Hiscox Ltd (HSX) Earnings Call Transcript & Summary

Good morning, ladies and gentlemen. My name is Robert Childs, and I'm Chairman of Hiscox. We have before you the normal panel of executives, but we have a newcomer this year. We have Joanne Musselle as our Chief Underwriter. If you remember, she's taken over from Richard Watson, who retired last year. I've been at Hiscox quite a long time. And it's very satisfying having been here this length of time to see that our long-held strategy of balancing the big-ticket business with retail business has paid off. So it allows me this year to announce a very resilient financial performance. Our gross premium is up 8% in constant currency. PAT, $53 million, and the dividend, we're going to put up by 3.6%. It's very important to say we have a robust balance sheet to support growth, and we're well reserved at 9.5% above the actuarial estimate. We have opportunities ahead, which you'll hear later. It's the third year of rate rises in the London market. In reinsurance, we're disciplined, and we're ready to capture any upside. And we're investing for retail growth on already what is a very good business at $2.2 billion. Anyway, to tell you further details about the financials, I'm going to hand you over Aki Hussain, our CFO. Aki?

Thank you, Rob. And good morning, everyone. I'm Aki Hussain, Group CFO. 2019 has been a challenging year for our business, but with the changes we made and we are making in our retail division, and with the positive pricing momentum in London market, we're optimistic about the future. Now turning to our financial performance for the year. Across the group, we increased our revenues by 8%, with each of our divisions delivering growth. We have seen a material increase in loss costs, and this has been driven by a combination of weather-related events and a more general increase in claims. We delivered a profit before tax of $53.1 million, and we are increasing our final dividend by $0.01 to $0.296 per share. Now as usual, I'll take you through the financial performance of each of our divisions in turn, beginning with retail. Our retail revenues increased by 7% in the year on an accelerating trend with the second half increasing by 8%. In the U.S., as we advised last year, in the second half, we saw a significant uptick in revenues. Now as the full year is up 7%, the second half alone is up 11% in terms of revenue. Europe had another fantastic year with revenue growth of just under 16%. And in the U.K., we delivered a steady headline growth of 4%. The 4% in the U.K. masks a couple of underlying trends. In our Commercial Lines business, we increased revenues by 9%. This has been partly offset by the reduction we saw in our underwriting private client and high-value home business as a result of taking underwriting and pricing action, which has been successful during the course of the year. We did see lower customer retention. That has now stabilized. As we look forward into 2020, we expect U.K. business to pick up growth rate. The loss ratio in our retail business has increased by 5 percentage points to a respectable 49%. The factors driving this are almost exclusively in our U.S. retail business, and they essentially boil down to 3 key items: firstly, as we advised last year, we have strengthened the reserves on our private market D&O book. We've also, in response to slight lengthening of claim settlement periods and increased legal fees, we have [ look at ] a more cautious approach to reserving. And finally, we are seeing a slight increase in severity and frequency on our small-ticket U.S. general liability book. Now of course, we've not been idle in the face of these emerging constraints. We've taken significant action. On D&O, as you know, we started taking actions from 18 -- or 15 to 18 months ago. And in terms of revenue [ reduction], we've reduced the revenues for the D&O book from a peak of $80 million to under $20 million in 2019. Around that retained book, we're also seeing significant rate increases in excess of 10%. We're also enhancing the capabilities and increasing the resourcing in our U.S. Claims Department, and this will reduce external legal fees as we go through. And of course, we're always taking underwriting action, and you'll hear much more from this -- on this from Joe in a few minutes. And for the first time in a number of years, we're also seeing material rate improvements in parts of the U.S. book. Now these factors, in particular, the reduced exposure to D&O, is already leading to a reduction in our current year loss ratios in the U.S. business. We're reporting an overall combined ratio of 98.7%. This is in line with our revised guidance, albeit outside of our normal target range. Notwithstanding this, the upward trajectory of profitability in retail has continued and at $178 million, we're reporting a profit increase of 22%. Moving on to Slide 5, our London market division. We've seen continuous and persistent positive pricing momentum in our London market division, and we've grown into that rising tide of prices. All across the full year, we're up 11%; and in the second half, the growth rate accelerated to 16%. We have experienced higher loss costs in our -- in this division, in part driven by weather-related events, in particular, Hurricane Dorian and some adverse development in the prior year, as we reported in June or July last year. The most significant effect in London market has been as a result of less favorable experience on our property binder portfolio. Now this is something I spoke about last year. This is an area of the business where we have been taking action. The action is having a positive effect, but not in line with our expectations. As a result, we're taking further action in 2020, where we're pushing further rate through this book, and we will be reducing our capacity on offer to the property buyer business. We've also had some large losses on Alder's account in D&O as well. Moving on to reinsurance and ILS on Slide 6. Here, we're reporting a 7% growth in revenues. The pricing recovery here has been more modest. It's been in the single digits, and our growth is reflecting that. And as we go into '20 -- into the 1/1 in 2020, again, the pricing recovery, whilst on an upward trajectory, has been somewhat more modest, and we will remain disciplined. In 2019, we experienced significant increase in claims, primarily as a result of our exposure to Japan, where we've set aside net reserves of $130 million for typhoons Faxai and Hagibis. We've also strengthened reserves on health care, which is a line of business we exited in 2017. As a result, we're reporting a combined ratio of 164% and a loss of $94 million. Now for those of you who've been following the business for some time, you'll note that the Reinsurance and ILS division is at the more volatile end of our business and it's much more cyclical in nature. So it's worth considering the performance of the business over a longer period of time. And over the last 10 years, inclusive of the loss in 2019, this division has provided around $700 million of profit to the group. In terms of our [ ILS ] fund, where we continue to report assets under management of $1.5 billion with deployable capital of $1.3 billion. The gap -- the $200 million gap, reflects the additional buffer capital we hold back for paying claims. Over the last week, we have received a redemption notice from one of our major capital backers. We expect around $240 million to redeem from one of our higher-risk ILS funds. This redemption will take place over the next 15 to 18 months. The capital is still available to be written against for the next few months. We expect some of this capital to flow back into one of our lower-risk funds, and we also expect to offset some of this outflow with new money that we're attracting during the course of the year. Moving on to Slide 7 and reserves. We continue to report resilient reserves and a positive aggregate reserve development of $26 million. This is significantly down on the prior year and as you can see on this chart behind me, that is coming through as a result of less favorable experience on more recent vintages. We're also disclosing a buffer above our actuarial estimate of 9.4%. This equates to around $300 million. If you move on to the next slide, Slide 8. This provides a great deal more insight into the source of reserve development. And as you can see, we continue to report significant positive reserve development from our retail business of $46 million. Again, this is down on the prior year, but most of this is explained by the deliberate action we've taken to slow down the recognition of favorable experience. The most significant impact on reserve development has come from our big ticket businesses. And this, in essence, falls into 3 categories: firstly, the natural catastrophes. If you recall, in 2017, following Harvey and Maria, we set aside prudent reserves from which we saw significant releases in 2018. In contrast, in 2019, we had some reserve -- adverse reserve development. The combined effect is a year-on-year reduction in reserve releases of $90 million. We've also strengthened [ in terms ] of health care, and we're seeing less favorable experience on some current lines of business. As we look forward, we expect the normal reserve development pattern to emerge over the next 2 to 3 years. For 2020 specifically, our expectation is of positive reserve development of between 3% and 5% of opening net reserves. That equates to around $100 million to $150 million. On Slide 9, we have our capital -- our usual capital disclosure. And as you can see, we are well capitalized on all bases. Now to provide some more detail on our regulatory capital position, I'll remind you that our group regulator is the Bermuda Monetary Authority, the BMA. And our capital regime is the Bermuda Solvency Capital Requirement, which is equivalent to Solvency II. You remember from my updates last year that the BMA is strengthening the standard formula. And as a reminder, we are a standard formula company. That strengthening of the standard formula is taking place over 3 years, 2019 being the first year, '20 -- and then the subsequent 2 years. The strengthening is on a linear basis with an equal 1/3 strengthening in each of those years. Following adoption of that first stage of strengthening, we are reporting a solvency coverage ratio of 205%. This includes or incorporates an 11-point reduction as a result of that strengthening, which we've been mostly able to offset through ongoing capital optimization. If we were to adopt the full strengthening, which is yet to take place over the next 24 months, if we adopted that on our 2019 year-end balance sheet, it would result in an estimated 20-point reduction to our coverage ratio to around 185%. Of course, we expect to offset all of this through ongoing capital generation and capital optimization. So as you can see, we are well capitalized. On to Slide 9 and our investment return. 2019 has been a great year for investment returns. And as you can see from the chart on the bottom left, as a result of this surge -- continuing surging global equities, we were able to report a return on risk assets of -- in excess of 14%. We've also benefited from falling bond yields. If you look at the chart on the top right, you can see that the mark-to-market gains have effectively offset the mark-to-market losses over the last couple of years. Now falling bond yields, as you know, is a double-edged sword. It does result in lower reinvestment yields. And if you look at the chart on the bottom right, you can see that our reinvestment yields or yield to maturity has, in fact, fallen from 2.4% at the end of 2018 to 1.6%. And as a result, as we look forward to 2020, our expectations of investment income are significantly moderated. In fact, our central scenario for 2020 is of investment income of $115 million. Our investment portfolio continues to be conservatively positioned and short duration. As you know, we have been investing in our business, building for growth given the vast opportunities we see in each of our franchises and in particular, in our retail division. Over the last 10 years, we have invested in excess of $0.5 billion in marketing, in building our brand in the U.K. in the continent and in the U.S. We expect that investment to continue. We've also been investing in replacing our technology infrastructure. To date, that's been in excess of $300 million. 2020 will be the final peak year of investment, where we expect to invest a further $100 million or so. Now by the end of 2020, we expect most of this technology replacement -- the technology replacement to be complete, and much of this will be operational and in place. So over the next few years, we expect to sweat these assets and generate operating efficiencies. And as we look forward, our ambition is to reduce the expense ratio for the group from its current 46% to the low 40s and to reduce our retail expense ratio from 50% to the mid-40s. We don't expect the reduction in our retail expense ratio to drive a sustained improvement in margins. We continue to target a combined ratio of 90% to 95% over the medium and long term. But the reduction in expense ratio will provide additional financial flexibility. It will allow us to increase our addressable market and provide further fuel for the growth of the retail business. And finally, as we look forward to 2020. Our expectation for our Reinsurance and ILS division are that we will see a reduction in our gross revenues as a result of a modest recovery in pricing and less third-party capital available for deployment. In our London market, we expect the growth to be fueled by continuing positive pricing momentum. That growth will be moderated somewhat as we continue to underwrite and prune the portfolios that are underperforming. And then finally on retail, and it's worthwhile just reflecting a moment on the compounding nature of our retail business. Over the last 10 years, the retail business has more than doubled in size. Ten years ago, the revenue was around $1 billion. We're now up to $2.2 billion and over that period. We've delivered profits of $1.2 billion with most of those coming in the last 5 or 6 years as the business reached scale. As we look forward into 2020, we expect the retail business to continue to grow between -- in the middle of our 5% to 15% range, and to achieve a combined ratio of 96% to 98%. And now, we'll hand over to Joanne Musselle, our Chief Underwriting Officer.

Thank you, Aki. Good morning all. I'm delighted to be here as Group Chief Underwriting Officer for my first results presentation. So many of you will be familiar with this slide -- chart on Slide 14. It's a chart of our rated index back to 2010 and on a rolling 12-month basis for our 3 major segments: Retail, Catastrophe and Reinsurance, and the London Market. And it's an improving picture, particularly in London market, which is the blue line, we've seen rates up 11%. We've experienced the ninth consecutive quarter of rate increase, and that's an aggregate [ $0.30 ] rate up since the low of 2016/2017. And this is being driven by much-needed discipline in the London market, driven by a reduction in capacity and the Lloyd's Decile 10. And it's not just rating that has shown discipline. We are managing to address the terms and conditions and commission creep is often a feature of the soft market. So overall, rates in 14 of out our 15 lines have seen an increase in 2019. And some lines, like D&O -- public D&O, we're seeing an increase of 60%, and we're seeing this as a good opportunity for growth. In Reinsurance, which is the red line rating, rate correction has been less pronounced with overcapacity still a feature of that market. However, we have seen rates up 6%, more in loss-affected accounts. And we've not seen the Japanese portfolios renew. They will renew on 1/4, and we are expecting a material rate rise after a couple of tough years. From a retail perspective, the retail pricing is less cyclical, with our homeowners and our business customers much more expecting a more consistent premium from 1 year to the next. I think the other thing to note on our retail portfolio is this is a collection of about 50 different portfolios across 10 different geographies. So what you see in terms of the downward trend is as much a function of mix as it is in terms of rate. [ Rather ] healthy. U.K. and Europe are flat and we've seen an uptick in our U.S. rates works. We've experienced a 5% uptick in our errors and emissions portfolio and a 13% in our private Director and Officer. So overall, my assessment is that 90% of our portfolio is operated in markets where the rating supports the results and the returns that we're targeting. So another familiar slide to most will be Slide 15, and this is a different view of the world. This is a segmental view, looking at the different parts of our portfolio. I think the most pleasing thing about this slide is, despite exiting $200 million of business in 2019, of underperforming business, we've managed to grow each single one of our subsegments. The one on the far left, our biggest subsegment, is Small Commercial. And that has experienced double-digit growth, and it's now nearly $1.6 billion. That portfolio has doubled since 2015. And we continue our investment in data, technology and people to both fuel and support that growth. We take the opportunity in that small segment as the premiums of most of those portfolios pay less than 1,000 in premium. So that's $1,000, EUR 1,000, GBP 1,000 and we automate our underwriting -- we automate the underwriting of the small and straightforwards, freeing up our underwriters' time to focus on the more complex and to grow that portfolio through adjacencies. Another area of growth you'll see is on the far right of the slide, and that's global casualty, and that is up 28% in the year. The most important thing to note here is whilst premium is up 28%, exposure is flat. And that's because of the rate rises that I talked about earlier through our Lloyd's business. So in summary, we've got 28% more premium but we're covering the same amount of risk. And lastly, on this slide, Reinsurance. So Reinsurance is up 11% from a growth point of view. But from a net point of view, it's actually a 3% reduction, and that's driven by 2 things. The first is an increased participation with our [ quota ] share providers. And the second thing is we -- as Aki had mentioned, we've exited health care, and that was a net retained portfolio. So looking into some of the segments in a little bit more detail. So firstly, this is the picture for retail over the last 5 years, and we segment our portfolio into 3. The red is what we call course correction; blue is hold, where we're growing, but margins are slimmer, so we're more disciplined; and the green is grow, where we're seeing good returns for our markets. And I'd say the vast majority of our portfolios are actually in the green and the blue. So 90% of our portfolios are actually in both [ways. ] We always expect about 5% to 10% in our portfolios for a variety of reasons. Things change. So as markets change, as markets emerge, therefore, we have to slightly correct our portfolio with regard to risk or rating or exposure. And also, as we enter new things, as we start growing our portfolio, we don't always get it right the first time. But the key takeaway from this slide is, 90% of our portfolio is in that grow and hold of retail, and we see great opportunity. So London market. So as you all know, London market is far more cyclical with regards to this business, and we are coming out of a softening environment. And that softening, not just in terms of rates, but in terms and conditions and commission and, of course, losses. And we have maintained discipline during this period. We've exited $400 million of premium since 2016. And we've come out of certain lines altogether, like aviation, political risk and health care. Now we've not always got the timing of that exit absolutely spot on and what we see this year is some subdued reserve releases from some strengthening in some of those exiting lines. But I'm pleased to say that the position as we leave 2019, as we go into 2020, the outlook is really good. Some of those lines, like D&O, is a line that we've referenced with a significant uptick in terms of rates. That would have been in the red at the start of 2017. But markets change and market change considerably, and D&O is now one of our of growth lines as terms and conditions have reduced and rating has increased. So some of the areas that are [ still in this, ] so course correction. It was referenced earlier. So our portfolio -- our binder's portfolio, property portfolio. And we've seen an uptick in terms of attritional loss ratio. Plus, we have updated our view of risk. We've taken some corrective action. That corrective action will continue into 2020, and we don't expect to see the results fully earned through until 2022. But in aggregate, I'm really pleased with the discipline that we've shown, and I think we're really well positioned as we enter 2020 with the upturn in rates, rising rates. So from a London -- from a Reinsurance point of view, our focus is on disciplined underwriting, but poised to capture any upsides. And we've already shown this discipline on one, where we've reduced some of our portfolio where the risk-reward was not there. We've also refocused our portfolio, exiting $90 million of health care and casualty reinsurance. So back to [ our focus ] and re-underwriting our risk portfolio. As I mentioned, we are expecting to see significant rate rise in our Japanese portfolios, which renew at 1/4 and this is because we've updated our view of risk, both for Florida and for California wildfire, but also for the typhoons -- the typhoons in Japan. So that's why you see, as we enter into 2020, a significant more of our portfolio is in the hold phase as we wait to see how that market reacts. So looking at Japan in a little bit more detail. As Aki mentioned, the 2 of the largest Japanese wind storms in history happened in 2019, and that was Faxai and Hagibis. And we have $130 million reserved for those 2 events. We've picked loss estimates at the upper end of that range. So we're not expecting any adverse development from that $130 million. Japan has been a long-standing relationship for us. We have been there for many years, and we increased our presence following the 2011 Tohoku earthquake. For context, we've had 7 profitable years preceding the 2 loss-making years of '18 and '19. And to Aki's earlier point, this is the business we're in. In aggregate, the reinsurance account has delivered $700 million of profit over the cycle, but it is more volatile, and we do have -- we've had 7 profitable years and 3 not. And as I said, in terms of the updated view of risk, we are now -- that is leading us to now to expect a material rate rise at 1/4, and we'll wait to see how the market reacts. But be sure we'll implement our discipline if that's not there. And lastly, I thought I would just show you how we think about climate change. Our broader ESG framework is in the appendix, but really, we think about climate change in 3 different ways. Clearly, as underwriters of catastrophe reinsurance, we've had a long history of researching and modeling climate change. And we've invested heavily over that in our history, and we continue to do that. We're bolstering that investment with the addition of 2 climate scientists in 2020. The second way we look at our -- the climate change is on the portfolio. So as our understanding of climate risk change and emerge, so does our understanding of the impact of that on our portfolio. And not just in property and reinsurance, but our broader portfolio, so looking into casualty. So as an example, there is no doubt that D&Os will face a greater scrutiny for how they manage their climate risk. And thirdly, as insurers, it's our job to help and anticipate our customers' needs. And so to evolve our product suite so that we help our customers and their businesses become more resilient. And a couple of good examples that we have here is our suite of flood products and our new -- and our wildfire modeling. So with that, I'm going to hand over to Ben Walter, who's going to talk more about our retail customers.

Thank you, Jo, and good morning, everyone. I'm Ben Walter. I'm the CEO of our Retail division. And I'd like to spend a bit of time this morning talking about how we think about the longer-term opportunity in retail, particularly in the context of this year's results and the investments that we made. You all have seen where we ended up on the combined ratio this year at just under 99%. Clearly, that's below our long-term ambitions and expectations due to a little bit of volatility on the loss ratio. What we could have done, but we didn't, was to curtail our investment in the division and taking a bit of upside on the expense ratio in order to deliver a lower combined. We didn't do that. We don't plan to do that because we see the long-term opportunity as significant, and we want to continue to invest behind that. And that investment is particularly focused on the small and micro commercial space, an area that is one of our largest now and that we view, and I certainly view, as one that's likely to follow the form of -- the way the auto insurance market transpired, particularly in the United States, where that trend is still in its infancy. And as context, if you rewind the clock, say 25 years ago, you'd would have seen auto insurance being largely manual and a hugely fragmented market with lots of market share strewn across the market in various players. That's changed drastically in the last 25 years. Most of that market is now placed electronically. And as a result of that, you've seen a few scale players emerge, and indeed, the top 5 players in that market now lead with 55% market share. We think something similar is happening right now in micro and small commercial insurance, and our goal is to be 1 of those 5 winners. And we are indeed well on our way. Looking at Slide 22. You can see the growth of that business over the last 10 years to $2.2 billion, as we've grown our original franchises in the U.K. and Europe, and as we've planted flags in the United States and now in Asia. But if you look at the right side of the slide, you can really see why that begs so much investment in operational scale. The business, like I said, has doubled over the last 10 years. But in just the last 5, the number of customers has almost tripled. So in that same time period as the business has grown by only 50% in terms of premium, you've seen a tripling of the number of customers. It takes operational scale to be able to run a business with that many customers. As Jo said, these are people who buy $1,000, GBP 1,000, EUR 1,000 euros at a time, sometimes less. So it's very operationally intensive, investing in automation and scale becomes very important in order to be able to drive a competitive moat between us and our competitors. So before I talk about that scale, I'd like to talk a little bit about what's in the book and give a little more of a breakdown. On Slide 23, you can see retail overall broken up by its component parts. On a personal level, when I moved to the U.K. a couple of years ago, and people would ask what I did, and I said I run retail for Hiscox. People would say, "Oh, that's the insurance you do for rich people." And it is, that's certainly a part of our portfolio, and it's one that we love. It's the cornerstone of our brand, particularly on this side of the Atlantic. We're going to stay in that business. We love that business, and we love what it does for our brand. But today, it represents, even at $400 million, only about 18% of the portfolio. The balance is all commercial, and that commercial really drives -- breaks up into 3 broad buckets. The first is on the upper right. 25% of the book is small commercial package. That's sold mostly through direct and partnerships, and that is a combination of covers, both liability- and property-driven, that we sell to the smallest of businesses. The second bucket is our historically core professional indemnity franchise that we have all across the world. These are in specialties and everything ranging from design, to consulting, to media and entertainment, areas that we know well, products that are tailored to the specific verticals that we choose to serve, claim service that is best-in-breed. And then finally, the last quarter on the bottom left is our collection of what we call the weird and wonderful. So we still are a specialist insurer. We're not a generalist, and that means we will continue to have certain niche markets in which we play, where we have the expertise and we have the history and the distribution to play in those markets. That's a product view. I'm going to put up now a slide -- on Slide 24 that you've seen before. This is a size view. And this really speaks to what's driven the chart that you saw a couple -- on a couple of slides ago about that growth in customer numbers, where we talked -- Jo called this volume versus value, which I think is a really great name for it. And what you can see here is just how much of our customer count is at the bottom end of that. So under GBP 5,000, EUR 5,000, $5,000 represents 90% and of the volume that we do. We do write some larger companies in retail. However, our appetite -- our underwriting appetite gets more and more restrictive as you go up the pyramid, and that's what helps us constrain the volatility relative to the big-ticket businesses. But broadly, the bulk of the business, from a customer perspective, from an operational intensity perspective, sits at the bottom of the pyramid and that's where we are doing most of our investment today. So let's talk about that investment for a bit. There are 3 big areas that you have to scale, particularly in a direct-to-consumer business. The first are the customer contact centers. That's really the coal face to our customers. Those are the people who talk to our customers every day. In that area, we've made great progress. We've seen double-digit improvements in productivity each year for the last 4 years, and we do expect that to continue. The second area that you can really achieve scale is in marketing and distribution. That's the chart that you see here. You can see the purple line represents that our distribution costs in the direct and partnerships business as a percent of premium. And you can see that was declining for a long time, although it's ticked up in 2019. That was a very conscious choice. We did that intentionally because we see more competition coming in, particularly in the United States. We think we still have very small market shares and room to grow, and we want to put more distance between us and our competitors. But scale does come eventually. If I can reference the auto market, again. One of the numbers I like to give is that both GEICO and Progressive, 2 of the biggest players in the market, spend less than 10% of their premium on marketing. However, for each of them, that less than 10% represents over $1 billion. So you do reach scale where you're everywhere, but it comes at a very large number. We will continue to invest in that area, although I do expect, in time, that will start to achieve scale as well. And we're doing some interesting things on the marketing front. So we're doing, in addition to the classic paid search and Internet marketing that lots of online players do, we've done some creative tie-ups with other entities, such as Major League Baseball. And we've done some experiential marketing that's paid dividends in the -- particularly in the area of cyber in the U.K., which you may have seen. And then the final area is technology, and that's where we really haven't begun to achieve scale. We are still very much, as Aki said, in the investment phase. That peak year will happen this year in terms of cash spend. But then, obviously, that will start to flow through the P&L in the form of depreciation. And at full run rate, we expect the total spend to amount to about $30 million to $40 million a year in terms of depreciation running through the P&L. However, once those systems are built, we can really start to sweat those assets as premium continues to grow. As Aki said, we're going live this year with -- we're already live with our U.K. implementation. We're going live this year in the U.S., particularly for directed partnerships and the European project, though smaller, is just starting. But I wanted to give you a sense in the meantime of what these systems and these systems that are built around it can do to give you a sense of how we achieve true operational scale in a customer-intensive environment. There are really 3 areas we've invested pretty heavily in. So I'll start with robotics. Robotics are everywhere. They're all across the business, and they're already having a huge impact. We do over 2 million transactions a year and that's growing double digits every year. The more of those we can automate, the more we can grow without having to add headcount. And we've already automated hundreds of thousands of transactions through robotics. Repetitive tasks that required people to do them are now being done by the software. The second area, and one that I'm really excited about, is natural language processing. We've adopted a platform that's backed by artificial intelligence that can dynamically interpret messages coming into Hiscox and do things like dynamically route them. So as an example, in the U.K., we now have a system in place that can take an e-mail that comes in from a broker in the traditional way they would have sent it to us. It can read that email. It can dynamically learn, based on experience, what that e-mail wants us to do and route that to a person who can complete that task in one fell swoop. And increasingly, we can now put that technology together with robotics to have something go straight to the machine without any human touch at all. Second is -- sorry, third is APIs. You're going to hear a lot more about APIs across the insurance ecosystem in general. APIs are really just the way that one machine talks to another but they're rapidly becoming the standard by which insurance will be traded in an electronic format. To put that in context, 5 years ago, in all financial services, ex insurance, there were 3,000 industry standard APIs. Today in insurance, there are fewer than 300, and it's growing every year. So the insurance industry is slow to catch on, but is catching up fast. And in the U.S., we already have over 30 partners live on our API platform and our U.K. platform has gone live already with its first partner. And then finally, portals and pre-price proposals. This is a new way of trading, particularly that's caught on in Europe, and it now represents over 50% of our submissions, and it helps us again process large numbers of transactions without the people intensity that it used to. So huge investment across the portfolio in automation and straight-through processing and APIs and in various technologies that help us scale the business. And why are we investing so much? Because we see a huge opportunity. This is why I think I have the most fun job in insurance. Today, that small business -- small and micro business market across the world represents over $80 billion of GWP. I'm thrilled that we have $1 million of them, but I'm far from satisfied because that implies very low market shares. And just as critically, these are really fast-growing markets relative to the general economy. Most studies put it at roughly twice the rate of general GDP growth. So it's an attractive large segment with good loss ratios, sub-50%, that when we can run at scale, we think there's huge opportunity for us to gain market share and drive the expense ratio down at the same time. I'll leave you with the picture on the right, which Bob Thaker, who's here today, who's our new CEO of the U.K. business likes to use. That's a jar and every black marble represents one piece of the small business insurance market, and the red marble represents Hiscox. And our goal is just simply to go from 1 marble to 2 marbles. And if we can do that, we'll go from a slightly over $2 billion business to a slightly over $4 billion and on its way to $5 billion, while still being a relatively small player. With that, I'll turn it over to Bronek for our business outlook.

Thank you, Ben. I think at this point of the presentation, it's always helpful just to take a step back and look at the overall shape of the business. So here on Slide 29, you can see the growth year-on-year of the different major segments. Retail grew in constant currency by 7%. And as you heard from Aki, that was 6% in the first half and then 8% in the second half as the impact of some of the corrective action took place and worked its way through the system. We thought it would be interesting to show you rather than a geographic split, which is still in the appendix, a channel split. And in the broker channel, this is in nominal currencies. It was pretty flat overall year-on-year. But if you look below the surface, again, that was a shrinking in the broker channel in the first half as we took some corrective action in the U.K. and the U.S. and then growth in the second half, which is stronger. If you put it in constant currency, it's more positive than that sort of 0.4%. But the other interesting segment, we thought -- other segment we thought you'd be interested in is, in fact, the direct and partnerships. It's now $0.5 billion business, grew -- and it grew last year by 25%. It's the area where, as you've heard from Ben, we have invested a lot in technology in order to replatform our U.K. business. We will be going live in the first 6 months in the direction partnerships business in the U.S. And I know from a investor perspective, that's seen as a lot of execution risk. But in our view, there's an even greater execution risk and in not doing the investments because it will restrict our future growth. And as you can see, we're pretty ambitious for that. In terms of the London markets, it grew by 11%. Again, that was 6% in the first half and over 16% in the second half, which means that if you look back to the chart which Jo and myself put up, we are growing strongly when the rates are rising the most. And again, we think that sets us up well for the future. And in reinsurance. In reality, there's no first half/second half split, as over $700 million of that $867 million was written in the first half. So all in all, we think that the group has done well. But you -- as you've heard, we're pretty optimistic going forward. Rob started the presentation by effectively saying how the diversification has paid off for us. It certainly has. It's a strategy which has been at the core for Hiscox for over 20 years. It's steadily growing that retail business and expanding and contracting in the big-ticket business. And that has allowed us to have a very resilient financial performance, but we have to be frank. The results for 2019 are below our ambitions for ourselves and your expectations of us. And for that, we apologize. And you can hear today that we've taken a lot of action to address that, which we hope will pay off during 2020 and 2021. And we've seen, as Aki said, some emerging positive trends, whether the actuaries will allow us to recognize it, that's what we have to wait and see. But ultimately, we're not happy, and we've done something about it. The one thing our past conservatism has given us though is a robust balance sheet. The P&L hasn't been great, but the balance sheet is actually in pretty good shape. And that really, to me, provides a very good basis from which to launch into 2020. And that, from a shareholder point of view, means we can be well-funded to take the returns that we see out there. In terms of the business, as you've seen from Jo, disciplined underwriting still remains a core. We are ambitious to grow, but if underwriters have the choice between good underwriting and growth, we always want them to take the former, not the latter. I think it's probably worth just touching on 2 of the issues of the day, coronavirus and the flooding and storms. In terms of coronavirus, we don't see this as a material impact for Hiscox. The areas where we're most exposed are obviously in the events business, and then to a much lesser extent, travel and business interruption. On the events business, buying pandemic cover is an extra cover that people have to buy, and less than 10% of our customers in events have actually bought that. And in terms of the ability to trigger the cover, that requires government action to actually stop the event. So we're not that worried about the event exposure as we look forward. We have had some small travel losses, obviously, and we've had a little bit of business interruption-related expense. But again, neither of them, particularly in business interruption, pandemic cover is not a core part of the offering. So whilst it's very newsworthy, we don't think it's going to be material from a financial perspective. Operationally, Hiscox can actually run the entire business remotely. We've got capacity to run over 3,000 people through remote dial-in. One area we haven't yet cracked is call recording for our U.K. service centers. But I'm sure the PRA is in the room and the SA should be in the room. I'm sure that if the whole world starts working from their home, there might be a little of regulatory easing in that perspective. That's a request. The other area, clearly, are the storms and the floods. As of Friday, we had just over 100 flood claims. Some of those are going into FloodRe, the rest we'll retain. But if you take the storms and the floods together, we are into our reinsurance program, so our losses are kept at GBP 10 million as of today. So we feel, again, pretty good about that. One of the broader issues in terms of the coronavirus that people are beginning to worry about is clearly the impact on global economic growth. It's at times like this that I'm pleased that we're in the insurance business. When times are really good in the economy, we don't see the big surges of growth. But when the economy slows down, we don't see a material reduction either. We're not a luxury goods. If you have a home, if you have a car, if you're in business, you have to care on buying your insurance. And indeed, if you look back to 2008, '09, 10, in the last sort of material slowdown, we actually saw an uptick in the number of those micro businesses that we insure because people who are made redundant from larger companies, often through -- not through choice, but through necessity, set themselves up as independent consultants that then need to buy insurance, and they come to us. And on the claims front, there's -- was a small uptick, but that happened over 2 or 3 years, not to some big overnight surge and was well within what we thought was manageable at the time. So I think from an investor perspective, actually, if there is a global downturn, general insurance, as a whole, and I've got to say Hiscox specifically, is a good place to be. As we look forward in terms of the London market, clearly, we're into our third year of price increases, and we're seeing positive price movement across virtually every line. And clearly, we are trying to focus our resources where the price movements are the most attractive. In terms of reinsurance, we have to be more selective picking our way through the marketplace. Our reinsurance volumes were down at the 1st of January versus the 1st of January last year as we felt we weren't being paid for the risks that we saw. But that means we have more capacity to deploy during the course of the year. The 1st of April for Japan is clearly a material decision date for us. If the market pays the sorts of increases that we're looking for, well then, we will deploy our capital. If it doesn't, we'll shrink. And that's fine. That's consistent with that long-term strategy of growing the retail business and only deploying our capital in the big-ticket business, when we think it's warranted. We don't have a growth ambition, particularly for that area of business. On the other hand, in the retail business, as Ben outlined, we are pretty optimistic and ambitious. In the U.K. business, we want to go from one red marble to a second red marble. That will take a number of years, but the market is there, and we think we've got the ability to capture that. Our European business is now about half the size of our U.K. business. And I can remember when that was 10% of the size of the U.K. business. It is growing fast. And certainly, our ambition is that in the first instance, that the European business should be the same size as the U.K. business. And as we look at the market structure in -- particularly in France and Germany, in the longer term, I personally see no reason why each of those countries shouldn't equal the U.K. in size. In a funny way, Brexit has been good for us because it's created a separate legal entity for the European business with greater focus, greater independence from the U.K., and you saw that in the 15% growth that they delivered in 2019. And when we look at America, as Ben said, we see a small business market, which is very fragmented at the moment. It isn't a winner-take-all market, but it will consolidate over time. And our ambition for 20 years or however long it takes, is to be 1 of those top 5 in the marketplace. We have the product, we have scale, we have the knowledge to be able to do that. So if you take a step back and look across all 3 segments, we are positive. It's never a complete one-way street. But be under no illusions, we're determined to do better in 2020 and beyond than we did in 2019. And we think that we have the capability to capture the upside. And with that, we'll take any questions. Before we go into questions, we are on -- being recorded. So if you can -- there's a mic in your seat next to you. And if you could just say who you are and so the people remotely know the questions.

It's Kamran Hossain from RBC. Three questions. The first was just on the reserve release guidance for next year. So you pointed to 3 to 5 points of net reserve releases next year of opening reserves versus the kind of historical kind of normal. I know it's very difficult to give a historical normal for around 12. How should we think about the transition between the 2 numbers over time? Do we step back to that next year? Or how do we think about that? The second question is just on the expense ratio and I guess, the low 40s ambition. Obviously, it was retail growing substantially and the suggestion that you, I guess, deploy that additional expense margin into growth. Should we expect the overall group expense ratio to materially benefit or say, see a slightly smaller benefit than if you were kind of as at today? And the third question is, could you talk about any impacts on the brand from, I guess, the issues in the U.S. last year? Are customers noticing? Or have there been any kind of NPS impacts there?

Okay. So I'd suggest that Aki takes the first 2 questions on reserve releases and the expense ratio, and then Ben can talk about the United States. Aki?

Thank you, Bronek. In terms of reserve releases, the guidance for 2020 is between 3% and 5%, as I said, $100 million to $150 million. I mean, over the longer term, the trend has been, if you can call it a trend, somewhere between sort of 8% and 12%. Our expectation is that we would trend back to that over the next 3 years. So not only immediate snap-back in 2021. I'd let it -- I'd run the trend line for the couple of years. It's important to remember where the reserve releases are coming from. They're coming from our actuarial estimate and for the margin that we hold above that estimate. The margin has remained intact. What we have done during the course of 2019 is increased the actuarial estimate. Part of it is deliberate, and this is the slower recognition on favorable experience in retail. That is going to take another 18 months to come through. So that's a natural governor on the amount of reserve releases that we'll see over the next couple of years. In terms of expense ratio, the current expectation is that in the business, the retail business is reaching scale. As you heard from Ben, we are generating efficiencies in our -- in the marketing expenditure. But the big efficiencies are going to come from the nearly $400 million that we've invested in replacing technology around the group. By the end of this year, most of that will be operational. And our expectation is that going into 2021 and beyond, for the next few years, we would expect to see the retail expense ratio fall by around 1% per annum. That should translate into 3/4 or 1% into the group expense ratio also reducing. That's not a forever trend. But I would say, for the next -- beyond 2020, for the next 3 or 4 years, you could assume that.

Ben?

Yes. My short answer is no. The longer answer is, if you look at where we've had to retrench, it's mostly been in D&O and Media, which are broker-only lines. And in most of those cases, either specialist brokers focused in those areas, who know the issues very well, largely wholesale-driven. So most of our brokers in the U.S. are wholesalers. They're used to a more, I'll call it, [ warp and weft], like you would see in London as opposed to a sort of small-ticket retail broker. We haven't had to make wholesale changes like that in our Direct and Partnerships business, which the most sort of direct-to-the-consumer and customer-facing So our NPS scores remain as and when they were. Our brand numbers actually hit an all-time high in the U.S. among that small business population. And then the other thing I would say, the other place where we were -- that's always at risk is in our claims service. I mean, as we reported to you, we have invested more than we'd like in some cases, in outsourcing some of the some of the legal expenses that come through. But part of that has been, as the business has grown, we have not been willing to back off of what we still consider an industry-leading claim service, and we've invested in that. So our customers who do have claims from us have a good experience, and we believe that pays dividends in the long run, and it's what the customer is buying. So broadly speaking, I suppose if you're a niche D&O broker who lost a bunch of business, you're not too happy with us, but you're a professional, and you understand how the market works. General customers, we've seen that continue to go north.

So next question.

It's Ivan Bokhmat from Barclays. Two questions, please. On the retail, could you speak about the pricing you achieved in connection with the claims inflation that you have seen, maybe separately for U.S. and the rest of the businesses? And I'm wondering what you saw in 2018 and '19, whether the numbers are materially different and what you think would be the kind of the long-term trend for the BREC business in particular? And secondly, Joanne, maybe a question to you. You spoke about growing in D&O right now. Just wondering whether you are seeing -- what risks you are seeing to that strategy? Which geographies you focus on? And what kind of reinsurance protection you've got?

I think -- thank you. I think both of those actually are for you, Jo: some of the terms of the retail claims inflation versus pricing and then the D&O.

Sure. So I'm just talking about -- more generally about the retail increase in price. What we've seen is about a 5% uptick on our error and emission portfolio and 13% on the refocused private D&O portfolio. As Aki and Ben both alluded to, we've scaled back that private D&O portfolio from $60 million to $20 million. We've picked the $20 million of core, which actually had been profitable over that period. And we are seeing an uptick of 13% because the wider market is increasing. More generally, we're not immune. Whilst we're not directly impacted by social inflation, we're not immune to that, to sort of the casualty stories in the market. And so therefore, we are seeing a general uptick on our smaller business. I think from a particular social inflation point of view, there's a lot of schools of thought around social inflation, but this is really looking at the sort of the jumbo awards, particularly focused in general liability, high access and umbrella. And actually, across our business, even including our bigger-ticket business, we've got less than 2% of our premium, which is in that area. So the -- so what we're seeing is a general uptick in our -- on our premium, which is good news, and we'll -- and we're seeing that continue into 2020. And then particularly on the London market D&O. So this is the public D&O particularly focused in the U.S. And as I said, 3 years ago, that market was in the course correction. It -- the prices were deflated. Claims were higher than the prices that we were achieving. And also, we were in a soft market, where terms and conditions were very wide. But what we've seen is a complete turnaround in that market, obviously driven by losses and driven by a reduction in capacity. When those 2 things happen, then obviously the market changes considerably. And we're seeing a sustained increase in rates. We talked about 60% in 2019. We're seeing that continue into 2020. That possibly will slow down during the course of the year. We've not seen that yet. I suppose the big thing is the exposure that I talked about. So whilst we've increased our premium, the actual exposure for D&O has gone down. So we use a measure called Rate Online, which is basically a measure of premium versus the exposure. And our 1% rate online has gone to 3% rate online. So you can see that as our premium increases, our -- actually exposure has decreased. So we absolutely see that as a market that we want to grow and develop in and that's within our plan for 2020.

A small follow-up, if possible. In November, we spoke about the red part of the portfolio. There was quite a particular focus on that. I'm just wondering, over those 3 months, where do you think is that as a percentage of the book, where you need corrective course overall? And where have you made most gains, you think?

Sure. So in aggregate, across the whole of our portfolio, 90% is in grow or hold. So 90% of our portfolio is not in the course correction area. And that slightly varies by the different parts of the portfolio, but that's the view -- [ that's the ] thing in aggregate. So we feel like we have come out, particularly London market, come out of a soft market. We have been disciplined. Since 2016, we've exited $600 million of underperforming business, $100 million in retail, $400 million in London market and $100 million in reinsurance. So therefore, we feel now the portfolio is positioned. And what you see with those graphs at the end of '19. We feel that, that portfolio is well positioned as we go into 2020 to take the opportunities as those arise in the different segments.

Okay. Great. Let's take one from the middle, and then we'll go to here, and then we'll go across the room, otherwise. So who's in the middle? Go, Jonny. Sorry, you're from the home team, so you should know how to work it.

Jonny Urwin, UBS. So firstly, on the reserve buffers. Thank you very much for the disclosure. That's great. The 9.4%, where does that compare to your kind of comfort range or target level? Just to get a feel for how happy you are there? And then secondly, on the in-sourcing of the legal function in the U.S., how's that going? What's the timing?

Okay. Great. I'll take the first one to Aki, and then the second one to Ben.

Sure. In terms of the 9.4% coverage or the buffer, that's $300 million, roughly, and that's pretty much in line with where we expect it to be in a kind of normal sort of operating year.

Ben?

On claims, I think it's important to say we had never outsourced our claims function, even in the U.S., are you -- it has always been in-sourced. But that business grew at a 30% clip for 4 years in a row. And the reality was sometimes then the claims don't come in right when you grow, so they come in spurts. And at times, we would outsource components of that from the internal team to keep up. We've started reigning that back in. We've hired over 20 people just in the last few months, rehashed our training, and that's all going well. It will take time, but we can already see the numbers of loss adjustment expense coming down.

All right. Over here. Andreas then Andrew. Sorry.

Andreas van Embden from Peel Hunt. Three questions, please. One on your binder portfolio here in the London market. You mentioned that further action is underway. Could you maybe comment on the size of the book and what actions you're taking? Is this a material shrinkage of property binder? The second question is on capital and the optimization that you're doing. What type of optimization is this? Is this part of sort of the derisking maybe of the reinsurance book? And finally, on pandemic risk. You disclosed $175 million sort of exposure in worst case scenarios, Spanish flu in your appendix. Is that all personal accident? And is this a realistic scenario given what you commented on before about your exposures to travel and business interruption? Is this something separate, which is really a cat event rather than what we're seeing now?

Okay. So why don't we go Jo on the binders and maybe the pandemic and Aki clearly on the capital. Maybe Aki, you should go first, on the capital.

Sure. Sure. So in terms of capital, just as a reminder, our solvency coverage ratio is 205%, and that includes an 11-point strengthening on the BSCR. So from last year, where we reported 210%, absent any of the optimization, where you've seen that coverage ratio reduced to 199%. The improvement has come through from a couple of areas. One is just not anything specific, but just the natural evolution of the reinsurance program and how it evolves over the year. But the more deliberate action has been that as the business has grown, both in terms of geography and the types of products we write, there are optimization opportunities, which are technical in nature, so I won't go into the whole depth of them. But there is premium diversification coming from both the types of products we write and the geographies in which we write. We've not taken advantage of this in the past, and that's something that we have done in 2019.

So Jo?

Yes. And then moving on to the binder portfolios. So as we mentioned, we did do some corrective action in 2018. And whilst that had an effect, there was an underlying increase in trends. So it didn't have the required effect that we wanted, particularly attritional loss creep on the -- and that is obviously through [ its own, a ] benefit. And we've increased our view of risk. In terms of what difference it will make, we are deploying a $4 billion less aggregate in Florida as well as changing some of the terms and conditions of the individual binders. As I said, that this will take a while to earn through and we won't see the full effect until 2022, but we will start to see the benefit this year and to next year. And then in relation to the pandemic, yes, you're absolutely right. There's an RDS at the back of the appendix, which has a gross loss of 350 and a net of 175. I'll just give you some of the scenario that we've used, just so you can see the relevance to the current potential pandemic. So we've picked an open-source bridge risk pandemic. It's very similar. We called it -- it's very similar to sort of Spanish flu. So in terms of what we look at, it affects about half of the population and it has a death rate of about 25 million.

Half the population of the world, that's right. Okay.

So half the population of the world and the death rate is 25 million, and it wipes off $17 trillion from GDP over a 5-year period. So you can see, it's a pretty extreme event. And obviously, in comparison to what we're currently seeing, clearly coronavirus would have to have a major impact in the sort of Western world for us to be anywhere near that. The majority of those losses that we built in the scenario are driven by sort of events cancellation, et cetera. But yes, in terms of this type of event that we've chosen it is a RDS, a disaster scenario.

We're a long way from that level now.

We are.

So I think you can relax a little bit. I saw that and I asked the same question. Andrew?

It's Andrew Ritchie from Autonomous. Four questions, I'm afraid. I think they're quick ones. On the U.S., I think Aki, you said the current yields ratio is now improving -- sorry, U.S. retail. I'm a bit confused. I thought the current year loss ratio was increased in '19 to reflect some of the issues, and you're kind of going to hold it there. Obviously, there is a bit of noise from PYD as well. But just understand, is the current year at U.S. retail already improving? Current year, not just current accident year and I guess, would you expect that to continue improving in '20? I guess you would because you think rates ahead of loss cost, but clarity on that. Secondly, Ben, you said that you increased the marketing spend a bit because of more competition. Is that specific? I know what Berkshire launched a new small business. Is there anything slightly more scary about recent competitors? I appreciate the market opportunity is huge, but do you think some of your -- the new entrants are getting a bit better? Thirdly, I think you've got 4 -- at least 4 closed books that I can think of. You're holding a big margin reserve best estimate, presumably on some of those books. I'm sure there's a lot of people out there who'd willingly take those books for a lower margin of a best estimate. Is there possibility of exiting some of those closed books through other people's balance sheets? And finally, 3 people mentioned the reinsurance profit over 10 years. It feels like you're trying to justify the reinsurance business or you feel somehow, you need to. I guess, by any point, would be, maybe you look at return on capital because it does consume a lot of capital, not just the profit? Do you think the reinsurance business is covering its cost of capital over time?

Okay. Well, that's quite a comprehensive list. So why don't we go with Aki first on the U.S. loss ratios. Ben, clearly, in the U.S. competitive environment. Jo on the -- what we aim to do in terms of our runoff books. And I'll take reinsurance at the end. So Aki?

Okay. So on U.S. retail, you're right, we have previously said we've increased the loss [ picks. ] But that is not across the board. That is on some selected lines. When you strip out the effect of prior year reserve development and the cautious approach we've adopted to reserving, the underlying loss ratios for the regional business are improving. They're improving, largely driven by the action that we've taken on the D&O book, which was a high-loss ratio business. And as you've heard earlier, we've significantly reduced our exposure and working rate on the retained book. The other -- the remaining management actions we've spoken about, whether it's further in-sourcing of the legal component of claims or the rate increase that you've heard from Jo, those are not yet reflected in those issues. Those will come through over the next few years.

In terms of the competitive environment in U.S. Small Business, it is absolutely heating up. There's no question about that. We've expected that to happen for a long time. My broad view remains that, that's good for us as opposed to bad for us on balance because it's validating direct end partnerships as a way to buy the product, and it's growing -- that distribution segment is growing as a share of the overall pie. In terms of specific competitors, sure, Berkshire's out there. There's a big new one called NEXT, which is -- has done a good job, and they're backed by MunichRe. But as I like to tell people, they -- when they sorted $80 million of run rate premium, MunichRe bought them at a $1 billion valuation with a 25% share. So I think that's a pretty good multiple on this type of business. But broadly, the space is maturing, and that means there will be more competition. I do expect some of the big traditional players eventually -- or direct players eventually to cut on and start being part of it. We're prepared for that. And our goal is to have the biggest head start we can in terms of brand and distribution capabilities and product scope before that happens.

Jo?

Yes. So with regard to this loss portfolio transfer that you alluded to in some of the other -- in runoff. I mean yes, that absolutely is something that we consider. What we've done that -- historically, we have a few that we've actually done that with if the deal is rightly rewarded. Obviously, it has a benefit of capital as well. But it is something that we look at on a case-by-case basis and we look at the merits on the individual portfolios to see if it's attractive.

I'd also add, we have done those in the past. We did one for a U.K. PI portfolio D&O portfolio. How many years ago was the deal on?

So we've done about 3 years ago.

Yes. So we're not, even when we are reserved, we're very happy if somebody gives us a bid we think is worthwhile. Or actually, we just want to cap it and move on. We're very happy to get into those sorts of transactions. In terms of the reinsurance, the way I always think about the reinsurance capital. If you look at the -- I always sort of think that in reinsurance, we're probably allocating for every dollar of net written premium, $2 of capital. Last year, we had $212 million of net written premium, so it's about $450 million, $500 million. And on that basis, $700 million over 10 years is a pretty attractive return on equity. What it goes to is actually the benefit for each part of the business from diversification gained from the other parts. If reinsurance was a stand-alone business, it'll probably need $4 per $1 of net written premium. So being part of the broader group, that's where the capital efficiency of the diversification pays off. People focus on the P&L impact. We're very aware of both of those. In terms of defensiveness, yes, don't get me wrong, I'm pretty irritated with that performance in the last 3 years in terms of that. And if you recall after 2017, I tried to lead the market up because I thought after that year, prices needed to go up. A lot of people didn't agree. So prices didn't go up, but prices need -- we've seen our reaction in the London market business and the reinsurers are being squeezed. So I think inevitably, economics is going to play through. On the one hand, we've had some redemptions in the ILS fund, which is from an asset management perspective, a bad -- and fees and profit commission is a bad thing. But from a market discipline perspective, an even better thing. So to that extent, we have within our armament, the ability to change course and adapt to changing capital flows, and that really is what the business can do really well. So Ben. And then we will come across Paris and then...

Ben Cohen at Investec. I think Andrew answered -- asked most of my questions. Just to follow-up on the reinsurance side. Where -- what sort of current level of margin do you think that business is capable of earning? And if it's a sort of a multiyear story to get it to the sort of target margins that you would like, over how many years do you think that will likely happen?

I mean, I think the reality of reinsurance is a far more 1 year repricing deal. That's why we've been quite clear that if they don't hit margin this year, we will shrink. So I do think that last year, if you took an expected loss ratio norm on a long term, reinsurance was an attractive place. You have this -- the other thing which does muddy the waters a little bit is we have internally updated Hiscox's view of risk, which is a moving feast as well. But we think that the book will end up within calendar 2020, if we have a normal loss year, will be attractive margins. So we don't think it will take a long time. It's different for the retail business and the speed with which you can reprice that. Sorry. Paris?

Yes. Paris from Exane. A few questions. So firstly, on ILS fees. $25 million impact this year. You do flag some redemptions. Basically, from next year onwards, what kind of payback are you going to get or how much of this $25 million you are actually going to be making back? Then I need to come back on this buffer over actuarial best estimates. You are flagging $300 million as a comfortable level, but you are saying that over the long term, you are targeting this 8% to 12% reserve releases, which should mean that from next year, you are actually expecting this $300 million to go up because we are releasing less than the 8% to 12%. And then lastly, just a clarification on the retail business. When we last met you at the 9-month trading update, I think you were saying that you are going to review the whole portfolio. And because you have been growing quite a lot, there were certain lines of business that you might have to relook. You are only flagging D&O and Media. Is there anything else that you think you have grown too fast or profitability doesn't really go your way and you will be taking a step back?

Okay. I think for the first 2, on the ILS fees and the buffer, Aki, that's for you. And then on the portfolio for Jo.

Okay. So let me start with the ILS fees. Now there's a kind of irony here. So as money leaves the investment fund -- the ILS funds, and new money is attracted, the new money is -- there's much higher probability that new money pays PCs, profit commissions, because the old money that's left, that was subject to a high watermark and so on. So actually, money recycling out of the fund will accelerate the pace at which we go back to earning $25 million, $30 million, whatever the number is for profit commissions. In terms of how you should kind of think about modeling this. The thing to note is, not all of the ILS -- some of the ILS funds are below what we call a high watermark, but some are not. Some are earning profit commissions today. So I would say the best way to do this is assume a trend that we get back to a normal profit commission level again over the next -- probably the next 3 years. In terms of the buffer and the reserve releases. You're right, we do have a buffer of $300 million, and that equates to about 9.5%. But when you think about reserve releases, they, in essence, come from 2 sources. They come from a favorable experience relative to actuarial estimate. And also as the margin is released on all the years of account, okay? Now, as we've mentioned earlier, we are -- we have refined the reserving approach for a retail business, where we're taking longer to recognize the favorable experience. That is acting as a natural governor on the pace at which we can release from our reserves. That does not necessarily mean the margin will go up because that favorable experience is held within the best estimate.

And with regard to the retail portfolio. I think if you recall, we've got 90% of that portfolio in grow and hold. So there is about 10% which is in the sort of corrective action. I suppose what we -- I would talk about is that we constantly course-correct portfolios, the risk change, the market changes. So we're always slightly tweaking our portfolio. What we referenced in some of the more extreme course corrections, so the downsizing of portfolio from $60 million to $20 million on the D&O account is what I call more extreme course correction. There is always more subtle and temporary course correction that happens in our portfolio. A good example would be the U.K. homeowners portfolio. At the back end of 2017, we saw an uptick in -- with the rest of the market in terms of escape of water, and so that portfolio is being corrected with regard to rate rises over the last 12 months to account for that uptick in escape of water. And obviously, that portfolio is now back in a hold phase. So picking up the U.S., particularly, we haven't got any extreme course correction going through. What we do have is more subtle. So we have -- I've mentioned on the slide, GL in our blue-collar segments. Again, this is not a segment that we don't think is profitable. We absolutely do. But this is some tweaking and changing into maybe our question set and just slightly tightening up some of the things that we do from an underwriting point of view, but it's certainly not at the extreme level, the course correction that we have in retail is constant, and it's much more around that sort of settlement temporary rather than this of the extreme in-or-out that you might see on some of the bigger ticket.

I mean, I think, to use like, a sailing analogy, it's like changing small -- changing when the wind changes, which is what we call course correction. What we did in D&O in America is what I call a crash jibe. And we far prefer, for those of you who know sailing, is to do small course corrections than crash jibes. And so that's our ambition, is to -- lots of little rather than sell them and often. So sell them in extreme. And so occasionally we get it wrong, and that was clearly the case in D&A. But most other times, there's always -- a new exposure comes, a new loss cost comes, a new opportunity comes and you're changing the wordings to reflect that. Edward?

Ed Morris, JPMorgan. First question, just on the homeowners book. The FCA, I think, is scheduled to publish its market study in Q1. Just wondering how you sort of view your product offering, which is a little bit more at the premium end? Do you see any specific questions that it may pose? And have you done anything on pricing in the last few years that is worth thinking about here? And second question, just a quick one on the catastrophe budget. So you mentioned that your view of risk has changed, but also that the renewals you pulled back a little bit. So how has the cap budget changed in June?

Okay. I think the first one, Jo, would you like to comment on the FCA and Aki on the budget?

Sure. Yes. So we are awaiting the market study, as you know, in terms of pricing. Clearly, we've been involved in that for a while and we're keeping abreast of what would look like -- could look like. I think the most important thing to say from a point of view is some of those pricing methodologies that have been talked about is not something that we do. The sort of low-to-high -- we don't have targeted low-to-high pricing, and we certainly don't target vulnerable customer groups. So in terms of some of the philosophies that have been talked about, that's not something that we engage in. However, we are not immune to the market. And obviously, it does depend on what they come out with regard to some of the remedies. And obviously, that could affect. We are looking into that. And -- but as I say in terms of the high level strategies that are being talked about, that's not something that we do.

Yes. In terms of the cat budget, there isn't a material change year-on-year, a slight increase. For London markets, we expect the rating environment will offset the increase in our view of risk. And for Reinsurance and ILS, as you heard from Bronek and from me and from Jo, in fact, is a bit of a wait and see. Our view of risk has gone up. And if the pricing follows, then we'll be writing more business.

Great. Are there any more questions? So thank you very much for coming. I hope that's given you a degree of confidence in Hiscox. We're pretty confident about where we're going. And I hope that in a year's time, we'll have nice numbers to announce to all of you then. Thank you very much.