Every answer and its explanation appears here once you have finished the path. Each one then links to the matching glossary entry, where the concept is set out in full with its worked example.
1. A cedant replaces a quota share with an excess of loss treaty and is surprised its margin requirement does not move. What must be explained?
An excess of loss treaty does not reduce volume, does not free capital the way a proportional cession does, and pays no commission: there is no pro rata ceded premium but a premium specific to the layer
This is the exact mirror of the quota share: one transfers volume, the other severity, and what you buy is not interchangeable. The three other answers concede the effect and argue about its terms, which is the most natural way to go wrong here. The one year lag describes what happens under a proportional cession, where reserves do leave. The priority level changes the price and the residual exposure, never the nature of the operation. And offsetting by the cost presupposes relief that never occurred.
Glossary entry · exces-de-sinistre2. Why is an excess of loss treaty expensive per unit of capital protected, and why is that logical rather than abusive?
The reinsurer sells a rare exposure whose cost it cannot offset with volume: it ties up capacity during the years it is not used, and those years bring in only a premium
The price of a layer pays for capacity held available, and a layer touched one year in ten ties up nine years for one premium. The same reasoning explains why a peak layer places at two percent of its limit and a working layer at forty. The answer on brokerage names a real and marginal cost. The one on reserving the whole limit describes an accounting practice that does not exist, a commitment not being reserved while no loss triggers it. The one on a proportional regulatory margin invents a rule and confuses economic capital with reserves.
Glossary entry · rate-on-line3. A layer of ten in excess of five is hit by a loss of thirty. What happens to the fifteen above?
They fall on the cedant if no higher layer exists: the layer saturates at ten and answers no further beyond fifteen
A non-proportional treaty shares nothing: it allocates slices, and whatever exceeds the last slice bought returns to whoever did not buy the next. The answer on an overrun clause invents a mechanism. The one settling pro rata imports proportional logic into the family that is defined against it. The one on reinstatement makes the most instructive error: reinstatement restores the limit for SUBSEQUENT losses, it never reopens a saturated layer for the loss that saturated it.
Glossary entry · epuisement-de-tranche4. A cedant budgets its reinsurance premiums without providing for reinstatements. When does it discover its real spend, and why does the timing make it worse?
On the day it can least afford it: the reinstatement premium is called right after a loss, pro rata to the amount consumed and often to the time remaining
The simultaneity is what makes the omission expensive: the cash that must pay the reinstatement is the cash that has just paid a loss. The three other answers all push the due date back, and that is exactly the belief that drops the provision from the budget. Year end settlement and building it into the renewal describe account practices that change nothing about when it falls due. The one waiting for final settlement confuses the amount consumed, known as soon as the first payments are made, with the ultimate cost of the loss.
Glossary entry · reconstitution-garantie5. A first layer is placed very high, above a large retention. How is it characterized, and what does that govern?
By what it sees, not by its rank: placed high, it can perfectly well be a peak layer, and it is then priced on an exposure model rather than on experience
The vocabulary is defined by how often the layer is hit, never by its position in the stack, and the consequence is practical: a working layer is priced on experience because there are losses to count, a peak layer on a model because there are none. The answer reasoning by rank is the most natural and confuses stacking order with exposure. The one reasoning by peril keeps a frequent correlation and takes it for a rule. The one proposing a limit to priority ratio manufactures an arithmetic test where an observation is required.
Glossary entry · working-layer-cat-layer