Economics & ClaimsAdvanced

IBNR

Q: IBNER (as distinct from pure IBNR) refers to:

Expected future development on already-reported claims — IBNER is the "not enough reported" component — case reserves that will strengthen. On excess layers it is often the larger and more dangerous of the two.

Q: Why do reinsurers typically carry a higher IBNR-to-reported ratio than their cedents?

They sit later in the reporting chain and on higher layers — Notification passes through the insured, the cedent and often a broker before it reaches the reinsurer — and losses only reach an excess layer after significant development.

Incurred but not reported

The reserve for claims that have happened but that nobody has told you about yet.

Open the interactive version → definition, quiz, structure diagram and progress tracking

Definition

IBNR is the reserve held for losses that have occurred but have not yet been reported, together with IBNER — incurred but not enough reported — the expected future development on claims already known. In reinsurance it matters more than anywhere else: a claim must be notified to the insured's broker, then the cedent, then the reinsurer, so the reporting lag through the chain can run to years.

Reinsurers are structurally later in the information chain and hold proportionally more IBNR than their cedents.
Excess-of-loss layers suffer severe IBNER: a claim reserved at $8M against a $10M attachment contributes nothing today and everything if it develops to $14M.
Estimated with chain-ladder development factors, expected loss ratio methods, or the Bornhuetter-Ferguson blend of the two.

BF ultimate = Reported incurred + (1 − % reported) × (Earned premium × Expected loss ratio)

IBNR is then the BF ultimate less the reported incurred.

Worked example

A casualty account with $150M of earned premium and a 68% expected loss ratio has an expected ultimate of $102M. If only 55% of losses have emerged and $61M is reported, the Bornhuetter-Ferguson ultimate is $106.9M and IBNR is $45.9M.

Scenario · figures in USD

Three methods, three answers, one accident year

Casualty accident year: earned premium $150M, expected loss ratio 68%, reported incurred to date $61M, and 55% of ultimate losses expected to have emerged at this maturity.

Earned premium	$150M
Reported incurred to date	$61M
Development factor (1 ÷ 0.55)	1.818
Chain ladder ultimate ($61M × 1.818)	$110.9M
Expected loss ratio ultimate (68% × $150M)	$102.0M
Bornhuetter-Ferguson ultimate	$106.9M
IBNR on the BF basis	$45.9M
Range across the three methods	$41.0M – $49.9M

So whatBornhuetter-Ferguson deliberately sits between the two: it trusts the emerging data where there is enough of it and the a priori loss ratio where there is not. At 55% reported, that blend is worth roughly $4M of reserve.

Check your understanding

IBNER (as distinct from pure IBNR) refers to:

Expected future development on already-reported claims. IBNER is the "not enough reported" component — case reserves that will strengthen. On excess layers it is often the larger and more dangerous of the two.

Why do reinsurers typically carry a higher IBNR-to-reported ratio than their cedents?

They sit later in the reporting chain and on higher layers. Notification passes through the insured, the cedent and often a broker before it reaches the reinsurer — and losses only reach an excess layer after significant development.

Word problem

A reinsurer holds an excess-of-loss account: earned premium $80M, expected loss ratio 72%, reported incurred $24M, and 40% of ultimate losses expected to have emerged at this maturity. Calculate the Bornhuetter-Ferguson ultimate and IBNR, and compare it with the pure chain-ladder figure.

Show a hint

Expected ultimate first, then apply the unreported proportion to it and add the reported figure.

Reveal the worked answer

Expected ultimate = 72% × $80M = $57.6M
Proportion unreported = 1 − 0.40 = 60%
BF ultimate = $24M + (60% × $57.6M) = $24M + $34.56M = $58.56M
BF IBNR = $58.56M − $24M = $34.56M
Chain ladder ultimate = $24M ÷ 0.40 = $60.0M; IBNR = $36.0M
Difference between the methods = $1.44M

BF ultimate $58.56M with IBNR of $34.56M, against a chain-ladder IBNR of $36.0M. At 40% reported the two methods are close because the reported experience ($60M implied) is running near the a priori expectation ($57.6M). Had reported incurred been $34M instead, chain ladder would have projected $85M while BF held at $68.6M — and the gap between them is the reserving judgement the actuary is actually paid to make.