case adjustment, exclusion of affected diagonals from factor selection, greater weight to the paid triangle (which is unaffected by case adequacy changes), use of Bornhuetter-Ferguson for affected years (which reduces leverage on the distorted development pattern), or a combination. Each choice has trade-offs, and the report should explain the reasoning.

4. If you applied Berquist-Sherman, what severity trend did you select? The trend should be documented and supported. Ask for the ultimate under at least two alternative trend assumptions (one higher, one lower) to see the sensitivity.

5. If you did not adjust, why do you believe the unadjusted triangle is reliable? There are defensible reasons: the shift was immaterial, the paid triangle was used instead, or the BF method anchored the estimate to an expected claim ratio rather than the development pattern. But “we did not observe a shift” is not a defensible answer when you, as the buyer, know one occurred.

What to require in documentation

A reserve report for a program that experienced case reserve strengthening should include:

A narrative describing the case adequacy change: what triggered it, when it took effect, and how it was identified.
Supporting data: average initial case reserves by accident year, paid-to-incurred ratios at early maturities, and reported development factors for the affected diagonals.
If Berquist-Sherman was applied: the selected severity trend with supporting data or benchmarks, the adjusted and unadjusted triangles side by side, and a sensitivity analysis showing the ultimate under alternative trends.
If Berquist-Sherman was not applied: the rationale for relying on the unadjusted triangle, and any alternative treatment (weighting to the paid triangle, using BF, excluding affected diagonals).
A comparison of the reserve indication with and without adjustment, so you can see the dollar impact of the case adequacy change on the projected ultimate.

If the report does not address case reserve adequacy at all, you are left evaluating a number without knowing whether one of the most common operational changes in the self-insured market was factored in. That is not a reserve analysis. It is a mechanical projection. For the broader set of documentation requirements that apply to any reserve report, see the diagnostic review guide.