David Stevens - Group Chief Executive Officer Geraint Jones - Group Chief Financial Officer Cristina Nestares - Chief Executive Officer, UK Insurance Milena Mondini - Chief Executive Officer, Europe Insurance

Greig Paterson - KBW Thomas Seidl - Bernstein Arjan Van Veen - UBS Andrew Crean - Autonomous Dom O'Mahony - Exane BNP Paribas David Bracewell - Redburn Waja Rizvi - Deutsche Bank Andreas Embden - Peel Hunt Nick Johnson - Numis Iain Pearce - Berenberg Dhruv Gahlaut - HSBC

Good morning, and welcome to Admiral's 2017 Full Year Results Presentation. Thank you for battling through the elements to join us today. And thank you to Bank of America Merrill Lynch for hosting this morning. Now Admiral started in January 2000 - January 1993. And so we're now 25 years old and the 25 logo that you'll see throughout the presentation reflects that. To celebrate that, we've decided not just to take the normal six to 12 month backward looking, forward looking perspective on the business, we've also thrown in some longer term perspectives as we go through the presentation, and I'll start with one of them. There's a picture of the founding management team when we were 57 staff, one country, one brand, new customers, slightly stiff and dour looking bunch. And there we are today, nine and half thousand staff, five and three quarter million customers, eight countries, 18 brands. So as you see a much bigger management team both collectively and sadly individually. Today's presentation will be starting with myself and then Geraint Jones, the CFO will give a group overview before Cristina Nestares, the UK Insurance CEO talks about our UK results. And I will come back to talk about the US, both insurance and price comparison, before Milena Mondini, Head of European Insurance talks about European Insurance and price comparison and then I'll come back for a brief wrap up. Over to you Geraint.

Thanks, David. Good morning, everyone. I'm going to talk to some of the highlights of the results, covering the turnover, customer growth profits, move on to talk about capital and dividends. And I'll finish with a brief update on Admiral loans. This is our usual highlight slide to kick things off. On the top we show customer numbers turnover. We're going to go into some detail on both of those figures shortly, so I'll just highlight double-digit increases in both. Again, all parts of the group growing particularly in households in the UK and generally things outside the UK. Our profitability, you'll have seen we reporting £405 million of pre-tax profit, earnings per share of 117.2 pence. Those are both record figures. Of course, the comparisons are clearly very impacted by Ogden, the hit on 2016 and so the plus 43 and plus 49 need to be seen in that context. But it is good to get back to an upward trajectory and profitability. A similar story for return on equity, 55% for 2017 a very positive number, back in line with our usual historic performance. And our full year dividend is 114 pence per share, that's 11% up on 2016. If you exclude the additional capital return we paid in 2016. And finally in the middle, solvency ratio remains very strong still up over 200%. Will be a bit lower than it was at the end of 2016. These are the top line figures from across the group. The picture looks pretty similar to the recent reporting periods, strong growth in all segments, lots of green arrows. I particularly point to nearly 4 million motor customers in the UK, which covers cars and vans, as comfortably gone past 4 million since the end of the year. We see more than 40% growth again in household customers and turnover, really pleasing figures again from household. And in our international insurers, the turnover was up by nearly a quarter heading towards £0.5 and customer numbers were up by a fifth, breaking through 1 million customers outside the UK, continued strong momentum, and more than 10% growth in the turnover from the price comparison businesses as well. Moving on to talk about profits. David mentioned Admiral turned 25 earlier this year. And so I wanted to show the full profit trajectory since that first year back in 1993. Hopefully you'll agree, it's an enviable track record. Profits have increased in all by two years. And of course, one of those years which was 2016, was due to the Ogden impact. And you can see that in the - in the red dots 2017 profit of 405 was over twice as big as the profit 10 years ago and was a huge multiple of the number 20 years ago in 1997. But let's look in more detail at 2017. Here we compare the results for each of our segments, compare to the previous year. The UK profit as you can see £466 million is a record. As I said earlier, the comparison is distorted by Ogden [ph] so I'll talk more about the UK separately. The international insurance results improved again to a £14 million loss compared to £19 million the year before. That's mainly a factor of higher profits at ConTe in Italy. A fourth year in a row in profit for ConTe. A lower loss in the US, and that latter result coming despite the hit from the hurricane in Texas in 2017. The price comparison results improved to a £7 billion profit from a £3 million profit. The main features there being reduced profits at Confused.com. Great results from Rastreator and LeLynx in Europe, and reduced losses at Compare.com, which it's marketing breakeven target for the year. Last up is our other segment and you can see that that was up a bit in 2017. Variety of reasons and £9 million of the movements you can see explained on the slide. That's the first year of the Admiral loans P&L, which is negative four and 2016 included an unrealized gain on foreign exchange forward contracts. There is a breakdown of those costs in the appendix if you are interested. As I said 2016 is not a good basis for comparison, so this slide looks back a year further to 2015. As you can see, the UK Insurance profit is £22 million higher than two years ago. £3 million of that improvement comes from household, where we have a much bigger book, a better expense ratio and a better loss ratio, which is a pleasing combination. And therefore, Motor profit is £19 million higher than 2015 despite earned premium being by 20%. Key point to remember, of course, as you can see on the slide that the 2017 results is prepared on the basis Ogden stays at minus 0.75%. There was a continued drag on profitability in 2017 as a result of that change, that's about £40 million. 2015 obviously had Ogden at plus 2.5%. So the Ogden change is a key factor in understanding lower profit in 2017 is £20 million or so higher than it was in 2015. This slide also helps illustrate progress in International Insurance, where since 2015, we've doubled our turnover and reduced the loss by a third. More detail on our performance of our various segments shortly from my colleagues. Change in subject to look at capital. Familiar looking slide here, shows our solvency ratio still very strong, again, over 200%. The main movements since middle of the year, as you'd expect, the second half profits deducting the final dividend and some increase in the capital requirement, which is mainly due to the growth in our balance sheet, but also the expected volumes we see in 2018. There is no change to report in the basis of the calculations. We're still on the standard formula, plus the capital add-on, and we are making progress towards our internal model application later on this year, hopefully, getting approval in 2019. We still think at this stage 150% is a good indicator of the upper end of our target range once we get that internal model approval. Of course, we do recognize that 200% is somewhat north of 150%. And to that point, I repeat the message of six months ago, which is until we've got certainty over the internal model capital position, we won't talk more about any potential surplus capital position. And what I mean by certainty over the internal model position is getting the internal model approved. Let's take a look at dividends. As you can see, we're proposing a final dividend for 2017 of 58 pence per share. In line with our guidance, we've looked to hit a payout ratio in the 90s, and 58 pence per share is 97% of the second half earnings, still leaving us with a 205% solvency ratio you saw on the previous slide. We repeat the formal dividend policy on the slide here, and I'd also reiterate that unless or until we say otherwise, payout ratios in the 90s is a good guide for our dividends going forward. You'll need to be on the record on the 11th of May to get paid on the 1st of June. The half-year, David talked about our aspirations to grow in areas beyond car insurance, and I'll just give a brief update on one of those areas that we hope will be an important source of growth and profit in the medium term, that's of course, Admiral loans. You can see that yearend we had nearly £70 million of loans receivable on the balance sheet. Recently, we've gone through £100 million of lending to our customers. Given the relative size and maturity of the business, we're not going to go into too much detail at this stage, but some of the key messages in terms of our products and our progress settle on the slide. We expect future - further growth, sorry, in the balances during 2018, assuming we remain happy with the performance in the KPIs. We are, of course, aware that there are potentially uncertain economic times ahead and our focus is on building a prime book. We are projecting a material change in the P&L guidance for 2018 is between £5 million and £10 million, up from £4 million in 2017. That's it for me. I'll leave you with some of the key messages, continued strong growth in all parts of the group. Record profits, they're still held back in the U.K. by Ogden. Very pleasing early progress for Admiral loans. And we've maintained a very strong capital position, whilst paying out basically all of our second half and actually full year profits to our shareholders with dividends. I'll hand you to Cristina who will talk us more about the U.K. business.

Good morning, everybody. I will talk to you today about the results of the UK Insurance operation. I will focus on the Motor results, and talking also about the market. And I will also cover household. But let me first start with a bit of historical perspective. You can see here at ad, a yellow pages ad from Bell from many years ago when actually Yellow Pages was our biggest source of business. Now, about 20 years later, you can see a very different ad for Bell. Bell is now a Telematics only brand that we sell via price comparison and it's helping Admiral to stay as a Telematic leader in the market. Now let's look at the highlights of the UK Insurance in 2017. It has been another good year of growth. In terms of units, our motor book has grown by 8%. And that includes a 5% growth in our car units, reflecting a slowdown in the first quarter of the year due to price increases after the Ogden announcement. And it also includes 120,000 vans that we put in the book at the yearend in the first year of operations for van. In terms of turnover for motor, it has grown by 13%, and that includes an increase of 11% in car, reflecting our price increases. In the case of household, both customers and turnover have increased by 41%. As for profits, Geraint has already talked about them. Moving on to claims in the market. The frequency of small bodily injury claims registered in the MoJ portal has continued trending down. You can see here in the graph. It went down by about 12% in '17, following a 19% reduction in '16. So why is it going down so much so much? Well, we think it's a combination of safer cars, but also change in lawyers' behavior. The prospect of a reform might lead to some lawyers to choose to focus in other sectors and less on motor. For 2018, it's hard to predict, but we'll work on the basis that we are not going to see such reductions in frequency. On the right, you can see a graph about accidental damage in the market. And you can see that the trends for '17 and '16 has been very similar, flat frequency, an increase in severity, reflecting mostly an increase in parts. For '18, we're expecting a continuation of this trend. So this is the market data. In the case of Admiral, our own claims experience for both small BI claims and accidental damage has been consistent with the rest of the market. In terms of large claims, it's actually much harder to comment on what is happening in the market because Ogden is causing a lot of disruption. What we can tell you is that for Admiral, 2017 frequency of large BI claims appears to have been okay, especially in the context of Ogden. So let's talk about market premiums. First graph is the ABI data, reflecting increases but also taking into account that it includes an IPT increase in the middle of the year. The graph on the right shows Admiral times top index to the last quarter of 2016. So what you can see in this graph is that depreciation at the times top of Admiral at the end of 2016 and at the end of 2017 is actually quite similar. This indicates that Admiral has put prices up in line with the rest of the market. However, the timing of those price increases is slightly different as we put prices ahead of the market at the end of 2016. Now before talking about what our price is going to do into 2018 in our view, I want to talk about access of loss. As we all know, the cover of access of loss has increased its price this year. And Admiral, what we did in 2017, as a reminder, we put more cover to cover for the shock of Ogden because the prices in the market were filing our view. So in 2018, we have gone back to our long-term average of around 8 million to 9 million but at higher cost. So that's in terms of access of loss. Let me talk about prices for 2018. Now three things that could affect what the market is going to do. First, access of loss. So I just mentioned that prices in the market have gone up. However, for what we have seen in the market, we haven't seen any significant increase in market prices so far. Second thing that could affect prices is the claims environment, we have seen relatively benign in 2017, and we expect a continuation during 2018. And finally, we think that the prospect of the reforms, as and if they become more concrete, could have an impact on prices and some player might start decreasing prices ahead of the reforms. So overall, we don't expect a continuation of the price increases in 2018. Moving on to the ratios. This is a slide that we have been showing for a number of years. You have Admiral data in blue and the market in red. However, as the market data is not being updated, it's becoming obsolete. So we might stop showing this comparison in the future. You can see the data for 2016, which we have taken from the load estimate. So for Admiral, the loss ratio have been booked to minus 0.75 discount rate, and as you can see, there have been some development in 2016, '13 and '12. These decreases are less than what we have traditionally shown. And the reason for this is that Ogden is causing disruption in the settlement partners of large claims. In this environment of uncertainty, we prefer to take a cautious view. In terms of expense ratio, you can see that we continue to hold a significant advantage over the market due to our focus on cost control. Reserve releases. Well, U.K. car reserve releases continue to be a strong feature of our results, with several releases in '17 were 21% ahead of the long-term average of around 15%. The margin in booked reserves remain prudent and significant, but it's slightly smaller in relative terms than the one at the first half of 2017. And going forward, we expect a continuing significant reserve releases if claims develop as expected. So this is it for motor. But before moving on into household, another bit of historical perspective. In this slide, you have the Other Revenue per vehicle. Only a few years ago, it was £84 per vehicle and now it stands at £64, quite stable in the past few years. So the reasons for the decrease are twofold. First, the ban of referral fees. And secondly is the fact that we have improved our products, which actually means a better outcome for our customers, but lower margin for Admiral. So in terms of households, I think the most important feature of our result has been the strong growth of 41% in the time where the rest of the market is not showing a lot of growth. So what are the reasons for this increase? I think there are three. The first one is the fact that price comparison, which is our main channel, has continued to grow, as you can see on the graph. The second reason is that there have been price increases in the market. Admiral has also increased the prices of its household book, but it has lagged the market, which has allowed us to grow. And the third reason is that beyond price comparison, we have increased our direct offering. We offer a multi-cover policy that allow us to cross-sell household policies to our motor book. So strong growth in units and also improvement in our ratios. You can see on the graph on the right, our expense ratio improved from 34% to 30% during the year due to increasing scale and a bigger renewal book. And I think it's particularly significant that we hold an advantage versus the market, which we believe stands at 45%. In the case of the loss ratio, we have seen an improvement during 2017. The good weather has helped, definitely nothing like today. So going forward, for household, we expect the continuation of the growth, possibly not at 40%, and we also expect an improvement in our ratios. So I would like to finish by talking about the customer, because these results are only possible because of the trust that the customer give us. Something very concrete that we have done to improve the customer experience during 2017 is to improve the feature of our products. And we are very proud to say that today all of our products hold a five-star de facto [ph] rating option. So that includes our car, our telematics, our household and also the two products that we launched this year, in '17, that is bank and travel. So in summary, 2017 has been another good year of growth for the Admiral UK Insurance business. The car insurance market has seen increases and Admiral has put increases in line with the rest of the market. For 2018, we do not expect prices to continue increasing in the market. And we have enjoyed a very good year for our household book with good growth in units and also improving our ratios. So that's it for the insurance business. Now over to David.

Thank you, Cristina. So before diving into the U.S. I just want to take a longer-term, bigger picture perspective on international. 15 years ago, we were active in one country, now we are active in eight countries. I'd like to take this opportunity to apologize to the people of Sardinia [ph] for their exclusion and assure them that they can buy policies from ConTe. Going on to the U.S. Elephant had a year of growth, 8% up in terms of number of cars covered. And a year in which the losses shrunk both in absolute and percentage terms, Admiral share of losses falling from $21 million to $16. What's driving this? Well, the - primarily reason for the improving result and surprising is continued improvement in combined ratio. So here you see the Admiral combined ratio compared with the market in red. And Admiral's improved from 143% three years ago to 119% last year, against some degree of headwind from the market, which has been unusually difficult for the U.S. auto insurance market, with an increase from 103 to 109 over the two years 2016, and I think it's probably about 107, 108 in 2017. So Elephant's on a journey towards profitability and our projection show it coming to profitability and making a decent return on the investment to date. However, we do always run stress scenarios at the end of each year, and there are certain stress scenarios in which Elephant would not fully return the investment we've made. And so we have chosen, as part of our conservative approach to our business, to reduce the book value of Elephant on our book from 125 million to 100 million. Big part of the improvement in combined ratio has come from an improvement in loss ratio, which has been something I've talked about over the last two or three presentations. And I think you can see here the progress that has been made. I've pulled out Texas because Texas represents 50% of the business we write in the U.S. We are the blue line. And again, you can see a situation where we've made progress, notably in 2017, against a difficult market context, where the home market in the U.S. was going in wrong direction in 2016 Hale adversely affected the Texas market. 2017, our 76 was achieved despite hurricane Harvey, which would have affected us and actually the market as a whole. So as well as improvements in loss other, other reasons why I am very positive about Elephant around a refinement in our strategy. At the beginning of the year, I asked Henry to go to the U.S. full-time, along with Alberto Schiavon, one of our U.K. managers to see if we could sharpen the strategy in the U.S. And the conclusion of the work that they did is that we should focus more on one specific segment, and that's what we've done. One of the ways of looking at the U.S. market is around propensity to stay with you around loyalty. The U.S. market is very polarized between customers who move in and out of insurance, are very active shoppers, but often struggle to maintain a policy throughout the life of that policy, and so they cancel and take it up again. In the other extreme, there are multi-car relatively affluent customers who typically shop much less often but are much more loyal. And historically, Elephant has appealed across the board. On the left hand side, you can see the mix of business that we wrote in Q4 2016, where the red block at the top represents the low propensity to renew and the green block at the bottom is the highest intensity to renew. And as you can see over the year, through a series of changes in marketing message, marketing mix and customer proposition, we've moved the mix of business substantially away from the lower retaining books of the types of business to the higher retaining types of business. And our estimate of the impact on that in terms of ultimate retention will be that the book as a whole will have a 17% better retention as a result of this change in mix. Now that's great, but you would expect to pay a price in much higher marketing costs to appeal to the ultimately more valuable customer. So we are very pleased to have combined that switch with an actual improvement in cost per vehicle sale, which was immediate cost of vehicle sale, which was over $260 and is now under $230. Henry has now returned to the U.K., working part time. Alberto Schiavon has stepped up to become the Elephant CEO. Over to the comparison business. Andrew was here at the half-year, talked more about the business. It's been a good year. We've gone below $10 million in terms of our own loss. The panel continues to evolve, and I've highlighted two names, one is obvious Travelers, big brand, U.S. and globally and Encompass. You might be wondering why I'm highlighting Encompass, small branch, you've probably never heard of. It is actually the third brand in the Allstate's table. So there is Allstate's insurance and by far the smallest brand Encompass, which signed on and is now active in a large number of states. Not necessarily material per se, but maybe a sign of some progress in terms of persuading one - at least one of the big four that it's worth engaging with Compare. Top rank [ph] you can see the year has been a great one in terms of marketing efficiency with cost per quote having despite volumes up by a quarter and the quality of those quotes improving more than translating into buy clicks and ultimately buys. Still work to be done there and we are guiding to a loss in 2018 between $5 million and $15 million for Admiral. So continuing improvement in Elephant's results, a refinement in the strategy that gives us increased confidence in its long-term health, a reduced loss at Compare and a milestone achieved of marketing breakeven. Over to Milena to talk about Europe.

Hello, good morning, everybody. I'm here today to talk to you about our insurance operation in Italy, France and Spain on our price comparison site across Europe. Price comparison site is our major distribution channel nowadays also in Continental Europe. But it was almost inexistent 10 years ago. I will then focus more on ConTe, our Italian and biggest operation, while there are more detail about Admiral Seguros and L'olivier in the appendix 2017 was a mixed year for price comparison site. It was a year of investment in Confused. We invested in the driver win campaign to position our ourselves as the go to place for drivers. We also invested in other project as car buying and selling online and car financing. We believe that this new strategy will help us to differentiate ourselves. The market is still highly competitive, with more than £110 million of investments of the top four players in one single year. Profit decreased year-on-year as a consequence of the investment. In the last month of 2017, we started to see some benefit of our investment. And we've noticed a good engagement with the brand, but it will take time for the benefit to fully deploy. In continental Europe, we experienced another year of strong growth, both in France and in Spain. The starting position is quite different. In Spain, where Rastreator at 70% market share and is leveraging on its dominant position to diversify the business with particular focus on the financial vertical. This is a bit more mature market compared to France. In France, LeLynx is also one of the two top player in the market, but the market is less mature and there are new entrants that are in investing in TV. We do believe that this investment will help us the channel to grow. And it's important to remind that for us its' very important that the market move direct also support of our operation, insurance operation there. Moving now to insurance in Italy, France and Spain. 2017, as 2016, was another year of very strong growth. We grew 60% in active policy base and turnover, reaching 309 million revenues in '17 and 854,000 customers. We are consciously investing in growth to reach economy of scale in each country and the skill that we need we believe will help us to deliver material profit in each single business. It is important to notice that we did achieve this growth despite the cycle is not favorable yet, particularly in Spain and Italy, where price are expected to increase. At the same time, even more important, reduced losses overall. Losses went combined in Europe from 10 million to 5 million to 2 million in '16 and '17. Main driver of this was ConTe that had a year of record profit, but I think it's important to mention that Admiral Seguros is not far from reaching sustainable profitability either. And in L'olivier, we are still investing to reach scale, but this investment is not large at all. In this slide, you can see the key evolution - the evolution of the key ratio in Europe. The first thing you may notice is that loss ratio is in a very good range, while expense ratio is over market average. So why expense ratio is over market average? There's a mix of reason. The most important one is that we have a higher percentage of the new business of renewal book and we don't have the scale yet, particularly in France - in Spain and particularly in France. This is why, for us, reaching a decreasing expense ratio is crucial. And is critical and one of the most notable results in my opinion for the year is a strong reduction of five points combined in Europe, while growing at the same time. We are indeed reaping the benefit of our investment in brand awareness. We are realizing that material internal efficiency through more process automation and optimization. And we are reducing the cost per policy in every single country. Reduction of our expense ratio was indeed realized in every single country and we remain positive to continue this trend and in this direction in the future. Loss ratio combined in Europe has been between 70% and 80% in the last four years. We believe that's very good results. It's important to us because being able to transfer our competitive advantage in underwriting and pricing was one of the key assumption for us when we decided to launch internationally, and we think we achieved these results. In 2017, in particular, the loss ratio deteriorated, as you can see, but please remind that this number include reserve releases and 2016 was a year of extraordinary reserve releases, particularly for ConTe, while the underlying loss ratio is relatively stable. It's also worth to remind that we adopted a more conservative approach in booking in France and Spain as we are already doing in Italy and U.K. Results for the year was also positively impact by reinsurance contract, in 2017 in particular. Moving now to ConTe. ConTe had a great year. They had record profit. It was profitable for the fourth year in a row, it broke even on written basis for the first time, while in the past, we had to wait few years to realize the profitability on written basis because we have to wait for the loss ratio, ultimate loss ratio estimate to improve. We had a record expense ratio with six points decrease, and this was mainly driven by internal efficiencies and despite the strong growth and despite we are in an adverse part of the cycle, with the average premium in the market decreasing and now down 22% compared to five years ago. We also went live with the new IT system, Guidewire [ph] that is now supporting the majority of Admiral Group. And ConTe was scored second-best place to work in Italy for the second year in a row. But the most notable results of all is definitely passing the mark of 5 million customer, what a great way to celebrate the new year, 600 happy colleagues celebrating, 500,000 happy customers. I'd like now to conclude with a comment about Europe in general. Italy, France and Spain are somewhat different but very similar among themselves if compared to U.K. They show similar challenges and somewhat similar opportunity. And we did work in the last two years to make sure that we were increasing the synergies and we were sharing more best practice. In summary, trying to make sure that the value delivered by European Union, Continental Europe combined was greater than the sum of the individual country. One of the most notable effort has been investment in technology. We have a company in Spain called EUI General Service, that support the three operations in Europe. But as you may expect, the most important project is preparation to Brexit. We selected Spain as location for a new insurance company that will underwrite business for our operation in Continental Europe from 2019. The project is going well and according to plans. We appreciate that given the size of European insurance is increasing, there is interest in knowing more about those business, so we will be very pleased to address this later in the year in a dedicated Investor Day. So to wrap up, here investment in Confused, very strong growth both on price comparison side and insurance company in Continental Europe. Improving results in each single business, in each operation, with reduced combined losses overall and record profit for ConTe. And I shall now hand over to David that will talk to us about group strategy. Thank you.

Thank you, Milena. One last long-term perspective. We've talked a lot about things that have changed from 25 or 20 or 10 or 15 years ago. One thing hasn't changed, and that's our commitment to making Admiral a place that people enjoy working at. And this is demonstrated in this exhibit. In 2001, the first Sunday Times Best Companies to Work for survey came out, and we came in at number 32. We are one of only two companies that have come in the top rankings of that survey every year thereafter. And the latest survey was particularly gratifying, seeing us voted as the third best big company to work for. And last year, we managed 23rd on the Global Best Companies to Work For award, so it's lovely to see that culture being exported internationally. I'm just going to finish then with a quick restatement of our strategy. Very simple, it's the same that I described two years ago when I first became the CEO, it's to ensure the Admiral remains one of the best insurers in the U.K., to demonstrate that we could be a great insurer beyond the U.K. and to develop sources of profit and growth beyond insurers, briefly expand on each of those. So in the UK it's around serving more customers, which we've done spades [ph] in 2017. That's both in car and through new products like van and travel and household. It's about knowing our customers better, which we partly due through product innovation like multi-cover and telematics. And it's about keeping them longer, and that's through product evolution, the de facto five-star, the improvements in the product quality of the ancillaries and the improvements in the customer service delivery. Demonstrating that Admiral can be a great insurer beyond the U.K. is around understanding what it is about Admiral that's exportable and making sure that we also adapt to local market conditions. And what's proved to be exportable is talent and risk selection and a low-cost expense culture. Beyond that, we've had to do a lot of adapting, and I think the Elephant positioning towards particularly segments in the U.S. market through direct marketing is an important example. Claims handling in Italy is another example, where there is a lot that we've learned over time and I think we're now ahead of the market on claims handling in the Italian market. We are focusing on the existing countries in which we have insurance operations. We are not planning to launch into new countries. And we are seeking to accelerate the switch in consumer behavior, which as you saw from Milena's graph is a long-term structural trends towards direct and price comparison, which seem to accelerate that in certain markets by our investments in our own price comparison site. Lastly, in developing sources of growth beyond insurance, again, we seek to develop on our strengths. So we are looking for intangible services sold at a distance ideally. We are looking for products sold through price comparison sites ideally. If we are looking for products which have a high degree of risk selection is an important success criteria. And we are going to test relevant products in that universe, with the most obvious test being our launch into loans, most recently our expansion from unsecured personal lending into car finance, and that's going to be an important initiative as we go forward. So thank you for that. Here are your team 10 years ago, and I'm going to open up for Q&A. I know the question on everyone's lips is how the four of us managed to avoid the ravages of time so successfully after that last 10 years, but can we keep the questions to business, please. Before I take my first question, I just like to take this opportunity of thanking Karen Maguire, our Investor Relations Manager, for whom this is the last set of results and who - and the last roadshow. Sterling work over the last few years, she's moving on up to be the CFO at Confused. Congratulations, Karen, and thank you for all your help. The Queen is dead, long live the Queen, Mauricia [ph] is our new Investor Relations Manager, Mauricia Kosna [ph] a very competent person that will you have all hopefully get to know over the next couple of years. So over to Q&A.

A - David Stevens: If we go straight to the back. Sorry, Karen. No, no, you are in charge.

Greig Paterson, KBW. Two numbers theme, please, and observation. Could you just give us what the - as you typically do the December year-on-year rate increases in U.K. market [ph] including new business and renewal together? And could you just clarify exactly what your retention was on excess of loss in 2017 versus 2018? I'd also interested to know what your plans on, what you'd been doing in terms of automated and connected vehicles? And finally, just an observation. My view, your business model is very cryptic, the financial structure and by dropping, if I heard correctly, you're going to drop the ultimate disclosure, I think, that makes a cryptic business model more cryptic. There was a comment about dropping the ultimates.

Market comparison…

Then I take back what I said. So obviously not a cryptic business model.

Okay. Let me do the two easy ones and I'll hand the price increase over to Cristina, that's very difficult. On our excess of loss, Page 44, it's full detail of the history of our excessive loss. In terms of autonomous vehicles, we invest a lot in touch with what's happened technologically. We are active with start-up account and plug-and-play. We have one of our senior managers embedded in a tech-orientated private equity fund. We do a lot in cooperation with some of the start-ups. We also stay close to some of the tests which are going on around the U.K. in terms of autonomous vehicle. So far, we have felt that it is not commercially advantageous to invest beyond our investment in that area. We feel we can follow the developments in that technology adequately from the work that we do elsewhere. Cristina, would you like to comment on pricing?

Yes, so prices in the market, according to the ABI, have gone up by 8%, and that includes 2% IPT. We have put prices similar to the market.

That's December to December

That's December to December.

So 80%?

Around similar to the market, yes.

[Inaudible]

Yes.

Perfect.

Thomas Seidl, Bernstein. First question on home, great progress. You still run a 20 percentage point higher loss ratios there so is there any reason why you shouldn't have reached market average 50%, 55% loss ratio was increasing diversification in knowledge and data? And secondly on U.S. price comparison, you talked about the progress you made, can you talk about the market there? Are switching rates going up? I think it's about 75% retention, whereas in the U.S. switching rates going up. Those people who do switch use more price comparison? And how do the competitors of Compare.com fare?

You'll do home, and I'll do the U.S.?

Yes. So our loss ratio in the household book is much higher than the market. Several reasons. First, we are still learning of in four, five years, we're collecting more data, improving how experience - or sorry how we handle our claims. So we expect to improve our loss ratio. The second reason is in the market, the retention is very high. So a lot of the competitors have very large renewal books. So we might not be able to get very quickly to the average of the loss ratio in the market. The third reason is price comparison, it's only 50% of the market. As the share of the price comparison grows, more people will be switching, there'll be more comparison in prices and that would allow us to get to the loss ratio in the market.

And sorry U.S. comparison, pure true price comparison is tiny time in the U.S. There is a lot of noise, and one of the challenges for Compare, and from point of view is a lot of noise from people who advertise as if they were comparison operators but sometimes deliver an outcome as simple as a list of people to get through, sometimes collect a couple of pieces of data and then parse you off to someone, show an indicative estimated price. So the actual pure peer comparison market is very small but growing fast, I mean really, really small. In terms of shopping activity, it is a bit higher than it was two years ago, but I think we'll have to see whether that's structural or whether it's a response to the fact that prices have been increasing in the U.S. market substantially over the last two years in response to the adverse loss experience. That tends to stimulate more shopping in any market.

Arjan Van Veen, UBS. A couple of questions on the European insurance business, if I may. Firstly, the loss reduced to 2 million, but it was negative 5 million in the first half, plus 3 million in the second half. You mentioned some reserve release out of ConTe [ph] so just curious is the underlying trend improving that much and can we expect that going forward? Or is it a bit of distortion in the second half? Secondly, you mentioned price increases in Italy and France. If we look at the annual data, I think 2% price increases in this fourth quarter of 2017. A lot of your competitors' bring prices up towards the end of last year. So can you give more color of magnitude of price increases, particularly in France, please?

Okay. So going to the difference between first half and second half, we do have - there's some volatility in the results. And this is driven by reserve releases, by the fact that we have some business that are still with a small book, so there is some volatility in loss ratio. I mentioned that one of the reason why - one of the positive impact on 2017 is there is also some distortion of reinsurance contract with some deferred profitability on 2017 versus previous years. So there is some distortion, there is some volatility. We did have an interesting reserve release in the second half. This happened also in the past because we tend to have a more deep revision of the results in other year end. But also I think is very, very important to look at the years combined and you don't extrapolate to the trend. And so you look at really the trend year-on-year rather than looking at the difference between second half and first half looking forward. On the price increase, the reality is that in Italy, we're suspecting a quite strong price increase in the last actually couple of years. And this has been postponed mainly because good underwriting results. So the increase that we have seen are smaller than we expected. And we are very concentrated in the very last part of the year. Year-over-year, there was not material price increase. And all the price increase that you may see are only on traditional period. The direct market actually is not up at all.

France?

France is relatively stable with very, very small price increases and there is many claims inflation.

Morning. It's Andrew Crean, Autonomous. Could I go into a bit more detail on your U.K. major results. You're signaling I think flat pricing for 2018. Could you compare that with what you expect for claims and cost? And then I mean, I agree with Greig's point. I've found your results now increasingly obscure. The results you're actually reporting this year, I mean, to what degree are they reflecting underwriting years '14, '15, '16, '17. It seems to be now sort of three or four years behind what's actually going on. So that's the second question. Third question is, you say that your times top has gotten better, but I noticed that you actually put on more car policies in the first half than in the second half. So should we be looking for an acceleration in the number of car policies on the books in '18 rather than '17?

I'll comment on the profit and loss accounting the recognition of results. The philosophy of our accounting for claims is that we take a cautious approach, and we release reserves in the profit when we have a good amount of confidence that there are going to be releases and that profits are going to be profits. The business that we are earning in 2017, we are accounting for as roughly breakeven on a underwriting year basis and there is a cost profit from ancillaries and fees and things like that which gets recognized basically in the year. And reserve releases 21% of the premiums and associate profit commission is what drives the underwriting results in profit commission. And that comes from the underwriting you can see in the appendix. So we show you the loss ratios booked by underwriting. We show you the accident year best estimates. We tell you how much money we make on other revenue per vehicle. And really I think you can get to how much money we're making. Some of the structures particularly on profit commission are complex and was somewhat hampered because the terms are confidential, so we can't tell you what they are. And we try our best.

In terms of the evolution of prices going forward, do you want to talk to that, Cristina?

Yes. So you talked about our prices improving, but actually the graph we show is more about flat, so at the end of '17 and the end of '16, our times top is the same. In terms of claims evolution during 2018, we think it's going to be a continuation of the benign trends of '17, an increasing accidental damage and a continuation of the reduction in the small BI claims. What happens to large claims, I think it's harder to predict. So overall, it's going to be benign, but it's hard to predict very concrete. So we think in December I mean, prices are going to be flat, they could be a few points up or down, depending on how the claims evolve.

[Inaudible]

I don't think so, but that's our view at the moment. Having said that, what happens to the government reforms, do they become concrete? Is it clear they are going to happen in April? And the extent of the reforms is going to have a significant impact on the second half of the year. So it's hard to predict exactly what is [indiscernible] to prices. Then your question about the growth in '18, I think there are two components to growth, one is new business and the other one is renewals. And some of the things that you might be seeing around improvement in our retention rate. Dom O'Mahony: Dom O'Mahony from Exane BNP Paribas. Three questions for me, all on U.K. car. The first is on the Ogden, the 40 million that you put on this year, is this part of the drag that you flagged this time last year? And is it from PYD [ph] or is it from insurance commissions or combination? The second is you disclosed the terms for the car insurance arrangement with Munich Re have changed slightly so that you get the instalment income. Could you just remind us of the arrangements for the ancillary income as well there? And also whether actually this reflecting any broader changes in the terms of the agreement they're recognizing of course that if there is an inventory you can say given the confidentiality? The third is on prudence in the U.K. book. At half year, you said that while of course you continue to have enormous amount of prudence, you've relaxed business slightly and again you relaxed on H2. Could you just give us a sense of how much is contributed to your PYD, sorry not just to PYD but your accident year loss ratio? And is this a reflection of a new policy? Is this - are we expecting sort of still very prudent but not quite as prudent as before from now on? Thank you very much.

Geraint, do you want to take all three?

I will give it a go. Ogden, the £40 million hit this year mainly affects profit commission, it is the ongoing drag that we talked about in the 2016 results, effectively I think we said £200 million pretax impact just about and over 150 of that now we've taken. Second, instalment income change in the arrangements, there's been a slight shift where we get all of the instalment income. But of course, we therefore make less profit and so we get less profit commission. So really that's allocating around the lines in the income statement, and there aren't changes in any of our reinsurance arrangement really in course of the year. And the final one is prudence in the reserves. We talked at the half year about bringing the size of the conservatism relatively down. If you go back pre-Ogden, we were saying that we would expect to bring level of the conservatism down over time, have things gone back to normal times. Ogden obviously disrupted that significantly. And so we increased the size of the margin at the end of '16. We brought that down in the middle of '17. And down very gradually in the second half of '17 as well. Reserve releases for the year were 21% premiums, it's higher than the long-term average. I think had we kept the margin broadly the same as the end of the previous year, we'd have been back in that sort of normal range, so I think in that order.

And ancillaries in the U.K. are all accrued to us on car. Typically abroad, they are shared with the reinsurers.

Hi. It's David Bracewell, Redburn. Two questions. One on the ultimate loss ratios in the U.K. and haven't changed in '13, '14, '15 I think or '14, '15, '16. My understanding was the Ogden claim is settling at better rate than the minus 0.75. So I would have expected perhaps as I came through the numbers that the actuals would have improved there you realize there. I know you talked about different settling patterns, so maybe a bit more detail that would be useful. And then the second question, just on the new products you mentioned right at the end. I mean, are these new product going to be kind of car-related products or actually you're looking to expand into any different types of products unrelated to car going forwards? Thanks.

The new products aren't necessarily car-related, they will be related to our strengths. If you look at lending, we are doing car lending, but we are also doing lending for other purposes. But wherever we do lending, with selection and low expense is relevant and so we feel it's an interesting product and increasingly price comparison distribution is also very relevant to across a number of lending products. Maybe I'll do Ogden as well. What you are seeing is some gumming up of the works. It's very difficult both for insurers and lawyers to understand what is an appropriate settlement when there's so much uncertainty around the actual financial context. And so it is very difficult to - and that affects the evolution of claims. So they are developing in a different way because fewer of them are actually reaching settlement. It just creates more uncertainty. And in that context, we choose to take a relatively conservative view of how that might unwind.

Waja Rizvi, Deutsche Bank. Just going back to the question on accident loss ratio. So at half year stage, you said that loss was slightly - it reflects some conservatism because of Ogden uncertainty. And the full year, it has come down quite a lot. And so I was just wondering if you could break down, break that improvement down between what's coming from benign claims, price increases and actually what's the conservatism which you have released, that would be the first question. The second one would be on reinsurance. So you have increased your deductible to 8 million. Just wondering in terms of are there any changes to the terms of the contract, as well in terms of the Ogden rate moves, would the price change as well with it? And what would that represent as a cost of your premium, not for the market? And just finally, some of your peers are still talking about escape of water claims inflation in home and your price increases have lagged to the market. So do you genuinely believe that you have underwriting edge when it comes to escape of water claims? Or what's happening there? Thank you.

Geraint [indiscernible] reserving, and maybe Cristina escape of water.

If there was any other change to the terms of the contract in the ancillaries…

The first one was the improvement in the current year loss ratio from half year to full year. I can't remember exactly what the position was at the half year. The full year has gone basically 88 to 85, so 3 point improvement. And I think we are not going to break down it into its various components. I would say that we take quite little - very little credit for potentially more favorable large loss outlets that would be implied by minus 0.75% Ogden rate. And so I would say that our projection of where 2017 will end up at this point is a conservative projection.

And the final question was around escape of water in household. And we did experience some inflation of this type of claims during 2016 and took action both in claims management and in pricing. And during 2017, escape of water hasn't been a big issue for us.

Just for sake of clarity, although the terms of XL haven't changed, the cost has. And for someone with our type of retention, that's between 1% and 1.5% premium. Q -: Good morning. Andreas Embden, Peel Hunt. Just going back to Confused and your strategy there, could you just maybe update a bit more detail the investments you're making this or last year in 2017 and how you think that the profitability of the business will improve in 2018 and '19? It seems to me you're still lagging your competitors in terms of investment. What are you exactly doing to turn the business around? Thanks.

Well the rationale of the strategy on Confused is you've got four players which in terms of their offering are relatively undifferentiated. And we are seeking to differentiate Confused going forward as the price comparison site for motorists. So there have been two types of investments. One investment has been in product range, so we developed the first car finance comparison offering in the U.K. We are developing car buying and selling options on Confused. So there's been some investment in product. There's also been an investment in communicating the repositioning towards motorists solely and driver wins through the James Corden campaign. That has caused us in terms of profitability. In terms of its long-term outcome, it's still uncertain. What we have seen is a better outcome in the second half than the first half, and that's carried through into the first couple of months for Confused. But this is a long-term repositioning and not a short-term issue.

Morning. It's Nick Johnson from Numis. Just a question on U.K. car unit growth in terms of customer numbers. I think you previously said that there's still a lot of market share to go for given your current market share. Could you just comment a bit about perhaps the competitive landscape? Is it getting harder as competitors increase their footprint and operate pricing models, et cetera? Or is Admiral managing to stay ahead of the competition? At least just give us a feel for that. Thank you.

Well, we've said for a number of years that the U.K. market is characterized by - increasingly characterized by a small number of, I think, very competent, quoted, focused, normally direct personal lines players and the market is moving towards those players of which obviously we are very much one of those players. That will continue to happen. And we have to work in that context. So the focus on remaining one of the best insurers in the U.K. is a reflection of the importance of being able to match the progress that the likes of Hastings direct line insurer are making and being better operators than they were 10 years ago. Having said that, to be able to grow the business as we have 5% for the core car insurance business and much more substantial if you look across the whole portfolio of products, it suggests that we are able to do that competitively and produce a very good return.

So Andrew Crean again. One of the things which you could do, which might be helpful he just asked for, on the sensitivity issue [ph] one point improvement in the ag claims ratio which you got in the appendix, doesn't reflect changes when the terms change, particularly on Munich Re contract. If you could give us a proper sense to it, because this essentially is highly misleading, that would help us, I think, in being able to reject the profitability as you improve your underwriting years?

You mean sensitivity a range of different movements rather than just one either way?

Exactly. And it shouldn't compromise Munich Re because there's a lot of moving parts in there?

Thank you, Andrew.

Hi. Iain Pearce from Berenberg. Can you just clarify one thing, whether that has actually been any effects on PYD profit commissions from the fact that Ogden claims haven't been successfully in minus 0.75. I just think we need some further clarity on that, please?

Is the question - has prior development been affected by the fact that claims aren't settling at minus 0.75? I think there's very little of the impact on our projected ultimate loss ratios have come from the fact that claims haven't been settling minus 0.75 because the settlements patterns are disrupted, and the very largest of losses aren't settling and then sorts as you don't expect them to settle that. So there has been some claims settle at a rate or implied rate at least that isn't minus 0.75, so at least some of that spread through into the results. But I think as Cristina and David have mentioned, the patterns have been quite disruptive this year, so I don't think there's a large impact and we certainty haven't banked a lot credit for parts of other claims not settling at that rate.

Sure.

Excess of loss premium, I wonder if you can just give us an actual number for '15, '16, and '17, as maybe I wasn't that smart in trying to write down your answer? And second thing is if you can give us sort of an indicative inflation rate in motor. I know you said it was benign, but what sort of rate, what sort of rate range happened in 2017? And then you made a comment that you set up an insurance license in Spain. I assume that's right next to Gibraltar so that you can share staff? Or does it indicate a much more committed view to the local Spanish market? I just want to say, one of the views I have is that at some point you're going to have to leave Spain. And if you set up the license, does that indicate it just put us next to Gibraltar because Brexit? Or are you showing a renewed commitment to the Spanish market?

Great, you are over reading the entrails. We don't - you don't chose Madrid because it's close to Gibraltar. But there are a number of very strong reasons around the fact we've got an existing relationship with the regulator. It's a nice place to be, and there are regular flights from Cardiff. In terms of the excess of loss, it's not that material an item in our cost structure. So I wouldn't overly worry about it. We will see whether we're comfortable with further disclosure.

[Inaudible]

It's an indicative increase for people who have a structure similar to ours, with quite a high retention.

[Inaudible]

Yes. What we show in the appendix, Greig, is cost of an '18 program from 10 million to 2.5% of premiums, program is slightly - cost slightly less than that, there is very much at that level. And we also show you the cost of that type of program the year before, which is 1% of premiums. The year before your cost more cap. So we paid something less than 2.5% of premiums in 2018. Sure. Claims inflation, you can see from the numbers we've shown there's more BI than the AD, and I'm sure your model is sophisticated enough to incorporate that into historic claims inflation. The challenge is what's going to be the ultimate claims inflation on BI. So it's not easy to say, look, the actual result in 2017 was x or y. And it's very hard to also project forward. The really interesting issue is do those small BI trends continue, flatten off, reverse. So it's going to be very much watch the space as the market evolves.

Something positive?

Some positives and lots of uncertainties.

[Inaudible]

Unlikely, unlikely that there will be negative inflation. And can I ask if there's anyone in the phones who's interested in asking a question?

We do have a question from the telephone line from the line of Dhruv Gahlaut from HSBC. Dhruv, please go ahead.

Good morning. Thanks for taking my questions. I have three questions. Firstly, could you update on the reinsurance contract, I am assuming that must be coming for renewal on the 35% or 38% on the U.K. motor book. And do you expect any increase there? Secondly, on the investment income side, you still are sitting on almost 50% of your portfolio around cash, money market fund, et cetera. Should we expect any changes there? And what's the yield you are getting at this point? And third question in terms of your PPO propensity, could you remind us in terms of have those been adjusted down and what interest rate do they reflect? Are they still based on the negative 75 or are they around the 1% mark?

Yours, Geraint. Thank you, Dhruv.

Quota share contracts for 2019 will be extended there. I think are now fully signed. Same terms and condition, same price. So that will roll on into 2019. Second one I think was investment return. The asset allocation hasn't changed very much over the course of the year, I wouldn't expect it to change materially in the future. These running yield I think is something in the order of 1.4%, 1.3%. If you multiply that by £3 billion of cash, hopefully you get to our investment income. And PPO propensity WE haven't changed that over the course of the year, and obviously nothing is settling as PPO you might expect in this Ogden environment. And hence, partly the comments earlier on distortion of settlements patterns. But no, we haven't changed the PPO propensity since the end of last year or since the end of '16.

Maybe one more question on the phone, if there is one.

We have no more questions on the telephone lines.

Okay. Lovely. Thank you all for coming. See you all in six months.

Or before.