QBE Insurance Group Limited (QBE) Earnings Call Transcript & Summary

Hello, and thank you for standing by. Welcome to QBE AASB 17 briefing. [Operator Instructions] Please be advised that today's conference is being recorded. It is now my pleasure to introduce Group CFO of QBE Insurance, Inder Singh.

Thank you, and good morning, all. I appreciate you taking the time to join us this morning. Before we begin, I'd like to acknowledge the traditional owners of the many lands on which we meet today and recognize their continuing connection to land, waters and culture. I also pay my respects to elders past, present and emerging and extend this respect to any First Nations people joining us today. So we'll spend the next half an hour or so stepping through 3 key areas in relation to the new accounting standards that we've adopted as of 1/1/23. The first is to really give you some background on the standard. Secondly, go through some of the new accounting principles and associated impacts. And then we'll wrap up by walking through how we intend to present our financial results going forward. You've hopefully seen the data pack we've released today. It's a bit extensive, and it's got a bit of detail, but happy to take you through that through the course of this presentation. That pack includes our restated 2022 results applying AASB 17. I'll start on Slide 3 and make some introductory remarks. So AASB 17 has been in the makings for a number of years. Its ambition is to drive greater accounting consistency across the full spectrum of the insurance sector, including property and casualty, life and health insurance. Our key message today is that the adoption of the new accounting standard will not materially change the economics of our business nor the way in which we manage performance. However, the statutory financial statements as presented under the new standard will look quite different, and we will spend some time today walking through these and, in particular, how these reconcile back to the key metrics through which we will continue to manage the business going forward. The core principles behind AASB 17 are very similar to AASB 1023, our previous accounting standard here in Australia. Some of the key pillars behind the new standards include discounting of claims liabilities, which we currently do; holding a buffer above the central estimate on reserves, which we've been doing in the form of what we call the risk margin; and then a test for unprofitable business, which is similar to the liability adequacy test under 1023. Most of the regions where we operate will be adopting the new standard this year with the exception of the U.S. For many of these regions, the new standard is quite a material change, and one of the more positive outcomes of the new standard is that QBE will be more comparable to other international P&C companies that will now move to an accounting basis that is similar to ours. Moving to Slide 4. Before we get into the accounting concepts, I will make a few high-level observations around the economics of our business. Strategically, nothing changes under AASB 17, and there is no change in the relative fundamentals of our various businesses. Profitability will remain consistent. There will be no material change in the level of profitability for QBE nor the shape of profit recognition. We've made some pleasing progress in our financial position in recent years. The new accounting standard will have no impact on our capital position or capital generation, gearing or dividend policy. Finally, all our key performance indicators remain in place with our combined operating ratio and return on equity broadly consistent across the two standards. I'll now step through some of the key AASB 17 accounting principles, starting with profit recognition on Slide 6. AASB 17 is centered around two profit measurement models, which effectively determine the shape of revenue and profit recognition. For most P&C businesses where revenues from our policy are earned over a 12-month period, the premium allocation approach, or PAA, will be the primary measurement model, and it will apply to the vast majority of our business. PAA is consistent with profit recognition under 1023 and there will be no change to the shape of our earnings. The only notable business of QBE where the general model will apply is our lenders' mortgage insurance business here in Australia. The adoption of the new standard will extend the earnings pattern for this business from around 10 years to around 15 years, albeit there'll be minimal impact on overall QBE Group earnings in any 1 financial year as we'll still continue to earn the majority of LMI revenues across the first handful of years. Most of the extension in the earnings pattern is in the tail of the business where we are dealing with a relatively modest amount of premium. Moving now to reserving on Slide 7. Reserving will be consistent across the two standards. Our definition of the central estimate and approach to reserving remain unchanged. At our recent full year results briefing, we spent some time discussing the risk adjustment. AASB 17 requires a version of the risk margin called the risk adjustment. The underlying concepts are very similar, albeit the calculation of each is different. Going forward, the risk adjustment will operate within a target range, which we've set at 6% to 8% of central estimate. This target range is determined and calibrated with reference to the cost of capital allocated to QBE's reserve risk. We entered 2023 with our risk adjustment balance at 8%, which is at the top end of this range and equates to a confidence level of 90%. This means that the absolute balance from risk margin to risk adjustment has not changed at 31 December 2022. Over time, where we reside within the 6% to 8% target range will depend on a multitude of factors. However, just as we tended to hold a fairly steady probability of adequacy under the old risk margin regime, we would expect similar outcomes to the risk adjustment and shouldn't regularly adjust up or down through this range. I would note that QBE is a well-diversified business, both across geography and product, and we get some benefit for this diversification as we calibrate the 8% risk adjustment to a 90% confidence level. This confidence level is calculated on a similar basis as the old probability of adequacy. This is worth keeping in mind as you observe where our risk adjustment and associated confidence level may land relative to less diversified peers. Finally, for capital, under the APRA PCA basis, we will continue to get capital credit for the risk adjustment above the 75th percentile. So this will remain unchanged. Moving now to claims discounting on Slide 8. AASB 17 requires you to discount claims liabilities in the same fashion as we do today. The mechanics of claims discounting will continue to work in the same way as they do today, but there will be some nuances around the presentation of the different components. So it's worth just walking through each of these different components. Firstly, as we write new business, there is a benefit to the P&L from the initial discount applied to the new reserve we establish. For instance, for a policy with a 3-year term to settlement, we establish a reserve and then discount that reserve back to present value and recognize the full amount of that discount in year 1. We'll call this the initial discount. What then ensues is the gradual unwind of that discount as we approach the expected claims settlement. Separately, as risk fee rates move between balance sheet dates, any associated change in the discounted value of the outstanding claims liability flows through the P&L. So this process works in the same way across both standards, though presentation is slightly different. Under 1023, all three claims discounting elements sat within net claims. AASB 17 introduces a new line item in the P&L called insurance finance income, which sits below the underwriting result. The impact from changes in risk-free rates and the discount unwind, essentially items 1 and 3 in the slide, move into the insurance finance income while the initial discount benefit will remain in the claims component of the insurance services expense. There's a lot to absorb on this point, and we'll return to it later in the session when we spend more time on the result presentation. Moving on, beyond presentation of discounting, the only notable change in AASB 17 relates to the illiquidity premium. Going forward, the rate we use to discount our reserves will include an illiquidity premium which, in essence, represents the spread of our risk-free to capture the frictional cost of trading a portfolio of insurance liabilities. We have established an illiquidity premium of 32 basis points. This is broadly referenced to the historically observed illiquidity premium embedded in corporate credit spreads across our fixed income book. We'll review the illiquidity premium at times where we see significant changes in our business mix or long-run average credit spreads, though our base assumption is that the 32 basis points will remain stable for the foreseeable future. Moving on to the onerous contract slide on Slide 9. The standard requires the pull forward or early recognition of underwriting losses on in-force insurance contracts, where we have enough information to reliably estimate this. We see onerous contracts as similar to the liability adequacy test under 1023. The LAT test assess the sufficiency of net unearned premium to cover premium liabilities plus the risk margin and any deficit was recognized upfront. In both instances, we are trying to project future profitability and how it may develop. And in both instances, there is no change in ultimate profitability, but rather a change in the timing of loss recognition where any expected loss is taken upfront. The onerous contracts test will be conducted at a more granular level than the LAT test. Under the LAT test, we really tested at the divisional level. And with onerous contracts, the assessment will be calibrated across half a dozen portfolios for each division. For instance, here in Australia, we'll distribute 20-odd sells down across 6 portfolios being, in this case, commercial lines, CTP, personal lines, New Zealand, Pacific and Australia other cell. In theory, the more granular testing could result in greater likelihood of an onerous contract provision than a LAT deficit. However, we believe we can establish a good process around this to ensure we manage volatility within a reasonable range and ensure that movements in the onerous contract provision over time reflect the commercial dynamics of the business. Importantly, we have a well-established planning and performance management rhythm around the cell reviews and the analysis from these reviews will underpin the onerous contract assessment. On transition, we needed to retrospectively test for onerous contracts across 2021 and 2022. Entering 2022, we had an onerous contract provision of around $30 million, which is an increase to around $106 million as we enter 2023. The business segments that gave rise to the balance of 2023 were predominantly Australian home and private motor, certain runoff segments in North America, plus the middle market business in North America. In Australia, the returns in our domestic personalized franchise had been challenged by the sharp increase in inflation over the last 12 months. And at the point of assessment, rate increases were not keeping pace with the claims inflation we were observing, resulting in the onerous contract charge at the start of this year. Moving now to transition adjustments on Slide 10. This bridge brings together most of the concepts we've just touched on. As we transition to AASB 17, we are required to retrospectively apply the new standard to our 2022 accounts. Any changes as a result of the new concepts will be booked as a transition adjustment to net assets. So moving from left to right to touch on a few of the key drivers, the risk adjustment impact is neutral, where I noted earlier, there was no change in the absolute balance from risk margin to risk adjustment. For discounting, the introduction of the illiquidity premium means a slightly lower discounted claims liability, which increases net assets. The onerous contract provision I just discussed around $106 million decreases net assets. LMI shifting to the general model results in a slightly longer earnings pattern. In this instance, we need to reverse our previously reported profits back into unearned profit or what's going to be known as the contractual services margin. Collectively, these changes result in limited change to our net assets, which was consistent with the comments we made in February. Moving to Slide 11. I want to conclude this section by reiterating that the changes that come with the new standard will have limited impact to our profitability, capital or balance sheet. Our restated FY 2022 return on equity is broadly in line after accounting for the COVID risk margin release and the establishment of an onerous contract provision, which we'll discuss further shortly. Gearing is broadly unchanged, as is our definition of gearing. And finally, there is minimal impact on capital, and we'll continue to calculate the PCA on the same basis. APRA has signaled they don't expect AASB 17 to result in any change in capital for the sector and have recently released an updated standard with limited changes. Moving now into the presentational aspects of our results, starting on Page 13. This is a summarized version of our statutory results under the new standard. As you can see on this slide, the statutory presentation looks quite different to what you've been accustomed to under the previous standard. The key takeaway from this section is that while statutory presentation will change for at least the medium term, we intend to present a management P&L, which preserves the key KPIs and metrics we speak to today. This slide steps through a few key features of the statutory presentation. Firstly, AASB 17 is focused on a gross view of insurance performance, i.e., gross of reinsurance. Revenues and the underwriting results are presented on a gross basis which is, of course, different to the net view under 1023. Insurance revenue is the primary revenue metrics and very consistent with gross earned premium under 1023. Presentation of GWP is not required under the new standard. Insurance claims, commissions and other expenses are aggregated on a gross of reinsurance basis into insurance service expenses. These first two line items then form the basis of the gross insurance result. The effects of reinsurance are then presented separately as a stand-alone reinsurance result. So what you see here as reinsurance expenses largely represents QBE's outward reinsurance expenses, i.e., what we pay for our reinsurance program. Reinsurance income is then largely reinsurance recoveries. Under 1023, we would have presented claims expense net of these reinsurance recoveries. The sum of the first four lines equate to the insurance services results, which is the AASB 17 underwriting results. Moving to nonattributable income and expenses. AASB 17 requires that income and expenses associated with activities that don't directly relate to the fulfillment of insurance contracts be disclosed separately from the insurance services result. For QBE, these numbers will be relatively small and, on the expense side, largely represent head office expenses and some technology and change costs. Income here represents fee income, which we generate in certain functions, for instance, where we might support the administration of a government managed workers' comp scheme. Under 1023, these items were, of course, included in underwriting expenses and within the underwriting result. For investments, accounting is very similar and will continue to provide a split of our policy and shareholder funds. We realized the insurance margin is an important metric in our home market, though just as a reminder, unlike our domestic peers, we manage our investment book as a single portfolio with a single investment strategy. Moving to claims discounting. This is an interesting theme that comes with some pros and cons. We discussed the new insurance finance income line earlier. The impact from changes in risk-free rates on outstanding claims will now reside in insurance finance income. We like this feature as it takes the mark-to-market volatility from risk free rates out of the underwriting results, removing the need to maintain any discussion of a combined ratio ex-discount rate movement. Conceptually, we are less enthusiastic about splitting the initial and subsequent claims discounting elements I spoke about earlier. Both currently sit within the ex-cat claims under 1023, and AASB 17 proposes keeping the benefit from new discounting within claims, but moving the unwind to insurance finance income. For us, these elements tend to offset each other reasonably well over time such that the net of the two is usually contained to only a few minor impacts on the claims ratio. So we think reporting them separately will introduce volatility for items, which are difficult to predict and/or guide to. And that's probably a good segue into Slide 14. So we intend to begin presenting a more formalized management view of the P&L. This will form the primary basis of our new investor report and how we narrate performance going forward. The rationale here really is twofold. Firstly, we want to maintain a number of the key KPIs we speak to today, for instance, the ex-cat claims ratio and the net combined ratio. Secondly, we want to ensure that we present our performance externally on the very same basis that we assess performance internally. We hope these decisions will ultimately be additive to your ability to assess and analyze our results. And these diagrams are a good way to visualize the adjustments we'll be making. It's important to note that these adjustments will be recurring and are really geared around two categories. Firstly, the desire to maintain a net view of our results, both our revenues and underwriting profitability. We have set performance internally on a net basis. And we use significant quota share reinsurance across the group and certain segments like Crop, and these would be quite complex to present on a gross basis. The second category of adjustments encompasses a handful of items we're making to promote the clearest basis for performance assessment going forward. I'll now set through each category in turn, starting with the net view adjustments on Slide 15. This table forms the basis for the stat to management reconciliation we'll be providing going forward. It essentially highlights how we rearrange the statutory presentation of the insurance service results into a net view presented as net insurance revenue, net claims expense, net commission and expenses. We're effectively taking insurance services expenses line under AASB 17 and breaking it up into its three component parts. We think that giving you this additional granularity is important. Net insurance revenue is very similar to net earned premium under 1023 and across the other categories, there is also a limited change. On Slide 16, we then make a series of adjustments geared to give you what we believe is the clearest basis of assessing our performance. So briefly taking each of these adjustments on Slide 16 in turn. Firstly, we'll include nonattributable expenses and income within underwriting expenses and in effect, maintain these items within our combined ratio. [ Consequently ], we see all head office expenses, including technology and marketing expenses, as ultimately supporting the enterprise in its ambition to provide insurance and support our customers. In relation to claims discounting, the discount unwind element I noted earlier will be maintained within claims expense. Turning to underlying prior year development. This change is less to do with AASB 17 and more about improving the transparency of our accounts. Some of you will recall we often discuss an underlying view of prior year development, which can differ from our reported number. We do this to account for certain classes of business where PYD is neutralized by other underwriting line items. For instance, in Crop, where we have PYD, it's generally offset at the NEP line due to payments to or refunds from the federal scheme and the net impact of these items is profit neutral. There are similar versions of this across other classes and going forward, we plan to more formally adjust the results for these profit-neutral elements, which will alleviate the need to discuss two versions of prior year development. Finally, the accounting for loss portfolio transfers, such as the $1.9 billion reserve transaction we announced earlier this year, will be different under the new accounting standard. Accounting for the upfront cost of an LPT is much more straightforward under the new standard with the net cost included within reinsurance expenses. This is an improvement on 1023 where the commission -- where the composition of the net cost would be presented across reinsurance expenses, prior year development, risk margin and commissions. And this old presentation resulted in large distorting impacts across our P&L and underwriting ratios. So going forward, we don't need to make the upfront adjustments on LPTs as we go forward. The less attractive feature of AASB 17 for LPTs is that beyond day 1 and the recognition of the upfront cost, LPTs could have a large impact on the P&L. As claims settled within the loss portfolio transfer, you have those claims settlements flowing in on a gross basis through the claims line with an offsetting reinsurance recovery. This process continues right through until all claims are settled. So if you think about the transaction we just announced, having roughly $1.9 billion in gross claims settlements rolling through the P&L over the next few years is going to create a fair amount of noise for something which is neutral to profit. For management reporting purposes, we'll exclude the ongoing impacts of these transactions, removing the offsetting claims and reinsurance expense elements. Moving now to Slide 17. This then is the end product and how we intend to present our new management P&L. We believe this gives you the best basis to assess our performance. In the first section of the P&L over the first 8 rows or everything above what's titled Analysed As, we present our management result in the AASB 17 format. For those interested in the new statutory presentation, including the various new line items and gross metrics, we are giving plenty of prominence to the disclosures required under the new standard. Effectively, then everything below Analysed As takes into account the presentational adjustments I just referenced. And outside of a few naming conventions, should look quite familiar. You can see the insurance operating result of $632 million is consistent across the two segments of the P&L. At least for the near term, we'll primarily speak to performance based on the second organization of the insurance operating result. Below the insurance operating result, the final item I wanted to flag is the line we've added below net insurance finance income. As a reminder, net insurance finance income for management reporting will represent the impact from changes in risk-free rates used to discount outstanding claims. This, in effect, is taking the mark-to-market volatility out of the underwriting result. As you're aware, our ALM strategy is based on running a fully matched balance sheet to neutralize the P&L impact of changes in risk-free rates. With the liability risk free rate impact now out of the underwriting result, we intend to report the asset risk free rate impact alongside net insurance finance income and a new line item called unrealized gain or loss on fixed income securities. In FY 2022, we experienced a heightened degree of basis risk in our ALM process given record interest rate volatility, resulting in a mismatch loss. But over time, you'd expect that the net of these two line items should be close to 0. This ultimately means that volatility driven by changes in risk-free rates on both sides of the balance sheet is now taken out of the underwriting result and investment result with the impacts contained to this section of the P&L. Moving on, I wanted to conclude this section by noting that we aren't introducing any new management definitions of NPAT here. NPAT in our statutory accounts will be the same as the NPAT we report to management purposes. And these adjustments setup we just walked through are purely presentational or geographic in nature. We think we've landed in a good place for now. Based on our consultation with the industry, both domestic and abroad, we expect a wide range of initial approaches to presentation. Over time, we expect the market will ultimately drive some convergence around AASB 17 presentation and KPIs but that will take time and may differ by region. To this end, we expect our presentation will evolve with the market over the next few years. Hopefully, however, you can appreciate that we are trying to do this slowly and minimize disruption. Moving now to Slide 18. This slide highlights how we'll be presenting claims under the new standard. All claims discussion and presentation will continue to revolve around the three categories we spoke to today being ex-cat, cat and prior year development. Under the new standard, there will be two key changes to note. Firstly, the impact from onerous contracts will be recorded within ex-cat claims. As we discussed earlier, we're not expecting onerous contracts to introduce any material volatility. Hence, we don't intend to have a separate onerous contracts line item and it will reside within ex-cat claims. If we were ever to have a large onerous contract impact, we'd likely call that out when we present and analyze performance of the ex-cat ratio. The other key change relates to how we'll classify the risk adjustment. Under 1023, in our claims table disclosure, we provided a line item for the net movement in risk margin. Essentially, as new policies are written, we established a new reserve for expected claims and a risk margin above that reserve. As that policy approaches claims settlement, the risk margin is gradually released. This initial establishment of risk margin and subsequent unwind of the risk margin all played out within this one line item under the old accounting standard. While the mechanics of the risk adjustment will play out in the same way, we will present these impacts slightly differently under the new standard. Going forward, the initial establishment of the risk adjustment will reside in the ex-cat claims ratio. This will always be a headwind or a drag on the ex-cat line. The subsequent unwind of the risk adjustment will now reside within the prior year development where we'll effectively be reporting prior year development on total reserves. The PYD line item will then become the sum of the two components, any development on the central estimate like we report today and the unwind of the risk adjustment. As the risk adjustment unwind should always be a benefit to the P&L, what will likely occur is we'll always have a degree of favorable prior year development provided there are no meaningful top-ups to the central estimate. You can see here in the middle of the slide that we should -- that we would have reported favorable prior year development in both first half and full year 2022. We realize this is quite a change and appreciate that having some assessment of development around the central estimate remains important for the market. While our reported PYD will be the sum of these two elements, in the broader disclosures, we'll continue to provide a disaggregation of this PYD similar to the chart here, either in the investor report or in the investor presentation. This approach looks very consistent with what our peers are adopting across Europe and the U.K., and therefore, how we report PYD should be directly comparable. The logical question here is, obviously, to what degree of favorable development should we expect going forward. On the central estimate, as you'd expect, we plan for 0 development around this. And it's clear in this slide that the risk adjustment unwind was worth around $500 million or 3% in FY '22. We'll likely need to go through a few reporting periods before we get a good sense of what a normal level of risk adjustment unwinds looks like for QBE. So now if you think about our risk adjustment balance of $1.3 billion, and an average term for settlement of around 3 years, roughly 1/3 of the risk adjustment should roll off every year. Things which could influence this outcome include growth in reserves, movement in reserves due to interest rates, the timing of large cat events and the duration of reserves. The other side to this discussion is that our restated ex-cat claims ratio is higher due to the initial risk adjustment strain. This added around 4% on our restated numbers. Similarly, we will provide some additional context in the investor presentation so that you can see the impact of the risk adjustment strain on the ex-cat claims ratio to give you the best basis to assess underlying trends. This line item is, of course, going to be driven by growth and many of the same influences I just noted for the unwind of the risk adjustment. Moving now to Slide 19. I want to briefly outline where we are seeing some differences in our restated FY 2022 numbers compared to the equivalent numbers under the old standard. You'll notice that both our restated FY 2022 NPAT is lower and our combined ratio is higher than the metrics we reported in February. One feature explains almost all of this variance across our KPIs and relates to the $160 million COVID risk margin release, which occurred in the second half following the domestic business interruption ruling. The $160 million release isn't included in our restated numbers. This is due to very technical circumstances related to the different definition between risk margin and risk adjustment. The risk adjustment represents the compensation required or cost of capital allocated to reserve risk whereas the risk margin represents an allowance for inherent uncertainty within the central estimate. We're getting a bit academic here and it's a bit of a technical issue. But ultimately, in how we've retroactively applied the standard in 2021 and 2022, the $160 million risk margin release would not have occurred in the P&L. As these charts highlight, we think the more sensible thing to do is consider the restated numbers adjusted for this release. And on that basis, there is very limited change and largely contained to the establishment of the onerous contract provision and some other minor elements. As noted earlier, with an onerous contract provision now established at around $106 million and with what remains a relatively favorable outlook for underwriting margins and investment income, we expect this item should be more stable going forward. Moving to Slide 20. I won't talk to all these numbers today, but we've outlined a number of other small changes, which you'll likely pick up as you start to work through the numbers in more detail. These are generally minor. And the only item I wanted to briefly touch on is the fact that the quota share ceding commissions will now be recorded within the reinsurance expense line rather than the commission expense line. For those businesses which use material quota share reinsurance, such as Crop, while this won't impact profitability, it will impact the composition of the combined ratio. Now to the balance sheet on Slide 21. AASB 17 somewhat simplified the presentation of assets and liabilities related to underwriting activities into insurance and reinsurance contract assets and liabilities. Our restated balance sheet is available in the data book we released this morning. And that really concludes this section. And before we move on to Q&A, I just want to reiterate a few of our key messages. So we see no impact to our fundamentals from the change in AASB 17. It will not impact our strategy, our profitability, capital, gearing or dividends. While there are a series of changes in presentation, hopefully, you followed our approach to management reporting and appreciate some of the familiarity we're trying to preserve here. Finally, to Slide 24 to close out with outlook. We restated our outlook for the new standard. There is no change to GWP -- to the GWP outlook we provided on May 12. For our combined ratio, you can see that the new standard will not impact the planned combined ratio of 94.5%. And our outlook, of course, will be based on the net combined ratio per our management results. This continues to exclude the upfront impact of the $1.9 billion reserve transaction recently announced. I will conclude here. I've got Ash Dalziell, our Head of Investor Relations, with me and we're happy to take any questions. Over to you, operator.

[Operator Instructions] And our first question comes from the line of Kieren Chidgey with Jarden.

Thanks for all the data and consistency in disclosure metrics, I think, with previous disclosures, a significant positive. Still sort of a lot to get our head around. I just wanted to start on onerous contracts, a couple of questions around that. Firstly, just interested if you could sort of give us a little bit more color on what drove that in FY '22? And then if we are to think about that as sort of just a timing delta in recognition of earnings, should essentially the pull forward of those losses not be generating a positive benefit into FY '23 that doesn't seem reflected in new combined ratio?

Yes. Look, good question. Just -- so in terms of what drove the charge, as I referenced in the remarks, Kieren, the -- there are a few areas within North America. We've got a -- some of these run-off programs. And as we look at the unearned premium component of that, that's unprofitable, and hence, that's really pulling forward some of the losses relating to those contracts in runoffs. The middle market business, which we've sort of said for some period of time, has been a focus for us in terms of growing that book. We're still not making underwriting profitability in that book. It's a combination of some of the loss activity we've seen through the second half and also that business is still building towards operational scale. So that should -- that's not surprising as a feature in that onerous contract charge. And then here in Australia, it's really the personal lines, home and motor books. As you know, we've got a relatively small presence in those books. And as we've seen a significant uptick in inflation through the course of last year, the written rate has just lagged that. And so then when we look at the unearned component of that at the end of last year, we felt it made sense to take an onerous contract charge for that. Clearly, we're pushing really hard to make sure these books either run off more profitably or we continue to put more rate through the book to get them into a better place. So in essence, what will happen is as we go through the course of this year, we'll be assessing, at both reporting periods, that onerous contract provision so to the extent the profitability improves, that should be reflected in the provision as we go forward. But clearly, there's a risk that other portfolios may emerge on initial assessment that could add to that. So -- look, I think our objective here is to -- and you'll see this from the theme of the entire presentation today. We are trying to be sensible about not adding to the number of moving parts in our results, and we obviously spend a lot of time thinking through the initial provision that we establish around onerous contracts. And we want to make sure that things that move in the onerous contract provision should be things we should be talking about from a performance book point of view, right? So these are all the cells that -- where profitability is either challenged or we've taken some actions to address that. And so therefore, there should be a level of consistency between what you hear from us on performance and what you see come through on the onerous contract provision. And the cell review framework that we use to run the business gives us a very good opportunity to make sure that that's tightly managed.

Okay. But is it fair to say that roughly $100 million of losses you thought might occur this year have been pulled back to last year, so we should have a stronger profit basis for this year as a result?

I mean, in essence, what you're looking at is the unearned element of that profit at the end of last year, and we're sort of impairing that in essence, right? And so what will happen is it will depend on how that provision gets adjusted, Kieren, because that might be the case, but other things that might be getting not enough rate, et cetera, as we exit this year into next year might fall into that. So we'll be very open with what the moving parts around that provision so you can see that. I think it will be important to look at that year-on-year rather than just trying to look at the $106 million as unwinding into 2023.

Yes. And secondly, just two other questions. The risk margin going back to your comments there that sort of December '22, you're at the top end of your 6% to 8% range. Should we be thinking about 7% as where the business will head at some point? Interested in any thoughts around potential timing.

Yes. Look, I think the -- what we're really trying to do is as we think about the risk appetite around reserving and the prudence we want to have around that risk margin, we've thought about the 90th percentile as being the target around the confidence level or the probability of adequacy in the past that corroborates to the 8%. Now mathematically, as you look at the capital under the new standard, that is attributable to reserve risk and be imputed cost of capital. Against that, you could support mathematically a number of 7%. We just thought it was sensible, again, to limit the number of moving parts from transition to keep that at the top end of the range. And also just given the outlook for inflation at the moment, you think about the sort of risks going forward and we think it makes sense to be at that 8%. But mathematically, sure, you could support something that's lower in the range, hence, the rationale for having the range.

Okay. And just one final question on NEP. As we think about it for '23, I appreciate you cannot talk specific numbers, but conceptually, with particularly, I guess, the revenue recognition change around LMI, all else equal, is there much of a change on a go-forward basis? Or is it a pretty mild negative under the new accounting?

Yes. It's Ash here, Kieren. I wouldn't get too caught up on the LMI earnings pattern in terms of the medium-term outlook for net insurance revenue. Probably the only thing to be aware of in the restated management '22 accounts is that the net insurance revenue does include the upfront impact of the E&S transaction, which was circa $70 million. That was previously adjusted out of the 1023 management results due to the change in accounting treatment of these LPTs, which Inder touched on, we don't see the need to adjust that upfront impact anymore.

Okay. So very little change, all else equal, from an NEP point of view?

Look, I mean the only other thing to be aware of, Kieren, is with the quota share cede commissions now in reinsurance income and NEP effectively, to the extent that we were to dial up or dial down one of the quota shares on, for instance, Crop or LMI, you could potentially have some influence on net insurance revenue line item, but we're quite clear about the size and the proportion of those quota shares going into '23. So it definitely won't be a feature for this year.

Okay. All right. Okay. But just to clarify the...

[Operator Instructions] Your next question comes from the line of Siddharth Parameswaran with JPMorgan.

Just -- my first question actually just relates to the discretion you have around risk margin, that range of 6% to 8%. Is it mathematically calculated? Or -- I mean to end up at the bottom end of that range, is there a basis on which you will make a decision to be at the bottom end versus the top end, Inder?

Look, I think -- yes -- I mean most of it is mathematically derived, right? And the key inputs really relate to the cost of capital and the inputs around the cost of capital, ultimately, the profile of reserves, right, the profile reserve changes, the mix of reserve changes, the duration changes and then really with the reference to external transactions such as LPTs, et cetera, the traded value at which those reserves might trade and what that imputes as cost of capital. So there are a few moving parts. And clearly, there's always going to be an element of actuarial judgment in these. But it's fair to say there's probably less discretion and the way you can cross check that is -- our references to the fact that things like COVID, for example, where we put up that additional risk margin, it'd probably be a slightly different set of things for us to consider around how that might play into a risk adjustment change going forward, right? So I think it's probably more tightly defined. But as always with these things, there's an element of actuarial judgment around the risk profile, around the duration, around the mix, around the cost of capital and you'd be pretty familiar with some of those moving parts.

Yes. But just to be clear, the 6% to the 8%. So you've given us a range there. So the 6%, what does that reflect versus the 8%?

So if you think about the -- the other way to calibrate that is to look at the confidence level. So the confidence level, which is akin to the old probability of adequacy is at the 90th at 8%. And it's probably kind of as it's constructed at the end of 2022, that 6% is probably closer to the 84th or something like that, right? And therefore, that just gives you a bit of a sense on what that 6% to 8% looks at. And obviously, you could impute that into a range around the cost of capital as well.

Yes. Okay. But just to be clear, like if you wanted to run, so it's -- the 6% reflects something that you think that you can change to a 9% in a reasonable way. That's the discussion you have. That's -- there is discussion going from 8% to 6%.

Yes. Over -- say, over the medium term, right, we shouldn't really be seeing a lot of change in that risk adjustment, right? Again, the plan here is not to have lots of moving parts. Mathematically, you could easily support 7%, which is the midpoint of that range. We're just saying over time, this is where we see that risk adjustment sort of operating at within this target range with a view to not having material changes. And every time we change it, we will obviously explain it to you very clearly as to what's changed.

Yes. Okay. No, that's clear. Okay. And similarly, the illiquidity premium, is there any discretion around that?

Yes. I mean this is very clear principles that you need to apply to calculate that. Now the inputs to that could change over time. Again, we're not trying to set this up Sid so that the 32 basis points will change it to 28 to 27 to 40. We don't want to create unnecessary point. What we've done is we've looked at -- it's very difficult to impute what we think is an illiquidity premium that reflects the illiquidity of a book of reserves, right? I mean it's not an easy thing to be able to calculate. We've looked at illiquidity premium imputed in credit spreads over a very long period of time, and that's given us some proxy to that. So yes, there's some judgment, but we're not intending to turn off with that moving around all the time, right? So the 32 basis points should be a number we stick to over a medium term and then really need to explain to you why that's moved. There's been a fundamental reset in the market or we've seen a big movement in corporate credit spreads, et cetera. We'll explain that to you. But the intent is to try and keep some of these moving parts as stable as possible given the -- whilst applying the principles as a standard.

Yes. No, that's quite clear. It looks like basically, there's -- it seems like there's a bit less discretion than before certainly around the risk margin and not that much discussion on the illiquidity premium line. Just on the onerous contracts, again, just following on from what Kieren asked. Just -- I wasn't quite clear on the chart that you were showing on Page 9, where you're explaining the different splits at which the level -- the calculations will be done. It seems like according to that chart, you're effectively moving from 3 buckets to 9 to do the assessment. But I think the way you described the explanation, it seemed like you've done a lot more. I mean it was much more granular than that. So I'm just wondering -- I think you did give some numbers, but I just wasn't quite sure how many portfolios you're actually doing this at. Maybe you could just give us an idea around that. And also, just relating to this, it seemed like, correct me if I'm wrong, but did you say that the number went from about $20 million to $106 million? So it was actually lower in the previous year? I was a bit surprised at the starting balance -- I mean it seems like we've deteriorated over the year despite very, very strong rate increases and what I thought would have been a better outlook.

Yes, that's a fair question. I mean the chart on Slide 9, it is illustrative in nature. As you know, we've had a very well-established rhythm around cell reviews. We have circa 70 cells across the company, which then provide the -- will determine under the standard facts and circumstances against which we assess performance and the need for making any charges around onerous contracts. So it's a bit nuanced and I don't want to get into the weeds of it all. But in essence, you look at each of those cells at that level of grain, right, versus what you used to look at a regional level. However, as you look at -- so that's the initial recognition point. And then the subsequent measurement point is more done on a portfolio level, right? And so -- but in the 70 cells we might aggregate, as I was given an example in Australia, to a few set of portfolios, right? So the 70 might be, call it, 18 or 20 or something like that. And so in essence, you've got two different bases on initial recognition and subsequent measurements in the way you assess the onerous contracts piece. What that tends to probably mean is that once you've established a provision, hopefully, there's less volatility because you're then aggregating across portfolios and assessing subsequent measurement against that. In terms of why it's moved from $30 million to $106 million, so that was at the end of 2021 to 2022, the key areas. Whilst we're getting a ton of rate, in some books, inflation really spiked up in the second half. And I referenced the personal home and motor books here. And also in North America, where we made that assessment to put a number of these programs into runoff because we felt inherently the profitability even despite 20, 30, 40, whatever points of rate, we didn't see a sensible path to profitability. So at the end of the year, it results in an onerous contract charge because we made those decisions during the course of the year to exit those programs. And similarly, with middle market, whilst we're getting a lot of rate, we -- and whilst we're confident about the execution on that, it's still not screening as profitable when you look at the unearned premium outstanding at the end of 2022, which is obviously a higher number than 2021 because we've been growing that book.

Yes. Okay. And just very quickly, just APRA data, what are we getting for the first quarter -- we saw the first quarter data, but what did we actually get? Was that on the old basis?

Yes. Yes. So I think that switches second half, Sid, [ month of July ].

[Operator Instructions] And our next question comes from the line of Nigel Pittaway with Citi.

Just first of all, just a clarification. You're effectively going to provide two combined ratios, if I got that right, one management, one statutory with the main difference just the head office expense allocation. Is that correct?

No. We're just providing one combined ratio.

Right. Okay. So the statutory combined ratio. So you won't actually give it. So will that be management or statutory?

That's management. I mean there's no such concept as a combined ratio under the statutory view of AASB 17.

Right. Okay. So that will still be based on -- so what will be the denominator for that? That will still be -- will that be -- yes. So it's still basically...

You take your underwriting results, right, divide it by your net insurance revenue, right? So it's the same basis of looking at underwriting profit over a net insurance revenue and premium property. And then we're seeing all the ratios that are done on a consistent basis, right? So expenses, ex-cat claims, et cetera. So these are all, in essence, management views.