Conditional Expected Loss
Definition
Conditional expected loss is the average share of principal an insurance-linked layer loses in the years in which it is hit. It equals the expected loss divided by the attachment probability. A high figure means that when the layer attaches it tends to be wiped out; a low figure means partial losses are more likely.
Why it matters
Expected loss alone mixes how often a layer is hit with how badly. Investors holding a portfolio of cat bonds care about both: frequent partial losses smooth into returns, while rare total losses arrive as a single drawdown. A conditional expected loss near 100% marks a cliff-like layer.
Common misconceptions
- •It is not the probability of loss; it is the average size of the loss once one occurs.
- •A high conditional expected loss is not a sign of a bad deal; it describes a narrow or stepped layer.
- •It is a model output, so it inherits the model's uncertainty.
- •The label is only a starting point. Investor outcomes usually depend on the governing documents, cash-flow priority, valuation policy, disclosure quality, and manager discretion surrounding the term, not just the shorthand definition used in marketing materials.
Technical details
Formula
Conditional expected loss = expected loss / attachment probability.
Bounds: exhaustion probability / attachment probability below, 100% above.
Jamaica 2024: floor 0.84% / 2.46% = 34%, actual 65%.
Reading it
Near 100%: once hit, usually exhausted (narrow layer or all-or-nothing payout).
Near the floor: losses usually stop well short of exhaustion.
Stepped payouts (30%, 70%, 100%) pull it toward the step levels.
How investors apply the term
Investors usually apply this term by tying the definition back to the controlling documents and then asking which cash flows, rights, restrictions, or valuation assumptions it changes. A useful review looks at the term alongside fees, leverage, liquidity, reporting frequency, and manager discretion rather than treating it as a standalone concept.
The practical diligence test is simple: identify who benefits if the assumption is wrong, who has control when conditions change, and what evidence the investor receives after closing. Terms that cannot be reconciled to documents, cash movements, and monitoring data should be treated as marketing shorthand until proven otherwise.
Frequently asked questions
- How do you calculate it?
- Divide annual expected loss by annual attachment probability. For Jamaica's 2024 World Bank cat bond, 1.61% / 2.46% is about 65%: when the bond is triggered, investors lose about two thirds of principal on average under the model.
- What does it look like across deals?
- From the World Bank case studies: Chile 2023, 1.00% / 1.48% = about 68%; Mexico 2024 class A, 0.90% / 1.17% = about 77%; class B, 5.84% / 8.30% = about 70%; class D, 4.09% / 6.26% = about 65%.
- Why does it matter if expected loss is already known?
- Two bonds with the same expected loss can behave very differently: one hit often for small amounts, another rarely but completely. Conditional expected loss separates frequency from severity.