Insurance Loss Reserves: Estimating Future Claims

Key Takeaways

• Insurance loss reserves consist of case reserves for known claims and IBNR reserves for losses that have occurred but not yet been reported, together forming the insurer's largest balance sheet liability.
• US P&C insurers released reserves in aggregate for 18 consecutive years through 2023, consistently flattering reported combined ratios across the industry above their sustainable level.
• A common mistake is treating reserve releases as a persistent earnings boost; sudden large releases that close weak quarters are a warning sign of earnings management rather than disciplined reserving.
• Long-tail lines like workers' compensation and asbestos-related liability can produce adverse reserve development 20 or 30 years after the policy was written, making them the highest-risk segment to analyze.

What It Is

An insurer records a reserve for every policy obligation that has not yet been paid. There are two main categories:

• Case reserves: estimates for claims the insurer already knows about, file by file.
• IBNR reserves: incurred but not reported. These cover losses that have occurred but have not yet been notified to the insurer, plus further development on claims already reported.

Together, case reserves plus IBNR plus loss adjustment expense reserves form the loss reserves on the balance sheet. They are the largest liability for most P&C insurers and the primary source of float.

The Intuition

A policy written this year can generate claims that take years to report and decades to settle. A worker injured today may not file a claim for months. A product liability lawsuit tied to a policy written in 1995 could still be litigated today. The insurer has already taken the premium, so the cost of those future claims has to sit on the balance sheet as a liability.

Because the ultimate number is unknown, reserves are estimates. Every quarter, actuaries revisit those estimates in light of new claim data. When reality comes in better than the old estimate, the insurer releases reserves (a boost to current earnings). When reality is worse, the insurer strengthens reserves (a hit to earnings). Reserve adequacy is therefore both a capital question and an earnings-quality question.

How It Works

Actuaries estimate ultimate losses using claim triangles, which arrange paid or reported losses by accident year (rows) and development year (columns). The two most common methods are:

Chain ladder. Project future losses by extending the historical development pattern. If accident-year 2021 went from $100 million reported at 12 months to $130 million at 24 months, that 1.30 development factor is applied to 2022 at 12 months to estimate its 24-month value. The method is purely data-driven and works well for mature books with stable patterns.

Bornhuetter-Ferguson (BF). Blend the chain-ladder projection with an a priori expected loss, usually derived from an expected loss ratio applied to earned premium. The method was introduced by Bornhuetter and Ferguson in their 1972 paper "The Actuary and IBNR." It is the standard approach for recent accident years, where claim data is too sparse for pure chain ladder to be reliable.

The high-level formula for BF is:

Ultimate Loss = Reported Loss + (Expected Loss x (1 - Reporting Pattern))

Where the reporting pattern is the share of ultimate losses expected to be reported by the current development age. Early-stage accident years lean heavily on the expected-loss term; mature years lean on reported losses.

Long-tail lines (workers' compensation, general liability, asbestos, medical malpractice) develop over decades and carry the most reserve risk. Short-tail lines (property, auto physical damage) settle within 12 to 24 months.

Worked Example

An auto insurer's 2024 accident year has $400 million of reported losses at 12 months of development. Historical patterns show 80 percent of ultimate losses are typically reported by 12 months. The a priori expected ultimate loss ratio is 68 percent applied to $800 million of earned premium, giving $544 million of expected ultimate losses.

Chain ladder projection: 400 / 0.80 = $500 million ultimate. BF projection: 400 + 544 x (1 - 0.80) = 400 + 109 = $509 million ultimate. IBNR reserve (BF) = 509 - 400 = $109 million.

If a year later claims come in light and the 24-month triangle shows ultimate trending to $480 million, the insurer releases about $29 million of reserves, which flows through the current period's loss ratio as favorable development. Reversed, if inflation pushes ultimate to $540 million, the insurer strengthens reserves by $31 million, hurting reported earnings.

Common Mistakes

1. Taking favorable development as a sustainable trend. U.S. P&C insurers released reserves in aggregate for 18 consecutive years through 2023, which flattered reported combined ratios across the industry. That streak was exceptional and drove some recent analyses to assume continued releases, which has proven unsafe in commercial auto and general liability lines.

2. Ignoring social and economic inflation. Jury awards, litigation funding, and medical cost inflation can shift ultimate losses well above old estimates, especially in casualty lines. Reserves set in a low-inflation environment can prove inadequate even if claim counts behave as expected.

3. Missing the tail. Long-tail exposures (asbestos, environmental, pharmaceutical) can produce adverse development 20 or 30 years after the policy was written. Historical examples wiped out capital at carriers that thought the liability was fully reserved decades earlier.

4. Reading one year of development in isolation. Reserve adequacy is best judged across multiple years in the Schedule P section of statutory filings, where accident-year losses are re-stated at successive valuation dates. One benign year proves nothing.

5. Treating reserve releases as a free earnings boost. Consistent large releases can indicate either disciplined reserving in the past or current-period earnings management. Sustained single-digit release rates are normal; sudden multi-point releases that close an otherwise weak quarter are a warning sign.

Frequently Asked Questions

Q: What are insurance loss reserves in simple terms? Insurance loss reserves are the liabilities an insurer records today for claims it expects to pay in the future. They include case reserves for known open claims and IBNR reserves for losses that have occurred but not yet been reported to the insurer.

Q: How do insurance loss reserves affect investment decisions? Reserve adequacy is an earnings-quality question. When reserves prove excessive, the insurer releases them, which boosts reported earnings in a later period. When they prove inadequate, the insurer strengthens them, which hits earnings. Sustained large releases often inflate reported combined ratios above their true economic level.

Q: What is a real-world example of insurance loss reserves? In the worked example, a BF actuarial projection of $509 million versus a chain-ladder result of $500 million differs by $9 million from the same data. If claims later track to $480 million, the insurer releases $29 million into earnings. If they track to $540 million, the insurer takes a $31 million charge. The spread is entirely in the estimation.

Q: How can investors use insurance loss reserve analysis? Review Schedule P in statutory filings, which re-states accident-year losses at successive valuation dates. A pattern of reserves strengthening year after year signals chronic under-reserving. Also watch the spread between favorable development rates and industry averages, outlier release rates deserve scrutiny.

Q: How are insurance loss reserves different from the allowance for credit losses at banks? Both are forward-looking estimates of future losses, but insurance reserves cover a highly uncertain range of claim outcomes while bank ACL is modeled on a loan-by-loan or segment-by-segment basis under CECL. Insurance reserves are set by actuaries using statistical methods; bank ACL is set by CFOs using macro forecast models with different disclosure requirements.

Sources

1. S&P Global Ratings. "U.S. Property/Casualty Insurance Maintains Reserving Discipline Amid Higher Inflation" (2024). https://www.spglobal.com/ratings/en/research/articles/240531-u-s-property-casualty-insurance-maintains-reserving-discipline-amid-higher-inflation-13127158
2. Addactis. "Reserving in Insurance: Bornhuetter-Ferguson and Loss Ratio Methods." https://www.addactis.com/blog/reserving-bornhuetter-ferguson-loss-ratio-methods/
3. SIGMA Actuarial Consulting. "The Bornhuetter-Ferguson Method." https://www.sigmaactuary.com/2019/02/14/the-bornhuetter-ferguson-method/
4. Carrier Management. "After 18 Straight Years, Favorable Loss Reserve Development Streak Could End." https://www.carriermanagement.com/news/2024/01/18/257899.htm

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.