Internal:Training/IFRS17/Fulfilment cash flows

🔗 Recall. In the previous page, you learned how IFRS 17 decomposes an insurance liability into four transparent building blocks: fulfilment cash flows, discounting, the risk adjustment, and the contractual service margin. Now we take a close look at the first and most foundational of those building blocks: the cash flows themselves.

🎯 Objective. In this page, you will learn:

• Which types of cash flows an insurer must include when measuring an insurance contract, and which it must leave out.
• What the contract boundary is, why it exists, and how it determines the point at which future cash flows stop belonging to the current contract.
• How probability-weighted estimates work, and why the insurer must keep its assumptions current at every reporting date.

~*~

Which cash flows to include

💰 Starting with the basics. Fulfilment cash flows are the foundation of IFRS 17's measurement model. They represent every future cash flow that the insurer expects to arise as it fulfils its obligations under a group of insurance contracts. The idea is intuitive: if you want to know what a promise costs, you need to list all the money you expect to flow in and all the money you expect to flow out because of that promise. IFRS 17 requires the insurer to consider both directions, inflows and outflows, as part of a single integrated estimate.

📥 Cash inflows. The most obvious inflow is the premium the policyholder pays. But premiums are not always a single upfront payment. A motor insurance customer in Germany might pay monthly instalments over the year, so the insurer must include all expected future premium receipts that fall within the contract. Other inflows can include recoveries the insurer expects to receive, for example salvage from a written-off vehicle or subrogation rights that allow the insurer to reclaim money from a third party who caused the loss. Every euro the insurer reasonably expects to receive because of the contract counts as an inflow.

📤 Cash outflows. On the other side, outflows include everything the insurer expects to pay. The largest category is claims: the compensation paid to policyholders when insured events occur. Beyond claims, the insurer must include acquisition costs such as commissions paid to brokers, claims handling costs such as the salaries of claims adjusters and legal fees, and ongoing administrative expenses directly attributable to the contracts. For example, if AXA underwrites a portfolio of home insurance policies in Belgium, the fulfilment cash flows must capture not just the expected storm and fire claims, but also the broker commissions, the cost of running the claims department, and the policy administration costs tied to that portfolio.

⚠️ Common misconception. A frequent mistake is to assume that fulfilment cash flows include only claims. In reality, IFRS 17 casts a wide net: any cash flow that would not exist if the contracts had never been written must be included. This means acquisition costs, maintenance expenses, and claims handling costs are all in scope. Leaving them out would understate the true cost of fulfilling the insurer's promise.

🤔 Think about it. If fulfilment cash flows must include all future premiums and all future costs, where exactly does the insurer draw the line? At what point do future cash flows stop belonging to the current contract and start belonging to a new one?

~*~

The contract boundary

🚧 Drawing the line. The contract boundary is IFRS 17's answer to a critical question: how far into the future should the insurer look when estimating the cash flows of a contract? The boundary marks the point beyond which the insurer is no longer compelled to provide coverage or accept risk. Cash flows inside the boundary belong to the current contract measurement. Cash flows outside it do not, even if the insurer fully expects the customer to renew. This concept prevents insurers from bundling the economics of future, yet-to-be-written contracts into today's measurement.

🔑 The key test. IFRS 17 defines the contract boundary by asking a practical question: can the insurer, at a given future date, reassess the risk of the individual policyholder and reprice or reject accordingly? If the insurer has the right to change the premium to fully reflect the policyholder's updated risk, or the right to terminate the contract, then the boundary falls at that point. Everything beyond it is considered a future contract. Consider a one-year motor policy in Spain that renews annually. At each renewal date, the insurer can adjust the premium based on the driver's updated claims history and current risk profile. The contract boundary for that policy is one year, because at renewal the insurer has full repricing power.

⚠️ Common misconception. Many people confuse the contract boundary with the legal term of the contract. They are not the same thing. A contract might have a legal duration of one year but include a guaranteed renewal clause that prevents the insurer from repricing for three years. In that case, the contract boundary extends to three years, because the insurer is obligated to provide coverage at the current price for that entire period. The test is about economic substance, specifically the insurer's practical ability to reprice, not the date printed on the policy document.

🏥 A longer example. Imagine a health insurance policy in France where the insurer guarantees renewal at community-rated premiums regardless of the policyholder's health status. Even though the policy renews annually on paper, the insurer cannot adjust the premium to reflect a particular individual's deteriorating health. The contract boundary may therefore extend well beyond one year, potentially across the policyholder's expected lifetime, because the insurer lacks the ability to reprice for that individual's specific risk. This creates a much longer stream of cash flows to estimate, which has a significant effect on the measurement of the liability.

🤔 Think about it. Now that we know which cash flows to include and how far into the future to project them, a new challenge arises: the future is uncertain. How does the insurer handle the fact that it does not know exactly how many claims will occur, how large they will be, or when they will be paid?

~*~

Probability-weighted estimates and keeping assumptions current

🎲 Embracing uncertainty. IFRS 17 does not ask the insurer to predict the future with certainty. Instead, it requires a probability-weighted estimate, sometimes called an expected value. The insurer must consider a range of possible outcomes and weight each one by its probability. Suppose an insurer like AXA is estimating claims for a portfolio of property insurance in southern France, a region exposed to both minor weather events and rare but catastrophic floods. The estimate does not pick a single "most likely" scenario. Instead, it blends together many possibilities: a calm year, a year with one moderate storm, a year with severe flooding, and so on, each multiplied by how likely it is to occur. The weighted average of all these scenarios produces the probability-weighted estimate.

📊 Why the average matters. To see the logic, think of a simple example. A portfolio of 10,000 home policies in the Rhône valley faces two broad scenarios this year: a 90% chance of a quiet year with total claims of €8 million, and a 10% chance of a major flood causing claims of €40 million. The probability-weighted estimate is (0.90 × €8 million) + (0.10 × €40 million) = €11.2 million. This figure is not the outcome in either scenario, but it represents the average expected cost across all possibilities. IFRS 17 uses this average as the starting point for the liability, because over many contracts and many years, the average is the best unbiased predictor of the insurer's total outflows.

⚠️ Common misconception. A common error is to confuse the probability-weighted estimate with the "most likely" outcome. These are not the same thing. In the Rhône valley example above, the most likely outcome is €8 million (the quiet year), but the probability-weighted estimate is €11.2 million because it also reflects the tail risk of flooding. Choosing only the most likely outcome would systematically understate the liability, ignoring the impact of severe but plausible events.

🔄 Keeping assumptions current. IFRS 17 adds one more crucial requirement: the insurer must update its estimates at every reporting date. Assumptions made when the contract was first written do not remain frozen. If new information emerges, such as a change in claims trends, updated weather models, shifts in inflation, or new legal rulings affecting bodily injury awards, the insurer must revise its fulfilment cash flows to reflect current conditions. This "current estimate" philosophy means that the balance sheet always shows the insurer's latest and best view of what it will cost to fulfil its promises, rather than a stale figure locked in at inception. It keeps the financial statements honest and transparent, even when the world around the contract changes.

~*~

Takeaways

📌 Key takeaways.

• Fulfilment cash flows include all future cash inflows (premiums, recoveries) and outflows (claims, acquisition costs, administrative expenses, claims handling costs) that arise directly from fulfilling a group of insurance contracts.
• The contract boundary determines how far into the future the insurer projects those cash flows; it ends at the point where the insurer can reassess the policyholder's risk and reprice or reject, regardless of the legal term of the contract.
• Cash flow estimates must be probability-weighted, reflecting a full range of scenarios rather than a single "most likely" outcome, and must be updated at every reporting date to incorporate the latest available information.

~*~