AM Best believes that on balance IFRS 17 will be more meaningful, transparent and consistent across territories than current financial reporting standards. IFRS 17 should better assist the universe of financial stakeholders in identifying and supporting (re)insurance activities with good economic returns and create more effective market discipline. The new reporting standard is not anticipated to have a direct effect on credit ratings, although the data, terminology and many of the KPIs used will change. Challenges will include discontinuities in data records and KPIs, possible comparability questions within territories, at least initially (including between IFRS and local GAAP reporters), and also costs and disruption over the implementation phase.
As IFRS 17 moves from the standard setting to implementation stage, a new chapter has opened, bringing with it fresh uncertainties. Stakeholders are starting to debate what users of (re)insurance company financial reporting will do with the new data and what the likely key performance indicators (KPIs) will be under IFRS 17.
This report explores some of the challenges and opportunities that AM Best has identified at this point as stemming from the new framework, as well as some of the KPIs it believes will be critical for both (re)insurers’ financial stakeholders and (re)insurers themselves, including:
- The calculation of net of reinsurance combined ratios, claims ratios and expense ratios
- Non-life deferred acquisition costs (DAC) in capital models
- Use of the contractual service margin (CSM)
- Capital in participating life funds
- Return on equity (ROE)
AM Best does not expect IFRS 17 to have a direct impact on credit ratings, as the economic reality assessed by credit ratings will remain unchanged. Nevertheless the content of analysis, the KPIs used, the terminology and the conversations with insurers will all change, and rating committees will be making their assessments utilising the new data.
Increased market discipline, resulting from greater transparency, should mean that insurers will be less likely to write uneconomic policies.
Perhaps less obviously, more business may be written for products where insurers currently struggle to communicate performance effectively to financial stakeholders despite good potential for profit. Participating life business with low guarantees could fall into this category under favourable conditions. Similarly, AM Best believes that greater transparency around profitability will in general promote product innovation.
Combined Ratios
Though IFRS 17 is a principles-based standard, it is prescriptive on the treatment of reinsurance commissions.
AM Best’s analysis indicates that the new standard creates a significant likelihood of change to combined ratio calculations for many non-life (re)insurers.
In straightforward terms the result of the treatment of reinsurance held under IFRS 17 is that:
- Ceding commissions received will be netted off against reinsurance premiums paid under IFRS 17, instead of being a negative expense as is usually the case under IFRS 4.
- Profit commissions are added to claims recoveries from reinsurers under IFRS 17, thus reducing net claims, instead of being a negative expense as is usually the case under IFRS 4.
- The reinsurance result is presented separately rather than its components being netted off against the various gross of reinsurance amounts.
The consequence is that existing net of reinsurance expense and claims ratios using IFRS 17 data would produce different results.
At the same time there is an alternative way to calculate these ratios that is suggested by the form of IFRS 17 accounts and which, in AM Best’s view, has some clear advantages. Both sets of ratios will involve a degree of discontinuity with existing net of reinsurance ratios, while gross ratios will remain unchanged.
The choice for claims, expense and combined ratios is essentially between two calculations:
Source: AM Best data and research
The article is reproduced from a Best’s Special Report with the kind permission of ICMIF Supporting Member AM Best. To access a complimentary copy of the full report, please click here.
Maintaining existing ratios but using IFRS 17 data. We refer to these as net components ratios. They sometimes are called “Net/Net” ratios for short. However, the expenses used for the net ratio will be the same as the gross expenses (as commissions on reinsurance held no longer reduce expenses under IFRS 17).
A claims ratio which takes a numerator of gross claims less the reinsurance result and a denominator of gross revenue to represent the net of reinsurance claims position. There is only one expense ratio using gross numbers. These ratios are sometimes called “Net/Gross”.
AM Best’s analysis is that in general claims ratios under the Net/Gross method will be slightly higher than existing ratios, while under the Net/Net method they will be slightly lower.
Expenses ratios generally move in the opposite direction to leave combined ratios little changed from those under existing practices using IFRS 4.
However the results will vary with the type and amount of reinsurance purchased. A heavy buyer of quota share reinsurance that receives substantial ceding commissions is likely to see a significant effect with, in most cases, a distinctly higher net expense ratio under the Net/Net method than would currently be stated.
The different influences at work
The two calculations are somewhat different in nature. Nevertheless the different factors at work can be analysed as follows.
Using IFRS 17 data, the treatment of ceding commissions and profit commissions will normally lower net claims ratios when using the same approach as currently (ie, Net/Net claims ratios are lower under IFRS 17 than under IFRS 4 or local GAAP reporting). Ceding commissions increase net premiums and expenses by equal amounts when compared to existing reporting. Similarly the treatment of profit commissions decreases net claims and increases net expenses by the same amount.
Claims ratios can currently benefit from what we would call a “$-swap” effect whereby reinsurance with little to no economic impact may nevertheless act to make net ratios lower than gross ratios.
For example, consider reinsurance which is expected to operate as working layer reinsurance where, simplistically, claims equals premiums and there are no expenses. If the gross claims ratio is less than 100, then deducting an equal amount from premiums and claims will result in a net claims ratio which is lower than the gross ratio.
While this effect may be familiar to more sophisticated users of stated ratios, and likely also to underwriters managing reported ratios for their business, it can be unhelpful to an easy understanding of the dynamics of a (re)insurer’s business. In contrast, Net/Gross claims ratios should not change at all due to an introduction or expansion of a $-swap element of reinsurance as it is neutral for the reinsurance result.
The absence of this benefit to stated ratios will normally push Net/Gross claims ratios up compared to currently stated claims ratios. Most other reinsurances can be viewed as a mix of a pure reinsurance loss (being the reinsurance loss element of the reinsurance premium) and a $-swap effect, with one offsetting the other in claims ratios under existing reporting. For quota share reinsurance the two effects will closely offset each other.
AM Best has summarised some of the reinsurance effects (see below).
IFRS 17 - Impact of reinsurance factors on Net Claims Ratios vs existing ratios
| Reinsurance factors | Net components Claims Ratio (Net/Net) using IFRS 17 data | Net of reinsurance result (Net/Gross) Claims Ratio under IFRS 17 |
|---|---|---|
| Treatment of ceding commissions | Lowers claims ratios as net premiums are higher | No impact as neutral for reinsurance result |
| Treatment of profit commissions | Lowers claims ratios as net claims are lower and expenses higher | No impact as neutral for reinsurance result |
| Reinsurance loss | Same as for currently calculated ratios. Impact vs gross ratios often more than offset by $-swap effect. | Directly increases claims ratio |
| “$-swap” effect of working layer XoL reinsurance | Same as for currently calculated ratios. Lowers net claims ratios when compared to gross. | No benefit from this effect. Net claims ratio therefore higher |
Source: AM Best data and research
In general AM Best’s analysis is that the Net/Gross ratios appear less susceptible to influence from economically neutral transactions which move a mix of premiums, claims and expenses without changing the reinsurance result or underlying profit. Primarily for that reason, they may appear to be a preferable set of ratios, though how market practice will evolve is still unclear.
The claims ratio on the Net/Gross basis may be stated in two parts - that is the gross ratio and the impact of reinsurance (reinsurance result/gross premiums).
It clearly is the case that the Net/Gross set of ratios comprise a departure from current practice and, to that extent, are a discontinuity. However, in light of the discontinuities if Net/Net ratios are used under IFRS 17, users of the data will not have a smooth data transition in any event.
The question of comparability with insurers not reporting under IFRS 17 needs to be part of decisions on performance measures, but, ultimately, is subject to similar discontinuity considerations. Fortunately, as noted above, combined ratios are normally quite similar across the currently calculated ratios, IFRS 17 Net/Net ratios and IFRS 17 Net/Gross ratios.
For many insurers, there is a significant likelihood IFRS 17 will provoke a fundamental shift in the reporting of non-life performance measures in the direction of net-of-reinsurance result (Net/Gross) ratios.
IFRS 17 reporting splits claims expenses into claims incurred on coverage provided in the year (or “pure-year” claims) and adjustments to liabilities for incurred claims. A pure-year claims ratio will therefore be available automatically.
It also is important to highlight that the claims ratios referred to in the table above will normally be overlaid by a general trend for ratios to be lower under IFRS 17 compared with currently stated ratios, due to the discounting of claims reserves. AM Best expects discounting to reinforce market discipline as underlying profitability will become more directly comparable across lines. This will especially be the case for longer-tail casualty lines as the 100% combined ratio level will become a clear indicator of whether or not business is profitable.
Deferred acquisition costs (DAC)
IFRS 17 takes a different approach to DAC than is usual in current financial reporting. Essentially the acquisition costs are treated as just another cash flow for reserving purposes. This does not distort profitability as a realistic view of all future cash flows is taken.
What currently is an acquisition costs asset on the balance sheet is netted off against the liability for remaining coverage (LRC) under IFRS 17. In other words, it is a negative liability instead of an asset.
However, in addition:
- Some acquisition costs may be allocated to future renewals or new business under IFRS 17. This will usually be a new feature for a (re)insurer’s financial reporting. The amount is not treated as an asset but is, again, netted off against the LRC as for conventional DAC. This quantity, which for present purposes we call IFRS 17 DAC, is required by the standard to be disclosed in financial statement notes.
- When IFRS 17 DAC is amortised, it is amortised into the conventional DAC quantity which we call here unamortised acquisition costs (UAC). The UAC is amortised into the profit and loss over the remaining duration of the existing business, as under current reporting.
In a static book, the UAC should be similar to DAC under current reporting, while IFRS DAC will be an additional quantity.
We expect that the UAC will normally be available, though it is not a required disclosure under IFRS 17. Indeed insurers will have to maintain a record of the UAC as it is the LRC gross of UAC that is amortised into revenue. In addition the UAC is amortised into expenses.
Users will have to decide how to treat these new quantities in capital models. AM Best believes one option would be to deduct IFRS 17 DAC, net of associated tax, from equity before entering the equity figure in capital models. This would normally maintain continuity with current treatment. Meanwhile the UAC, which may be expected in broad terms to be similar to existing DAC, could be treated similarly to existing DAC.
Use of contractual service margin (CSM)
AM Best expects that the CSM in IFRS 17 reporting will be used to assess the net economic value due to long-term business as referenced in AM Best’s Available Capital and Holding Company Analysis criteria. The risk adjustment may also be considered in the assessment. Value of in-force from embedded value reporting was used for many years to derive the net economic value due to long-term business for some life insurers. However the discontinuance of this reporting by many insurers has led, for some European insurers, to Solvency II data being used for the calculation. Through the CSM and the risk adjustment, IFRS 17 should provide a more uniform and publicly disclosed measure across territories.
IFRS 17 requires records of profitability by annual cohort, in the form of the CSM, to be maintained separately for each group of policies. The groups must comprise policies that are all issued in the same 12-month period. IFRS 17 sets a specific treatment for loss-making groups (referred to as onerous policies) in that the estimated losses are recognised in profit and loss when they are identified, rather than being amortised over the duration of policies as with the CSM.
AM Best notes that while IFRS 17 prescribes the treatment of loss-making groups, it does not require the contribution by cohort of profitable contracts to CSM amortisation to be disclosed. It does require that some detail be given for the run-off profile of the balance sheet amount.
However we expect users will want to know the contribution to current profitability, in the form of CSM amortisation, from recent and older cohorts. AM Best anticipates that the estimate of profitability made when the business was written - as represented by the CSM from business written in the year in roll-forward tables - will be unsatisfactory on its own. Users will want to see if this profit is delivered. Indeed, the absence of any clear line of sight to whether estimated new business profit was ever delivered became a fundamental drawback of embedded value reporting.
In the view of AM Best, that dissatisfaction was a foundational force behind early support from users for IFRS 17. The history of cohort profitability and the narrative that management can supply around the record, should be a pivotal contributor to improved transparency under IFRS 17 for life insurers.
Capital in participating funds
Under IFRS 17, the options for representing capital in participating funds on the balance sheet widen. Rather than simply a choice between liability and equity, the new standard offers three alternatives: liability, CSM or equity. There are, as before, different slices of participating funds to consider.
The capital profit share segment represents amounts that would become shareholders’ were the capital all to be distributed to policyholders (and shareholders) under the usual terms for bonus payments.
The expected payout to policyholders will be a liability. The shareholders’ profit share which arises on these payouts to policyholders will pass through the CSM. Timing differences may mean that a portion of the profit share has already been amortised into equity at a balance sheet date - before the associated payout to policyholders.
However AM Best sees a degree of uncertainty at this stage as to how other sections of participating funds will be represented in IFRS 17 balance sheets.
Practices may vary between territories and depend on the substance of participating contracts. For example, consider the capital profit share in a territory where there is a degree of ring-fencing of participating funds. It may be classified as CSM in which case it would be an “orphan” CSM in the sense that there is no expectation it will be distributed out of the participating fund and become available to meet liabilities elsewhere in the group.
Similarly if the capital profit share is considered a liability then, unless it is separately identified, it will be indistinguishable from the policyholder liabilities and its role as capital will not be apparent. If it is classified as equity then, again, it will be indistinguishable from equity elsewhere in an insurance group that will be far more fungible to meet losses across the group.
Ultimately AM Best will want to identify the sum of the two capital segments and the profit share, as these amounts will be considered appropriately to their role in an insurer.
For KPIs and participating business, we expect users will want to know the most current view of the profitability of recent cohorts, and not be limited to an estimate of the profitability made when the business was written.
It will be disappointing if this up-to-date cohort profitability is not available for some European life insurers should the European Union (EU) “carve-out” annual cohorts for significant segments of participating business from its adoption of IFRS 17. The history of cohort profitability is a crucial marker of the health and management of participating funds.
Return on equity
Return on equity (ROE) is an important indicator for performance assessment and will be calculated in the normal way using IFRS 17 reporting. However AM Best envisages that additional ROE-type KPIs may be also be calculated. For example the CSM may be regarded as value that has already been created at the balance sheet date and ROE might be assessed on the basis of a denominator that includes the CSM. In this case the numerator would be
IFRS 17 net profit + (taxed) unwind of the CSM + (taxed) CSM on business written in the period – (taxed) amortisation of CSM.
Another variation may involve adjusting CSM amortisation for loss-making contracts. In principle some insurers who normally write profitable long-term business accounted for under the general measurement model (GMM) might occasionally write business that turns out to be significantly loss making, perhaps due to varying market conditions and wider economic circumstances. If this were to happen, say, once every five or 10 years then AM Best foresees that CSM amortisation will eventually be perceived only to represent the “good news”.
The extent of any bad news arising in, say, one year out of five would always be missing from CSM amortisation as the losses are taken immediately and therefore do not impact the CSM.
Once this is appreciated, a view of underlying performance that captures the average per annum loss from occasional loss-making years in CSM amortisation becomes a natural next step. An ROE that uses such a CSM in the numerator might be regarded as a preferable, indeed powerful, measure.
Fortunately an average per annum loss from loss-making groups over the policy durations is explicitly reported in IFRS 17 under the GMM in the form of loss component amortisation.
Details of accounting presentation are not the central subject of this note but loss component amortisation appears as a negative insurance service expense in IFRS 17. AM Best envisages that an additional ROE may be calculated by deducting a (taxed) loss component amortisation figure from the numerator.
The ROE calculation also would recognise, for consistency, a relocation of the balance sheet loss component amount away from being a deduction from equity (arising when the losses were taken “up-front”) to being a deduction from the CSM instead. A (taxed) balance sheet loss component amount would therefore be added to equity to provide a new denominator.
It may be several years until loss components become a significant feature. AM Best therefore envisages that, despite the attractions of using an adjusted CSM in ROEs as above, the use of these ROEs may develop only over a period of time following the introduction of IFRS 17.
IFRS 17 ROEs - an improvement on existing measures for the life (re)insurers
AM Best considers that ROEs derived from IFRS 17 profit will be far more meaningful for life (re)insurers than currently calculated ratios.
Whereas users of financial statements might normally expect a relationship of broad equality between ROEs and internal rates of return (IRRs) on a company’s products, this relationship is largely broken for life (re)insurers. It is, at least anecdotally, a commonplace for life (re)insurers to quote estimated IRRs that are not obviously related in a systematic way to ROEs calculated from accounting data. Indeed these IRRs are usually presented as an explicitly additional data point.
ROEs calculated from IFRS 4 data normally reflect some mix of a measure of equity which is often too low to be an economic measure and widely varying reported profit profiles. With significant exceptions, life insurance profit profiles are often reported late in the policy duration (particularly for participating business), often referred to as “back-loaded”.
Back-loaded profits would normally lead, in a mature industry, to accounting ROEs that are considerably higher than IRRs. On the other hand life (re)insurers also frequently have to explain the low ROEs (at least when compared to stated IRR estimates) that arise from value-type reporting formats such as embedded value, or Solvency II in Europe. Here the issue is that the profit is significantly “up-fronted”, inflating the “E”, so that surplus and profit do not provide a run-rate view of profitability.
Nevertheless the actual relationship between ROE using reported data and achieved IRRs will vary widely, being influenced by a range of factors, many of which relate to reporting conventions, that in general are inadequately signposted by the financial statements.
In any event, accounting ROEs under existing reporting, even though they may be a first-port-of-call for users as a performance measure, do not provide the expected data point of an achieved IRR for business on the books, or may do so only rarely (and even then most often by an accidental mix of factors and reporting conventions).
However AM Best sees the more level profit profile over the duration of policies under IFRS 17 as leading to ROEs that will, for the first time, be expected to provide a meaningful indication of profitability for business currently on a life (re)insurer’s books. ROEs themselves should also be far more level through a policy’s duration under IFRS 17 and, on that basis, AM Best’s current expectation is that they will in general approximate far more closely to IRRs. However there are clearly many moving parts in the process, not least of which is the level of capital retained in the company.
Life segment reporters are therefore likely to be called upon to discuss the relationship between ROE and the profitability of their products with a reduced reference to the vagaries of financial or actuarial reporting. Underlying achieved product profitability should become the focus to a much greater degree. We hope this is not too optimistic a view. Discussion of profitability would, in this way, assume a far more recognisable shape, at least for users that are familiar with the reporting to external financial stakeholders of businesses in other economic sectors.
In summary, a fundamentally closer relationship between accounting ROEs and underlying profitability should enable more conventional and easily understood conversations between life segment (re)insurers and users of financial reporting. AM Best views this aspect of improved transparency as one of the principal benefits of IFRS 17 for users, and hence also for (re)insurers. It should help enhance and widen (re)insurers relationships with the universe, both existing and potential, of external financial stakeholders.
