The recoverable is decided by clauses, not by the loss
After a cyclone makes landfall or a monsoon flood inundates an industrial cluster, a property insurer adds up its damaged risks and assumes the reinsurance recoverable follows the loss. It does not, at least not directly. What the catastrophe treaty pays back is decided by a handful of clauses that were agreed at the last renewal and that most direct underwriters never look at closely: the hours clause, the number and terms of reinstatements, and the event definition.
These clauses turn a pile of individual losses into a number of recoverable events, and the number of events, not the total damage, drives what comes back. A flood that runs for several days could be one event or several depending on a time window. A bad season with two storms could exhaust the cover or sit comfortably inside it depending on how many reinstatements there are. This post takes each mechanic in turn and shows why a direct underwriter who understands them writes accumulation differently from one who does not.
The hours clause: turning a span of damage into events
A catastrophe rarely strikes in an instant. A cyclone tracks for days, a flood builds and recedes over a week, a riot flares across a city over several days. The reinsurance treaty needs a rule to decide which of those losses belong to one occurrence, and that rule is the hours clause.
An hours clause groups losses from a catastrophe into a single event using a time window, commonly cited as 72 hours for windstorm, 168 hours for flood or riot, and 24 hours for earthquake. Losses occurring within the stated window from a single catastrophe are treated as one occurrence under the treaty; losses outside it fall into a different occurrence.
Why the window changes the recoverable
The length of the window directly shapes the recoverable. Take a monsoon flood that damages a property book over several days. Under a 168-hour flood clause, all the damage inside that one-week window is a single event, so the insurer makes one claim against one retention and one limit. If the flood instead straddled the window, or were tested against a shorter clause, the same damage could split into two events, two retentions and potentially two reinstatements consumed.
The direction of effect cuts both ways. One large event means the insurer absorbs only one retention but is capped at one event limit. Splitting into multiple events means more retentions to absorb but access to more aggregate limit. Neither is simply better; what matters is that the underwriter knows which way the clause cuts for the perils its book is exposed to, because that is what its net result will actually look like after a bad season.
Reinstatements: how many times the layer can be tapped
A catastrophe excess of loss layer is not an unlimited bucket. Once a loss erodes the layer, the cover for the next event depends on whether the layer reinstates, and on what terms. This is where reinstatements come in.
Most catastrophe excess of loss programmes provide a limited number of reinstatements, commonly one to three, at pre-agreed terms. The number of reinstatements caps the maximum recoverable across the treaty year at the layer limit multiplied by one plus the number of reinstatements. A layer with two reinstatements can pay up to three times its limit across the year (the original limit plus two reinstatements), and no more, however many further events occur.
This is the clause that decides whether a book survives a bad season or runs out of cover partway through it. In a benign year with one event, reinstatements are irrelevant; the layer pays once and resets. In a year with several catastrophes, the number of reinstatements is the difference between continuous protection and a book that exhausts its catastrophe cover and carries the next event net.
Reinstatement premium and the event definition
Reinstating a layer is not free, and the way it is priced affects the economics of a heavy loss year.
What it costs to put the cover back
Reinstatement premium is typically charged at a stated percentage, often 100 percent additional premium, pro rata to amount and pro rata to time, to restore the layer after a loss payment. Pro rata to amount means the cost scales with how much of the layer was used; pro rata to time means it scales with how much of the treaty year remains to be protected. So an early-season event that uses much of the layer triggers a larger reinstatement premium than a late, partial one. For the cedant this is a real cash cost arriving precisely when it is also paying claims, which belongs in any view of the net cost of a catastrophe year.
What counts as one event, and the clash question
The event definition in the treaty does the deeper work of saying what a single occurrence actually is, and it governs clash, the situation where losses from different lines or sources are aggregated into one event. The hours clause is the time test inside that definition; the wider event definition decides whether geographically or causally linked losses combine. How widely or narrowly an event is defined determines whether several pockets of damage roll up into one occurrence (one retention, one limit) or stay separate. For a book exposed to widespread perils like flood and cyclone, the event definition is as consequential as the limit itself, because it decides how accumulation crystallises into recoverable occurrences.
Why a direct underwriter should care about treaty clauses
These clauses can feel like the reinsurance department's problem, several steps removed from the underwriter deciding whether to write another factory in a flood-prone cluster. They are not. The treaty mechanics set the real boundary on how much catastrophe-exposed risk a book can safely accumulate, because they decide what actually comes back when the season turns bad.
The practical reading for a direct underwriter is:
- Think in events, not totals. The hours clause and event definition convert damage into recognised occurrences, and it is occurrences, against retentions and limits, that drive the recoverable. Accumulation should be assessed in the events the treaty will recognise.
- Respect the reinstatement count. The number of reinstatements caps annual recovery at the limit times one plus the reinstatements, so a book that could plausibly see several events in a season should not be accumulated as if the cover were inexhaustible.
- Price in the reinstatement premium. A heavy loss year carries the cash cost of putting the layer back, pro rata to amount and time, which is part of the true net cost of catastrophe writing.
- Work within the 1 April terms. The clauses are fixed for the year at the principal renewal, so the season is run on terms set before it began, and the underwriting through the year has to live inside them.
An underwriter who understands the hours clause, the reinstatements and the event definition writes catastrophe accumulation with a clear view of what the treaty will and will not return, rather than discovering the limits of the cover only after a cyclone has tested them. Reading those treaty terms against the property wordings and accumulations a book actually carries is where structured market intelligence earns its place. Sarvada gives commercial insurance brokers structured, searchable access to insurer policy wordings and the intelligence around them, so catastrophe accumulation is managed against real wording and capacity detail. Request Access to bring treaty-aware discipline to your catastrophe-exposed property book.
