Hi, everyone. Welcome to the Admiral IFRS 17 briefing session. Good to see so many of you in the room face-to-face after such a long time of Zoom calls and VCs over the last few years. Really pleased to have such a good turnout. I'd like to start off by apologizing on behalf of our CFO, Geraint Jones. He is on the agenda to present, but unfortunately, he's not here today. He is off ill and sends his apologies to everyone. I think I know most of you, but for those of you who don't know me, I'm Marisja Kocznur. I look after investor relations for the group, and I'd like to introduce you to our two presenters for today. We've got Rachel Lewis, who is our Director of Group Finance and Chief Actuary, and we've got Katie Baggott, our IFRS 17 Reporting Manager. It is Katie who's been driving the implementation of the standard for us over the last few years. I'd like to take this opportunity to thank her and her team on behalf of Geraint and our team for their hard work and doing a really fine job so far. In terms of today's presentation, we aren't going to cover general IFRS 17 training. I think there's plenty of content available on that, we'll give you an overview really of the impact on Admiral. In general, you know, we are probably not as complex an insurance business as many others, the impact might not be as significant. There certainly are some changes, primarily presentational changes and some potential changes in the timing of profit recognition, which we'll touch on. In terms of the outline, we'll take you through our accounting choices. We'll have a talk through the group equity position and the impact on transition at transition. We'll take you through those presentational changes, as I mentioned, the potential profit recognition timing impacts, and we'll also spend some time on the reinsurance accounting. As you probably expect, but I'll point that out, and hopefully it's understandable, we haven't given profit guidance or detailed numbers, but we have given some illustrative examples, and hopefully that gives you an idea of the impacts from an Admiral perspective. The presentation will be about 35 minutes, and then we'll open up to Q&A. Without further ado, I've got the outline of the presentation, and you should all have one in front of you. Without further ado, I will hand over to Rachel to take you through it in more detail. Thank you, Marisja. Good morning, all. For those that I haven't met before, I'm Rachel Lewis, and I'm Director of Group Finance and Chief Actuary at Admiral. I've been with the group for 16 years, having joined from KPMG back in 2006. Today, I'd like to take a brief moment to recap on the couple of fundamentals on IFRS 17 before moving to our key accounting policy choices that we've made and then highlighting the main messages as to the overall impact of the standard for Admiral. We are quickly approaching the January 1st, 2023 IFRS 17 implementation date. At Admiral, we've been working towards this for a number of years. We are well prepared, on track, and expecting a smooth transition. The new standard impacts all of our insurance businesses across the group. Our presentation today focuses on our largest business, U.K. Motor insurance. We'll give some work examples and so on. It can be assumed that any impacts for our other lines of business, both in the U.K. and internationally, will be largely immaterial. Our non-insurance businesses, such as Admiral Money, are unaffected. Our 2022 year-end results in early March next year will include disclosures on the expected impact on transition as at January 1st, 2022, which is the opening position for the first comparative period. In our presentation today, we'll give some high-level directional insight on this. The first reporting under IFRS 17 will be our 2023 interim results, and that in the announcement in August next year. At that point, we'll present IFRS 17 income statements and balance sheets for 2022 and for the first half of 2023. One final point here is that a number of insurers will talk about the January 1st, implementation of both IFRS 17 and IFRS 9, the accounting standard for financial assets and liabilities. Admiral has already adopted IFRS 9, but as we'll highlight today, certain IFRS 17 accounting policy choices are consistent with those previously made under IFRS 9, and those choices limit the potential for mismatches and volatility. I want to give you a quick reminder of the principles of IFRS 17 that are most relevant to general insurers. The aim of the standard is transparency and comparability through the following, which I'll talk through in turn. Firstly, the measurement of market consistent insurance liabilities and assets, which are based on a discounted best estimate cash flows plus a risk adjustment and a contractual service margin that reflects deferred profits on earned business. Many general insurers, including Admiral, will use a premium allocation approach simplification for earned business. IFRS 17 results in an increased level of granularity in the accounting, meaning the groups of loss-making contracts can no longer be offset against those of other profitable business. We will see separate accounting treatments and presentation of gross direct insurance business and ceded reinsurance, which in some instances could lead to a mismatch in the recognition patterns of gross profits and of reinsurance costs and recoveries. The standard gives accounting options for the treatment of acquisition costs, which result in either the acceleration of cost recognition or the deferral of costs over a longer period. There are also options available in respect of discounting. The calculation of the discount rate and also the allocation of income and expenses arising from changes in the discount rate. The presentation of the income statement and balance sheet changes significantly under IFRS 17, with the introduction of new disclosures as well. More on that a little bit later. Here we set out Admiral's key accounting policy choices, which have been made with the aim of reducing both complexity and volatility in the accounting. As I just mentioned, we intend to use the simplified premium allocation approach for accounting for all contracts, both gross direct insurance business and ceded reinsurance. This means that the accounting for unearned business is not too dissimilar from the unearned premium reserve approach under IFRS 4. When discounting cash flows, we'll use a bottom-up approach for calculating discount rates, starting with a risk-free rate and then adding a premium for illiquidity that's appropriate for the liabilities that we're holding. Once a discount rate has been set, we refer to this as the locked-in rate. Subsequent changes of that locked-in rate because of movements in the risk-free curve, for example, will be recognized in other comprehensive income rather than the income statement. This is consistent with our IFRS 9 treatment, where the movements in the value of most of our financial assets are also reflected in other comprehensive income. Risk adjustment. We intend to maintain a conservative approach to reserving in line with the group's risk appetite. We expect our policy on risk adjustment to set a corridor around the 90th percentile confidence level. On transition, we expect to be at the higher end of that corridor, but we'd expect to move towards the middle of that corridor over time, should our view of reserve uncertainty allow. The remaining decision to highlight is the intention to expense acquisition costs. This, again, is a conservative approach which is transparent and less complex than alternative options. What impact does this, the standard and these accounting decisions result in for Admiral? Our main messages are set out here. IFRS 17 does not change the ultimate profitability of the business that we write or the reinsurance that we place. It is a matter of timing and presentation. Similarly, there are no significant impacts on group strategy, on dividend policy, cash generation or on solvency. Katie will take us through the high level impacts on transition shortly. As part of that, she'll show a reduction in equity on transition. We don't expect this to have an impact on dividend policy at a point of transition or in subsequent periods. I've mentioned already. It's worth repeating. In line with our risk appetite, we'll continue with our conservative approach to reserving. The two final points we'll also explore further shortly. There will naturally be some change to profit recognition patterns. We'll give some illustrative examples. We'll also talk to the nature of enhanced disclosures that we'll be making in a set of IFRS 17 financial statements, which will provide more granular information on which our business may be modeled. I'll now hand over to Katie to take us through the transition impact and to give some insights on profit recognition patterns. Thanks, Rachel. Good morning, everyone. As a quick introduction, my name is Katie Baggott. I've been at Admiral for just over five years now, leading the group IFRS reporting team. For the last year to two, I've been focused on leading our IFRS 17 implementation. In the next section, I'm going to talk about what the accounting changes mean for Admiral on transition. As a quick reminder, when we talk about transition, we mean an opening balance sheet as at the first of January 2022. That's because it will enable us to produce our comparatives for the 2022 income statement. The transition balance sheet that we produce is calculated using a fully retrospective approach. That means we calculate that opening balance sheet as though IFRS 17 had always been in place. Finally, the impacts that we're presenting here are illustrative, in draft and unaudited, they are subject to change. As Rachel mentioned, at transition, we're seeing a reduction in equity. Before I explain why that happens, let me repeat that despite that reduction in equity, we don't expect it to impact our dividend capacity. That's because our dividend capacity is constrained by our solvency ratio, and we don't expect our solvency ratio to change when we move to IFRS 17. The main drivers for the reduction in equity are set out in this waterfall. You can see that they're comprised of, firstly, a favorable movement related to our claims reserve revaluation under IFRS 17. Secondly, an adverse impact in relation to our quota share reinsurance assets. Finally, a number of other impacts that overall are adverse, but are driven primarily by that decision to expense our acquisition costs and therefore to write off our deferred acquisition cost asset as a transition. I'll take you through more detail on that over the next few slides. This slide sets out a comparison of the insurance balance sheet items under IFRS 17 and IFRS 4. You can see from this that there are lots of similarities, but there are also some key differences. It's those differences that are driving the transition impact that we'll see over the next few slides. That first impact, the change in claims reserves as that transition, which is a favorable impact on equity, is driven by a number of factors. Firstly, our best estimate under IFRS 17 becomes a probability-weighted best estimate. That means that we include an allowance for Events Not In Data, which is a similar concept to Solvency II. Secondly, we have a change in our approach to discounting under IFRS 17. The standard requires that we discount all claims, and also that we apply a different discount rate, which is derived by that risk-free curve, as Rachel mentioned, similar to Solvency II, plus an illiquidity premium. It's worth noting that the combination of that probability-weighted best estimate and the discounting gives us our ultimate claims cost under IFRS 17 and our ultimate loss ratios. Finally, IFRS 17 requires the addition of a specific risk adjustment. There's no prescribed methodology or range for calculating that risk adjustment. As Rachel's already mentioned, we intend to maintain a conservative approach to reserving, and as a result, to set our risk adjustment in a corridor around the 90th percentile. At transition, we expect to be towards the upper end of that corridor. Even at that higher level, you can see that is lower than the booked management margin as at the end of December 2021. Overall, those factors therefore lead to that reduction in claims reserves as at transition and the favorable impact on equity in relation to claims. The next impact that I mentioned is a reduction in our quota share reinsurance assets. There are a couple of main factors that are driving that, which you can see in this waterfall. The first one of those factors is the most significant, it's due to a changing in the phasing of when we recognize our expense recoveries from those quota share reinsurance contracts. In our current accounting, which we'll go to in a little bit more detail later, we recognize the majority of our ex-expenses, i.e. all of our non-acquisition expenses, up front in the first underwriting year in which they're incurred. That means that under IFRS 4, we also recognize our related quota share expense recoveries at the same time, so that the timing of the gross incurred expenses and our expense recoveries from quota share contracts is aligned. When we move to IFRS 17, any expense recoveries that we get from our quota share reinsurance contracts are required to be netted off the ceded premium and earned over that contract in that underwriting year. That means that a particular point in time, like the transition date, will effectively have earned less of those expenses than we have under IFRS 4, and that's what's driving this first adverse impact. That does mean under IFRS 17, that there is a slight misalignment between our gross expenses, which we'll recognize when they're incurred, and the quota share recovery of those which we'll recognize over the full underwriting year. The other impact that you see relates to changes in how we recognize our claims recoveries under IFRS 17, and in particular, they relate to that reduction in gross claims reserves that you've already seen, but also the fact that we'll also discount our quota share reinsurance claims recoveries as well. As I mentioned, the biggest other driver of the adverse movement in equity at transition is the write-off of those gross deferred acquisition cost assets as at transition. There are a couple of other items that I'd just like to quickly mention in relation to these impacts. The first one of those is that you'll see a small onerous loss component net of reinsurance recoveries as at transition. You might be a little bit surprised to hear a profitable business like Admiral talking about onerous losses. The reason for that is really driven by our conservative approach to reserving. The loss component under IFRS 17 is actually fairly similar to the unexpired risk provision under IFRS 4. There are a couple of key differences. The first of those is that you calculate the loss component at a higher level of granularity. For example, splitting into new business and renewal contracts. The second of those, most importantly for us, is that you also include the risk adjustment in the calculation of the onerous loss component, and it's that that's driving that recognition of an onerous loss component as that transition. The other item that I'd just like to cover briefly is the impact on coinsurance profit commission. We anticipate that under IFRS 17, coinsurer profit commission will continue to align to the development of our loss ratios on a gross net of excess of loss reinsurance basis. Given that we're seeing a reduction in those claims reserves as that transition, we'd expect to see a small increase in our coinsurance profit commission as that transition. To sum up, we're seeing a reduction in equity as that transition, but that won't impact our dividend capacity or policy. The next section that I'm going to talk to you about is the impact of IFRS 17 on post-transition profits. As we said, there's no change in the economic profitability of the business that we write or the reinsurance that we place as a result of implementing IFRS 17. However, because of the changes that IFRS 17 brings about, so the acquisition costs, discounting, risk adjustment, and the changes to those quota share reinsurance accounting, there are changes in the phasing of the earnings profile between the current accounting and IFRS 17. We've used a number of illustrative examples in the next section, primarily focused on U.K. Motor, to show how and why the timing of that profit recognition changes. We've split those impacts between that gross net of excess of loss reinsurance part of the business and then the quota share contracts. I would note the numbers that we've used here are purely illustrative and directional, and they shouldn't be used for modeling purposes. You'll see in the packs that in the appendices, we've also provided some further details for you to have to understand these examples. Firstly, I'm just going to quickly talk through this example because it's used as the starting point for the later examples where we go into the phasing of the profits. It's based on gross written premium of GBP 2,000, a 20% expense ratio, and a 65% ultimate loss ratio. For simplicity, we've assumed that all of the claims are settled after five years and that there's no excess of loss reinsurance in place. This slide sets out the profitability of that contract after five years. Firstly, under IFRS 4, you can see that the profit that we've recognized on the gross basis is GBP 300, which aligns to the combined ratio of 85% on the written premium. On the quota share side, we've assumed a 60% quota share reinsurance contract on a funds withheld basis with a 2% reinsurer margin on the premium that's ceded. After five years, you can see that the ultimate cost of that reinsurance contract is the 2% of the reinsurance premium of GBP 1,200. Under IFRS 17, you can see that the overall result after the five years is exactly the same as IFRS 4, i.e., there's no change in the ultimate profitability of that business. The reinsurance cost is the same, as well as the gross cost, gross profit on that contract. There are presentational changes. As you can see, the claims cost that's presented is the discounted claims cost, with the unwind of that discounting being presented as a finance expense that will go through in the net investment results. We know that the ultimate profitability is the same under IFRS 4 and IFRS 17, but there is a change in the phasing of how we recognize those earnings. This slide looks at the gross side of the results. Firstly, as a recap under IFRS 4, you can see that in year one, we recognize a loss on an individual underwriting year basis. That's because whilst we earn our premiums and claims, we recognize the majority of our expenses up front in that first year. In year two, you see a significant profit. That's again because we continue to earn our premiums and our claims. There are minimal expenses being recognized in that second year. The profits in years three to five, are reflective of the claims reserves that are being released and the development of our loss ratios from the booked loss ratios down to the ultimate loss ratios. Under IFRS 17, the phasing is quite similar. You can see that there are some differences. To start with, in year one, we see a higher loss. That's primarily driven by that decision to expense all of the acquisition costs in year one. There is a little bit of offset because we also have our claims costs that are discounted under IFRS 17. In year two, we see a higher profit than under IFRS 4, and that's again because of the expensing of the acquisition cost that's already taken place in year one and also again the discounting of the claims cost. In years three to five, you continue to see the profits being generated by reserve releases, but they are offset to an extent by the unwind of the discounting benefit on the claims that have been incurred, so they're slightly lower than are seen under IFRS 4. When you combine the underwriting years, if a business is growing rapidly, you can see that the choice to expense acquisition costs will result in the profits of that business being slightly delayed compared to the current profile under IFRS 4 because of that decision to expense acquisition costs. When the acquisition costs are stable, the difference in profits recognized is dependent on the balance between the discounting and the risk adjustment that's applied in years one and two when those claims are incurred to the release of the risk adjustment and the unwind of the discounting as the prior years develop. The next couple of slides focus on the related quota share reinsurance costs. I will warn you, these are quite technical, please bear with me. Bit of a recap under IFRS 4. The first example that you can see here is linked to the gross example that you just saw on the previous slide. You can see that we've got that loss in year one offset by a profit in year two on the gross side. We recognize that loss in year one, we recognize under IFRS 4 a corresponding gain on our quota share reinsurance contract in year one of 60% of that loss. In year two, we recognize an offsetting cost on the quota share reinsurance contract to take us back to the ultimate cost of that contract being the reinsurance margin. It's worth noting that as soon as that year is cumulatively profitable on an earned basis, then the full cost of that margin is recognized. In this case, that's over the first two years. In this example, you can see that the underwriting year isn't cumulatively profitable on a gross earned and booked basis until after 36 months. Here, we again recognize our reinsurance recoveries as being the proportional share of the gross results, with the total cost of the contract being exactly the same by the end of year three. Just going into that in a little bit more detail, this is the simplified IFRS 4 income statement after 24 months for that example on the right-hand side, example two. You can see that in this example, after the 24 months, there's a gain on the reinsurance contract of GBP 24. At the start of year three, we may choose to commute that contract, even though at that point the gross position isn't profitable on a booked basis. That's because whilst the gross position isn't profitable at 24 months, we expect it to be ultimately profitable. When a contract in these terms under a funds held arrangement is ultimately profitable, we don't expect to make any further recoveries on that reinsurance contract. If we don't commute the contract, there's an additional cost to that reinsurance contract. In those circumstances where we choose to commute whilst the gross position is still loss-making, we would realize a loss on commutation. That loss on commutation is equal to the difference between the gain that's been recognized by the end of year two of GBP 24 and the ultimate cost of the contract of GBP 24, i.e., a loss of GBP 48 on commutation. So in IFRS 4, we typically recognize a large gain in year one and then an offsetting cost in year two and possibly year three until we get back to the ultimate cost of the contract. Under IFRS 17, the phasing of that ultimate cost is quite different. That's because under IFRS 17, we consider the recoveries that we expect to make on a written underwriting year basis at each balance sheet date. You can see what that looks like here, for example one. We've got a funds withheld contract and at the end of year one, we expect gross business is profitable on a written risk-adjusted basis. Therefore, we would simply recognize the cost at the margin, the earned margin in year one. Similarly again, that's the cost that we would recognize in year two. In the second example, at the end of year one, we would expect to make some recoveries on a written risk-adjusted basis because the combined ratio is well over 100%. That means at the end of year one, we'd recognize recovery. By the end of year two, we don't expect a recovery under IFRS 17 on that risk-adjusted basis. That's because of a few factors, but primarily the fact that the quota share recoveries are discounted. You can see the comparative income statement under IFRS 17 at the end of year two for the same example. You can see that in this example, the full cost of that reinsurance contract has been recognized under IFRS 17. That means that typically we wouldn't expect to recognize a loss on commutation under IFRS 17. Overall, the main things that I would highlight from here is that ultimately, as you can see, the cost of that contract is the same under IFRS 4 and IFRS 17. The timing is clearly different, whilst there's a big change in the accounting under IFRS 17, in some ways it could be considered as simpler. If a year is forecast to be profitable on a written risk-adjusted basis at the end of the first year, the cost is simply the cost of the earned margin in that period. If it's not forecast to be profitable at that point in time, we'd expect a small recovery. Typically, by the end of year two or by the point that we do commute, we'd expect that full cost of the reinsurance contract to be recognized so that typically under IFRS 17, we wouldn't expect to see a loss on commutation. When the underwriting years perform consistently under IFRS 4 and IFRS 17, as well as both the underwriting year result being the same and being the same ultimate cost, the result in each financial year would also be the same. This example shows what happens when those underwriting years don't perform consistently. In here, we've moved between our illustrative example one, which is more profitable, to illustrative example two, less profitable, and then back again. What you can see from the line at the bottom is that there is a difference in the phasing of those reinsurance costs in that circumstance. Again, the ultimate cost of the contract is the same at GBP 24 in each underwriting year, but the timing of that cost in each financial year is quite different. In years two and three, we recognize an increased cost under IFRS 17 compared to a much reduced cost in this example and in year four. Overall, in terms of post-transition profits, there's no change in the economic profitability of the business written or at the cost of our reinsurance contracts. You can see that the phasing will be slightly different due to those impacts of acquisition costs, discounting, risk adjustment, and also the changes in the quota share reinsurance recoveries. I'm now going to hand back to Rachel to talk through presentation and reporting. Thanks, Katie. As Katie just said, let's take a look at the presentation and reporting impacts of IFRS 17. The IFRS 17 income statement will look significantly different to the IFRS 4 format that we're used to. There are examples of each on here, IFRS 4 on the left and IFRS 17 on the right. I'll pull out a few high-level points. We'll focus on a couple of the lines in a little more detail. Under IFRS 17, we'll report an insurance service result, which is effectively the net underwriting result. Within this, insurance revenue and expenses, including claims, are presented on a gross basis. The reinsurance result is presented separately in one line rather than the typically proportional allocation to individual premium claims and expenses lines that we're used to under IFRS 4. Katie has already shown this earlier with the illustrative example. Insurance revenue will include some elements of revenue previously presented as other revenue under IFRS 4. These elements are primarily revenue generated from the portion of our insurance business that is underwritten by group entities. We'll take a closer look at that in a moment. The costs that relate to this revenue are included in insurance service expenses. If you move down the income statement, the net investment results will include the impact of the unwind of the locked-in discount rates applied to insurance assets and liabilities. It will no longer include the impact of any changes in allocation of investment income to quota share reinsurers. We'll also take a closer look at this in a moment. Towards the bottom of the income statement, other income and expenses will include revenue and costs resulting from the portion of insurance business underwritten by external insurers. It'll include profit commission from coinsurance arrangements, the remaining elements of other revenue and expenses that relate to that external coinsured portion of the business, as well as revenue from ancillary products that are underwritten outside the group. Let's take a look at the top line in a bit more detail, insurance revenue. This waterfall analysis starts at the IFRS 4 net in premium and then steps across to IFRS 17 insurance revenue. The first step takes us from net to gross. This revenue will be a gross measure under IFRS 17. As I mentioned, it will then include a couple of additional elements. The first of those is gross premium earned from ancillary products that are underwritten by group entities. The second is fees generated by the portion of the insurance business that are underwritten by group entities. As I mentioned a moment ago, the costs in relation to the additional elements of revenue will also be included in the insurance service result as part of the insurance service expenses. Fees and other revenue generated by the portion of the insurance business is underwritten by external co-insurers, as well as any commissions from ancillary products not underwritten by the group are captured, as I've said, within other income and expenses. This comparison of investment income under IFRS 4 and IFRS 17 shows a couple of differences in the components of the net investment result. Investment income under IFRS 4 included the change in allocation of investment income to U.K. Motor quota share reinsurers. This either increased or decreased investment income depending on the development of individual underwriting years in any particular financial year. This isn't captured in this line under IFRS 17. That will all be within the reinsurance results within the wider insurance service results. Under IFRS 17, we do, however, see the inclusion of the unwind of the locked-in discount rate with a net investment return. We expect this to be a reasonably significant reduction in the overall investment income. Given that we're well-matched in terms of asset and liabilities, we'll use our asset portfolio to determine the illiquidity premium. We must also remember that the original benefit of that discounting is reflected in the insurance service result. Here we give a brief overview of the new or enhanced disclosures that will be made under IFRS 17. I won't go through them all in detail, but the key ones to call out here are naturally the risk adjustment disclosures. IFRS 17 requires the disclosure of the confidence level of the risk adjustment, as well as qualitative and quantitative information on the changes in risk adjustment in the period. I'd also highlight the best estimate or ultimate reserves where we should see more granular detail than we've disclosed previously. We'll give details of discount rate curves at different time intervals for U.K. and international businesses separately, and maturity analysis of undiscounted claims reserves, and finally, sensitivities to changes in best estimate loss ratios, risk adjustments and discount rates. What does all of this mean for key performance indicators? We're not expecting wholesale changes, there will be some subtle differences to the underlying calculation of certain metrics. Group KPIs, Earnings Per Share, Return on Equity, Solvency Ratio, they all remain relevant under IFRS 17, and we don't expect significant changes in any of these. We'll continue to report revenue metrics and Turnover, which has long been a measure of the size of our businesses will remain as currently defined. Profit metrics will remain largely consistent, but there will be some changes to the calculation of loss, expense and therefore combined ratios, which I'll give some more information on in a moment. To incorporate the results of proportional reinsurance, we'll introduce a metric that reports the overall margin from underwritten insurance and reinsurance business. This final slide sets out a brief illustrative example of how the calculation of loss and expense ratios will change. Katie has already taken us through some examples to show how the phasing of gross insurance and reinsurance revenue and expenses will change. For this example, we put that to one side. For simplicity, we assume that there's no change to this phasing under IFRS 17 so that we can focus on the calculation of the metrics. The first point to highlight is that whilst under IFRS 4, the reported loss expense and combined ratios were reported on a net of all reinsurance basis. This will change under IFRS 17, and we'll exclude the results of the proportional reinsurance from that calculation. The second point to make is that the denominator for this calculation will be gross premiums earned. For example, U.K. Motor premium earned. We'll strip out those additional elements of revenue that we referenced earlier within insurance revenue, and we'll add them back in to the other revenue per vehicle metric, which we've reported historically and which therefore won't change under IFRS 17. Finally, you'll see on the right-hand side at the bottom that Admiral and indeed other insurers will introduce a new metric, insurance service margin. This will bring in the net proportional reinsurance result, with the denominator continuing to be gross earned premium. We don't have a comparable metric under IFRS 4. That's it for the main content of this presentation. To wrap up, let me repeat some of the key points that we hope you take away today. Admiral is a reasonably simple personal lines insurer, so thankfully, it was definitely right at the simpler end of the spectrum when it comes to impacts. IFRS 17 doesn't change our ultimate profitability, cash flow patterns, solvency position or dividend paying ability. We'll continue to set reserves conservatively and expect to continue to recognize profits over time. There are some changes to presentation and potentially to timing of recognizing profits and reinsurance costs. I'll leave it there. Thanks very much for your time, and I'll hand back to Marisja to open up for questions. Good. Yeah, it sounds that way. I think we're happy to open up to questions. We've got Carl and Laura who are going to bring the mic around to you. If you can please wait for the mic before asking a question. If we can try and limit it just per person to start off with because there's a lot. I realize you had sort of one to two questions. If you can start off also by stating your name, please, so that. Yeah, so you know. Go for it, Greg. Morning, everyone. Greg Paterson, KBW. you know, currently, and I think a lot of analysts, they have models where they're taking the loss ratios that you've done, and you usually put in your appendices, adjust them, and then push them through, call it black boxes, to produce quota share and profit commission, right? Those are complex models when you understand what's going on underneath and how and very important for understanding because if you actually look at the profit of Admiral, it's not very correlated to what's actually going on in the profitability. There's a big buffer there. Now, this complicates it endlessly, to the degree where I think all those models will have to be redone. My question is, will you be producing historical loss ratios/combined ratios under the new regime so we can push it through some models to produce, once we've figured out what are the mechanics of the reinsurance work, would you be restating all those triangles for us so that we can actually have at least a chance of trying to model it? Yeah, Katie. As we said, we won't go back further than the point of transition. By the time we get to the middle of 2023, you'll have some development through the first half and the second half of 2022 and the first half of 2023. We're not planning on sort of revising and going back beyond transition. How's one to model the profitability of the reinsurance and work out the stock that you've been holding back? Well, we've hopefully given some useful sort of approaches to modeling key principles of how the reinsurance will work and that. You'll have a couple of periods of development of those ultimate loss ratios that you'll be able to work from. That won't be happening. It's up to you. It's not that bad. I mean, to me, the information was kind of stock just drops through the roof. Need to give a mic. No problem at all. Listen, we're happy to spend a bit more time with you. You know, we recognize that we've presented quite a lot of some technical content and detailed examples today, so we're happy to spend some time with you over the next coming weeks and months to work through those as well. Freya. Hi. Freya Kong from Bank of America. With the investment results, do you plan to split out the unwind of the discount rate so we can see it over time? Secondly, on the loss ratio, expense ratio, combined ratio, will the denominator be the gross earn premium, or will we have some element of ancillary and installment income in that as well? Should I go? Sorry. Absolutely, yes. We'll split out the unwind of the discount rate within the net investment results. The denominator in the calculations of the ratios will be excluding those additional elements, so other ancillary revenue and so on. It'll be just the gross earn premium fo r the core business or, you know, for whichever business we're modeling. Hi there. Faizan Lakhani from HSBC. My first question is on onerous contracts. Generally, I've always seen that when you book your sort of first underwriting year, it tends to be booked above 100% combined ratio. I'd assume the onerous contract piece would be a much larger for you guys even though it's profitable over time, so can you explain that one? Second one might be a little bit of a silly question. When I look at your equity waterfall, the risk adjustment's positive in that. Is that basically a suggestion that you're releasing reserves in that to top up the equity? Will we see a accelerated amount of profit reserve releases at the full year? Katie, do you want to take the first one t hen Rachel can follow. You're right. In terms of the book loss ratio position in the first underwriting year is generally conservative. You're right, that would lead to a combined ratio of over 100%. The reason that we're not seeing significant onerous loss components is because we'll have taken all of the expenses in relation to that combined ratio. The onerous loss component is really only. You only need to consider the unearned premium and the unearned claims on those, on that premium. There's only a very small impact of expenses because of all those expenses being taken in the first year. By the time you reach that balance sheet date, you've expensed all of the expenses that are within that combined ratio. Sorry, just to understand that. If you had the underwriting year 2023. In the full year 2023, the calendar year, would the last year not be onerous at that point though? Because you're still, e ven if you're front loading, you're still above 100%. No? If you have, at the end of 2023, you've written premium of, say, GBP 100, and you've got a combined ratio of 110% that's based on a loss ratio of 90% and expense of 20%. You've taken the 20% of expenses. You'd only have unearned premium and the loss ratio of 90% to consider on that onerous loss component. In that, you know, example, you wouldn't then recognize that onerous loss component because the claims ratio isn't high enough, isn't over the 100%. Okay. Thank you. Okay. Should I take the reserving question? Yeah, you're right. You've seen within that overall reduction in equity, there's a positive impact from the movement in risk adjustment. That bar itself is on a gross basis. There are some reinsurance effects built into the other steps. We wouldn't really give guidance on ongoing level of reserve releases. You know, we've said that we've got a conservative reserving approach to reserving, we continue to expect to recognize profit over time, you'll still see reserve releases coming through. You're basically taking the impact of the reserving change at transition, but you're not going to release it through higher earnings next year and sort of just release it through P&L. It will be through the balance sheet itself, the reduction in reserve margin. The movement on transition is a one-off impact, so it sort of resets the balance sheet as at the January 1st, 2022. You know, from there, obviously to model the ongoing impacts of reserve releases over time, there will still continue to be those. In a nutshell, lower reserve releases long- term though. Is that a correct way of thinking about it? What did you say? Just in a nutshell, lower reserve releases in the long -term. I wouldn't necessarily say that. We wouldn't give guidance as to the level of reserve releases today, Mursh. I think that's fair to say. Sorry. Thank you. Ivan? Thank you very much. Looking at the earnings recognition patterns, this is Ivan Bokhmat from Barclays. Sorry for not. When I'm looking there at those slides, obviously the slide 18, which shows that you have a kind of a negative GBP 20 million or, let's say like a 1% worse result on the group gross basis and 3% better on year two. Then, you know, similar on slide 20, it kind of suggests that all else being equal, you're going to sort of amplify that loss in year one. Am I reading it right that I should, you know, there's a certain, I guess we'll maybe to make it simple and to not bog down the numbers? I know you can't commit to any particular numbers. If we were to think about it, what's the percentage that your year one profits are going to be lower and your year two profits are going to be higher? If there's a simpler way of thinking about it, 'cause there's clearly some numbers that we're going to work out later. It really does depend on that balance. What I would say is in slide 18, that is for an individual underwriting year. You obviously need to add together, there are multiple underwriting years going into any particular financial year's result. It really does depend on that balance between the risk adjustment and discounting that's being unwound, or added in on the incurred claims and the unwind on the prior year releases. We're not going to give any guidance as to the percentage on that. Directionally, am I right in understanding is that when we think about 2023 earnings, all else being equal, it should be a little bit lower than under IFRS 4 and then subsequent years a little bit higher or no? No. We'd need you to conclude. Yeah. We won't comment on 2023. I think that this gives you some idea on any underwriting year. You can adjust your assumptions accordingly. Sorry, Ivan. James Pearce, Jefferies. I'm just trying to get my head around why, I guess reserve releases before the impacts of discounting wouldn't come down, just given the fact that your claims reserves are coming down, but your ultimate underwriting profitability isn't changing. I would have thought the implication is that reserve releases should just come down as a result of that. Then also just trying to get my head around kind of the net impact of so you're basically recognizing more profit commissions up front in terms of Admiral's taking their share of profit, presumably have more profit commissions, but then less reserve releases. I'm just trying to get my head around what the net impacts on the economics are because, you know, the mechanics of how you make money from profit commissions and reserve releases are quite different? I'm just trying to get a sense of that as well, really. No, it's fine. Listen, we're not expecting you to sort of see these examples for the first time and fully get it. We recognize that you're going to need to take them away and, you know, take a look at your models and adapt and so on. I wouldn't say that, you know, it absolutely we're seeing lower reserve releases over time. There are a number of factors that impact that, the confidence level of the risk adjustment and how that moves over time being one of them. You know, you can take that away. I don't know whether you want to comment on the profit commission one. Yeah. I think the profit commission is important to think about it in two tranches. There's the co-insurer profit commission related to the business that, you know, is not on our, on our risk, in our book, and the reinsurer profit commission. I think the co-insurer profit commission, we'd expect to develop largely in line with those book loss ratios, under IFRS 17. Whereas the reinsurer profit commission, there are some changes which we went through. They'll be incorporated as part of the overall consideration of the claims recoveries, and we'll provide further information on that as we get towards 2023 half year as well. Hi. Thank you. Thomas Bateman from Berenberg. Just coming back again to the kind of profit commissions on the Coinsurance treaty. I feel like they're going to be higher in year one because the discounting, then you have an offset from the higher costs. Net-net, I think you're saying the profit commissions are still higher in year one. I think that there's a number of technical factors that go into it, and it's not like the impact of discounting on those profit commissions isn't necessarily the same. We'd expect it to broadly align with our IFRS 17 book-loss ratio development. It is accounted for under a different standard. Okay. Then just on your 90% reserving corridor. How big is that corridor? Because I think you're saying you're towards the top end to begin with, and you're going back towards the mid-end of the corridor. Is that 10, 20 points? Is it five points? Yeah, we can't talk about that yet at this point in time. You know, when we get into giving you IFRS results and so on, and details of our accounting policy, you'll be able to see. Fortunately, not sharing that today. Okay. Just last question very quickly. On the gross earn premium, I couldn't understand if you were saying there's no additional ancillaries in there or there is some ancillaries in there. Also what was the point you were making on those written within the Group and those outside of the Group? There'll be no premium from the ancillaries within our gross earned premiums in terms of the commission income. I guess the complicating factor in IFRS 17 is that the ancillary income where we underwrite that within the group, we have to recognize all of that as premium. Where the business is underwritten outside the group, we would continue to recognize that as commission income in the same way that we do currently. Where it's underwritten within the group, we have to recognize it as premium. Okay. Understood. Gross earn premium should go up a little bit. What's the split at the moment between internal and external ancillaries? We haven't shared that detail, but there's underwritten is the larger proportion. Yeah. By far. Okay. On the slide, it looked like 50/50. Is that fair or a bit high? I think it's the higher. Higher underwritten. Thank you. I'll be quiet now. Okay. Ben. Thanks. Benjamin Cohen, Investec. I just wanted to ask at a sort of practical level, I mean, I realize you're sort of, y ou're reconciling the two different standards back from one to each other and the other way. Is this change actually going to drive any sort of change in terms of how you run any parts of the business? I suppose particularly around reinsurance buying, how you would see that to Thomas' question in terms of, you know, what you would look to underwrite yourself versus third parties. Is this essentially, you know, a giant bonfire and a lot of money for the accounting profession? I'll take that. I mean, they're not our words. No, we don't see any impact on the business more generally or things like reinsurance buying strategy. I mean, there are a number of factors that we take into consideration when purchasing reinsurance. You know, the capital efficiency, the solvency impacts, the protection, profit impacts are one of those. You know, at this point, we're not expecting any material change in that or any other parts of the business as it were. from Peel Hunt. Once you've reset the balance sheet and we just, you know, carry on writing business, when I look at your gross unearned claims reserves, you've got a breakdown here, and you say, "Well, we're going to disaggregate new business and renewals." How will that look? Thanks. I think that split is performed for the underlying accounting and particularly to look at whether there are any onerous loss components. You wouldn't necessarily see all of that split coming through in the disclosures. It's required in order to model those claims reserves, to understand if there is an onerous loss component. Once the business is earned through, effectively the process is the same. So, we won't be able to see whether new business or renewals is profitable in the breakdown? You'll be able to see in the most recent underwriting year if we're recognizing an onerous loss component. We'll provide further information on why that's being driven, so which pocket of business it is, that's driving the onerous loss component. Would that be by, let's say, motor cohort in the U.K. or how should we think about that? Becuse this is the only indication of whether new business and what you're doing is going to add value or not over time. Is there a way we can sort of get a good view of the new business you're writing and how profitable it will be over time, or do we have to wait until you reset the balance sheet at year-end? I don't think that the point of necessarily the standard or, you know, what our intention is to give you that increased granularity of disclosures on new business versus renewals. It's a topic that's quite relevant for the onerous loss component, we know that others in the market are thinking about it in this way. You might argue that, you know, post FCA reform and so on, it becomes even less relevant that, you know, the splitting profitability between those two parts of the book. If there is an onerous loss component being driven by one of those two, we'll highlight it and you'll see it. Otherwise, I wouldn't expect lots of granularity and ongoing disclosures between new business and renewals. There's a hand at the back there. I can't see who it is. Yes. Thank you. Thanks. Will. Oh, Will, sorry. That's all right. Probably some silly questions, if I'm honest. The cohort, how granular... I know you tried to answer it there, but is it going to be literally new business renew book, or is it going to be, you know, drivers under X, etc.? I mean, how granular do you go when you look at an owner's contract? Second one, really silly question or just a silly question. Looking at this helpful illustrative example of the profit and recognition on slide 18, it gets to us thinking, you know, Sure, there's year one, there's year two differentials, but typically maybe after 24 months, it's a bit earlier profit recognition overall than previously. I guess under what scenario after 24 or 36 months, wherever you want to pick your point, would it be a delay to profit recognition? Just thinking about it the other way around so I can try and get my head around that side of things. That's probably it for the time being. If I take the first one of those around how granular do we get in terms of the split of the business. The standard and the simplified approach just requires you to look for any particular facts and circumstances that might indicate you've got onerous pockets of business. The interpretation of that is generally that unless you're deliberately pricing at a lower rate in order to, for example, grow, where you're willing to write unprofitable business that you wouldn't recognize onerous loss components. We're looking at a slightly higher level. It's really where are there known areas of business where potentially there's differences in profitability. Shall I take the second? I think you asked, well, after 24, 36 months, what could cause a deviation or a delay in profit recognition? I think as Katie explained in the example, when she gets to that point, actually it's large. If I certainly take the growth example, it's that release of risk adjustment that's largely driving the recognition pattern. It's just a question of, okay, what does that look under IFRS 17 versus IFRS 4. It's as simple as that, really. Back to Faizan again. Hi, thanks. I'm patient speaking. Sorry, a few more questions. The first one's sort of following back on Will and Ivan's point. Just trying to understand, if you start growing quickly, would that mean given that underwriting year one is loss-making versus sort of year two, would that suggest that the profitability could be suppressed relative to a year where maybe growth is not so high? Second question is on, you know, it's clear that cash generation isn't changing, but will you still maintain a payout ratio for your dividend policy? Yeah. Yeah, those are two questions for now. Yeah, in terms of that first question, where a business starts to grow quickly because of that policy to expense acquisition costs, you would expect a slight delay in the profit recognition compared to IFRS 4. Obviously, you know, typically those businesses are of a certain size before they start to grow that quickly. I think it's fair that you'd expect a slight delay, from year one to year two, driven primarily by that expensing of acquisition costs. On dividend, yes, we'll continue to take a payout ratio approach. Sorry, can I ask one more question just quickly? Yeah, of course. You're using a 90th percentile for your reserving. What is it if you were to maintain the same level of margin right now? We're not giving guidance on the information. Got you. Another one at the back there, Carl. Is that Will again? Yeah. Sorry, guys. another silly one. coming back to Grace's question at the beginning. Presumably on the triangle, so at the date of the balance sheet, the transition date, do we get, let's say, you know, the triangle for the past 10 years on this basis? Because I couldn't work out from your answer whether we literally get, you know, out of that 2022 and then to 2023, but obviously the history for the 10 years. Will we get that historical triangle, whether it be on a different-? Not to the granularity that IFRS 17 requires. That's fair, isn't it? You'll see a couple of periods of development, but we won't be going back 10 years in that. In reality, we'll probably only be getting sort of 2021, 2022, 2023, that side of things. Yeah. Okay. Thanks. Greg? Over. Sorry, go ahead, Laura. Thank you. It's Anthony from Goldman Sachs. Just two quick questions. Firstly is follow up on the dividend. I think you mentioned the payout ratio, the policy is the same, the payout ratio. Does it mean that if you decide to pursue higher growth in a year, then you have higher acquisition costs, which you expense all of them year one, so that would have lower dividends in that year? Secondly is on the IFRS 9, basically on the assets. Are there any change to your assets recognition? Say you do the same amount in OCI or you're going to put more in fair value through P&L? Thanks. Let me take the first one. I didn't quite catch the second, let's revisit that in a second, if that's okay. Sure. On the first one, yeah, continue to maintain the payout approach. I think it everyone understands that our dividend policy is a normal dividend of 65% of, you know, post-tax earnings and then a special dividend, which is informed by a number of factors, solvency being some of them and a couple of others. We'll continue with that approach. I can't say explicitly, you know, if you've got a change in profile because of growth or, you know, high level of acquisition costs, then you'll have a, you know, like for like change in dividend amount 'cause there's a number of factors that go into that decision. I would say, you know, continue with a payout ratio approach and wouldn't expect sort of significant step changes in that. Cool. Thanks. Could you repeat the second question? Sorry, I didn't quite catch it. Sorry. It's just the comments on those, slide seven. The change in discount rate, between P&L OCI to match movement in fair value of assets. I think, one of European peer, last week, I think they decided to put more assets as fair value through P&L. There could be some net income volatility. Just wondering if there's any? No. I mean, we're not changing our approach under IFRS 9, which has already been adopted. As we said, this choice certainly to put changes in this locked-in discount rate over time through other comprehensive income aligns really well with that choice that we've made previously under IFRS 9. That reduces the volatility in the income statement and also within the other comprehensive income, if you think about it in that way. Thank you. Shall we pass to Greg over here and then Carl? Was it Andreas? Sorry. Yes. Thank you. Greig Paterson. For slide 27, Well, I'm just looking at the IFRS 17 income statement. I'm correct that you will have a separate single line for quota share and a separate single line for Coinsurance? Yeah. The reinsurance results will be able to differentiate between those two? Yeah. Yep. Like you do currently, will you then show from which years that comes from? Yes. All right. Cool. Just on the discount rate, how are you going to set the illiquidity premium? Secondly, locked in, is this going to be locked in by underwriting year and it's going to stay locked? Can every underwriting year have a different discount rate? What kind of disclosure will you be giving on the discount rate? Is it going to be by underwriting year or a blended? Thank you. In terms of the illiquidity premium, we use a number of factors to set that, but it's primarily intended to reflect the illiquidity in the claims which we look at our own assets that are matching those claims liabilities in order to set. When we talk about a locked-in rate, the locked-in rate technically is supposed to be when the claims are incurred. You could therefore have daily locked-in rates, which is not very helpful. We're looking at quarterly locked-in rates. We will have a different rate for each quarter of when the claims are incurred. At the balance sheet date, we'll reflect the balance sheet date and revalue the claims to that. In terms of the granularity of the disclosures, we're still looking at the detail of exactly what we provide and what's needed. We set out, you know, some of the indicative disclosures here, but in terms of the full level of detail, we're still looking at what we'll do and what's helpful. You might find quarterly discount rates a bit less helpful, to be honest, with this diagram there. Yeah. We'll give to Ivan over here. Hi. Thank you. A couple more questions. On the investment income, you mentioned that this unwind of the discount will have a material impact on yield. Could you try to quantify it and put it in some context? The second question is, I know it should be unrelated, but has the migration to IFRS 17 affected the partial internal model project at all? If you take the first one, the way that I would think about the unwind of that discounting is that effectively it's like an interest charge on the value of the undiscounted claims liabilities, and you'll have the detail on those undiscounted claims liabilities. You'll have details on the maturities, the settlement patterns, and also on the discount curves in order to look at those. As we've said, we do expect it to be a reasonably significant proportion of the investment income because of matching that illiquidity premium and the discount curve against the assets that we hold. On the partial internal model, no. No impact. You know, I think we said that we'll come back when we've got some more news on timescales and so on, so that still stands. There were some sort of linkages, I guess, in the sense that, we're talking here confidence levels on a reserve risk distribution, which is aligned with the capital model, but no impacts on that one. The latest bit on the date? No. We're not commenting on dates at the moment. You know, we're working incredibly hard, as we've always said, and we'll come back and let you know when we've got some firmer dates. Tom. Hi. Thank you. Thomas from Berenberg. Just going back to the illiquidity premium, what's like a gauge at the moment you can give us for the spread that you're likely to book? You know, if we look at the yield that you're currently reporting, is that a good gauge or, is it going to be some kind of somewhere in between there and a risk-free rate? I mean, it'll be somewhere in between. The illiquidity premium, we won't give specifics today. It's not meant to include the kind of credit spread element. I think that's probably one thing that w e can say. You've got to strip that part out. Then if you say, well, it's somewhere between those yields and the risk-free, well, you'll get more information out over time, just not today. Okay. Just, just to be clear on the timeline. It's not until half year next year, we will get any IFRS 17 numbers? You will get an impact at transition. In our March 2023 reporting, we'll be reporting on the impact of transition. In that disclosure, we'll give, you know, either a small range or a point estimate of that impact at transition and the key drivers and key accounting policies that relate to that transition. Okay. The more you can give us at that point would be really helpful, but I think that period of like discontinuity, is a big issue for the whole sector. Understood. All right. Can we get Greg over here, please? I'm just, you know, thinking in my case, I'm going to have to rebuild the model from scratch. I'm just thinking, in March when you produce the, could you do an indicative income statement showing all, each one lines that you're going to have and some kind of sort of disclosure regime so we know what we can work with? You know, we don't want to build a model and then find in June we have to throw it out again and rebuild another model. Just the layout. Exactly with no numbers in, just exactly what your ordinary accounts will look like. Yeah. Look, I know I won't commit to anything today. I think we understand that, you know, information is necessary and will be helpful. Let us take that away and talk through it. If you know, the bottom line is the earnings of Admiral's heavily smoothed and don't take offense to the word manipulated by management over the last, y ou know, 10, 15 years. Smoothing earnings aggressively. I'm not saying it's they're not making money or anything. You know, we've invested a lot of time in trying to understand that. Now without the triangles, we have to throw all that knowledge out. That's why I'm saying it's. If we can give us some triangles or some history so that we can put some inputs into, that would make a huge difference. We'll take it away. I don't think you need to throw everything out. You haven't seen my model yet. Fair enough. You know, if you try, I'm tracking each underwriting year, Coinsurance, quota share, all the idiosyncrasies of each year and a feeling for what the reserve redundancy is on that year, et cetera. You know, to keep trying to stock now that can't be used again. We will give you what we think is helpful, what we can at that point in time and what we think is relevant. You know, we'd certainly take on board your comments. All right, cool. Thank you. Yeah. Greg, I don't think it's absolutely fair to say that we're manipulating our profitability because we've had quite a consistent reserving approach over time, and we'd say that that won't change. We will try to give you as much information as possible, but also practically it's not possible to do 10 years backwards in granular detail. We will think very hard about sensitivities. Please, if you guys, you know, if there's specifics that you think of that you think might be useful, drop me an email. Like I'm open and we can, you know, between ourselves take that to the table and consider how we can help you more if there's things that you maybe haven't mentioned today. I've got a suggestion. Hang on. Could you just wait a minute? If you go back about 10, 12 years ago, people were struggling heavily to try and understand the profit commission and quota share, and it was causing a lot of, how can I say? It was causing a discount in the price. Geraint actually produced a very simple indicative spreadsheet with illustrative examples with loss ratios in that. He sent it to all the analysts, and then they were, i t made a huge progress in people trying to understand that. You know, saying for instance, quota share expense, insurance results go back when it lost whatever 20 or lost five, six underwriting years and then showed all that. That might be helpful at some point. Let us take that away. Maybe just put it on the website so that. Because you want to reduce information risk. It's a bit black box-ish a nd people avoid that. You don't want that. I'm just trying to think about ways to try and make it more transparent. You know? That's helpful. Thank you. Any last questions? We're almost out of time. Mike, last quick half. Hi. Thanks. Just a quick one. Nick Johnson, Numis. Just wondered if you could say how you expect the volatility in the insurance margin to compare in a sort of real world to the volatility in combined ratio. Is there anything to say on that? Yeah. I mean, from my perspective, it's difficult. We, you know, we're not commenting on the wider market and what volatility others will see. I think, you know, a couple of years ago, IFRS 17, you sort of thought about it with a bit of trepidation, you know, complex terminology, lots of presentational changes, what will it do? I think as we've got closer now with all of us are starting to become, you know, much more familiar with the key concepts, the key principles. As more and more in the market talk about it, the hope is I think that actually, you know, there's not that increased level of volatility and certainly, you know, even the opposite in some cases. That, I mean, that's the way that I think about it in any case. Good. If no more questions, I think let's wrap up. Thanks everyone for coming. Hopefully, it's given you some insight, but you have my email address, so open to follow-up questions and discussions down the line. Good to see everyone today. Thank you. Thanks so much. Thank you.