Products and contracts

Reinsurance risk is transferred through contracts. A contract defines precisely which losses the reinsurer covers, under what conditions, and up to what amount. Understanding contract structures at a conceptual level is essential groundwork for the mathematical formalization in Quantification and the software modeling in Engineering.

The two families of reinsurance

All reinsurance contracts fall into one of two families based on how they share losses between the cedent and the reinsurer.

Proportional reinsurance

In proportional reinsurance, the reinsurer takes a fixed percentage of every loss (and receives the same percentage of premium). The sharing is simple and predictable.

The most common proportional structure is the quota share: the reinsurer agrees to take, say, 25% of all losses in a defined portfolio, in exchange for 25% of the premium.

Key properties of proportional contracts:

• The reinsurer participates in every loss, large or small
• The sharing ratio is fixed (or formula-based in some variants)
• Premium and loss move together — if the cedent collects more premium, the reinsurer gets more too
• Simpler to administer but less capital-efficient for the cedent
• The reinsurer must trust — or independently verify — that the cedent is pricing the underlying risk correctly, since the reinsurer inherits the cedent’s risk selection and pricing quality

Helios Re

In Helios Re’s portfolio, Contract 3 (European Windstorm Quota Share) is proportional: Helios Re takes 25% of all European windstorm losses from the syndicated cedent.

Non-proportional reinsurance (excess of loss)

In non-proportional (or excess of loss) reinsurance, the reinsurer only pays when losses exceed a specified threshold. The reinsurer is selling protection against the tail of the loss distribution — it collects premium in exchange for absorbing large losses, while small losses remain entirely with the cedent.

The most common non-proportional structure is the CatXoL (Catastrophe Excess of Loss): the reinsurer covers losses between an attachment point and a limit for a single catastrophe occurrence.

The chart above shows a CatXoL contract with $30M limit excess of $20M (written as “30M xs 20M”). Use the slider to explore how different occurrence loss amounts are split:

• Losses below the attachment point ($20M) are entirely retained by the cedent
• Losses within the contract layer ($20M to $50M) are ceded to the reinsurer
• Losses above the exhaustion point ($50M) are not covered — they fall back to the cedent

Key properties of non-proportional contracts:

• The reinsurer only pays for large losses that exceed the attachment point
• The maximum payout is capped at the limit
• Premium is typically a fraction of the limit — because the attachment excludes frequent, smaller losses, the expected loss within the layer is much lower than the total expected loss, and the premium reflects this
• More capital-efficient for the cedent: they get protection exactly where they need it (the tail)
• More complex to price: the reinsurer must model the probability and severity of losses in the layer

Helios Re

Three of Helios Re’s five contracts are CatXoL: Contract 1 (Florida Hurricane), Contract 2 (California Earthquake), and Contract 4 (Japan Earthquake).

Common Misconception

It is easy to conflate “retention” or “attachment point” with “deductible.” They are related but not the same. A deductible is the amount the insured pays before insurance kicks in. An attachment point is the loss level at which the reinsurer starts paying the insurer. The concepts are analogous but exist at different layers of the risk transfer chain.

Key contract concepts

Several concepts appear across all contract types and are essential vocabulary.

Attachment and limit

The attachment point is the loss threshold above which the contract starts paying. The limit is the maximum the contract will pay. Together they define a band of risk being transferred — this band is called a layer.

Term	Definition
Attachment point	Loss level where the contract begins to respond
Limit	Maximum payment under the contract
Exhaustion point	Attachment + limit — the loss level at which the contract is fully consumed

A contract described as “$30M xs $20M” means: limit of $30M, excess of (attaching at) $20M. The reinsurer pays losses between $20M and $50M.

Terminology: contract, treaty, layer, program

These terms are closely related and sometimes used loosely. Here is how the industry generally distinguishes them:

• Contract is the generic term for any reinsurance agreement — the legal document formalizing obligations between parties. This site uses “contract” throughout as the default term.
• Treaty is a contract that covers a defined book of business over a set period (usually 12 months, renewing annually), as opposed to facultative reinsurance which covers individual risks case by case. Industry practitioners may say “treaty” where this site says “contract.”
• Layer is a specific band of coverage within an excess-of-loss structure, defined by its attachment point and limit.
• Program is the entire set of reinsurance protections a cedent purchases for a given line of business. A program typically consists of multiple XoL layers stacked together, possibly combined with other sections such as a quota share, and represents the cedent’s overall reinsurance strategy for that risk category. A cedent may have several programs (e.g. a property catastrophe program, a casualty program).

In short: a program is made up of layers and other sections, each backed by a contract (or treaty), and these collectively define the cedent’s reinsurance protection.

Retention

The retention is what the cedent keeps. In a quota share, it is the percentage not ceded (if 25% is ceded, 75% is retained). In an excess of loss contract, the cedent retains losses both below the attachment point and above the exhaustion point — everything outside the contracted band stays with the cedent.

Participation

A single layer is rarely placed entirely with one reinsurer. Instead, the cedent (usually through a broker) offers participations — each reinsurer takes a percentage share of the layer. For example, a $30M xs $20M layer might be placed with three reinsurers taking 50%, 30%, and 20% respectively. Each participant receives its share of the premium and pays its share of any claims.

Why split a layer? The amounts involved are large — a single catastrophe layer can carry hundreds of millions of dollars of limit. Few reinsurers want that much exposure to a single risk, and few cedents want to depend on a single counterparty. Spreading participation diversifies counterparty risk for the cedent and concentration risk for each reinsurer. It also allows reinsurers to calibrate their exposure precisely: a reinsurer might want some Florida hurricane risk but not $100M of it, so it takes a 15% line instead.

This is distinct from a quota share: in a quota share, the proportional split applies to every loss from the ground up. With layer participation, the XoL terms (attachment, limit) apply first to the full loss, and only the resulting ceded amount is divided among participants. The layer responds as a single unit; the participation determines who funds it.

Helios Re

Helios Re is a participant on all five of its contracts — it does not necessarily hold 100% of any layer. The analytics in this site evaluate contracts from a single participant’s perspective: apply the contract terms first, then scale by participation share. For simplicity, all calculations assume 100% participation unless stated otherwise.

Reinstatements

When a CatXoL contract pays a claim, the limit is eroded. A reinstatement restores the limit for subsequent occurrences during the contract period, usually in exchange for an additional premium.

Helios Re

Contract 1 (Florida Hurricane CatXoL) has 1 reinstatement at 100%. If a hurricane causes $30M of ceded loss (fully exhausting the $30M limit), the limit is reinstated once — Helios Re would owe an additional reinstatement premium equal to 100% of the original premium pro-rata to the proportion of limit used. If a second hurricane also exhausts the limit, there are no more reinstatements, and Helios Re’s maximum exposure for the year is $60M (two full limit losses).

In Practice

Reinstatements introduce statefulness into the contract: the response to the second occurrence depends on how much limit the first occurrence consumed. This makes reinstatements fundamentally different from “stateless” terms like attachment and limit, and they are not the only terms with this property — aggregate deductibles, loss corridors, and profit commissions also carry state across occurrences. Any analysis of a contract with reinstatements must track the sequence of occurrences within each scenario, not just individual loss amounts.

Occurrence vs. aggregate

This is one of the most important distinctions in reinsurance, and one that is easy to get wrong.

• An occurrence contract applies its attachment and limit to each occurrence independently. If two hurricanes each cause $25M of loss, each is evaluated separately against the contract terms.
• An aggregate contract applies its attachment and limit to the total loss accumulated across all occurrences in the contract period (usually a year). An aggregate excess of loss is called an AggXoL — also known as an aggregate stop loss.

Contract Type	Applies To	Example
Occurrence (CatXoL)	Each occurrence independently	$30M xs $20M per occurrence
Aggregate (AggXoL)	Sum of all occurrences in the year	$35M xs $40M annual aggregate

This distinction matters enormously for analytics. An occurrence contract can be evaluated occurrence by occurrence. An aggregate contract requires summing across all occurrences in a scenario before applying the contract terms — it needs scenario-level state.

Helios Re

In Helios Re’s portfolio, Contracts 1, 2, and 4 are occurrence-level (CatXoL), while Contract 5 (US Multi-Peril AggXoL) is aggregate-level. The same attachment and limit figures mean very different things depending on whether they apply per occurrence or per year.

Watch out

Do not assume all reinsurance contracts work like a simple “deductible and cap” applied occurrence by occurrence. Aggregate contracts are common, and they require accumulating loss across occurrences within a scenario before determining the contract response. A system that only handles occurrence logic will fail on a significant fraction of real-world contracts.

Perspectives on loss

The same catastrophe produces different loss figures depending on whose books you are looking at. Throughout this site, all perspectives are from the reinsurer’s point of view unless stated otherwise.

Perspective	Definition
Subject loss	The loss entering the contract from the cedent’s book — the input that contract terms operate on
Gross loss	What the reinsurer assumes after applying contract terms to the subject loss
Recoveries	What the reinsurer gets back from its own outward protections — retrocession, ILWs, cat bonds, or other hedges
Net loss	Gross minus recoveries — what actually stays on the reinsurer’s books

Terminology across the chain

The industry does not use these terms consistently. The same dollar amount can be called “gross,” “ceded,” “subject,” or “ground-up” depending on who is speaking and from which side of the contract they sit. Conversations between insurers, reinsurers, and retrocessionaires can become confusing quickly — not because the concepts are hard, but because the vocabulary shifts with the observer.

To cut through this, it helps to notice that the Subject → Gross → Recoveries → Net cycle applies identically at every level of the chain. What changes is which upstream loss amount feeds in as Subject. The following table shows the same four-stage cycle for each participant, with arrows marking where one party’s output becomes the next party’s input:

Stage	Primary Insurer	Reinsurer	Retrocessionaire
Subject	Loss to the insured	Primary’s Gross	Reinsurer’s Gross
Gross	Loss assumed by the insurer	Loss assumed by the reinsurer	Loss assumed by the retro
Ceded / Recoveries	→ becomes Reinsurer’s Gross	→ becomes Retro’s Gross	→ (further cession, if any)
Net	Gross minus Ceded	Gross minus Ceded	Gross minus Ceded

The linkage rule is simple: one party’s Ceded is the next party’s Gross, and one party’s Gross is the next party’s Subject. The chain extends as far as risk is transferred — a retrocessionaire buying protection from a fourth party follows the same pattern.

Key Insight

The insurer’s ceded loss is the reinsurer’s gross loss. The reinsurer’s ceded loss is the retrocessionaire’s gross loss. Every cession creates the next party’s assumption. This transitivity is the mechanism of risk transfer.

You will encounter a range of alternative terms for these stages. Some of the more common ones:

This site uses	Also known as
Subject loss	Ground-up loss, underlying loss, inward loss
Gross loss	Assumed loss, loss to the book
Ceded / Recoveries	Outward loss, reinsurance recoveries
Net loss	Retained loss

Of these, “ground-up loss” deserves particular care. Strictly, it means the original loss to the insured before any contract terms — the primary insurer’s subject loss. But reinsurers often use “ground-up” to mean “loss before our contract terms,” which is actually the cedent’s gross, one step removed. This site uses subject loss to avoid the ambiguity.

Similarly, “ceded” and “recoveries” describe the same cash flow from opposite ends. The cedent says “I ceded $30M”; the cedent’s accountant says “We recovered $30M from reinsurance.” Same amount, different framing. “Ceded” describes the outgoing transfer; “Recoveries” describes the financial benefit received. Both appear throughout the industry.

Proportional vs. non-proportional

The clean linkage — one party’s Gross becomes the next party’s Subject — is most intuitive for proportional treaties (quota share and other proportional structures), where the reinsurer participates in a defined share of the cedent’s gross.

In practice, even proportional treaties scope to a specific book of business, and non-proportional (excess-of-loss) treaties add further filtering. The next party’s subject is whatever the treaty defines as its subject loss basis — which may be scoped to a specific line of business, net of inuring treaties, or net of certain deductibles. The four-stage cycle still applies at each level, but the arrow between levels passes through a filter defined by the contract terms.

This site’s convention

This site defaults to the reinsurer’s perspective unless stated otherwise:

• Subject = what the cedent’s book produces (the input)
• Gross = what the reinsurer assumes after contract terms (the output)
• Recoveries = what the reinsurer retrocedes or hedges
• Net = Gross − Recoveries = what stays on the reinsurer’s books

Each of these can also be viewed net of premium (loss minus premium received) or net of premium and expenses (loss minus premium minus brokerage, taxes, and fees). These financial perspectives become important in the Practice chapter; for now, the four loss perspectives above are sufficient.

Helios Re’s contract portfolio

Let us see how these concepts come together in the Helios Re portfolio. We introduced the five contracts conceptually; here is the full picture.

#	Name	Type	Structure	Peril	Geography
C1	Florida Hurricane CatXoL	Non-proportional	$30M xs $20M, 1 reinstatement	Hurricane	Florida
C2	California Earthquake CatXoL	Non-proportional	$25M xs $15M, 1 reinstatement	Earthquake	California
C3	European Windstorm QS	Proportional	25% quota share	Windstorm	Europe
C4	Japan Earthquake CatXoL	Non-proportional	$50M xs $25M, 2 reinstatements	Earthquake	Japan
C5	US Multi-Peril AggXoL	Non-proportional (aggregate)	$35M xs $40M annual aggregate	Multi-peril	USA

This portfolio illustrates the diversity of real-world reinsurance:

• Three CatXoLs (C1, C2, C4) covering different perils and geographies — these diversify each other because a Florida hurricane and a Japan earthquake are unlikely to occur in the same year
• One quota share (C3) providing broad proportional participation in European windstorm risk
• One AggXoL (C5) protecting against accumulation of US losses across multiple occurrences — this interacts with C1 and C2 because they cover overlapping US perils

Key Insight

Notice that C5 covers the same cedent (SunCoast Insurance) and overlapping perils as C1. In a year with multiple Florida hurricanes, C1 responds to each occurrence individually (occurrence-level), while C5 responds to the total annual accumulation (aggregate-level). You cannot understand Helios Re’s true exposure to Florida hurricanes by looking at either contract in isolation — the interaction between them is where the real risk picture emerges.

Beyond the basics

Real-world contracts include many additional features that we will formalize in later chapters:

• Sliding scale commissions — the commission rate adjusts based on the loss experience
• Profit commissions — the cedent shares in the reinsurer’s underwriting profit
• No claims bonuses — premium rebates if no claims are made
• Loss corridors — gaps in coverage between layers
• Inuring contracts — contracts that must be applied before another contract (netting order)
• Multi-year terms — contracts spanning more than one annual period
• Franchise deductibles — the full loss is covered if it exceeds the threshold (unlike regular deductibles which only cover the excess)

Each of these features adds analytical complexity, but they all follow the same pattern: define a rule, apply it to loss data, and track how it interacts with other rules on the same contract. The Engineering chapter formalizes this pattern.

---

We now understand the market, its participants, and the contracts they trade. But who are the people inside these organizations that build and use analytics? The next section introduces the key personas and what they need from analytics systems.