In this short discussion paper, we analyse the reasons for this, and the potential changes we might expect in M&A processes.
Will the approach to valuing companies in M&A processes change, and will some companies be valued at a different level due to the advent of IFRS 17?
In our experience, most acquirers of insurance companies in Europe value them based on the present value of expected future dividends, at a discount rate reflecting the shareholders’ required rate of return on capital.
Where Solvency II is applicable, it will be a constraint on dividends, due to the requirement to maintain an adequate solvency ratio. There may, in addition, be other constraints, which could include:
- Regulatory rules or guidance limiting dividends to accounting profits in some way.
- Internal guidelines linking dividends to accounting profits.
- Asset coverage rules (e.g., a requirement to cover certain types of liabilities with certain types of assets, valued in a particular way).
As discussed in more detail in the next section, we think that in most cases the introduction of IFRS 17 will not directly impact dividend payouts, either because dividends are not dependent on IFRS 17 at all or because any theoretical constraint is unlikely to “bite.”
We also note that there have been various public statements from companies that they expect IFRS 17 (and associated IFRS 9) changes to have no or limited impact on aspects such as: capital and cash generation, dividends, solvency and strategy. For example:
Aviva plc: “IFRS 17 does not impact our strategy, capital generation, dividend guidance, or capital return outlook.”1
NN Group: “No impact on strategy, targets or dividend/capital return.”2
Generali Group: “IFRS 17 and IFRS 9 implementation has…. no impact on cash and capital generation, net holding cash flow, dividends and Solvency.”3
Allianz: “…fundamentals unchanged…. Limited economic impact on Solvency II, Cash flow, Dividend.”4
On the other hand, earnings and earnings growth are still an important part of companies’ messages, and some have made statements linking growth in earnings with dividends.
If there is indeed no impact on dividends from the introduction of IFRS 17, we may need to understand whether the timing of IFRS 17 earnings may nevertheless be important to companies. We return to this question later.
Why we believe the introduction of IFRS17 will generally not impact dividend payouts
Applicability of IFRS 17
The extent to which IFRS 17 is compulsory, and to which it will replace local GAAP accounting, is far from uniform across the different European markets.
In most major European markets IFRS 17 is only compulsory for listed companies for consolidated group financial statements, although IFRS 17 can usually be applied voluntarily in these markets. With one or two exceptions, even where IFRS 17 accounting is adopted by an insurer, it is still required to produce accounts on a local GAAP basis.
It therefore seems unlikely that there would be a limit on dividends related to IFRS 17.
However, there are some markets, particularly in Southern and Eastern Europe (although notably excluding the largest markets like Italy, Spain, Poland and the Czech Republic) where IFRS accounting replaced local GAAP some years ago, and thus IFRS 17 is compulsory for all insurers. In these cases, there could conceivably be constraints on dividends related to IFRS 17 profits. However, it is important to stress that the fact a constraint potentially exists does not necessarily mean that it will be a biting constraint (i.e., one that can actually be expected to change dividends)—we explore this point further below.
Could IFRS 17 be a “biting” constraint?
1. Basic scenario
Let us consider the case where IFRS 17 profits do act as a potential constraint on dividends, as well as having to maintain a particular target capital under Solvency II.
For this purpose, we will consider IFRS 17 as acting as a constraint if it “holds back” distributable profits more than Solvency II at the acquisition date (thus ignoring differences in how various elements may run off under the two different measures) and excluding future new business in the first instance.
In most cases, under IFRS 17, there cannot be an immediate profit or loss at acquisition.5
Under Solvency II, on the other hand, the purchase price for the assets and liabilities should normally exceed the initial Own Funds less the required capital—typically expressed as the Solvency Capital Requirement (SCR) multiplied by the target solvency ratio—which we can refer to as the “free Own Funds”. This is because the free Own Funds capital can be distributed immediately and is thus valued at nominal value—in addition, there will be value attached to the release of the risk margin and Own Funds covering the required capital as the business runs off, plus any value placed on additional future capital generation from risk-adjusted real-world investment returns, in particular.
(See our paper, available at: https://www.milliman.com/en/insight/s2av-a-valuation-methodology-for-insurance-companies-under-solvency-ii.)
Therefore, net of the purchase price, distributable profits on a Solvency II basis are almost inevitably negative at the date of the acquisition, and in this case the IFRS 17 earnings would not therefore act as a constraint on dividends.
This is equivalent to saying that it is usually not possible for a transaction to be immediately self-financing on a Solvency II basis (which would be an illogical situation). This is a different situation from writing new business, which may be self-financing (initial Solvency II Own Funds greater than required capital), and which thus could potentially create a constraint under IFRS 17 (which doesn’t allow an initial gain); this is discussed further below.
If the contracts are acquired in a portfolio transfer, as opposed to a business combination,6then it is possible to make a loss at acquisition under IFRS 17. This would be the case if the purchase price for the assets and liabilities was greater than the value of assets less the value of liabilities under IFRS 9/17,7 hence consequently producing no initial contractual service margin (CSM).
If we assume, which is generally not unreasonable, that the best estimate value of liabilities (that is, excluding any risk margin, risk adjustment and CSM) and the value of assets are similar under Solvency II and IFRS 17, then the difference between the constraint on distributable earnings from the two Standards in this scenario depends on the relative level of:
- For IFRS 17, the risk adjustment (as noted above, there is no initial CSM in this case)
- For Solvency II, the risk margin plus the required capital (typically SCR x target solvency ratio)
Almost certainly the latter would be more of a constraint (indeed, in our experience, the risk adjustment will typically be smaller than even just the risk margin).
So, for the basic scenario considered above, it is very unlikely that IFRS 17 would act as a constraint on dividends compared with Solvency II.
2. Other scenarios
If we then consider future new business on its own, whether profits are held back under IFRS 17 more than with Solvency II depends, in particular, on the relative levels of:
- For IFRS 17, the CSM (driven by the basic level of profitability) and risk adjustment
- For Solvency II, the required capital and risk margin
As set out in our paper (available at https://www.milliman.com/en/insight/impact-of-ifrs-17-on-insurance-product-pricing-and-design), IFRS 17 is more likely to constrain profits for products with high profitability (on a market-consistent basis) and/or low capital consumption. The converse is true for products with low profitability and/or high capital consumption.
Therefore, in the case of a target where a material part of value comes from highly profitable new business, for which IFRS 17 may recognise earnings more slowly than Solvency II capital generation, a closer look at the impact of this on the expected future dividends (if applicable) may be merited.
A further point relates to contracts measured using the variable fee approach (VFA) under IFRS 17, where there will be a re-spreading of the impact of the shareholders’ share of projected real-world investment uplifts above risk-free over the future, via an increase in CSM (the “bow-wave” effect), creating a hold-back of profits. This means that the impact of such real-world uplifts will emerge more slowly under IFRS 17 than Solvency II (where they are recognised as surplus as they occur); however, this needs to be weighed against the additional required capital associated with the related market risk under Solvency II, which, all other things being equal, is likely to produce a bigger hold-back of profits than the bow-wave effect. (We note that any accumulation of real-world uplifts into CSM prior to the transaction is not relevant, as the CSM is reset at the transaction date.)
Thus, whilst a constraint is theoretically possible, we would generally expect IFRS 17 not to act as a constraint relative to Solvency II in respect of this point.
The arguments presented in the “basic scenario” above still hold true for short-term contracts measured with the premium allocation approach (PAA) under IFRS 17. The fact that, net of the purchase price, distributable profits on a Solvency II basis will generally be negative would still be true, and so the only possible constraining case is that where a loss is recognised on IFRS 17 (for onerous contracts). However, any initial loss on an IFRS 17 basis would be less than for Solvency II, in particular because the IFRS 17 risk adjustment will almost certainly be less than the Solvency II risk margin plus required capital. One particular area which could need further investigation, however, is that of material differences in contract boundaries between Solvency II and IFRS 17.
One other point to consider is post-transaction management actions, which accelerate distributable profits under Solvency II compared with the target on a standalone basis, for example related to capital synergies with the acquirer. If these actions had a very material impact they could actually create immediate distributable profits on a Solvency II basis (depending on the extent to which these actions are reflected in the purchase price). Please refer to our paper, available at https://www.milliman.com/en/insight/getting-the-best-from-buy-side-opportunities. We do not comment on this scenario further here, but such a case may need to be examined more carefully.
Overall, however, we think it will rarely be the case that IFRS 17 forms the “biting constraint” and we expect projected dividends normally to depend on Solvency II.
Why IFRS 17 may still be relevant
Perception and preference
We return here to the question of whether, if IFRS 17 has no impact on dividends, it may nevertheless be important to companies. For example, is it important because it impacts the perceived value of the company?
The question we should perhaps be asking is this:
If two companies have an identical solvency position, identical expectations for future dividends and are expected to produce the same level of IFRS 17 future earnings, but one would produce these earnings more quickly, then would an investor have a preference for this company over the other one?
It is tempting to conclude that a well-informed investor who fully understood the position should have no preference and that, if the market did have a preference for the company with the quicker earnings, it would create an arbitrage opportunity to sell shares in this company and buy them in the one with the slower earnings pattern but identical dividend prospects.
However, the effort insurers appear to be putting into decisions around items which impact only the timing of IFRS 17 earnings would seem to imply that they believe this will have some impact on the perceptions of investors.
Of course, investors will typically not have full information and understanding of reported earnings, and it is possible that they may perceive a different value of two companies with identical economic value but different reported earnings.
It is not even clear whether investors will have a preference for a company for which earnings emerge more quickly but with more volatility or one which has slower but smoother earnings. Again, the implications of these two alternatives may be hard for external parties to fully understand. A company might trumpet steady earnings growth as being something which can be expected to continue and is hence an indicator of long-term prospects for dividend growth—but it may be difficult for an external investor to see the extent to which this is due to real economic value creation (e.g., by selling profitable new business and increasing operating efficiency) or simply to choices taken in how to report earnings, which allows implicit and explicit margins to be held back and then released to create a rising pattern of earnings even if real value creation is flatlining.
Explicit margins would include the levels of the CSM and risk adjustment, whereas implicit margins could come from things like tending towards prudence when defining noneconomic assumptions such as lapse rates and dynamic policyholder behaviour assumptions for life businesses or ultimate claims costs for non-life. Even if assumptions may in theory be “best estimate” there can still be a lot of scope for different approaches, often with a material impact on value.
Even if IFRS 17 doesn’t directly impact approaches to valuation, companies which are reporting on an IFRS 17 basis are likely to be interested in knowing what impact an acquisition will make on its reported earnings.
A best guess for what might happen in the short to medium term in M&A is that buyers will be interested in understanding the IFRS 17 impact of a potential transaction but may not use IFRS 17 directly for valuation purposes or as a key financial parameter for M&A decisions.
As understanding of IFRS 17 reporting grows in the next few years, we may see an increased focus on it, but we still feel that distributable earnings will remain the main driver and these earnings will fundamentally be driven by the solvency position and capital generation.
There may also be different approaches depending on the type of buyer; a publicly quoted insurer may have a greater interest in IFRS 17 earnings, as they may feel this influences the share price more than in a private equity (PE) firm, which is more focused on the capital it has to put into the company and the dividends it can extract. Further, shareholders of a quoted insurance company will not have full transparency on underlying economic value and may therefore find IFRS 17 earnings to be a valuable indicator, whereas a PE firm should, in theory, have full visibility on the underlying economic value of an insurer it controls or is doing a thorough due diligence on.
Practical challenges in an M&A context
Given the complexity of IFRS 17, working out a full set of IFRS 17 numbers for a target would be challenging within the timeline of an M&A process (even more so if the target doesn’t currently report on IFRS 17, which will be the case for many targets, particularly in Western Europe). At least in the short term, such calculations are likely to be done on an approximate basis.
A key part of determining the impact on IFRS 17 profits will be estimating the CSM (assuming this is positive) and risk adjustment for the acquired in-force liabilities at the acquisition date. This is required to be calculated based on the fair value or consideration paid for the liabilities (regardless of any transition methods adopted by the target company), which in turn could be estimated from the purchase price paid for the transaction (eliminating the value of assets, and items such as the value placed on future new business and the impact of any management actions particular to the individual acquirer). To arrive at the CSM, the result of these calculations is then compared with the IFRS 17 fulfilment cash flows, which could be estimated from the Solvency II Technical Provisions, but taking the IFRS 17 risk adjustment as some suitable proportion of the Solvency II risk margin. There may be some splitting and proportioning needed here to give the required level of granularity.
An assumption then needs to be made about future CSM run-off into profit; some suitable proxy, for example based on Solvency II best estimate liability (BEL) and estimated outstanding duration may be suitable for this purpose.
The initial CSM on future new business in the year of sale could be estimated from information on new business value (NBV) on a market-consistent basis, with an adjustment to reflect the appropriate initial level of IFRS 17 risk adjustment. If such information is not available, it may be possible to construct simple models of new business. Again, an assumption about how the CSM runs off into profit needs to be made.
Allowing for the impact of shareholders’ share of real-world uplifts on CSM under VFA could be constructed based on an estimate of the amounts to be accumulated into CSM (based on a corresponding Solvency II projection), together with the assumed rate of run-off into profit, as above.
IFRS 17 implementation is reported to have cost the insurance industry around $20 billion globally to implement.8 If, as announced by Aviva plc, “IFRS 17 is purely an accounting change and does not affect the underlying economics of our business, our strategy or the way in which we operate,”9 then we might conclude that it will not impact which M&A transactions get done and at what prices.
Some commentators do say, however, that they expect it to improve transparency, stability and predictability of earnings and so they may expect it to lead to an overall higher valuation of the insurance industry.
In conclusion, although we don’t expect IFRS 17 to fundamentally change the way insurance M&A is done in the short term, the focus on results on this basis is likely to increase over time as companies and readers of accounts digest the Standard and become used to seeing how earnings develop over several reporting periods. This is what we saw with Solvency II, which was not used widely to guide companies on first implementation, but gradually came to be seen as an absolutely fundamental financial measure. This might lead to a bigger eventual impact than we see in the short term, but as yet it is difficult to predict exactly what this might be.
1 Aviva IFRS 17 transition update 19 July 2023 https://www.aviva.com/investors/ifrs-17-transition-update/.
2 NN Group investor update 17 Nov 2022 https://www.nn-group.com/article-display-on-page-no-index/nn-group-investor-update-17-november-2022.htm.
3 Generali press release 13 Dec 2022 https://www.generali.com/media/press-releases/all/2022/Investor-Update-Press-Release.
4 Allianz IFRS 9/17 overview 22 Nov 2022 https://www.allianz.com/en/investor_relations/conferences-presentations/IFRS9-17.html.
5 This is because a CSM is set up to eliminate an immediate profit, or recognition is made within goodwill or gain on a bargain purchase to eliminate an immediate loss (this elimination of an immediate loss doesn’t apply for a portfolio transfer—see below).
6 From IFRS 3 Appendix A – Business Combination: A transaction or other event in which an acquirer obtains control of one or more businesses. Transactions sometimes referred to as “true mergers” or “mergers of equals” are also business combinations.
8 The Actuary 21 June 2022 https://www.theactuary.com/news/2022/06/21/ifrs-17-forecast-cost-insurers-24bn.
9 Aviva IFRS 17 presentation 9 Dec 2022 https://www.aviva.com/investors/publication-of-IFRS-17-presentation/.