Internal:Training/IFRS17/Reinsurance held
🔗 Recall. In the previous page, you learned how the variable fee approach modifies the general model for direct participating contracts, where the policyholder's return is linked to underlying items. Now we shift perspective entirely: instead of looking at the contracts an insurer sells, we look at the contracts an insurer buys, specifically reinsurance contracts held to protect its own book of business.
🎯 Objective. In this page, you will learn:
- Why a reinsurance contract held is treated as a mirror image of the underlying insurance, with the insurer stepping into the role of the customer rather than the provider
- How IFRS 17 introduces deliberate asymmetries in the treatment of reinsurance, including the possibility of a day-one gain and the mechanism for loss recovery on onerous underlying contracts
- How the accounting differs between proportionate and non-proportionate reinsurance structures, and why that distinction matters
Reinsurance as the mirror image: the insurer is the customer
🔄 Flipping the perspective. Throughout this training programme, you have looked at insurance contracts from the insurer's side: the insurer collects premiums, bears risk, and holds a liability on its balance sheet. With reinsurance, the roles reverse. The insurer, sometimes called the cedant, pays a reinsurance premium to a reinsurer in exchange for the reinsurer taking on a share of the risk. From the cedant's viewpoint, the reinsurance contract is an asset, not a liability, because it represents a right to receive reimbursement if covered claims occur. Think of it like home insurance for the insurer itself: just as a homeowner in Bordeaux pays a premium to protect against storm damage, AXA pays a reinsurance premium to protect against the possibility that claims on its own book exceed a comfortable level.
📑 Measurement follows the same building blocks. IFRS 17 requires the cedant to measure a reinsurance contract held using the same fundamental framework as for underlying insurance contracts: fulfilment cash flows (estimated future cash flows, discounted and adjusted for risk) and a CSM. However, the direction of every cash flow is reversed. Where an insurance contract has premium inflows and claim outflows, the reinsurance asset has premium outflows (the cost of protection) and expected recovery inflows (the reinsurer's share of claims). The risk adjustment on a reinsurance contract held reflects the reduction in risk that the cedant enjoys by holding the contract, which is the benefit of transferring volatility to someone else.
⚠️ Common misconception. It is tempting to assume that the reinsurance asset simply equals a fixed percentage of the underlying insurance liability. In reality, the reinsurance contract is a separate contract with its own fulfilment cash flows, its own discount rate, and its own CSM. The two are related but measured independently. For instance, the credit risk of the reinsurer, meaning the chance that the reinsurer fails to pay, must be factored into the reinsurance asset but has no place in the underlying insurance liability.
🏗️ Separate grouping, separate presentation. Reinsurance contracts held are grouped into their own portfolios following the same logic as for insurance contracts: contracts with similar risks managed together, divided into profitability groups and annual cohorts. On the balance sheet, the reinsurance asset is presented separately from insurance contract liabilities. You will never see the two netted against each other. This separation is crucial for transparency: it lets the reader of the financial statements see both the gross exposure the insurer carries and the protection it has purchased.
🤔 Think about it. If reinsurance held mirrors the underlying insurance contract in structure, you might expect the accounting to be perfectly symmetrical. But IFRS 17 deliberately breaks that symmetry in two important places. Why would the standard treat the buyer of protection differently from the seller?
Key asymmetries: day-one gains and loss recovery
💡 Why symmetry breaks down. If IFRS 17 applied exactly the same rules to reinsurance held as to underlying insurance, some economically important information would be hidden. The standard therefore introduces two deliberate asymmetries. The first concerns what happens at initial recognition when the reinsurance contract is expected to be profitable for the cedant (that is, recoveries are expected to exceed premiums paid). The second concerns how the standard handles the interaction between reinsurance and onerous groups of underlying contracts.
🎁 The day-one gain. Under the general model for insurance contracts issued, if a group is profitable at initial recognition, the profit is locked into the CSM and released over the coverage period; no day-one gain ever appears in the income statement. For reinsurance held, the standard takes a different approach. When the cedant recognises a reinsurance contract that produces a net cost (the usual case, since reinsurance is a service the cedant pays for), a negative CSM is established, representing the cost of the reinsurance service that will be expensed over time. However, if the terms are unusually favourable and the expected recoveries exceed the premium, a day-one gain can be recognised. The reasoning is that the cedant is a buyer of a service: when you buy something at a bargain price, it is appropriate to recognise that benefit, unlike the seller side where IFRS 17 is deliberately conservative about recognising profit upfront.
⚠️ Common misconception. Many learners assume that a negative CSM on a reinsurance contract held means the contract is "loss-making." It does not. A negative CSM simply represents the net cost of purchasing reinsurance protection, which the cedant chose to buy. It is analogous to an insurance premium you pay for your own benefit: it is an expense, not a loss. The cost is released over the coverage period as the cedant receives the benefit of reduced risk.
🛡️ Loss recovery on onerous underlying contracts. The second asymmetry is perhaps the more powerful one. When a group of underlying insurance contracts becomes onerous, the insurer must recognise a loss immediately in the income statement. Without special treatment, the related reinsurance benefit would only be recognised gradually over the coverage period, creating a mismatch: the pain appears immediately, but the relief trickles in slowly. IFRS 17 corrects this by allowing the cedant to recognise a loss recovery component within the reinsurance asset. In practical terms, if AXA writes a group of property contracts along the Mediterranean coast and a severe reassessment of catastrophe risk pushes that group into onerous territory, the offsetting benefit from the reinsurance treaty covering those contracts is accelerated so it appears in the same period as the underlying loss. This ensures the income statement tells a complete story rather than showing only the bad news.
🤔 Think about it. You have now seen that reinsurance held uses the same building blocks but with key asymmetries designed to reflect the cedant's position. In practice, however, reinsurance comes in many structural forms. How does the accounting adapt when the reinsurance treaty covers a fixed share of every claim versus only claims above a high threshold?
Proportionate vs. non-proportionate reinsurance
📐 Two families of reinsurance. Reinsurance structures broadly fall into two categories. Proportionate reinsurance, sometimes called pro-rata reinsurance, means the reinsurer takes a fixed percentage of every policy in the covered portfolio. The most common forms are quota share treaties, where the reinsurer accepts, say, 30% of every risk, and surplus treaties, where the reinsurer's share varies by policy but is determined by a formula agreed in advance. Non-proportionate reinsurance, by contrast, only triggers when claims exceed a specified threshold. The most familiar example is excess of loss cover: the cedant retains the first €5 million of claims from a single event, and the reinsurer pays everything above that up to an agreed limit. The economic behaviour of these two families is quite different, and IFRS 17 reflects that.
📊 Proportionate reinsurance: a scaled-down mirror. Because a quota share treaty shares a fixed proportion of every underlying policy, the reinsurance contract held naturally mirrors the underlying insurance contracts on a scaled-down basis. The fulfilment cash flows of the reinsurance asset move in close proportion to the fulfilment cash flows of the underlying group. When new claims are incurred on the underlying book, the reinsurer's share is a predictable fraction. This makes grouping and measurement more straightforward. The reinsurance contracts can often be grouped to align with the same profitability groups and annual cohorts as the underlying business. For example, if an insurer like AXA cedes 25% of a Belgian commercial property portfolio under a quota share treaty, the reinsurance asset's cash flows at each reporting date will closely track 25% of the corresponding insurance liability's cash flows.
⚠️ Common misconception. Some learners believe that under a quota share, the cedant can simply net off the reinsurer's share and report only the retained portion. IFRS 17 does not permit this. The underlying insurance liability must be reported gross, and the reinsurance asset is shown separately. Netting would obscure the true size of the insurer's obligations and the extent of its reliance on reinsurers.
🌪️ Non-proportionate reinsurance: a different beast. Excess of loss contracts behave very differently. Because they only respond when claims breach a threshold, the reinsurance asset can sit dormant for long periods and then activate suddenly following a major catastrophe or an accumulation of large claims. The fulfilment cash flows of an excess of loss contract do not move in proportion to the underlying book; they are driven by the tail of the claims distribution, the low-probability, high-severity events. This makes measurement more complex and more reliant on actuarial judgment. The risk adjustment on the reinsurance asset may also differ significantly from a simple proportion of the underlying risk adjustment, because the reinsurer is absorbing a concentrated slice of extreme risk. In grouping terms, non-proportionate treaties may not align neatly with the underlying insurance groups, since a single excess of loss contract can cover multiple portfolios or lines of business simultaneously.
Takeaways
📌 Key takeaways.
- A reinsurance contract held is measured using the same building blocks as an insurance contract but from the buyer's perspective: fulfilment cash flows run in the opposite direction and the result is an asset, not a liability, presented separately on the balance sheet.
- IFRS 17 introduces deliberate asymmetries for reinsurance held: a day-one gain may be recognised in certain circumstances, and a loss recovery component accelerates the recognition of reinsurance benefit when underlying contracts become onerous, keeping the income statement balanced.
- Proportionate and non-proportionate reinsurance contracts differ in how their fulfilment cash flows behave: proportionate treaties track the underlying book in a predictable ratio, while non-proportionate treaties respond only to extreme events and require more actuarial judgment to measure.