Good morning and welcome to Admiral’s 2018 Presentation. Thank you very much for coming. Thank you to UBS for hosting us. Overall, a pleasing year, even a very pleasing year, rapid growth across large parts of the business, record profits. Let’s dig a little deeper, and it’s a year to an extent of yeses and buts. Yes, rapid growth across the business, 800,000 more policyholders, but the rate of growth in the core car insurance business did slow down in the second half versus the first half. Yes, record profits, but a lot of that profit growth is attributable to the government’s decision to partially unwind the change in the Ogden discount rate that cost us a lot of money a couple of years back. Yes, Confused, our biggest price comparison site, grew both market share and profits in 2018. But compare.com, our U.S. price comparison site, had a relatively more disappointing 2018. There are some unequivocal yeses, even yeses, and I’ll particularly point at the International Insurance result where we’re up 20% year-on-year in policy count and have practically eliminated our losses. So there’s the summary of the year. To talk more about that, Geraint, first of all, our CFO, will pick up on the key numbers. Then Cristina, head of U.K. Insurance, will talk about the core U.K. car insurance and home Insurance business, before Milena, our EU Insurance CEO, will talk about our continental businesses, both insurance and price comparison site. And Alberto, who has joined us from the U.S., the CEO of our U.S. Insurance business, Elephant, will talk about that business, before I wrap up with the U.S. price comparison business loans and a few finishing comments. We have also in the audience Scott Cargill, head of U.K. loans, and we have Colleen Benzin and Felipe Herrera over from Elephant. So if you have any questions you’d like to pick up after the presentation in either of those areas, they will be available. Geraint?

Thank you, David. Good morning, everyone. I’m going to give a high-level overview of another strong set of results, including the important favorable impact of Ogden. I’ll talk about maintaining strong capital position and a higher final dividend. And I’ll finish up by talking about U.K. motor loss ratios, some very positive back year movements, but a higher current year figure. And I’ll give a reminder of our cautious approach to reserving. First up, though, the highlights. As you can see, we continued to deliver very positively against these key measures. At the top, continued good levels of growth, nearly 800,000 additional customers in 2018 and double-digit turnover maintained. Profits, David mentioned, have been favorably impacted by the forthcoming partial reversal of the Ogden discount rate. We’ve changed the basis of our U.K. motor best estimate reserves to 0% from minus 0.75%. And so including the benefit of that change, pretax profit was up 18% to £479 million. Earnings per share was up 17% to £1.37 per share. Return on capital, at the bottom, was consistent and healthy, at 56%. And we’re proposing a final dividend of £0.66 per share, which will bring the full year total to £1.26, which is 11% up on last year. And finally, in the middle, Solvency ratio after that final dividend remains very strong at 194%. Next up, onto Ogden. Now as a result of the civil liability bill coming into law in December, we’ve changed the basis of the U.K. motor best estimate reserves. And this slide shows the impact of that change on some of the key numbers. Slide 47 in the back of the pack gives a bit more detail as well. We’ve changed the discount rate, as I mentioned, in the best estimates from minus 0.75% to 0%. And that 0%, of course, at this point is an estimate, possibly slightly cautious of what the actual rate will be when it’s announced later in the year. And the ultimate pretax benefit of moving to 0% would be somewhere between £120 million and £140 million. And you can see that 2018’s results is £66 million higher as a result of the change, and most of the balance plus or minus any differential to 0% if the actual rate is different, will flow into the next couple of years. I’ll explain what we’re doing in terms of the dividend shortly. Moving on to look at turnover and customer numbers. And once again, we’ve delivered good growth across the group, particularly from beyond the core U.K. motor business. Naturally, in some of the areas, those percentages are a touch lower than recent periods. In U.K. motor, the year-on-year position was positive, with turnover up 8%, customer numbers up 9%. Growth though, as David said, was more modest in the second half, as we moved our rates up in the car insurance business ahead of the market, as Cristina will talk about. In household, we saw more strong growth. Turnover up to nearly £150 million and a near 1/3 increase in the number of customers in that business to 870,000. Outside the U.K., we saw a continued strong progress, moving up to £539 million of turnover, plus 20%, and over 1.2 million customers now, plus 18%. And those were especially pleasing moves given the improvement in the results that we’ll see shortly. In price comparison, revenue was up 5% to £151 million. And our loans business continued to grow strongly, with the stock of loans balances reaching £300 million at the end of the year. Moving from the top line to the bottom line, what’s driving the increased profits. I’ve touched on the Ogden impact already, so I’ll focus here on what’s driving the underlying changes. At the top, you can see U.K. Insurance, profit was up £20 million to £485 million. And that’s made up, firstly, of a £30 million increase in the U.K. motor profit. We had the benefits of a bigger book, and very positive prior year development have offset a higher current year figure and expense base. And secondly, a £7 million fall in the household result, from a profit of 4 to a loss of 3. Now 2018, as you all know, is a very weather impacted year for household. Moving down to international. You see a very nice improvement in the result there, £13 million reduction in the loss to close to a breakeven result. And all businesses improving their results, and as we saw previously, continuing to grow very nicely. Three years ago in international, we made a loss of £22 million and ended the year with 670,000 customers, and in the 3 years since that loss has largely disappeared, and the customer base has grown by 80%. Price comparison produced a profit of £9 million, a couple of million up on the previous year, with confused.com standing out, profits up over 40%, turnover higher and an improved margin. Admiral Loans grew strongly. Its loss was in line with expectation at £12 million. Don’t forget, of course, the frontloading of costs in that business and the spreading of revenue. And finally, at the bottom, the other line, which was notably up on the previous year. The full details in the back of our pack. But that share scheme charges and some one-off central overhead costs were among the main reasons for the change. Moving from profits to capital. And also, to give an update on our internal model. We continue to have a very strong capital position, despite a very slight fall in the ratio from the half year to 194% at year end. Both the level of capital and the capital requirements moved up in the 6-month period. And there was a net benefit from Ogden despite the higher dividend that we’ll talk about shortly. For the time being, we remain on the standard formula with the capital add-on, and that’s consistent with recent periods. To give an update on the internal model. We’re making more good progress on developing the model, the validation, the documentation and so on. But we need to do more to bring those aspects to the standard where we and the PRA, of course, are ready for us to formally apply. Growth and the additional complexity of the group is also leading us to consider whether the scope of the model, purely U.K. motor currently, remains appropriate. Those points inevitably mean a delay to the process. And we now don’t expect to make the formal application in 2019, and possibly not next year either. We’ll update, of course, at the next results. Given strong profitability and a strong capital position, we proposed a higher final dividend of £0.66 per share. The approach we’ve taken at this year-end is to think of the dividend in 2 parts. The first is the dividend that we would have paid on the minus 0.75% Ogden basis. And then on top of that, we’ve paid out much of the benefit we’ve seen in 2018 from the Ogden change. And so the first element is the underlying dividend, in the light blue, the £0.55 per share. And that compares to £0.58 last year, with the move being in line with earnings per share. And then on top of that, as you can see, we are proposing paying out a further £0.11, which is about £31 million out of the Ogden benefits, and so the total is £0.66 per share. That £31 million, that’s about 60% of the post-tax Ogden profit benefit, and that’s roughly equivalent to how much of that post-tax profit benefit translates into additional capital surplus. That’s it for the group section. And so a quick recap would be good growth, higher profits, strong capital, higher final dividend. Changing topics to look at the U.K. Insurance business. And before Cristina takes over, I’m going to talk briefly about loss ratios, particularly the higher current year figure. And I’ll finish with the reminder of our cautious approach to reserves. The top half of the slide shows the most recent projections of the accident year ultimate loss ratios. And the figures in the brackets show the development in the second half of the year. And as you can see, it’s been a very positive 6-month period. As we said previously, 2017 is likely to be the peak of profitability for us and the market. And it looks as if it will be a very positive year, with a 4-point improvement over 2016, being due to higher premiums and generally a positive claims experience, maybe with the exception of damage inflation. 2018, shown on the bottom, has started higher at 81%, and so we’ve broken out the movement here into two blocks. The first part is, as we mentioned in August, higher large bodily injury claims experience than we would typically expect to see in a year, and large injury claims here means above 100,000. And secondly, in the purple, that’s everything else, which includes some premium deflation and continued damage inflation, which Cristina will come to talk about. Admiral, as you hopefully all know by now, tends to respond very cautiously in reserving for large injury claims. And whilst we regularly say that these projections tend to include a level of prudence, I would say that’s particularly the case in that early projection of 2018. Looking briefly now at our approach to reserving and how that impacts profits. As you know, we’ve got a cautious approach to reserving, and so releases from back years will usually contribute materially to current year profits. 2018 is clearly no exception. And you can see that without Ogden, the release would have been 21%, and including Ogden, it was up to 25%, which is the highest figure we’ve seen for some time. Despite that large release, we have increased the margin in reserves above best estimates since the half year, reflecting, firstly, the uncertainty around the actual Ogden rate, and secondly, the unusual big bodily injury claims experienced in 2018. As we’d usually say, we expect continued high levels of reserve releases in 2019 if things develop as we expect them to. So just to conclude on that strong set of results, good growth, higher profits, impacted by the partial reversal of Ogden, still record levels on an underlying basis, a very strong Solvency ratio maintained, with an increased final dividend. On U.K. motor claims, consistent very positive prior year development, though a higher than usual first projection of the current year. Thanks for listening. I shall hand you to Cristina to talk more about the U.K.

Good morning, everybody. 2018 has been a mixed year for the U.K. Insurance operation. We have seen a strong growth in customers, in profits, but we have also seen a high loss ratio, both for our car and our household book. In terms of the market, claims inflation has been higher than the increases in premium. Now let’s take a look at our results in more detail. Growth for our motor book has been close to 10% in terms of customers. And most of this growth is concentrated in the first half. And actually, the majority of this or a significant part comes from the growth of our bond book. Just as a reminder, we loan to direct operation for bank in 2017, and we have emigrating policies from the broker book to the direct book. And that explains strong growth in the first half for our bank book. In terms of the second half, we started to put prices up in August ahead of the market. And that resulted in a growth in the second half versus the first half of about 1% for our whole motor book. For household, strong growth, but also small loss due to weather. Now let’s take a look at what’s happening in the market in terms of prices and also for Admiral. During 2018, we saw overall softening of premiums. But as you can see in the graph, in the last months of the year, we started to see a change in the trend. In terms of Admiral, you have a graph on the right. It shows Admiral Times Top, and that is the number of times, the percentage of times that Admiral appears top in price comparison websites. The graph is indexed to January 2017. As you are aware, in Admiral, we tend to move prices ahead of the market because we focus on profitability over growth. So let me explain you the graph in more detail now and what has been happening since 2017 in terms of prices. So 2017, as you can see, during the year, we became more competitive. And this is explained because, at the end of 2016, we put prices up strongly because we were aware of the changes in the Ogden rate. The rest of the market started to increase prices during Q2 in 2017, and then we start to became more competitive. In the second half of 2017, we started to see certain reductions in premium in the market and Admiral also took the opportunity to reduce prices. Now let’s take a look at 2018. The market in the first half continued to put prices down, but Admiral kept its prices flat. And in the second half of the year, as we already told you in the half-year presentation, we started putting prices up in August and we continued throughout the year. So overall for 2018 we’ve seen a small decrease in prices, whereas for Admiral, we are seeing an increase in prices of low single-digit. Now let’s look at 2019. It’s difficult to predict what is going to happen. But in our opinion, we expect to see further price increases to offset the impact on claims inflation. A number of things that we need to take into account in terms of claims. One is the evolution of accidental damage inflation. Is it going to continue at the same level as previous years or are we going to start to see a reduction? But also the changes in the government reforms. First, when it’s going to be and exactly to which rate it’s going to be the Ogden rate to change. Secondly, how and when will the whiplash reform apply, which we think it will be in 2020. Now what we have seen so far in the first two months of the year is a small increase in prices in the market. And actually, Admiral is up a small amount year-to-date in 2019. And I think we are quite encouraged to say that we have seen in the past few days the collective discipline of the big portal players where they are favoring profitability over growth. So now let’s take a look at what has been happening in claims in the market. As I mentioned, one of the big features of 2018 is the strong increase in claims inflation around accidental damage, around the cost of repairing cars. The cost of parts, but also the increasing technology in cars, is driving this inflation. Frequency overall has remained flat, although we have seen a couple of increases throughout the year, but nothing too significant. The graph on the right talks about small BIs, small bodily injury claims, and it’s the year-on-year change in market portal notifications. You can see strong reductions in 2016 and 2017, partly driven by lawyers’ behavior around an anticipation of the whiplash reform, but we have seen a much more flat frequency in 2018. This is the trends for the market and they are very similar for Admiral, increase in inflation of the cost of repairing cars and a flat frequency in the small BIs. So a big change, as Geraint has explained, is that we have seen a big increase in the frequency of large BIs. Now in the next section, I want to talk about what makes a good direct insurer and why Admiral continues to be one of the best. So let’s start with claims management, which is the moment of truth for our customers. According to management estimates, we have a trend to outperform in the market, both in terms of speed to settle a claim, but also in terms of cost. And this speed and this short – or shorter time than the market in settling claims allow us to save cost, but also to serve our customers better. Another very important part about being a good direct insurer is around being good at pricing. We have talked many times about our very predator approach to pricing, our constant changes. We’ve also talked about the focus that we put in fraud prevention and detection. Today, I’m going to talk about also how we take our focus on data analytics beyond pricing. So you can see here an example of total loss claims model. This is a model that predicts whether a vehicle will be a total loss or repairable during the claims quote. And this allows the handler to take a different action. So what we have seen is a saving of around 20% in the cost of misclassification of a total loss. Now moving on another very important area is being an efficient player, efficient in claims management, in operations, but also in our marketing spend. As you know, we acquire most of our business through price comparison. But being a strong brand also helps because you convert more when you’re top, but you also convert more even when you are not top. We are very pleased to see that in the past few years, we had spent around half than the average of top 4 competitors. But we have seen the preference, we have seen the number of customers that choose Admiral as their preferred insurer increasing, and this definitely helps when it comes to converting sales. Second important thing is around efficiencies and our constant spend around digital, around technology, but also our improvement in processes, has resulted actually in a strong reduction in the number of calls that we get from our customers to the customer service department. In the past two year, we have seen a reduction of about a third of the number of calls that we get. So overall, being a good direct underwriter for us means being good at claims management, being efficient, and overall, being very strong in pricing and analytics. Now let’s park our motor business and move into our household business. The underlying performance of the household book remains strong. We have seen good growth, and we have also seen a further reduction in our expense ratio. However, I think the highlight of the year has been the impact of the weather events. We have seen a loss of £3 million. A number of things happened last year. First, exactly a year ago when we were here, we were covered in snow, it was quite beautiful, but we also had a spike in claims. Then we had floods in the East Midlands in May. And this was followed by a fantastic summer. It was even sunny in Cardiff, but of course, increased subsidence claims. So you can see the detail of our loss ratio in the graph on the right. I would highlight that the attritional loss ratio has remained flat despite its strong growth in our book, meaning we have a bigger proportion of new business. And we have seen 11 points impact of the adverse weather events and also eight points for subsidence claims. Now moving on as I mentioned, we have seen a strong growth in our household book. Now I want to highlight that this growth has been achieved despite us putting strong price increases for our household book. So why are we growing? Well, this is part because we have seen good retention in our increasingly large renewals book and also helped by our cross-sell ability to sell a household policy to our car customers. In terms of expense ratio, we are very pleased to see that we are able to replicate our efficiency that I was talking about in car to our household book despite its relatively immaturity. The market expense ratio is around 45%, and we have managed to have an expense ratio of 28%, better than the previous year. So overall, for the household book, we expect to continue the trend. We expect to continue growing the book despite at a lower rate than in the previous years. So in summary, growth continues in our motor book. We continue to focus on sustainable profit. And in terms of household, good growth, but impacted by weather. Now before I move on and give it to Milena, please let me indulge in highlighting how pleased we are in having been named the Best Big Company to work for in the U.K. Even more important to us is the fact that we are the only company that have been named one of the best companies to work in the U.K. by the Sunday Times since it has started more than 15 years ago. And this is important for us because it highlights the hard work that our staff do throughout the year to serve our customers. In Admiral, we always say, people who like what they do, do it better. So happy staff, happy customer, which is at the end, the goal that we try to achieve. So this is all for me, all for the U.K. Insurance. And now over to Milena to tell us all the successes of the international business.

Thank you, Cristina. Good morning to everybody. I’m very pleased to be here today to talk to you about European results in 2018. It was an outstanding year as we turned the corner of profitability. And I believe 2018 will stay memorable for achieving several milestones. The main ones, we were profitable on combined base for the first time. We reached 1 million customer. As you can see, our celebration in our Serbian office at very, very year-end. On top of that, we work on the launch of our first product outside motor, outside U.K. and the retail in-house Homebrella, household insurance in France, of which David will come back later. And last but not least, we successfully completed on time our Brexit project. That basically means that from January 1, all the policies sold in Italy, France and Spain are underwritten by Admiral España, Compania de Seguros, that is the new underwriting company we set up in Madrid. One million customer base was the results of a joint effort of the team, of all the three geographies. We have seen a substantial growth deliver in all the three countries, peaking with the 40% in France. On a combined base, we grew by 18% in customer and by 20% in turnover. So to deliver substantial and sustainable growth is one of our main priorities. We continue to focus on building our brands. And for example, in 2018, we progress from sponsoring the Soccer League Serie B in sponsoring the Soccer League Serie A. That basically means that if you’re going to watch one of the main league match in Italy, you are very likely to see, at some point in the game, ConTe brand flashing all around the field. We continue to invest in our product and improvement for our customers. And as another example, we deliver a fast quote process for new business in Italy. That means that if you try to get the price for ConTe, you can get that with less than half of the question than in the past. We’ll also continue to invest in our customer service and providing better and faster and more reliable service to our customer. As we were growing, we also managed to deliver £7 million of profits on a combined base. This is the first year we deliver combined profit for the full year, but is the third half year in a row. And I’m mentioning this because there is a consistent path of improvement in all the three geography on the current year and on the back years. As for turnover, also this has been a joint effort of the three countries, and the main driver were a reduction – a material reduction of losses from L’olivier and Admiral Seguros on one side and record profits on ConTe on the other side. For ConTe, 2018 was really a special year. It was our third year – sorry, 10-year anniversary. We were for the third year in a row in the podium of great – best place to work for in Italy. At the very, very year-end, we reached £1 billion of cumulated turnover since launch. And, and this is my favorite milestone of all. We reached cumulated profitability since launch at the very year-end. So now we recoup all the investment we’ve done so far in Italy. So moving now to the underlying ratio of European Insurance. The loss ratio was stable, slightly improved and stay at a pleasing 78%. So loss ratio is the area where we more clearly transfer our competitive advantage from U.K. into overseas. Now I explained when at European Investor Day we had in Rome in September, we look at the back years where the results are more mature. And we observe as our loss ratio is better than market average and stands something like eight, 10 points better than our direct competitors. If we look at France and Spain, we are more or less at market average as well if we account for the different mix of new business and renewals compared to the market. It’s also worth to remind that, in the last few years, we increased the level of conservatives on our reserves settling – claims reserves settling, that now mirror the approach that we use in U.K. and translates, as you can see in the graph on the right, in an ultimate loss ratio projections that tend to decrease over time as for U.K. So pricing and loss ratio have always been, and very likely will always be, our top one priority. We continue to look for new data, better way of using data. We continue to strengthen the team. And in the last year, we also increased a lot the coordination and collaboration among all the areas that affect loss ratio, pricing claims, underwriting, data analytics and antifraud. And on antifraud in particular, we create a center of competence in Italy where the function is more developed for market reasons to help and support the development of this function also in France and Spain. Now while we compare very well on the loss ratio versus the market, this is not the case yet for expense ratio. There, overall, on a combined base, our expense ratio is above market level for Europe as a whole. Having said that, we do have very good cost structure. We are very efficient, particularly in Spain. And if we look at Italy that has a bigger book, we are actually very close to market level. And the cost per policy in Italy in absolute term is lower than U.K. Unfortunately, average premium as well is much lower than U.K., so doesn’t translate necessarily in such good expense ratio. Now if you look at the graph on the left, you see that we improved eight points in the last two years. And this improvement came from one side from economy of scale, and from the other side, from internal efficiencies. And internal efficiency came mainly from three areas. We increased the percentage of interaction with our customer online. We improved and automated more of our internal process that led to higher productivity. And we increased the outsourcing of some activities in Italy. Also worth to mention that last year we increased our investment in data and technology to continue to support the evolution of our business toward the more agile and leaner model to better support customer expectation. So in conclusion, we believe we have strong foundation on which to continue to build upon in the future in European Insurance. Moving now to price comparison site in France and Spain, LeLynx – sorry, price comparison site. We continue to deliver on our ambition to be the relevant leader in market – in our market service comparison site where barriers to entry are low, but barrier to success are high. Confused had a very good year, with an increase of turnover of 10%, that was driven by increase of market share, both in motor insurance and home insurance. At the same time, profit increased by 42%. And this despite we are continuing to invest in people and in technology to continue to increase and improve our product and our service for our customers. Also, the profit margin increased by almost 30%. And this was also driven by a more effective spending in marketing and in media. Rastreator and LeLynx, our price comparison site in Spain and in France, have seen a small increase in turnover and a decrease of £3.2 million in profits in 2018. And this was driven by two elements. From one site, very challenging marketing condition. For example, in France, we have seen the entrance of a third player that put pressure on media spending for LeLynx. And on the other side, we invested more in several project, some driven by regulatory change, majority of those driven by willingness to continue to improve our product and also to increase product diversification. And you see a couple of example on the right on the slides. So we always believe that disruption is needed to improve customer experience. And in Spain, we put customer first and increase the reliability and accuracy materially of our pricing given to the customer in a market that is known for pricing accuracy. And we are pleased with the results so far in terms of market conversion rate. We’ve also invested in diversification. An example of that is the launch of energy vertical for LeLynx in France, and a mortgage comparison in Spain. So in conclusion, 2018 was a year of growth, both for price comparison site and Insurance. We are particularly pleased with the combined profitability in European insurance and improvement of all underwriting metrics in the different countries, and on those basic and on those foundation, we look forward to continue to grow in the year to come. Thank you very much. And now Alberto will complete the picture of our international operation, adding United States.

Thank you, Milena. Good morning, everyone. My name is Alberto Schiavon, I’m the CEO of Elephant, our U.S. insurance company. I joined Admiral in 2012. And I moved across in many departments, including pricing and operation, before moving to Elephant in 2017. What I would like to share with you today is how Elephant has made clear progress in our path towards profitability and how the results we have achieved in 2018 are a strong foundation for the future. As you know, Elephant is purely focused on car insurance for now. And our refined strategy of focusing on longer lifetime customer is paying off in multiple areas of our business, including growth in turnover and number of customers. This shift towards high-retaining customer has meant that our turnover figure from 2017 has actually delivered much more positively than expected last year. Therefore, our effective turnover growth is 16% year-on-year. Improvements in many underlying key performance indicators also translated to reduced losses of $10.1 million for Admiral, which is less than half of the losses only 2 years ago. Elephant is converging towards market levels across many several metrics, despite still being a subscale player. Our combined ratio improved by five points, mostly on the back of progress in our expense ratio. As Texas policy holder represent about half of our book, let’s look at their performances individually. Our loss ratio experience in Texas has been better than the market in 2017, with further improvements in 2018. Now bear in mind, Texas has had a very fortunate year on catastrophe losses in 2018, so it’s very reasonable to expect that the market will also improve. Outside of Texas, the loss experience hasn’t been as favorable, which is mostly being driven by sharp bodily injury increase. Now moving to persistency. I’m very pleased with the results that we have achieved this year. The continued shift towards high-retaining customer has meant that persistency has increased by 24% compared to the 17% projection that we shared with you a year ago. This increase in percentage of customers who stay with us longer has been instrumental in guaranteeing the sustainability of our growth. One of my favorite metric is vehicle per policy, which has also grown significantly since 2016 and has hit a record high in the second half of 2018. You can see this on the graph on the right. But also, you had seen this in the very nice T-shirts of our marketing department in our first slide. Why do I like VPP? Vehicle per policy is a good indicator of better efficiencies and likelihood of retention, as those policies cost us less to service per vehicle and customers with more vehicle tend to stay longer. This is also consistent with the experience of our U.K. business. While I do not expect this metric to continue to grow as rapidly in the future, the trajectory that we see gives me comfort that the improved customer lifetime will be sustainable for the long term. Part of our 2018 growth has also been driven by our marketing efforts. There are three main ways we attract business. First, through European-style price comparison websites, mostly three main players. They represent 12% of our sales. Second, lead-generation website, 38%, where we leverage our digital skills. Those are represented in the graph on the right. The remaining 50% of the business is attracted through direct means. And because direct is such an important and main source of traffic, it is essential that we have a strong and unique proposition that is appealing to our consumer. Elephant has pivoted this message to insurance that makes sense through our safe car and multicar value proposition. Safe car rewards drivers who invest in safety features, like lane departure warning or backup camera. Multi-car, on the other hand, rewards for buying in bulk, which also drive, my favorite metric, our vehicle per policy. In January 2019, we also launched a new brand, apparent, that focuses on delivering product features that are particularly beneficial for parents, like child seat and stroller replacement. We launched this brand in Texas, and we are pleased with early indicators. We believe we can provide our customers with a great product that fit their needs. And through their loyalty, these customers will generate a longer ultimate lifetime value for Elephant. We continue to see positive development on our expense ratio as well, as a result of many factors, the longer customer lifetime, as I mentioned before, good cost management practices across the business, and our decisive move towards digitalization and self-service. On this latter point, our customers are now able to make simple policy changes and file claims online by themselves without having to talk to any of us. In conclusion, Elephant has made good progress in 2018. Our refined strategy has had the desired effect of attracting high-retaining customers, which in turn is driving more sustainable growth. The operational efficiencies that come from insuring those customers, in conjunction with many other efforts, are having a positive impact on the bottom line. And branding efforts have supported this progress, and we plan to continue and increase those efforts to further accelerate our growth. I will pass it back to David who will talk about U.S. price comparison and loans.

Thank you, Alberto. So firstly, U.S. price comparison. So 2017 was a very positive year for Compare. 2018, relative to 2017, has been somewhat of a disappointment. And a lot of that is a function of the U.S. insurance cycle, which you can see on the top left. In 2017, we benefited from very substantial price increases being put through by U.S. car insurers as they responded to an unusually difficult year in 2016, and that stimulated a lot of shopping in the market, as you can see from the Google volume graph below. In 2018, insurers have returned to profitability. The rate of price increase has slowed across the market. That’s resulted in less shopping activity in 2018. It’s also resulted in insurers having a renewed appetite for new business themselves. That’s good news at one level. The interest in participating on Compare has risen. We’ve gone up to 86 partners. We’ve signed some important brands like USAA and Nationwide. The way you have a situation of reduced supply of shoppers and increased demand for shoppers, you see a price increase in the cost of acquiring those shoppers, and that’s what Compare experienced during 2018 relative to 2017. That has meant that we’ve been unable to do what we did in 2016 and 2017. We’ve been unable to reduce the losses in Compare in 2018 versus the previous year. What are we doing? Well, first of all, we’re writing down the carrying value of Compare by 50% to reflect the uncertainties around the business model going forward. Secondly, we are reducing overhead costs. In the fourth quarter, we took them down by about 15%. We are looking at diversifying our acquisition approaches away from those which are most competitive with a direct marketing activity by insurers. And we think those actions will help the sustainability of the business, but we are still projecting losses in 2019 for Compare. Onto the loans business. As you can see from the graph, the loans business grew quite rapidly during 2018. It started the year with £66 million of balances and finished with around £300 million. This is a massive industry, and this represents roughly 1.5% market share. What you can see on this graph also is that the pace of growth was faster in the first half than the second half. And this is because we have tweaked our portfolio in the second half towards increasingly prime parts of the market. In the first half, we were pretty prime. In the second half, we became more prime. And we did that in anticipation of the possible risk of economic disruption in 2019, particularly post Brexit. The loss we made in 2018 of £12 million was in line with our projections at the half year and reflects the fact, as Geraint mentioned, that costs, both acquisition and upfront provisioning, are front-loaded and revenue is back ended. The business is 95% unsecured personal lending at an average APR of around 8.5%, which again is a reflection of a conservative approach to portfolio management, and 5% car finance. We anticipate reduced losses in 2019, with the caveat that that’s subject to not being a material economic dislocation in 2019. So that’s covered off practically all of our major businesses. Before I finish, I’d like to talk about a couple of things which have been irrelevant to our P&L in 2018. But when we look back in five years’ time and think about 2018, we might think about these two things as material events in 2018. One of them, as Milena has alluded to, the launch of Homebrella, our French renters or contents product. Material for two reasons, mainly because it represents an extension beyond motor insurance underwriting beyond the U.K., so that’s an important first. But also because it’s a slightly different approach to acquisition from our normal approach. It’s a very mobile-friendly shortened question set, user-friendly product aimed at the urban youth of France. It’s bilingual. And a relatively simple insure techy type product, which is an interesting approach for us to develop in the French market and possibly beyond. The other one I’d highlight again, which is irrelevant in the P&L for 2018, is the contract we signed at the back end of 2018 to become Ford’s sole insurance provider from 2019 for five years. We don’t normally do white label, but we were particularly excited about signing the Ford deal for two reasons. Obviously because Ford is the biggest brand in the U.K. car market, but more importantly, because, with the advent of connected cars, we see the opportunity to develop very exciting customer propositions in conjunction with Ford towards the back end of that 5-year period. So there is a couple of acorns. Just a quick summary then of what we’ve said. A record year, with and without Ogden. Growth across the group. In my press release, I talked about yes, buts, an alternative way of talking about the year would be a year of £3 billion. It’s the year we went through £3 billion of turnover for the first time. And it was the year when our cumulative dividend payments at least as of the 31st of May when we pay out will have gone through £3 billion as well. So thank you for your attention. I’ll open it up now to questions from the floor and the phones. We have upgraded our technology, or to be more exact, UBS have upgraded the technology. You will find microphones in front of you. So if you have a question, if you can pick them up, and there’s a button to press, a light will go on, and then you are – you’re on. Q - James Shuck: Yes. Thank you. It’s James Shuck from Citi. On the subjects of technology actually. So my first question is really around your IT capabilities. So in the U.K. in particular, you’re implementing Guidewire. We’ve seen other companies start looking to integrate Guidewire and Radar Live as an integrated approach, opening up their systems more to cloud-based approaches, which allow much more flexibility. Could you just update on where you are with your own IT capabilities, use of outsourcing and cloud in particular, please? Second question is slightly complex. But if I just compare the booked loss ratio in 2018 with the ultimate loss ratio in 2018, there’s about an 11 point difference this year. If I look last year at the year one booked loss and the year one ultimate, that was closer to 13 points. I think you do allude in the release to the margin of the best estimate coming down. It looks like it’s come down by about two points. Is there anything kind of I’m missing in that? Is that a useful guide for thinking about that margin?

Can we do those two, and then come back? Cristina, do you want to do the technology? And Geraint, will you do the...

Yes. So in terms of Guidewire, we completed the implementation of the policy and billing centers. And we’ve run it with Guidewire for about two years now. In terms of Radar, we use a different rating engine, and it’s already integrated with Guidewire. And we are already in a plan to migrate certain parts of our business to the cloud.

In terms of the margin, you’re basically – what we said in the report is that the margin above the best estimates moved down slightly year-on-year and moved up slightly half year to full year. I don’t get too heads up on individual years. We think about the margin across all years. What you tend to see is a widening margin as the year becomes more recent. But yes, you’ve basically got the point. The margin is slightly smaller at the end of 2018 compared to the end of 2017, but bigger at the end of 2018 compared to the middle of 2018.

Can I come back if there’s time. We’ll just let some other people in now and we’ll come back if there’s time, which there probably will be.

It’s Edward Gunby from Goldman Sachs. Just a first question on the motor profitability. I think you said last year that you expected 2017 to be the best year in the cycle. This year, it’s obviously got a bit worse. It’s kind of hard to see from your presentation how much worse, so it would be great if you could just talk a little bit more about what you see as tempering, what you see as the underlying level. And also, looking ahead into 2019, how do you see pricing tracking versus claims inflation? And then the second question I had was on the PRA review into pricing practices. If you could just give maybe a view on what you see there and whether there is any risks or opportunities there. Thank you.

Do you want to do the PRA review, and maybe, Geraint, just in the interest of sharing that, doing a loss ratio in the cycle?

Yes. So it’s actually the FCA who has been doing a marketing price study. They started looking at the household book and now they are looking at all the practices. There is also a bigger research into different sectors in the economy looking overall at the – or long-standing customers. Now we are in close collaboration with the FCA, but it’s hard to predict what they are going to do at some point in the summer, what type of reforms they are going to announce. We think that any company with a large renewal book like ourselves might be impacted. However, to say that the majority of our customers come from price comparison. This is a very transparent industry. And to give you an example, our car customers, around 80% of them, engage with us every single year. So overall, we don’t expect a major impact. You also asked us about what the pricing is going to do in 2019 and whether we think overall for the market is going to catch up with claims inflation. The answer is we think that is the direction of travel definitely. But it’s going to be impacted by the Ogden rate, what is the actual rate that it’s going to be moved to. It’s going to be impacted by whiplash, is it going to be implemented in April 2020 or later in the year? And also, around the claims inflation as I said, I’m very encouraged by the fact that our competitors, at least the big quoted ones, they are taking a very rational approach to growth and to margins. So overall, if we continue to see claims inflation growing, we will see price increases.

Thank you, Cristina.

I would say on 2017, we’ve actually seen that ratio improve in the – over the course of 2018, so on Page 9, you saw the 2017 accident year was three points better in the second half of the year, one point of which was due to Ogden, so 2017 has got better. And we’ve released – we brought the booked loss ratio down three or four points I think in 2018. The 2018 loss ratio was worse than 2017 for the reasons we talked about earlier, large BI, damage inflation and some small premium reduction. Dominic O’Mahony: Hi. Thank you. Dominic O’Mahony, Exane BNP Paribas. So just three questions, if that’s all right. The first is there’s been prudent – there was a bit of relaxation in the first half, a bit more conservative in the second half, it sounds like net slight relaxation. Could you give us a sense of what PYD ex Ogden might have been evening some of that out? The second question is, you indicated that your Time Stop has come down since the beginning of 2017. But of course, that’s only price comparison. Could you give us a sense on the change of the portion of the book coming through price comparison versus direct? I’m very aware that obviously multi-car and so on changes that dynamic. And the third is it’s boring, forgive me, it may be more difficult to answer. If I look across your portfolio beyond U.K. car, but you look at home, you’ve implied that actually the loss ratio is higher because you’re investing in growth. The first year was less profitable than the second year. The same is true of international. And you have the compound effect of the loss ratio being higher for first year. You have the expense ratio being higher because of acquisition cost. If you were to stop growing, purely manage the book for cash, could you give us some sense of how much that would impact your earnings? I know it’s a difficult question, but some sort of direction would be very helpful. Thank you.

Cristina, do you want to do price comparison 2017 without net? Let’s start with price comparison. And then do Geraint on 2017. Milena, do you then want to do international if we stop growing, and Alberto, if we stop growing?

So I understood that your question was what happened to prices in the direct non price comparison and have the proportion changed materially? Yes? So in terms of volume, no significant change in the proportion of business that comes through price comparison versus direct. As you know, we have a multicar proposition and also multi-cover proposition, and that has been growing, especially around multi-cover because it’s new. But overall, I’d say we keep quite flat the proportion of customers that comes through direct. In terms of household, I think I didn’t explain very well actually what we were highlighting is that we grew by 30% customers, but what we call attritional loss ratio, so that is taking the adverse weather events impact out, has been flat year-on-year, almost implying that new business book has a higher loss ratio in general, and we have been able to keep the overall attritional loss ratio flat. So that’s quite encouraging. In terms of profitability, if it hadn’t been for the adverse weather events, we will have made a profit of about £7.5 million. And that is a continuation of the upward trajectory that we had in profit increases since 2017.

Geraint?

Prior development. So if you look at Slide 10 in the pack, that shows our releases over the past few years. We tend on average to release something in the order of 20% in terms of releases as a percentage of premiums. 2018 would have been very similar to that before Ogden. And the graph on the previous page shows you the impact of how those ratios have developed, and one point of that improvement is Ogden as well.

Stopping growing in international. Good idea?

Of course, all the three business will have better loss ratio if we stop growing and better economics overall in Europe. This is more the case for France and Spain because in those markets, the business is more front-loaded. There are high acquisition costs and the delta between loss ratio in new business and renewal is quite large. It’s not necessarily the case in the same extent in Italy. So in contained Italy, we’re already profitable. We will be more profitable, but the main difference will definitely be in Spain and in France. I commented in September that we were close to breakeven in Spain. So you may assume that we stopped growth we would be profitable, of course, with the volatility of a smaller business. In France, we are still investing growth, so it’s not the case yet, but that definitely will be the trend and will make us closer to our goal.

From a U.S. perspective, high-retention customer tend to have, exactly as you said, a higher new business loss ratio than renewal. So we believe that we will have better loss ratios in the year coming. Also, advertising costs are very high, and therefore, you’re able to spread them over a longer lifetime. If we said – took your argument to stop growth, we would probably be close to breakeven. However, we also have done some internal modeling of what’s the right balance between growth and profitability. And we believe that slowing down the business to achieve that breakeven wouldn’t actually be the optimal solution in a longer timeframe. Dominic O’Mahony: Just a suggestion.

Yes, yes. No, no. I mean interesting hypothetical question, but these are all businesses where we see real momentum from being bigger, household, international, all of them, not just through economies of scale, but also through data and economies of understanding the market. And so were we to slow down, we’d be putting aside a really interesting profit opportunity for four or five years hence. Greig?

Can you hear me? Greig Paterson, KBW. I was just a little bit confused. Just to sort of nail down the rates story. So rates, you said rates were down a bit and the other comment was rates were up a little bit. On a written basis for 2017, including the renewal and new business effect, what was the year-on-year rate increases? That’s question one. Question two is the burn rate for claims inflation in the U.K. Others are saying it’s circa 5%. Are you similar to that? And the third thing is I was a little confused. In terms of the ultimate written loss ratios, was – did you say there was an increase in prudence year-on-year from 2018 to 2017? I thought there was a comment that was made.

I’ll just do that one very quickly. There was an increase in prudence from the half year to the full year and a small decrease in the size of the buffer from the full year to the full year. Yes.

[Indiscernible] 2018 ultimate versus 2017 ultimate, and whether there was a change in the prudence or not?

[Indiscernible] in the level of prudence? What we’ve seen in 2018 is a bigger proportion of that cost is from larger bodily injury and our approach to those types of claims tends to be very cautious early on.

[Indiscernible]

Well, I wouldn’t – I wouldn’t bank it and I wouldn’t spend it, but what you normally expect to see on a ratio that includes a decent chunk of large BI you’d expect to see that to improve more over time.

[Indiscernible]

Greig, in terms of rates, two things. When we look at the – if I understood you correctly, almost an average rate increase for the whole year 2018 versus 2017, all that written. Yes, when we look at that I’ll say because we increased rates for 2018 low single digits, but that only affects August onwards, yes? When you compare this, what we see is a very small increase in rates from 2018 to 2017.

So 1%, 2%?

I’ll say something like that. Yes, that’s versus a market when you are comparing according to the ABI, the average written premium of 2018 versus the average written premium, you’ll see actually a decrease of around three points. So in our comparison, three points versus one to two. Yes. The burn rate, if you may, I’m going to split it in three parts. One is accidental damage, that is the cost of repair cars. Then it’s a small BI and then it’s large BI. When we look at the trends, the ABI data in terms of the cost of repairing cars and their frequency, we are in line with the market. Yes, when you look at the small BIs, similar to the market, it’s actually large BIs when there’s a difference. So I’ll see some competitors talking about the rates of three to five. I’ll say we are in the higher part of that range.

And so just another one, BIs spike in frequency?

Yes.

What’s happening? What’s your experience – do you think that’s going – going to see that trend continuing into 2019? Or should we just say that’s a lot of...

In the interest of brevity, I will take that and say that we’ve seen that large BI can be a very random thing. We saw – as we mentioned at half year, higher levels in Q2 and Q3, not so much in Q4. We’re always very cautious in that context, but it’s a reasonable hypothesis to say it’s not based on some sort of underlying change, but we would like to see that proven by the evolution in 2019. Can we go over there next?

Thank you. Ed Morris, JPMorgan. The first question is on just capital and dividends. So the Slide 48 is very helpful giving us a bridge of how your solvency ratio moved. Can you just talk a little bit about why capital generation was a little lower than dividends in the sort of ordinary part of those earnings? And if we see the trends going into next year of softer pricing, would you expect a similar dynamic on capital generation versus earnings and dividends? Second part of that question is if you could just clarify, I think you said the £0.11 extra dividend sort of offset the capital impact of Ogden. On that chart, it implies that there was an uplift from that. So if you could just clarify that. And then second question, I think second question. Loans, the loan book, £300 million is quite large now. And I think you said we’d have an update at year-end on how you plan to finance this. Just how big can it get without needing to resort to a third-party financing? And what happens next on that book? Thank you.

I’ll do loans and then hand over to Geraint on capital. We had expected to do a longer session on loans about six months ago we flagged that. We will do a longer session in six months’ time. What we found after our Investor Day is a lot of interest in Elephant and a desire for us to spend more time on Elephant. And there’s also a lot happening in the results as a whole. So what we thought we’d do is we’d park more detail on the loans till six months’ time and we’d go with bigger picture. If you want some understanding of how we finance loans to date, Scott is available for a conversation afterwards.

Yes. Page 48, the left half of that slide shows the capital generation on the old Ogden basis, technically, the current Ogden basis I guess. 21% is clearly lower than 29%, 29% the dividend, that’s obviously on an IFRS basis. 21% is the underlying economic capital generation and when you’ve got a loss ratio for that financial year that looks higher than preceding years, then you’ll see less economic capital generation. Over the course of a cycle, those things will even out, you’ll get some periods where economic capital generation is higher than the dividend, sometimes it’s lower, and technically that should even out. 2018 was a year with a higher loss ratio, hence, that green bar is lower than the red bar. On the right-hand side, Ogden. What you see with a change in Ogden rates, you’ll see an associated increase in our assumption of how many PPO claims we get, and that tracks a high capital requirement. And so the 16% is the post-tax capital benefit, the 6% in the middle is how much of that we’ve paid out and the 3% which is bar six is the increase in the capital requirement. We paid out £30 million in – we will pay out £30 million of the benefit we’ve seen in 2018 and we’ll pay out some more over the next couple of years. So ultimately, we’ll pay out approximately 60% of the post-tax IFRS profit benefit. I’ll talking through later. Does that make sense? So it’s not a pound for pound impact on capital compared to our profits because of the capital requirement associated with what PPO claims.

Nick Johnson from Numis. It’s a question on quota share profit commission. So on Slide 50, obviously, there’s quite a striking move in projected combined ratio for 2018 to 98%, getting close to 100%. Could you – and I take on board your discussion around loss ratio, the reason it’s not moving. Could you just remind us of the dynamics of the profit commission arrangements, it’s very complicated. First of all, is PC linear to combined ratio? Any sort of rule of thumb you can give us in terms of how you think about profit commission earning out over the next few years given that combined ratio projection move, that would be very helpful. Thanks.

Yes. Well, I’ll take the 2018. The way our quota share reinsurance contract works is effectively we pay a margin to the reinsurers. And anything beyond that margin, we take as profit commission. The margin is less than two, but think of it as two for simplicity. And so if that combined ratio is stuck at 98%, we get zero profit commission. I’d refer you back to the comments I made earlier about the loss ratios and how they tend to improve over time. The profit commission, therefore, after that 2% margin is 100% to Admiral. On the coinsurance, it’s slightly different, it’s tiered, and so as the business gets more profitable, we get more of the profit. We’re not allowed to disclose the exact terms, but for a combined ratio in the 90s, we get something like 2/3 of the profits. So a combined ratio, low 90s, we get 2/3 of the profit as profit commission.

In terms of how that pans out in the next few years?

In terms of the earning profile, what we tend to do early on is obviously book a loss ratio that will imply that an underwriting year is loss making and then bring that down over time. The profit commission, you see in the notes of the accounts where the current year of profitability is coming from. And we tend to recognize profit commission over the course of three, four, five years usually, but will be none in year one, very little in year two, and then it starts to flow in year three, four and five.

It’s Andrew Crean for Autonomous. Can I have a couple of questions? One, I think you said, for 2019, you think that pricing claims will move together in motor, U.K. motor, that’s after the best year in the cycle in 2017. Do you expect that now the pricing and claims patterns have reached their worst and that we can look for improvement longer term as opposed to further margin decline? And then secondly on the sort of scope of the overall business, you are year-by-year introducing new products, new plans and new areas time and time again. Remember when you started down this road, you said that it was going to be a shotgun approach and that you would take a fairly hardnosed view if you thought any of your launches weren’t going to be best-in-class businesses. I was just wondering as you look at your portfolio, whether there are any businesses now where a more hardnosed approach might be the best idea thinking possibly about Compare where you’ve taken down your percentage of share.

Do you want to talk first about the U.K. prognosis for claims costs? And then I’ll come onto culling the runts?

I’d say three main considerations around claims costs in 2019. First, is the inflation of car repairs, we think it’s going to continue in the higher three to five range. Then in terms of the small BIs, we think we’ll reach a flat frequency. Costs are under control but it is going to depend on what happens in the whiplash reform for 2020 and there might be some anticipation to that, because when it comes to MOJ portal claims, the small bodily injury claims, it’s not just the actual frequency, it’s also the lawyers’ behavior in anticipation on what we’ll have next. So you could have even a spike into the terms of the frequency of these claims for a certain period in anticipation of the whiplash reform. And the third change is the actual Ogden rate that they are going to announce shortly. And I’ll say, all in all, it’s what we need to take into account. Does that answer your question?

It was more a general question, it was whether we’ve reached the bottom of the pricing cycle. If you’re now saying that you hope or think that pricing and claims will match each other, I mean it’s not done so in 2018, whether you think now we’ve sort of reached the bottom of that sort of pricing cycle in real terms relative to claims?

We’ve seen encouraging signs so far this year. We have heard our competitors talking about a very disciplined approach. So I’ll say hard to tell, but possibly no. I think we will see a continuation of price increases.

So I think, possibly, we might have reached the bottom.

Yes.

One of the interesting other variables, of course, will be whether we return to falling frequency, which also might be a function of the economic climate we’re in. So one of the interesting things about car insurance is it tends to be somewhat countercyclical in terms of how it works, people have to continue to buy it and then they don’t use their cars as much. In terms of actions to address businesses which are succeeding less well than others, I think what you can see with us is a track record of doing two things. First of all, we do sometimes recognize that we’re not able to compete effectively. Obviously, a long time ago, we pulled out of Germany. More recently, we pulled out of China with price comparison. Unsurprisingly, we don’t tend to talk, revisit those things that we’ve pulled out of, but we are willing to do that when it’s necessary. The other thing that we consistently do is we look to minimize risk when we’re trying to do new and exciting things, often in the form of partnerships through reinsurance internationally. And in the case of Compare, for example, through partner shareholders that have reduced the risk of the businesses that they’ve been helping us on going forward. We will take a rational view of the potential of each individual business. And I think rationally when we look at Compare, we see a lot of potential upside as well as a lot – a lot of risk.

Thanks. Johnny Urwin, UBS. Two questions, please. So firstly, on the internal model, what would you need to see to be comfortable submitting the application where do you need to get to? And secondly, on the consumer side, so we run the survey across Europe looking at consumer sentiment and demands, and how you guys are doing versus those. The two interesting things relevant to Admiral this year, firstly, it looked like Continental European shoppers were shopping around a bit more. Are you seeing that, firstly? And secondly, on L’olivier, the French brand, actually, there’s a big spike in consumer awareness this year. Has there been any specific marketing campaign? Thank you.

Great. The two tickets to Paris are waiting for you outside. Milena, do you want to talk to those?

Yes. So first question is, are we seeing Continental European consumer shopping a bit more? I would say is it’s a mixed picture. In general, the reason underlying trend of more shopper, but with some bumps in the road. In Spain, for example, we’ve seen a quite weak first half of the year, but much more – a bit stronger second half of the year. In France, we have seen some increase. It was a pleasing year until the very last quarter, that was quite disappointing. L’olivier, the increasing awareness, I think that was your second question. And we’re very pleased with the results, to be honest, in terms of brand awareness. I would say everywhere, but particularly in France. And yes, we came up with a new campaign. I think it was the very end of 2017 and looking at new one, and we are quite pleased with the results. What makes a bit of a difference in a different country is the level of media spending of our competitors. So in Spain, for example, we have very strong competitors that used to spend tens of millions per year. And in France, the direct spend in the market is not as great, so we tend to have a bit higher return on media investment.

[Indiscernible]

On the internal model, I would say a couple of things. Firstly, we recognize it’s disappointing that it’s a delayed. I would say that it’s 100% knockdown to the efforts of our team who worked very hard on this and now are disappointed that we see a further delay. There are a few things that we need to see before we’re ready to apply. Firstly we need to finish the work on confirming that the scope is right. We expect to do that in the very immediate future. And then, secondly, we need to see process of cleaner validation, an improved set of documents to sit around the model and a number of other smaller things. Those are the key things. We’ll update further in six months.

Thank you. Andreas van Embden, Peel Hunt. Just coming back to the large BI claims. I just wondered, could you just remind me why that trend is continuing in the third quarter? Are more sort of settlements falling outside of the claims portal? And are you going more to court or are lawyers becoming more aggressive? And PPOs, just what is the trend? And just looking at reserving, is this sort of large BI inflation, is that just the new claims or really only frequency? Or do you also see some severity on your case reserves? Because if I look at the sensitivity to the ash index, you see it’s doubled, more than doubled. Is that to do with the large BI claims or just PPOs? And then finally, on your reserve margin, if this large BI trend continues in 2019, would you probably hold on longer to that margin over time and be more cautious on what you release? Thanks.

Okay. I’ll throw the ash one over to Geraint and the margin one. On the large BI claims, this is all about severity of 100,000 plus new claims. It’s nothing to do with settlements. And the 100,000 plus claims, the smaller ones of those will take two, three, four years. The bigger ones will take six, seven, even if that’s shorten 20 years, to settle. So it’s all about new notifications. And we have seen in the past that that can be – the volume of new notifications, these aren’t very big numbers. It can vary quite a lot quarter by quarter, year by year.

Actually, inflation, one implication of the change in Ogden discovery is a change in PPO propensity assumption, so we’ve got twice as many PPO cases, and therefore, the sensitivity is twice as big, assumed cases rather than actual cases, I would say. Margin in 2019, we need to wrap it up, I’m being told – the margin in 2019, if we see further volatility on large injury claims, then that would encourage us to retain a margin. If we see much more normal experience in 2019, which is more like 2017, then we’d bring that down further.

Sorry, we are running – run out of time, I’m afraid. There was no one on the phone? Okay. So thank you very much for your questions. We will be hanging around, some of us, for a while if you want to pick up anything you haven’t been able to pick up in open forum.