Admiral Group plc (ADM) Earnings Call Transcript & Summary

Good morning, everyone, and welcome, and thank you for joining us as we review Admiral 2025 year-end results. Today, we'll be announcing another remarkable year of financial results and strategic progress. So I will start with the key highlights before handing over to Geraint on the financials and to Alistair on U.K. Insurance and Costi on Europe. I will then come back to reflect on what we have achieved over the last 5 years and finally explain how the evolution of our strategy position us to create even more value in the years ahead. So let's start with the main achievement for 2025. We delivered a record profit of GBP 958 million. This was up 16% year-on-year, reflecting disciplined execution and growth across the group. 2025 also marked exciting progress across data, technology and AI and the evolution of our motor proposition, including the acquisition of Flock subject to regulatory approval. Today, we'll also outline the evolution of our group strategy. This strategy builds on a very strong platform, but more diversified customer base and the competitive advantage we already have to deliver higher long-term value for all our stakeholders. We will also cover our new capital distribution framework, including share buybacks. Geraint will take you through that later. So more in detail. As already mentioned, 2025 was a year of record. Our customer base increased 7%, while we continue delivering strong customer outcomes with the group Net Promoter Score over 50. Group profit reached a new high, driven by record U.K. Motor profit, passing now the bar of GBP 1 billion, following another record year in 2024. This was achieved in a challenging market environment, thanks to positive evolution of recent years and continued underwriting discipline across the cycle. Importantly, this was not just a U.K. Motor story. All parts of the group contributed. In the U.K., Other Personal Lines, Admiral Money combined delivered a profit of GBP 88 million. Europe also performed strongly with a fast return to profitability in Italy and great results in France, which Costi will cover shortly. 2025 was also a year of strong shareholder returns, supported by a 7% increase in dividend per share, a very strong capital position with a solvency ratio of 193% and another stellar return on equity of 53%. Beyond the financial results, 2025 marked an acceleration in our strategic progress. We are pleased with our rapid advancement on artificial intelligence, particularly with the value delivered by machine learning models and the new gen AI center of excellence to scale priority use case, train our people and provide them with the right tools. We are managing more than 150 gen AI initiatives across the group, including support to over 4,000 colleagues, some agentic models with promising initial results and more potential to come. Selling more product to our existing customers remains a key growth driver with our multi-risk customers now exceeding GBP 1.6 million. Across Europe, we continue to evolve our broker propositions with stronger segmentation and more customized offering, driving better margin, as Costi will explain later. We also continue to innovate in Motor. An example is our partnership with Octopus that positions us well in the fast-growing salary-sacrifice scheme for electric vehicles with a tailored risk-based proposition aligned with our ambition to support customers in making greener choices. And in Admiral Money, completing our first forward flow deal was an important milestone as it opened up a more capital-efficient growth path and support higher returns and lower volatility. On M&A, the integration of More Than is now fully completed and contributed positively. Elephant disposal is also completed. And finally, early this year, we announced our intention to acquire Flock, a company we had invested in since 2024. Flock offers a telemetry-based fleet proposition with an effective feedback loop to improve safety and performance. It's an excellent strategic fit with our U.K. Motor expertise with promising underwriting and claim synergies and it's closely aligned with our joint ambition to improve safety on the roads. And by combining Admiral data ambition to Admiral's strength with Flock technology, we see an opportunity to develop a differentiated fleet business in an underserved market. So in summary, 2025 was a record year for Admiral with strong profits, customer growth and progress in technology and strategy. Now before handing over to Geraint, I want to take a moment because this will be the last time that you joined me on stage to present results. And I think it's fair to say that the strength and discipline of the performance for about 2 years are a good reflection of his leadership, his judgment and his consistency over the last 12 years as CFO, a period during which Admiral tripled its turnover and grew profit from GBP 350 million to almost GBP 1 billion. And please join me to congratulate Rachel who is here with us today and will succeed to Geraint, bringing deep knowledge of Admiral, a strong track record within the group and a great skill set for the role. So thank you, Geraint, and congratulations, Rachel.

Thank you. Good morning, everyone. 12 years of not being conduct. One last time, let me talk you through the main drivers of an excellent 2025 result. Lots of positives, lots of good milestones. I'll cover the U.K. Motor loss ratios, the dividend, strong capital position. And as Milena mentioned, I'll talk you through the change in the approach to capital return that we've announced today. To start with though, let's look at the component parts of the group profits and the main ratios. The group combined ratio was very positive again at 80%. That was 3 points higher than 2024, though the impact of Ogden accounted for around 2 points of that difference. So in reality, only a very small change. And that, in turn, has made up of a slightly improved expense ratio and a slightly higher loss ratio. The latter as expected, due to the higher loss ratio 2025, underwriting year in the U.K. Motor having an impact. On to the results then. In U.K. Insurance, overall profit was GBP 1.1 billion. That's GBP 110 million higher than 2024, including Ogden, or GBP 180 million higher if Ogden is excluded, very big increase. The U.K. Motor results, I'll cover shortly, but the result there was a record profit, just over GBP 1 billion. And we're very pleased with a really strong year for the U.K. Other Personal Lines, Home insurance, Travel and Pet insurance, all profitable, strong growth. And the combined profit there was GBP 62 million, was nearly triple of 2024's result. In Europe, we're reporting a much better results, improving by nearly GBP 30 million versus 2024. We see growth in higher profits in France, small loss in Spain, impacted by new reinsurance arrangements and a recovery to profit in Italy. Good to see that happen so swiftly. And worth reminding that we continue to hold prudent booked reserves in Europe in the upper end of our range and the best estimates are also conservative. Admiral Money had a great year. Profit was double 2024's, benefiting firstly from good growth in the balance sheet. But also, as we talked about at the 2025 half year from profit generated from selling some back book loans in the first half and selling newly originated loans, which don't hit Admiral's balance sheet. That will be a continuing, we think, attractive feature of the Admiral Money business model. We continue to see good margins on the unsecured loans business, which makes up the big majority of the balances, but the results from car finance, which was relaunched in late 2024, are also encouraging. Credit loss experience remains very solid, and we hold an appropriately prudent provision for losses. There are some other comments on the page, which cover the movement in the share scheme costs and the other line, and you've got the usual extra information in the back of the pack. All in all, group profit was up 16% or 28% if you exclude the impact of Ogden on both years. Let's take a look at the very impressive U.K. Motor results. So this is a summarized income statement plus some of the key ratios and some commentary. Both years include the impact of the Ogden discount rate change. And so some of those year-on-year comparisons, you see look a little less stronger than they really are. We show the pounds and the percentage impact of Ogden in the table and starting with the top line. Customer numbers increased by 2% year-on-year, 50,000 added in the first half and around 80,000 in the second half, so 1% increases half-on-half. As Alistair will talk a bit more about later, we reduced our prices in H1 last year, and hence, average premiums have fallen. And so despite our bigger portfolio, turnover was down by 7% as the team took a disciplined approach in the competitive U.K. market and reflecting the claims trends that we were seeing. As a result of the reduced premiums and continued claims inflation, the current year loss ratio for '25 is 3 points higher than '24. And of course, we also don't see quite the same positive impact of Ogden in '25 than we did last year. And those 2 items are the main drivers of the higher combined ratio you see at the bottom, which is as we expected. The underwriting results improved by around GBP 40 million with higher earned premiums and a much lower reinsurance charge offsetting the higher net claims cost. You'll remember that we had much more limited quota share recovery assets coming into 2025, and we see a similar picture as we exit 2025 too. Net investment income was higher, up to a record level due mainly to higher invested assets at a similar rate of return. Profit commission was notably higher as we started now to recognize income on the high profitability 2024 underwriting year, though we still haven't yet recognized income on '21 to '23 or on 2025. We do expect to see revenue coming through on '21 and '25 very soon. I already mentioned the main drivers of the higher combined ratio we see at the bottom. But within that mix, reserve releases were 10% year-on-year, basically the same like-for-like. Next up, we'll take a quick look at the main U.K. Motor loss ratios, which, as always, are a key driver of this result. The chart shows the U.K. Motor discounted book loss ratios and there are generally positive and consistent messages to report here. We can see -- we see continued strong improvements in '23 and especially on '24 over the last year. 2024 is clearly a very good margin year on a very large premium base. In 2025, we see burn cost inflation around mid-single digits level and that's a small improvement in H2 versus where we saw things at the half year point. The first discounted booked loss ratio for '25 is at 78%. That's 7 points up versus '24 at its equivalent point. And that's again basically in line with our expectation. On an undiscounted basis, 2025 is 85% compared to 77% for '24. Now we expect '25 will be a good profitable year. You can see it looks healthy on the chart at the 12-month point, and it should develop positively from here, though obviously won't end as profitably as 2024. We maintained very high reserve strength. It's very close to the maximum percentile, and we expect that will reduce a bit further in 2026 towards the middle of our range. Overall, on claims, positive experience in line with our expectations, usual trends and there's more information in the back of the document. Moving now to look at the capital position. So this is the bridge of the solvency ratio from half year to full year '25. There's a couple of observations. Firstly, the capital generation in the second half is largely offset by the final dividend. And secondly, due mainly to pretty flat revenue in '25 versus '24, we see a much smaller change in the capital requirement in '25 than we did in '24 and particularly in the second half. And then the change in the capital requirements and the other items in the middle almost cancel each other out, leaving the group with a healthy -- very healthy, almost flat ratio of 193%. Short update on the internal model. Lots of hard work by our team, as usual, over the past few months since we last updated you. We now expect to make our application for approval shortly. Post approval, we'll target solvency coverage in the 150% to 170% range, probably at the upper end, in part to give us flexibility for smaller M&A opportunities. We'll give more information on the post-model approval capital position at the appropriate time. Speaking of M&A briefly, Milena mentioned earlier, the Flock acquisition. As we said in the press release, if that gets regulatory approval and completes in the second quarter, we estimate the impact on solvency will be a bit less than 10 percentage points and is, therefore, largely absorbed by the strong position. Next up is the dividend. So these are the details of the dividends, split between interim and final. And for 2024, we call out the impact of the Ogden change, which was obviously significant on the dividend for last year. The proposed final dividend is 90p per share. That brings the total for the year to 205p, over GBP 620 million, and that's 7% higher than 2024. The difference in the payout ratios year-on-year is due to us starting to use capital to purchase shares for the share schemes, which we said back in August would start in the second half of '25. You'll remember that historically, we issued new shares each year for those share schemes rather than purchased in the market, but we haven't done that since 2023. In the fourth quarter last year, the trust bought about 1 million shares for just over GBP 30 million. And the capital that we use for dividend and the share scheme purchases equated to basically the same percentage of earnings across both years, close to the 90% level. And in 2026, we expect the trust will buy around 3 million shares. Next up, we'll cover the change in the capital return approach. On the left, we show a summary of our capital allocation framework. Milena will talk a bit more about Point 1 later, which covers how we allocate capital to our businesses. And we're generally comfortable that around 10% of earnings is a fair guide of what we need to retain to fund and invest in growth. And that's meant an average dividend payout over the last 5 or 6 years of 90%. Step 2, we know the importance of strong cash returns to our shareholders. So the ordinary dividend remains at 65% of earnings. Step 3, as just mentioned, we purchased shares for the share plans. And final -- and Step 4, not finally, using some surplus capital is an option for funding M&A. And then that leaves the surplus capital and that's what's changing today. Historically, as you know, we've returned this to shareholders in the form of special dividends. But from the interim 2026 dividend, we'll change that Step 5 to be either buyback and cancel shares or pay a special dividend depending on what the Board believes is the best option. For 2026, subject to regulatory approval, we expect to buy shares at the interim and final dividend dates. The 90% guidance we've given out over the past few years to cover the ordinary plus the special or buyback plus the share schemes purchase should generally hold moving forward. And then one final slide for me to sum up. Looking back on 2025, clearly, it was a really strong year, record profits, record returns to shareholders, lots of positive results and developments across the group. For U.K., the Personal Lines and Admiral Money, great results, strong and swift turnaround in Europe, progress on the internal model, very pleasing stuff. And looking ahead, a few comments on what we might expect in 2026. On growth, in summary, we plan to grow everywhere. That's obviously subject to how the markets develop, in particular, when prices in U.K. Motor start to increase. For turnover, I'd expect a bit more growth in '26 than we saw in '25. And in general, of course, we expect faster growth from the newer businesses, U.K., the Personal Lines, Admiral Money and Europe. And then a few comments in respect to the group profit. Firstly, obviously, we will see more of an impact of the less profitable '25 underwriting year feeding into the 2026 results, but we will still benefit from good releases and profit commission coming through on 2024 and '23 and some of the earlier years too. Secondly, we project continued improvements in the results in aggregate for the newer businesses that we talked about. And finally, we expect group profit in '26 to be quite flat versus '25 after a really very strong last couple of years where profits have more than doubled. And all those comments, of course, subject to the usual caveats on markets, geopolitics, war and weather. That's it from me. I will hand you to Alistair now to talk to us about U.K. Insurance.

Thank you, Geraint. Good morning. I'm very pleased to take you through an excellent set of U.K. Insurance results. 2025 has been a record year across all our lines of business, underpinned by disciplined execution, customer centricity and strong operational delivery. Starting with the headlines. Customer numbers reached 9.6 million, up 9% year-on-year, with strong contributions from Motor, Household, Travel and Pet. We delivered GBP 5 billion of turnover and GBP 1.1 billion of profit, passing the GBP 1 billion profit milestone for the first time. We continue to deliver competitive prices, great service and good customer outcomes, which is recognized in customer feedback. We remain #1 in Trustpilot and achieve an NPS over 55. Importantly, 1.6 million customers now hold 2 or more products with us, a 14% increase year-on-year. Customers buying more products gives us better data to improve risk selection for all products. is a driver of our retention advantage in Motor and growth in new lines of business. Overall, an efficient source of growth that contributes to improved expense ratios. Recent announcements are leading to a more predictable regulatory landscape. Outcomes from the Motor insurance task force and premium finance review were in line with expectations, and the Home and Travel claims handling review is now complete, and we have no significant concerns. Let's turn to the Motor market. Starting with claims trends, frequency was largely flat following the marked decline we saw in 2024, and severity has returned to more normal mid-single-digit levels. Our expectation is that these trends continue, but the current macro environment introduces some uncertainty. The graph on the left shows a dark line for claims burn costs. Claims burn costs increased steeply through 2022 and then continue to increase but more modestly. The light line for market average premiums shows a lagging response to claims costs, increasing rapidly in 2023, outpacing claims costs and then declining. Both lines are indexed to 2021, and you see they cross in 2025 as increases in claims costs now exceed increases in premiums over the period. Let's focus on recent market prices. On the right, you see prices continue to decline through the second half of 2025, though at a slower rate than in the first half. We estimate average premiums declined by around 10% in 2025, broadly in line with movements reflected in ABI data. Since the start of '26, market prices are relatively flat with some differences in strategy between market participants. Market prices need to increase imminently. EY forecasts a Motor market combined ratio of 111% for 2026. This is on an earned basis, and EY assumed price increases through 2026. So delays in market price increases will put more pressure on this 2026 market combined ratio. Turning to Admiral U.K. Motor. In 2025, we focused on disciplined cycle management and maintaining our strong advantage in pricing, claims and customer retention. In 2025, we reduced rates by around half as much as the market. All the decreases were in H1. In H2, our prices were broadly flat. The left-hand graph shows that this led to a decline in new business market share in the second half of '25. Lower new business was more than offset by strong retention, resulting in modest policy growth, though lower average premiums resulted in a drop in turnover. In '26, we started increasing premiums with low single-digit increases at the start of the year to reflect the claims outlook and maintain good written margins. Taking a longer view, our disciplined approach results in varying growth through the cycle, but maximizes value and growth over the medium term. The graph on the right shows our year-on-year vehicle growth rate in blue, in yellow is our written loss ratio. Our loss ratio is consistently better than the market, but still fluctuates within a range due to the cycle. We respond quickly to claims trends, even if it results in slower growth in the short term. It then enables us to grow quickly when loss ratios are low, for example, by 15% in 2024. Since the start of 2020, our vehicles covered has grown at a CAGR of 5% and with an average combined ratio advantage of around 20% versus the market. We continue to invest in strengthening our pricing, claims and claims capabilities, including embracing predictive AI and gen AI, which Milena will talk more about. Electric vehicles is a great example of our pricing and claims focus. We lead in this growing segment. We're very competitive whilst delivering comparable loss ratios to high levels of reparability. Our overall approach is to be disciplined and grow when the time is right, whilst focusing on driving advantage in pricing, claims and customer retention. We're confident this will result in growth and maximizing value over the medium term. Let's move to our other U.K. Insurance lines where we've had an outstanding year. We welcomed 650,000 new customers, year-on-year growth of 21% and tripled profits across Household, Travel and Pet. In Household, market premiums softened further and subsidence claims were elevated in the second half of the year. Our own pricing remained more disciplined than the market and weather-adjusted loss ratios improved by about 2 percentage points. This, combined with top line growth meant that although prior year reserve releases normalize from the 2024 exceptional levels, we still delivered a record Household profit. The More Than integration is complete with around 380,000 Home and Pet customers transferred successfully. This has accelerated growth and enhanced capability, particularly in Pet. Travel grew customers by 29% and continued its positive profit trajectory. Pet grew even faster and reached breakeven just 3 years after launch. All 3 lines are now profitable with clear momentum and strong positions across their markets. So -- I'm going on too fast. So in summary, in 2025, we've delivered record profits. But in addition, Motor remains disciplined and well positioned ahead of the market. Pricing increases expected in 2026. Household, Travel and Pet are performing extremely well with growing scale and margin and customer satisfaction and retention are excellent with more customers choosing to buy more products from us. We enter 2026 with confidence that we'll continue to deliver sustainable profitable growth over the medium term. Milena will talk more about this shortly. Now I'll hand over to Costi for Europe.

Thank you, Al, and good morning, everyone. For our European operations, 2025 has been a year of consolidation. We have directed our efforts towards strengthening the operational core across our 3 markets, focusing on the fundamentals of discipline and optimization. It has been a positive period where we have made good progress on our strategy, providing a positive contribution to the group's ongoing diversification efforts. Moving to the financial results. The headline for the year is a return to combined profitability across the region. The business delivered GBP 39 million Motor profit on a wall account basis, of which GBP 11 million is the Admiral's share. Going back to the business performance, we closed 2025 with a good combined ratio of 94%. While this represents a significant year-on-year improvement of over 10 points, it is important to look at the individual market dynamics. In Italy, with ConTe, we have reached a small profit which is a GBP 30 million recovery from the previous year. This significant recovery was driven by strong actions taken on the expenses and a deliberate and disciplined pruning of the portfolio. We made the conscious decision to prioritize technical margins over volumes, leading to an expected vehicle in force reduction. With the business now on a more stable footing, we are in a position to look towards growth, always keeping the focus on its underwriting quality. Moving to Spain, where Admiral Seguros' reported results includes about GBP 8 million of one-off accounting impact related to a change in the reinsurance structure. Going forward, we have established new multiyear and large reinsurance arrangements at the European level with our historical partners. Effective from 2026, these agreements aim to improve capital efficiency and provide greater stability to our results. Excluding this specific item, the Spanish business is nearly breakeven. This is supported by the direct business, which provides a positive contribution to the results. While our diversification initiatives with ING Bank and brokers are showing very encouraging improvements while scaling up. Closing with France, where L'olivier has had a very strong year. We achieved double-digit growth in both turnover and profit with results reaching GBP 16 million profit and we surpassed 0.5 million customers. This performance demonstrates that L'olivier is successfully applying the Admiral model, maintaining a strong combined ratio advantage versus the market while driving growth through digital channels. Let's move to review our strategic progress, starting from the shift in distribution. We have focused heavily on our new brokers proposition in Italy and Spain. As the business mix indicates, we have moved away from an initial test and learn proposition towards this new one, which focuses on building long-term relationships with the intermediaries and also targets better risk segments and higher-margin business in line with our expectations. The early metrics from this shift are positive and provide a solid foundation. We are seeing solid improvement compared to our order book across all the key metrics like higher income per policy, lower frequency and lower cost per claim. And these improved fundamentals have contributed to a 9-point reduction in the overall loss ratio. While there is still more work to do, we expect these benefits to continue as more growth will come and the new proposition mature. In France, we are continuing to diversify through our household insurance product. We now cover over 100,000 risks, a 25% increase versus last year, which provides a meaningful second pillar to our French operation. Regarding efficiency, we have managed to steadily reduce the European motor expense ratio by 7 points since 2022. This has been a necessary step to remain competitive. And even in Italy, despite the reduction in turnover, we improved the expense ratio by 1 point through automation and more streamlined digital customer experience. These operational improvements are also supported by our new common data platform, which is now operational across the 3 countries. This asset allows us to deploy data futures and machine learning models across border with greater technical agility and quality, which is essential for maintaining our edge in a rapidly evolving market. To wrap up, we're very pleased by the progress made this year. Our European operations have reached combined profitability, giving us confidence in their future contribution to group's broader diversification strategy. Our objective moving forward is to leverage this stability to increase our scale and enhance our earnings. We have the right expertise, a solid data and technological framework and a disciplined path ahead. Thank you. I'll now hand over to Milena to talk more about the group strategy.

Thank you, Costi. So as you just heard, 2025 was an excellent year across the group. Now I would like to take a moment and step back with you and see what we have accomplished over the last 5 years. In 2020, we announced our 5-year group strategy based on 3 pillars: business diversification, Admiral 2.0 and Motor evolution. And today, we're extremely proud of what we achieved in this time frame. First, remarkable growth with turnover up nearly 90% and group risk and profit almost 60%. We returned overall GBP 3.2 billion to our shareholders. Second, we diversified the group with more than 50% of customers now coming from other lines of business or geographies and contributing close to GBP 100 million of profit. In addition, we developed new business such as Pet insurance in the U.K., Commercial insurance in the U.K. also, Household insurance in France and we extend our addressable target market with U.K. Commercial Insurance as just mentioned in B2B2C, in B2B and B2B2C in Europe by opening up the broker distribution channel. Third, we refocus our portfolio. We exit all of our price comparison sites and the U.S. insurance business to concentrate on the growth great opportunities we have in U.K. and in Continental Europe. The acquisition of More Than and of Flock are instrumental to strengthen our product diversification in the U.K. Fourth, we overachieved our Admiral 2.0 ambition. With cloud migration, new data platform, tech stack renewal, hybrid working, scaled agile delivery, predictive AI excellence at scale and the announcement of our multiproduct offer. Fifth, we made further progress in our Motor proposition, including market leadership in EV, as Alistair mentioned, growing telematic product, fast-growing subscription model and short-term insurance with our brand for the youngest Veygo. Last but most important, throughout this period, we maintained our historical and quite unique strength. More than 20-point combined ratio advantage versus market in our core business, a unique 30-point delta return on equity versus market, a group NPS above 50 and the legendary Great Place to Work status. So back to nowadays where this leave us. Our current market presents very significant growth opportunities. They are large, attractive, growing with a combined size of around GBP 130 billion. And today, our market share across many of these markets remains relatively modest, and this leaves us substantial headroom to grow. We're continuing evolving our offering to unlock further opportunity in lending with the first forward flow deal and a new car finance product in Europe, extending our distribution and product lines. Commercial Insurance and SME are also good opportunity to provide strong proposition to a large underserved market, experiencing similar trends to Personal Lines 20 years ago with more digitalization, pricing sophistication and automation, where we can deploy our competitive advantage. Organic growth in all these segments will be driven by market-leading expertise in price comparison site and digital distribution, channels that are growing faster than the rest of the market. Cross-selling and higher retention and increasing economies of scale; and fourth, automation and synergies across the group. Our plan is based on organic growth, but we will consider opportunities for selective accretive acquisitions to accelerate diversification. Importantly, the diversification also reduced over time our exposure to any single market cycle, making Admiral more resilient in time. So having delivered on our strategy, we now look forward starting with the market context that is fast evolving, but also presenting very interesting opportunities and tailwinds. The U.K. market cycle in Motor is expected to turn and the regulatory environment is expected to be more predictable as Alistair commented before. Market consolidation could create more rational dynamics overall. More importantly, the rapid evolution of AI and gen AI represents a major opportunity for us. Predictive AI is becoming the key driver of underwriting differentiation. And we already have 12 points of loss ratio advantage versus market and this is a big driver of it. Gen AI and automation offer efficiency potential of up to 30% in the long term for customer service area and may also disrupt distribution and proposition in the long run. What is interesting to us is also the potential to accelerate the transition to direct distribution in markets where direct has more room to grow. Another major trend is the advancement of car technology, another key pillar of our strategy since 2020 Motor evolution. In the short to medium term, the most impactful change will be the shift to electric vehicles, expected to reach around 80% of new car sales by 2030, where we already have underwriting and market share leadership, as Alistair commented before. This is followed by an increased penetration of advanced safety systems. These technologies have a positive impact on collision frequency, but this has been so far more than offset by an increase in severity. As for EV, our scale and sophisticated prices approach results in a competitive advantage. In the long run, we expect autonomous vehicles, now in their infancy, to grow in share and reach a point where frequency decrease will not be anymore offset by severity. For this to happen, we need to see technology, customer appetite, regulation, infrastructure, all to further develop across countries. It will take anyhow long time to scale with higher level autonomy expected to represent around 4% of the car park by 2035, and the overall market premiums expected to be continuing to grow for at least 20 years, supported by number of cars on the road and the mix. We remain very close to this evolution, having, for example, underwritten Wayve, an autonomous vehicles player in the U.K. since 2018. As AI and mobility trends evolve, our view is that the key winning factor will remain sophisticated data-driven decision-making, scale and a lot of good quality data at scale, an entrepreneurial mindset consistently looking for opportunity to innovate and cost efficiency. And those are all areas where Admiral has a structural advantage, including 8-point expense ratio delta versus market. So overall, we are strongly positioned to leverage those key trends. Now let me introduce our evolved strategy framework. First of all, this is not a discontinuity in our strategy. It's an inflection point where we start compounding what we have already built, a more diversified business, stronger platforms and proven competitive advantage. The focus is now making those trends to reinforce each other and more deliberately over time. I think about this strategy with a set of reinforcing layers, each layer supports the next and within each layer, the benefit compound as the business grows. The other layer of our strategy is where we compound performance. Our first pillar is scaled selectively and profitably. And this is about translating diversification into sustainable growth and accelerate margin in our newer lines of business as they mature. The middle layer is where we compound capabilities. Our second pillar is future-proof our competitive advantage and it is about leveraging on the strong capability we have built in data and technology and the multiproduct benefits to improve customer lifetime value and our structural edge. At the core, at the center, we are compounding our foundations. Our third pillar is amplify the Admiral DNA, and this is to ensure that our culture, our talent, the innovation and the impact continue to evolve, providing stability and long-term direction and resilience as we grow. So the pillars are interconnected. Stronger foundation are a driver of stronger capability and in turn, these are enabler of stronger performance. Let me now walk you through each of these pillars in more detail and explain what we intend to prioritize and what we aim to deliver for each. So first pillar, scaling selectively and profitably. We have a clear ambition to continue to scale all our business while increasing margins in our newer lines. In U.K. Motor, we'll continue to grow as we have done in every cycle since Admiral was founded with discipline and at the right time, as Alistair illustrated earlier. We'll continue to invest to maintain our market-leading margins. In other lines of business, U.K. Personal Lines, Admiral Money and Europe, we will continue to grow and at a faster pace of Motor on average, generating greater economy of scale and higher margins. A key focus will be transferring our underwriting and claims trends from U.K. Motor into other products and geography as well as creating more cost synergies across the group and leverage the benefit of multi-risk ownership. Overall, we expect to deliver strong revenue growth everywhere, including reaching top 3 position in Other Insurance Personal Lines in the U.K. and substantially higher margin for those business combined, more than doubling profit by 2028 and more profitable growth thereafter. Finally, we will continue to develop our new and still small U.K. Commercial Insurance business, building a stronger SME proposition and growing commercial motor starting with the integration of Flock. Now let's move to the capabilities that are the key enabler of the growth ambition that we just discussed. It's a virtuous circle that starts from our structural strength in data, customer focus and speed with the objective to increase customer lifetime value. And with higher customer lifetime value, we create optionality, the flexibility to reinvest in these capabilities or to invest and grow or to retain margins. On the left side of the slide, we see how this focus translate into better underwriting results and efficiency. We'll continue to extend our advanced predictive AI capability, increasing both the quality and the velocity of pricing across all the lines of business and geography beyond motor. We'll also increasingly leverage on connected vehicle data and predictive AI beyond underwriting into customer-based management. This model -- this predictive AI model already delivered over GBP 100 million of incremental loss ratio value, and we expect this to continue over time. At the same time, we see good potential from generative AI to improve customer engagement, to increase productivity in technology and service area, and in particular, to improve speed of settlement that is great for customer and in addition, correlates with lower cost of claims. It's a win-win. Combined with further automation and continued cost discipline, we expect more than GBP 100 million of annual efficiency benefit by 2028. That as I said before, we may decide to reinvest in existing capability. On the right side of the slide, we look at our customers. We are strengthening a mobile-first digital end-to-end experience, multiproduct ownership and retention, which is already above market and will further benefit from multiproduct customers retaining around 5 points better than the rest. Lower expense ratio, higher retention and multiproduct ownership are key driver of higher customer lifetime value. As mentioned, this creates optionality, but also a more resilience to long-term market trends, margin pressures and volatility. So moving to the third pillar, amplifying Admiral DNA. This is what makes Admiral Admiral and different. It's our culture, it's our approach and it's something we are deeply proud of. As the environment evolves, we are focused on ensuring that our DNA evolves too, starting with our people through reskilling, developing internal talent and strengthening diversity and mobility across the group. We had this year so many examples of senior leader moving across different area and geography, including the new CEO of Veygo and new Head of Claims, new Group Data Officer, new Head of Data and Tech in Europe. And this internal mobility allow us to get different perspective and cross-fertilization from one side, but also continuity and cultural fit from the other. Another strong feature of Admiral culture is relentless curiosity and innovation, and we'll continue to evolve our products and innovate for our customers. We focus on offering competitive price, inclusive product, affordable product, including for nonstandard risks. Safety and sustainability are central in our product proposition, whether through fleet safety proposition like Flock or through EV electric vehicle leadership or initiatives around flood prevention. We also want to increase our positive impact on communities, investing around 1% of profit into community initiatives with focus on employability and climate resilience. We are proud of the 45,000 volunteering hours delivered by our colleague in 2025 and remain committed to our net zero ambition by 2040. So that was the third pillar of our strategy. Our strategy is also supported by a simple and disciplined capital management framework that is designed to increase value over time. So how we allocate capital to our operation? In U.K. Insurance, we focus on optimizing returns over the medium term. As we've always done, targeting consistently high return on equity with no structural capital constraints. In other lines, we invest to support growth and margin expansions where financial orders are met or expected to be met in the near term. In newer hires, like commercial, for example, we allocate capital to R&D and early-stage investment while requiring a clear right to win and scalability in the medium to long term. A key structural advantage is our capital-efficient reinsurance model and this is a competitive advantage that is quite difficult to replicate as it stands as it is built over more than 20 years of strong track record. Geraint already talked you through the other steps of our framework, including the introduction of buyback as additional way to return surplus to our shareholders. Selective M&A remains an opportunistic tool to accelerate growth, especially on Other Personal Lines in the U.K. and in Europe, only where our financial hurdles are met. So in this slide, next slide, we bring all of it together. Our strategy and capital management framework designed to deliver strong value for our customers and shareholders. We already delivered strong earnings growth, exceptional return and resilience through the cycle with a 7.6% EPS CAGR over the last 5 years. Looking forward, our ambition is to sustain and build on that performance by scaling what already works, disciplined growing U.K. Motor, faster growth and margin expansions in other lines and continued optionality from capability improvements. Our model is quite unique in the sector, delivering at the same time, strong returns, growth and exceptional capital efficiency. Importantly, this is about quality of growth as much as quantity, retaining our competitive advantage, our capital discipline and our culture that underpins them. In short, we believe we can continue to deliver higher returns sustainably while staying true to what makes Admiral different. So conclusion, to sum up, 2025 was a record year for Admiral, record profits, record dividends and strong customer growth delivered through discipline in U.K. Motor and increasingly diversified contribution across the group. Second, we have fully delivered our 2020-2025 strategy. Admiral today is resilient and more diversified with proven competitive advantage that are difficult to replicate. Third, looking ahead to 2026, while U.K. Motor market remains competitive, we expect price to increase. Admiral is well positioned to perform strongly and remain disciplined and resilient through the cycle. Fourth, we have evolved our strategy to compound those trends, not change direction, but raising ambition while staying disciplined. We have strengthened our capital management framework, adding buybacks alongside dividends while maintaining a very strong balance sheet and flexibility to invest. We remain confident on our trajectory, on our ability to leverage market trends and continue to deliver even greater value to our customers and to our shareholders for the long term. Thank you very much for listening. And now we're ready to take questions.

[Operator Instructions] I think, first one, I saw it.

Darius Satkauskas with KBW. The first question is sort of a statement and a question. I appreciate the update to the capital return policy introduction of opportunistic buyback. I think one of the challenges with having an opportunistic buyback rather than the program is that when you do it, it's great. When you don't, it sort of signals in the market that management is saying the shares may be expensive. How are you going to deal with that challenge? And are there any hurdle rates you'd like to point for us to sort of gauge how you think about when we should expect buyback and when not? And the second question is, your Flock acquisition, where do you think you are in positioning for the potential liability shift to Commercial from Personal among your competitors? And who do you think is going to determine the win in the future? Is there a risk that a company like Allianz with a huge balance sheet simply takes the entire market in Commercial Insurance? Or do you think Admiral can appropriately compete 10 years down the line, 20 years down the line?

Geraint, do you want to take the first one, I'll take the second?

Yes. So buybacks, it will be based on what the Board assesses is the right thing to do to try and deliver the max return for shareholders over the medium to long term. I don't, certainly for the foreseeable future, expect it to be dip in, dip out. We'd expect to be doing it for 2026, and we'll give -- we -- the company will give an update on that at least annually, I suspect, as we move forward. But yes, I hear your point on opportunistic versus steady.

So your second question is about Flock and Commercial Insurance. So there are a few reasons why we're interested in this market. It's attractive as stand-alone market, but it's also a market where we see we can deploy a lot of our strength. And the fleet market is very competitive. So you do need to be a very good underwriter. Our claims and pricing strength can be transferred across nicely. But we also think it's a market that is -- it will be disrupted. And that's why we didn't want to really enter in the traditional way, but just focus on a proposition that we think is fit for the future, is the type of business that's going to grow in the future. So it's telemetry based. There's a lot of data from -- a lot of driving data, a sector in which we already have developed strong components to very large and growing telematic portfolio in Personal Lines. It's an interesting proposition because there is a strong feedback loop to driver to increase safety, to increase performance. And I would say it's also a building block for car of the future. The more the car become -- embed safety features and become autonomous, the more this type of skill set, data-driven pricing and underwriting and the feedback loop is important. So for us, it's a very interesting way to create and to develop a business that is interesting per se, but is also a way into the future. And I think it's a competitive market. We need to do it in the right way and with a proposition that is future fit. That's our ambition there. And we think the mix of Flock skill, technology and proposition, an Admiral amount of data, strength in pricing, data-driven pricing sophisticated telematic and also very strong claims management really can create something unique. Sorry, I'm going to go in order one.

It's Ivan Bokhmat from Barclays. My first question would be on the strategy into 2030. I just want to clarify, perhaps, did I interpret it correctly. So the slide that shows your 8% CAGR in the past 5 years, you're saying that you're trying to achieve that same growth into 2030. So as a statement, maybe you could just confirm that. And secondly, on the trajectory of those earnings, as far as I understand, for 2026, you're talking about flattish numbers and then it would imply more of a hockey stick trajectory in later years. So perhaps you could just talk a little bit about how this trajectory might look like, where the acceleration will come and maybe specifically on the U.K. Motor, that cyclical target where you would grow 5% through the cycle over time, when will that time frame apply in this particular case? And maybe one final small question. The partial internal model, the -- if you apply imminently, do you think you will get the regulatory approval by year-end? And what does it mean for some extra capital decisions?

Yes. So I think what we're seeing here is 3 things. As you know, we normally don't give very precise guidance on long-term or medium-term earnings. But what we're saying is that there are 2 very clear revenue for growth and profitable growth in the future. Our core market, that is U.K. Motor, is a market where we have a market-leading business, we expect to continue to grow across the cycle. We'll continue to do at the right time and with the right choice and the discipline around pricing. But we'll continue, as we've done in every single cycle since Admiral was founded. We'll continue to grow our U.K. business from a larger base and retaining very, very strong margin. We also have another leg that is our other lines of business, Personal Lines and U.K. Insurance, Europe and Money, and we're planning to grow across all of them indistinctively and also increase margin for those business combined. So if you take those 2 things together, we expect to increase shareholders' returns over time without having necessarily put a specific date because there will still be some cyclicality. But the other preservation is with increased contribution from other lines, although there will still be market model cyclically to impact our results. We think we are gradually, over time, reducing the dependence on a single cycle. So that's the key message.

Internal model? On the internal model, we do expect to submit our application for approval very soon. The time line for review of that is not fixed. And as you can imagine, it's not a short read. So I think we'd update on the outcome of that at the appropriate time rather than comment on how long we expect it to take. If and when it's approved, you can be sure we'll be trying to use it around the business to optimize and things like that. And again, we'll talk about that at the right time.

Sorry, you mentioned something also about the shape. And as I was saying, there is still crack in the market. So as Geraint suggested, we do see a different path across the next few years. So we'll grow through the cycle. But next year may have a different impact on our growth ambition than the year after that. So that's -- but that's very normal. That's what we have done in the past, as you've seen in the slide that Alistair projected. We tend to grow when it's the right time when underwriting margin are healthier and will continue to do so. Sorry...

Ben Cohen at RBC. I just wanted to ask a few things on the U.K. Motor business. Firstly, would your central assumption be that you would be able to match claims inflation through the course of '26? And could you make a comment as to what you've seen in the market reaction as you've tried to put through or you have put through some price increases at the beginning of the year? And the third element, could you just remind us what happened to claims inflation in Motor kind of post the Ukraine, Russia invasion, just to maybe give some sort of comparison with maybe where we are now in terms of the situation in the Middle East?

Al, take it the first and I'll, sorry...

Yes, sure. So in terms of managing through the cycle in 2026, we're expecting claims inflation, as I mentioned, to be towards more normal levels, so mid-single digits. We'll be looking at that. We'll be looking at elasticity within the market. We'll be thinking about average premiums and continuing to price with discipline. As you saw in 2025, we did that and we've -- as Geraint said, we're very happy with the profitability on that yet. It's not as strong as '24, but it's still strong. So that will be the same approach that we'll take to 2026. As Milena said, reiterated, we think that's the right approach to optimizing both value and growth over the medium term. So far, in terms of market at the start of this year, we're seeing different strategies from different players. But broadly speaking, I'd say that market premiums have been relatively flat. But as I say, we've started to increase our prices at the start of the year. In terms of claims inflation post the Russia invasion, there was a lot of disruption to the supply chain and that was one of the impacts that caused higher parts, vehicle inflation as well as supply chain constraints. We don't think it's a direct parallel to what we're seeing at the moment. In terms of the disruption that we're seeing at the moment is more about oil and fuel prices not directly related. But I think as you're inferring, it increases supply chain or the geopolitical instability increases the risk of that. So that's something we'll be watching very carefully through our supply chain.

Will Hardcastle, UBS. First of all, I'm going to embarrass you, Geraint. Thanks very much for your help over the years. There's been some journey in the current role. I've always really enjoyed our interactions, some of them quite lively. But you've always been really helpful. So good luck for future endeavors, including the Admiral roles. Next, on to the questions, I'll ask you the tough ones now. You booked 2025 under -- undiscounted booked loss ratio at the 78% or 85% undiscounted. That's an average number. So I'm assuming the exit was slightly worse, given the shape of the pricing last year. I guess prepricing in excess of inflation, pre-percentile shifts, how roughly, where is the starting point essentially for that '26? How much worse than the 85% should we be thinking? And then moving on to something a bit bigger picture, doubling of the non-U.K. Motor business. It's quite non-Admiral to give a target like this. I'm sort of intrigued as to the logic, the thinking behind being -- it must imply a lot of confidence behind it. Does it imply any slowdown at all of top line and sort of extraction of the benefits you put through? Or is this just a better hope and a direction of Travel from here?

I'll take the first one. So the first one was about the exit loss ratio. So as you pointed out, the undiscounted booked loss ratio is at 85%. It's not -- it's higher, obviously, than '24 that was an exceptional year, but it's not unusual if you look back at previous years, hence, the comments about good profitability. As I said, we managed rates through the year, paying very close eye on claims trends. And in the second half, we were flat. And I think that means that the exit loss ratio was slightly higher than the overall, but not significantly so. And as I also mentioned, as we started in '26, we made some adjustments to price to make sure that our starting point in '26 was in the right place.

The second point is about confidence about the other line of business. It's a mix of 2 things. First of all, is the momentum. If you look at where we are, momentum and maturity, if you want, of some of our other lines of business, we have fantastic 2025, doubling profit in Admiral Money, strong recovery in Europe and return to profitability with confidence in the prospect and the future. And other lines of insurance in the U.K. also deliver a stellar year. I think we are reaching a maturity in those areas that allow us to continue to grow and increase margin over time. And it's also, I would say, a reflection of our strategy because the strategy is also very much about compounding. And so what I mean is that we have a few things -- we're focusing a lot on is our proposition to multi customers, very important because customers with more risk have better retention, have better loss ratio and better NPS, tend to be happier and stick longer with better results and also better experience for them. So I think there is momentum in terms of the multi -- our journey to multi-products are probably later than some other player in the market because historically, we were very much a U.K. Motor story. But as this business grow, there's a lot of potential there. That's a very interesting opportunity, but also transferring some of the strength in Predictive AI, for example, across all the business is also another big driver of value. And so if you merge those 2 things, plus economy of scale, plus potential benefit, we do see a momentum that will allow us to both continue growing and deliver more profit. And so I think it's really very much a reflection of our strategy and a bit of the switch of focus from growing individual business that are stand-alone interesting to really compound the benefit. So it's just meant to be that over time. We'll continue to be disciplined. So we follow cyclicality. There's cyclicality in the U.K. Household market that will take into account, but we do think we can achieve both growth and higher margins.

Thomas Bateman from Mediobanca. Just a quick question on the reserving. I was surprised to see the PYD quite low, but then the risk adjustment percentile come down. Could you just explain now how that's working? And the second question is actually a follow-up to Will's on -- I guess, on Europe. I take your comment of Spain is breakeven now, but I guess you have been working on that for a while and there is more confidence there. And you alluded to AI platforms, et cetera. Have you been able to launch on any of those AI platforms, either in the U.K. or in Europe yet?

So do you want to take the risk adjustment? Costi maybe briefly comment on the confidence in Europe and Spain, and I will pick up maybe on the AI.

Actually, Tom, if you split out the Ogden impact on last year and compare year-on-year, you get the same percentage. So a fairly strong level of releases coming through year-on-year. The percentile was really ever so slightly down. I wouldn't say that we'd notably dropped the risk adjustment strength. So it's a very strong set of reserves and continued pretty consistent releases coming through basically in line, I think, with what we've guided in that kind of 10-ish range over the past couple of years.

So on Europe and then on the AI point, so basically, yes, as Milena mentioned, there is a good level of confidence. It's a large opportunity where we are making very good progress on several fronts. And what is giving to us the confidence is that we are keep trading at very good margins on the direct business and we are seeing very good progress coming from the distribution-diversification initiatives, which will help us to target much larger opportunities. And so once also those initiatives will turn into profitable ones in the medium term, we expect our overall margins to expand. In addition to that, we expect a more efficient reinsurance agreements to provide benefit in the medium to longer term. Clearly, there is also an element about the competitiveness on the expense side on the efficiency. And we're also there making good progress, and we are testing also some more advanced gen AI tools and models. On this front, it's more early days, but early signs are very promising.

I think more in general on AI, there is a lot of opportunities. And a lot of that is focused on improving efficiency internally. It's about improving automation, increasing speed of servicing the customer and so forth. I guess your question was more referred to the distribution element. So how customers interact and choose insurance. And if you think about Admiral from very early stage, we've always been a bit of a forefront of disruption and distribution, and we were among the first direct player in the market in the U.K. We were the first to have an Internet-only brand, Elephant, let's call it in U.K. We're the first one to embed price comparison site with confused.com and so forth. So it's obviously something that, as you may imagine, we're very close -- we're working very close to price comparison site as they may embed more gen AI technology in their way of interacting to customer and distribution and adapting our website. We also have interesting pilot in part of the business, like gen AI embed chatbot in Veygo and other initiative across the group. So something we're very, very close. Now if you ask me, do you think this is going to be a very big disruption in U.K. Motor in the short term? I personally don't think is the case. I think there will be a different way of interacting with the customer. But the value proposition to a customer, how much you can save by shopping on price comparison site on Motor insurance is huge. It's hundreds of pounds sometimes. So I don't think it's going to be the first market where we see a lot of change. I think it's early to say. Everything is very nascent at this stage, but I think there are markets where this could be an acceleration to direct and that could be the one where customers are more used to speak with an intermediary, for example. So it can be commercial lines, it can be Europe where direct is not picked. But at this stage, it's very early to say. As for us, we try to be close to everything and work and progress on all the fronts at the same time. I think we had 1, 2 and 3, yes.

Carl Lofthagen from Berenberg. Just the first one on the U.K. Home book. I think we've seen kind of continued expense ratio improvement as you've gained scale and you're now running the business at a combined ratio in sort of the mid-80s. Is that sort of the level that you're sort of happy with? Or are you willing to trade some margin to take market share as you've kind of said you want to be a top 3 player? And then the second question is just a clarification on the share count development. I think if I just look at the basic share count, which decreased by GBP 5 million from GBP 306 million to GBP 301 million in H2, but diluted went up GBP 1 million. Presumably, the shares you're buying back for the share scheme shouldn't impact the share count. Just I guess for modeling purposes, I mean, how should we kind of think about that, excluding sort of any sort of additional buybacks, et cetera?

Sorry, Al, you take the first. Geraint will take the second.

So on the Household expense ratio, we've had an advantage in terms of expense ratio for Household for some time. But as you highlight, it's an area of focus. As the book grows and we get more renewals to customers, that helps in terms of expense ratio, but we're also focused on driving improvements. For example, Milena talked about how we can use gen AI for both customer experience and efficiency. So those are areas of focus. In terms of the combined ratio range, we think about Household similar to Motor, where it's about optimizing for value over the medium term. But as you're alluding to, we're a bit more biased towards growth on Household than margin. But we -- I think the 80% is good. I think we talked a bit about a range when we did the deep dive, so we're not sort of sticking to a specific target. But we'll do that in optimizing for value and growth over the medium term.

On share count, Carl, if you look at the -- in the back of our accounts on Note 12, you got the update the number of shares that were in issue every year. It's been GBP 306 million odd for a couple of years since we stopped diluting for the share plans. The GBP 301 million is actually the number that's used in the EPS, and that excludes some of those shares that are held in the trust. The share purchases per share plans won't adjust the number of shares that were an issue, obviously. They will go to employees. The share buyback and cancel, obviously, that will reduce the number of shares in issue. So the purchase for the share plans doesn't change the number of shares. Buybacks obviously will.

It's Derald Goh from Jefferies. Two big picture questions, if I may, please. So the 4% sort of EV -- sorry, AV penetration rate by 2035, that's an interesting number. I'm just keen to hear what are the main variables that might sway that number. Essentially how prudent is that 4%? And maybe if you could also say what were your projections 10 years ago? How does that compare to what you had 10 years ago, let's say? And then secondly, going back to distribution, you mentioned there's potential to disrupt there. Maybe could you speak to your past experiences? I know you've been trying to push PCWs outside the U.K. Some places are more successful than others. What might be different this time that would allow you to be more successful with whether it's AI or changes in customer behaviors and what not?

Sure. I think I'll take the first and second, but Costi, if you want to add anything on distribution outside U.K., it would be great. So EV, this is referred to this is referred to relatively common forecasts that have been out like the World Economic Forum and a lot of other organization. And I think the number is quite aligned. You may look at car sales in terms of car park is 4% because there is quite a lag time from new sales to fit into car park. The average -- the median age of a car in U.K. is 16 and probably the average is 11 years. So it takes time as the new model gets released. I think you're absolutely right. So it's still very much of an estimate, and there are a lot of influencing factors. You need to get, first of all, regulation in place, infrastructure in place, technology and investment in place and customer appetite in place. So there are a lot of things that can contribute and can go both direction. If we don't see the simultaneous development of all these 4 areas, it's difficult to imagine a world in which people will really freely just use autonomous vehicles car on the roads. So I cannot tell if it's prudent or not. But I would say this is really the majority of the -- I think everybody agrees that it takes some time, both because there are some hurdles and because there is time for the car park to evolve. And I think also take the chance to remind that this is about L3+. L3 basically means when the driver can take the highs off of the wheel, but still need to get in, in 10 seconds. So it's not really full benefit of EV in terms of, for example, liability shift and so forth. Anyway, it's very nascent now. So you asked me how this was versus 10 years ago? I think this is a story that we see in all the technology disruption. If I go back 10 years, this was supposed to be earlier. And so what happened is that this projection tend to shift. If you ask me how it was compared to a few years back, like maybe 3, 4? What I would say is not very different, but we see a slightly later adoption, but probably faster. It very often happen with every technology now that it takes longer, longer, longer, but then it can be more. So it's difficult to say, honestly. It's very, very early stage, and the U.K. is a bit behind the U.S. in terms of regulation and so forth. Second question was on distribution. I don't know, Costi, do you want to kick it off on?

Yes. On distribution, the -- well, the European markets, as you know, are very different. When we started our businesses a few years ago, I think direct was about 5%. Now on average, it's getting closer to 20% -- between 15% and 20%, so direct is still growing. And therefore, if a more AI-driven disruption would happen, we could say that being a leading player in those markets that will put us in a nice place. At the same time, price comparisons, you're right. We try to educate the market and to push more digital growth to accelerate. It didn't happen at the speed we expected. And at the moment, as I said, direct is still growing, price comparisons are doing nicely, but not at a supe- fast speed. At the same time, traditional are still a very important channel, which predominantly is the main channel, which is linked why we decided a while ago to start to diversify the distribution and on how to win and how we can be confident basically because the right to wins are exactly the same of direct. So risk selection, customer experience and lean operations. And the moment you demonstrate that you can replicate those, then you can win in that market and our results that are coming through are making us confident more and more that we can achieve this.

Andreas van Embden, Peel Hunt. Two questions, please. First one is on ancillary sales. I saw that the average sort of revenue per vehicle in the U.K. has come down from GBP 76 to GBP 71. Just wondered what your outlook is for this in '26 and '27, particularly not only on the installment income because I assume you've lowered your APRs, which has brought down the average premium per policy there or per vehicle there. But also on the other ancillary sales, whether you're seeing any pressure on those fees and commissions? And on -- the second question is on price velocity. I think you mentioned that. I just wondered what would you exactly meant by that in the U.K. and whether extending it means changing pricing more. I don't know whether you do intraday pricing or not in the U.K. But whether -- with that velocity, by how much could you extend it? And how important is that to maintain your competitive position, particularly through PCWs in the U.K. as the market becomes more competitive?

So it's Al. I'll take the first one. As you mentioned, the main driver of the change in the revenue is premium finance. It's worth noting that there's 2 impacts there. So we did reduce our APR through 2025, in line with the cost of funds. But also, we saw our average premiums coming down through the year. So that also impacts on the other revenue per vehicle. So in terms of our APRs, they're very competitive at the moment, not necessarily anticipating any changes, but we'll continue to assess for fair value on that product. In terms of average premiums, as we've said, we're expecting the market to turn and that will lead to -- and we're increasing our premium, so that will flow through as well. I don't think there's anything else of significant note to call out on other revenue.

So on the price velocity, I think if we step back just a few years back, a lot of the pricing was done through SaaS or XL, and we could put price change in production overnight, we can have a meeting. If you all ask me, I'll let Geraint decide and make change almost overnight. That is still the case. And I think it's an advantage. And I think it's really rooted in the culture and the closeness of the management team to price into claims trend and that relevance that we give to loss ratio all around Admiral. But our pricing is more and more based on machine learning and predictive AI models. And this, of course, is not something that you change overnight because it's more complex, require more technology, the process to upgrade and renew the model just takes longer. And we've done a fantastic -- like a lot of work in the last year or 2 to really bring this time from ideation to change in production much shorter and shorter. I think there is still a bit we can go for. And so we're very close and we have a strong capability, more than 120 models in place, GBP 100 million of loss ratio incremental value. So we start from a good point. But I think there is more we can go to increase this even more. But the biggest opportunity in my mind is to extend this also more and to extend these trends more into other lines of business. So that's where I see the excitement. And to do that, we appointed this year a new group CDO. She did a great job in Europe to set up a new data platform. She just -- she's coming over and she came over actually a couple of months ago, and she's going to help to increase even more this ability. So I think we're really in a very strong position, but want to go. We're very keen to be as fast as we can.

[Operator Instructions] We will now go to the question. And the question comes from the line of Vash Gosalia from Goldman Sachs.

Hopefully you can hear me. First of all, apologies for not being there in person. I have 2 questions, please. The first one is just on something you mentioned on your 2026 profitability. So if I heard you correctly, you said you plan to move to the middle of your confidence interval through 2026. And you've obviously also said you expect flattish profitability. So can I read that as your earnings in 2026 actually being supported by PYD just to offset some of the weakness in U.K. Motor? And any sort of color as to how or why that might be different would be really helpful. And the second question, a slightly longer term or big picture question. So you've obviously sort of alluded to you having a lot of sort of advantage on the cost ratio front. And you can obviously leverage AI and gen AI to improve that. But I'm just trying to think if the technology essentially democratizes the use of AI, wouldn't that allow your competitors to actually gain advantage quicker and narrow the gap to you? So any sort of color or comment on that would be helpful as well.

Will you take the first?

Yes. So we would expect to start to release our -- reduce our risk adjustment percentile from its current near max level towards the middle of the range during 2026. I would reject your assertion of weak U.K. Motor. I think U.K. Motor profitability for '25 is strong, but slightly less strong than the extremely strong 2024. So we think there's good profitability to come on 2025. And obviously, that starts to feed into the accounts in 2026. PYD and reserve releases are a constant feature of our income statement and profitability. But you are right to expect as we reduce the risk adjustment to some extent, obviously, that contributes to profitability in 2026 versus 2025. And if you do the mix, flat profitability, but higher profits from other lines of business means slightly lower profits from U.K. Motor, but from a very high start point. So I think -- Vash, I think that was the nature of the question, right?

Yes. I mean -- and apologies if I said like weak profitability. I meant more direction-wise. But yes, you've answered my question.

No offense taken.

On the second point, it's a very good question. It's a very good question. I think every technological evolution, data evolution and if you think about digitalization, automation, migration to cloud, more and more like technology, it becomes more and more a commodity itself, but the way technology is implemented is very differentiating and it becomes more so as we move along. And so a big decision on AI is how you do it. A big driver are how much this is adopted. So you can deploy gen AI tool to have all the organization, but how much is adopted, how is adopted is a massive driver of how much efficiency benefit you can drive. I think we start in a great situation because we tend to have a very strong culture, very transparent and people with good expertise that are really, really keen to do what is right for the business. Governance is another differentiating factor, how you govern what you put in place and how you make sure that it's solid, is stable, how make sure that the model learn over time. I think an appetite for innovation, bottom up as well as top-down is also very important. So I think a lot of the element that plays here are a culture and also the ability to do it faster, better and cheaper than others. And it's still early to say, but I think we are well positioned to achieve that. We have a last question.

It's Shanti from Bank of America. You just touched a bit earlier when we talked about the internal model and how that could give you a bit of flexion for M&A. Historically, I guess, you guys have partnered with names via Pioneer or you've had a relationship with the existing names before you would kind of move forward with an M&A transaction. What kind of skill sets or regions, if you were looking at that, would you be thinking of?

It's the last question. Do you want to take it?

It sounds like I've explained this badly. I wasn't really referring to the internal model coming in, giving us more firepower for M&A necessarily. I understand the point of the question, but I think funding small M&A to retain profitability is one of the options I would talk about. I know you should cover where M&A might play a part of it.

Yes. So as I said, our plan -- our history of successful organic growth, and we're excited about the plan we have on growing organically. We will look into opportunity mainly to accelerate diversification, I would say. So as we've done with Flock as we've done with More Than, we'll look at accelerated diversification in other lines of business in insurance in the U.K. or in Europe where we need more scale. But we stay open, look and see, but also very, very focused on our organic growth plan and consider different option on how to eventually tackle the challenge. Thank you very much. Thank you for your question, and thank you for your time. And we'll be around a few minutes if that can help. Thanks a lot.