Internal:Training/IFRS17/The risk adjustment
🔗 Recall. In the previous page, you learned how discounting converts future cash flows into their present value, reflecting the time value of money. Now we build on that by turning to the next building block: the amount an insurer adds to its liability to compensate for the fact that those future cash flows are uncertain.
🎯 Objective. In this page, you will learn:
- What the risk adjustment represents and why it is a necessary part of the insurance liability under IFRS 17.
- How the risk adjustment can be measured using techniques such as confidence levels, cost of capital, and Value at Risk.
- How the risk adjustment decreases over time as risk expires, and how that release flows into the income statement.
What the risk adjustment represents
🎯 The price of uncertainty. When you learned about fulfilment cash flows, you saw that insurers estimate the probability-weighted average of all future cash flows. That average is the best estimate of what will happen, but it is still just an estimate. Actual outcomes could be better or worse. The risk adjustment is the explicit amount an insurer recognises on top of the best estimate to reflect the compensation it demands for bearing that uncertainty. Think of it this way: if two jobs paid the same salary but one involved perfectly predictable work and the other involved constant surprises and potential crises, you would want extra pay for the unpredictable one. The risk adjustment is that extra pay, expressed as a component of the insurance liability.
🛡️ An insurer's perspective. Consider AXA insuring 3,000 homes along the coast of Brittany against storm damage. The actuaries estimate that the probability-weighted average cost of claims over the next year is €6 million. But storms are volatile: in a mild year, claims might total only €2 million, while a severe season could push them to €15 million. The €6 million best estimate captures the average, yet it says nothing about how uncomfortable that volatility makes the insurer. The risk adjustment adds an explicit margin, say €1.2 million, to acknowledge that the insurer is exposed to outcomes far worse than average. This margin is not a hidden buffer or a secret reserve. It is reported separately on the balance sheet, visible to investors, regulators, and management alike.
⚠️ Common misconception. Learners often confuse the risk adjustment with a prudential margin designed to make reserves deliberately conservative. The risk adjustment is not about caution for its own sake. It is a principled measure of how much compensation is required for bearing non-financial risk. An insurer with a very stable, predictable portfolio would carry a small risk adjustment, while one exposed to catastrophic natural disaster risk would carry a much larger one. The size reflects the nature of the uncertainty, not a management preference for conservatism.
🔑 Non-financial risk only. An important boundary to note is that the risk adjustment under IFRS 17 covers only non-financial risk, that is, risks arising from the insurance contracts themselves, such as the uncertainty in the timing and amount of claims. It does not cover financial risk like changes in interest rates or equity markets. Financial risk is already handled through discounting and the choice of discount rates. By limiting the risk adjustment to non-financial risk, the standard avoids double-counting and keeps each building block focused on a single dimension of uncertainty.
🤔 Think about it. You now know what the risk adjustment represents, but how does an insurer actually put a number on something as abstract as "the compensation for bearing uncertainty"?
How to measure it: confidence levels, cost of capital, VaR
📐 No single prescribed method. Unlike some elements of IFRS 17 that follow strict rules, the standard does not mandate a single technique for measuring the risk adjustment. Instead, it sets a principle: the risk adjustment should reflect the compensation the insurer requires for bearing non-financial risk. How an insurer translates that principle into a number is a matter of judgement, but the standard does require the insurer to disclose the confidence level to which the risk adjustment corresponds. In practice, three techniques dominate the industry, and understanding each one gives you a well-rounded view of how the calculation works.
📊 Confidence level approach. The most intuitive technique is the confidence level approach. It asks: at what probability threshold do we want to be confident that our reserves will be sufficient? Suppose an insurer's actuaries model the full distribution of possible outcomes for a group of property insurance contracts in Belgium. The best estimate sits at the 50th percentile, the point where outcomes are equally likely to be better or worse. If the insurer sets the risk adjustment at the 75th percentile, it is saying: "We want to hold enough to cover outcomes up to the 75th percentile of severity." The risk adjustment is then the difference between the 75th percentile and the 50th percentile. A higher confidence level means a larger risk adjustment and a more cautious stance. Most European insurers using this method target a confidence level somewhere between 65% and 85%, depending on the volatility of their portfolio.
⚠️ Common misconception. It is easy to assume that a 75% confidence level means the insurer expects to have enough reserves 75% of the time and will fall short 25% of the time. While that is the statistical interpretation, it does not mean the insurer is planning to fail one year in four. The confidence level is a calibration tool for sizing the risk adjustment, not a prediction of how often reserves will be insufficient. In reality, insurers manage their risk actively through reinsurance, diversification, and ongoing monitoring, so the actual experience of insufficiency is far rarer.
💼 Cost of capital approach. A second common method is the cost of capital approach. This technique starts from a different angle: instead of asking "what percentile do we want to cover?", it asks "how much capital must we hold to support these risks, and what return does that capital require?" The insurer calculates the regulatory or economic capital needed to back the insurance risks, then applies a target rate of return to that capital over the remaining lifetime of the contracts. The risk adjustment equals the present value of that required return. For example, if a portfolio requires €20 million in capital and the target return is 6% per year over three remaining years, the risk adjustment would be the discounted value of €1.2 million per year. This method is popular among insurers who already perform detailed capital modelling for Solvency II or internal purposes, because it leverages existing infrastructure.
📈 Value at Risk and other quantile methods. The third approach is Value at Risk, or VaR, and its relatives such as Tail Value at Risk (TVaR). VaR identifies the loss threshold at a given confidence level: for instance, a 99.5% VaR says "there is only a 0.5% chance that losses will exceed this amount." The risk adjustment can be set as the difference between the VaR at the chosen confidence level and the best estimate. TVaR goes further by averaging all outcomes beyond the VaR threshold, capturing the severity of the tail. These methods are familiar to risk management teams and can be powerful for portfolios exposed to catastrophe risk, where the tail of the distribution matters enormously. Regardless of which technique an insurer chooses, IFRS 17 requires disclosure of the equivalent confidence level so that readers can compare risk adjustments across companies.
🤔 Think about it. Now that you know how the risk adjustment is measured at inception, what happens to it as time passes and the insurer's exposure to risk decreases?
How the risk adjustment releases as risk expires
⏳ Risk diminishes over time. An insurance contract does not carry the same level of uncertainty from the day it is written to the day it expires. As time passes, some risks simply disappear. Policies reach the end of their coverage period, claims are reported and settled, and what was once unknown gradually becomes known. Because the risk adjustment reflects compensation for bearing uncertainty, it must decrease as that uncertainty fades. Under IFRS 17, the reduction in the risk adjustment is recognised in the income statement as part of insurance revenue, rewarding the insurer for having successfully borne the risk during the period.
🏠 A concrete example. Imagine a group of one-year household insurance contracts written in Spain on 1 January, with a total risk adjustment of €3 million at inception. By 30 June, half the coverage period has elapsed. Many of the risks the insurer worried about in January have either materialised as claims or passed without incident. The remaining uncertainty is smaller, and the risk adjustment might now stand at €1.4 million. The €1.6 million reduction flows into the income statement as revenue for the first half of the year. By 31 December, when the contracts expire, virtually all uncertainty has been resolved, and the risk adjustment approaches zero. Each slice of release represents a period in which the insurer bore risk and is now being compensated for it.
⚠️ Common misconception. Some learners expect the risk adjustment to release in a perfectly even, straight-line pattern over the coverage period. In practice, the release pattern should reflect the actual expiry of risk, which is not always uniform. If a portfolio of contracts covers a region where storm risk is concentrated in the autumn months, more risk adjustment should release during and after the storm season than during the calm summer months. The release follows the risk, not the calendar.
🔗 Connecting to the bigger picture. The release of the risk adjustment works alongside the other building blocks you have already studied. As fulfilment cash flows are updated and discounting unwinds over time, the risk adjustment release adds a third layer of movement in the liability. Together, these movements tell a rich, transparent story on the income statement: how much revenue came from the passage of time, how much from bearing risk, and how much from delivering coverage. In the next page, you will meet the fourth and final building block, the contractual service margin, and see how the entire framework comes together to determine when and how profit appears in the insurer's results.
Takeaways
📌 Key takeaways.
- The risk adjustment is the explicit compensation an insurer recognises for bearing the uncertainty that actual cash flows may differ from the best estimate; it covers non-financial risk only and is disclosed separately on the balance sheet.
- There is no single mandated measurement technique: insurers may use confidence levels, cost of capital, Value at Risk, or other methods, but must always disclose the equivalent confidence level so that readers can compare across companies.
- As risk expires over time, the risk adjustment decreases and the reduction is recognised as insurance revenue, following the pattern in which risk actually diminishes rather than a simple straight-line allocation.