Beazley plc (BEZ.L) Earnings Call Transcript & Summary

Hello, everyone, and welcome to the Beazley's Interim 2021 Results Conference Call. My name is Emily, and I will be coordinating this call today. [Operator Instructions] I now have the pleasure of handing over to your host, Adrian Cox, CEO, to begin. Adrian, please go ahead.

Well, good morning, everyone, and welcome to the Beazley interim results. I thought I'd give a quick overview of what to expect this morning. We'll start with the highlights of our half year numbers. I am quite pleased with them. And following that, Sally will go through the numbers in a bit more detail, including the exhibits on capital and reserve strength that we know and love. We've been talking to a number of stakeholders this year and the 2 most frequent thematic questions that we get concern are our cyber business and our strategy given the new CEO. So after Sally, we will move on to the main event, and Paul will answer the big cyber questions. Cyber insurance is our biggest business. It's about 15% of our premium overall, and we launched our flagship cyber product, Beazley Breach Response back in 2009. And at that point, I was running specialty lines, which housed it and Paul was running the London International cyber business. So we've both been working on this together for a long time now, and we're both quite excited about the prospects for it. This year, we've also been thinking through our strategy. So lastly, I'll give a little detail on that. Spoiler alert, it is exciting, but not revolutionary, as well as some of the attributes of Beazley to which I'd like us to commit. And after that, we'll have some time for Q&A. Before that though, I'd like to begin by making a few introductory comments. At the start of the year, I wasn't expecting to be here doing this, but 113 days into my new role, I thought I'd share some reflections. The CEO role had always appealed to me, but I was genuinely enjoying being CUO and had certainly not expected this to happen this soon, but it's been fascinating getting exposure to different parts of the business and getting to know colleagues that I hadn't been spending too much time with before. And I've genuinely loved every minute of the job so far. The executive team continues to evolve, and I'm very pleased to welcome to Troy Dehmann, our new COO, a couple of weeks ago, and I expect to be able to announce the new CUO by the end of the summer. And I have to say, I'm pleased with our performance so far in 2021. We've done a lot of hard work these last few years to improve our business mix, to concentrate on what we're good at, remediate or stop where we weren't performing as we should have done, and to invest into opportunity. And I think that work is beginning to reflect in the numbers that we show today. And it feels like we're beginning to get the old Beazley back. My role, I think, is to get people here enthusiastic about the opportunities that lie ahead and ensuring that the business has the capabilities and capacity to deliver. But I also have to say that we did misstep last year during 2020. We had underestimated our exposure to a pandemic, assuming such a thing that would not cancel the events industry as completely and for as long as COVID-19 did. And it was also disappointing that we had to increase our COVID provisions last September. Unsurprisingly, I think that has impacted shareholders' confidence in us and our ability to do the hard things well. And one of my ambitions is to demonstrate to you over time, that we merit your trust, and that the Beazley you knew is the Beazley that you still have and that the future prospects for us are genuinely exciting. So with that then, let's move on to the results. Headline profit of $167.4 million, which generates a return on capital of 15%, which is our target cross-cycle returns, so that's good and a decent level of growth. We added about 4% of exposure growth in the second quarter. So although we continue to take exposures down in some areas, including cyber, it's being more than offset by the businesses that we're looking to grow. Combined ratio is in the mid-90s, which is what we guided to. And we have an accelerating rate increase, although with different drivers to last year, which Sally will go through a little bit later. Our COVID provisions remain unchanged at $340 million. The world continues to open up, albeit in somewhat of a staccato fashion, but that was the central assumption we had when we put together the $340 million. We intend to pay a dividend at year-end, all other things being equal, which will take into account the 2021 results as a whole rather than just the second half. And at that point, at year-end, we will also share the future dividend strategy. And with that, over to you, Sally.

Thank you, Adrian, and good morning, everyone. My name is Sally Lake, and I am the Beazley Finance Director. I'm going to get to the right slide. I'm going to take you through some of the more details on the financials that we're reporting today, focusing on investments, reserves and capital before I pass it over to Paul to speak in more detail about cyber. This half, we have seen an increase in reserve releases with the underwriting work performed in recent years beginning to show through. The investment return is very similar in dollar terms to this time last year. However, the drivers are rather different. And finally, I will give an update on our capital position, which remains in the top half of our preferred range. Before that, here is some more detail on the numbers. Adrian has already spoken about our top line growth of 22%. This is less after we allow for reinsurance as just over a year ago, we purchased some new proportional reinsurance on our specialty lines and CyEx books. We have also completed a loss portfolio transfer on our construction and engineering book, which we've stopped writing recently, and that's a one-off effect that happened in the first half of this year. Our net earned premium, on which our combined ratio is based, has grown by 14%. Our claims ratio improved significantly. At this time last year, we had incurred the significant losses around COVID-19 first-party claims. And the expense ratio has increased slightly, in part due to the reinsurance effects I just mentioned, but also given that the sterling has strengthened compared to the dollar and a large part of our costs are in pound. This has also put some pressure on our expense ratio. And now on to investments. So looking at our investment dispositions and how they changed during the first half of 2021. Most noticeably, we've seen an increase in our sovereign bond holding with an equivalent reduction in the investment-grade corporate debt portfolio. At the moment, we see the investment-grade spreads is relatively less attractive than in recent times. And so the maturing bonds are being reinvested into the highly liquid sovereign debt, improving our portfolio flexibility. Other changes at the summary level are modest. However, during the period, we have performed a number of management actions to maximize the return within our relatively prudent risk appetite for investments. So what has this led to in terms of performance? So $83 million or 1.2%, 2.4% annualized. The half one return is similar to last year. Low but rising yields created difficult conditions for bond markets and our fixed incomes produced a low, albeit positive return in the first half, and that makes up around 85% of our assets. Our capital growth investments did much better, returning nearly 8% year-to-date as risk assets rallied, owed by the improving economic outlook. Our tactical investment action has also added value. We reduced duration as yields rose, held more equity exposure for some of the period and have added significant inflation protected holdings to our fixed income portfolio as inflation expectations grow. Overall, we think this is a really good outcome in view of the market conditions and our risk appetite. In recent years, only 2019 saw a materially better first half return. Now moving on to reserve releases and how they have evolved over time. This graph shows a split by division. And you can see that we are very pleased to see reserve releases across all divisions for the first time since 2017. Since taking over in my role in 2019, I've really focused on ensuring that this graph gets to a better position, so this outcome makes me very happy. This is a result of the underwriting action that the business has been taking for a number of years. And as a percentage of net earned premium, the releases are around 7%. And as you can see on the dotted line, this number continues to go in a pleasing direction. So that's looking back but on to everyone's favorite graph, what we've got left in the tank. So to remind everyone, this graph compares the reserves on our balance sheet to our internal actuarial estimate, which in itself is not a best estimate and includes a level of margin, and we aim to hold 5% to 10% above this actuarial measure. The reason we show you this graph is to show how consistent we have been in our reserving strategy over time and 6.6%, which is the number at half year 2021 is very much in line with recent times. So it is great to see this continuing despite improved reserve releases and to show you that we are continuing exactly the same way as we reserve. Just before I head on to talk about capital, we just want to take a moment to review the rate change graph. You can see that the rate improvements have been impressive this year so far with a real acceleration caused primarily by the dislocation in the cyber market. The rate change is 20% across all divisions over the past 6 months. Most divisions are seeing double-digit rate increases with CyEx seeing 44% rate improvement. So whilst we are continuing to grow the premium, the benefit of rates are being seen when we look at our capital requirements. Since we last reported on capital, we have started allowing for the growth expected within our 2022 business plan within the numbers we show here. And whilst we continue to work on the business plan over the summer as usual, we are currently expecting similar growth in premiums next year as we planned for this year. And after allowing for this and allowing for other updates to the Solvency II positions, we are estimating a surplus of 23% above our lowest economic capital requirement, which is towards the top of our preferred range. Our capital resources remain unchanged with $550 million of Tier 2 debt, the banking facility of $450 million, of which $225 million remains unutilized. And as Adrian mentioned earlier, the Board remains committed to a dividend payment at the end of the year after taking the whole year into account. With that, I will pass over for the main event, as we call it, and Paul, on cyber. Thank you.

Thanks, Sally. Good morning, everyone. For those who don't know me, my name is Paul Bantick. I'm the Global Head of the Cyber and Tech business here at Beazley, and I have the privilege of leading both the underwriters and the service team that help us provide all the services to our cyber clients globally. I've been here, as Adrian said, about 16, 17 years and have been part of the cyber team, whilst it's been in slacks and also when it was previously SL when Adrian was leading itself. It's great Sally and Adrian have asked me to give you an update this morning. I plan to cover a brief history of cyber insurance, but also more importantly, its evolution and how it's evolved at Beazley. Talk a little bit about how we've seen the threats evolve and what we've seen in the past year, give a highlight of some of the underwriting actions that we've been taking, talk about some of the investments that we're making and then talk about what we see in the cyber opportunity as moving forwards. So I wanted to start by giving a bit of an overview of the cyber evolution. You'll see that this time line really shows 2 things. It shows what's been going on in the outside world and what's been going on at Beazley, as cyber evolved. Cyber really started in the U.S. back in 2009, which is when we saw the first real drive and boom of demand for cyber insurance, and it was really focused on data breaches. At the time, there was lots of large companies that held a lot of data that were being attacked and their data was being stolen. And this is really coupled with regulation and law changes at the time, gave clients a greater need for insurance. They wanted help with protecting the bottom line, and these things became very complex and very expensive. That complexity really gave us another opportunity. Clients weren't just worried about what these were costing when they had a data breach. They were saying, how do we manage one of these? How do we respond to one of these? And that's where at Beazley, we really saw this change in regulation and a change in what clients were looking for, and we developed a product and service around it. That's when we launched the BBR product, Beazley Breach Response. And we also created our Beazley Breach Response in-house team. That's an in-house team that we have that's focused on helping clients manage events when they happen. So BBR was born, and it was a product with a big native service around it. As the market continued to grow and we wrote more business, we decided that we needed to invest early into several things. Examples of this were, we invested in our international business. We now have locations all around the world writing cyber business. We have underwriters, claims managers and breach response people based around the globe. We also made big investments into risk management. One example is we created Lodestone. It's a cyber service security company based in the U.S. that's 100% owned by Beazley. Lodestone is now working with lots of our clients in the U.S. It's up to 40 to 50 employees and going very well. The BBR service team had a huge impact at our clients, but it also had a huge impact for Beazley. We were seeing that on our losses. We were saving 20% -- 10% to 20% compared to the market on the costs we were incurring. This is because of the in-house model that we have. This in-house model really enabled us to bring in-house what people were outsourcing to law firms, forensic firms. We had a lot of capability in-house to do it ourselves, and it's a model that we've invested in and continue to invest in. As '17 and '18 years rolled around, we saw data breaches became a little bit quieter. There were still some large ones that went on the high-level names that we also see now a little bit better. However, during this time, we continued to invest in that, because we knew cyber threats do change, and they do change come and go, and we knew that team would be even greater value moving forward. In recent times, as I'm sure many people know, the cyber effect has evolved. And ransomware in '19 and '20 really is the big shift that we've seen. And ransomware has done several things. Number one, it brings new exposures. If you have a ransomware event, your systems are locked up. Someone wants money to unlock your systems. You have to decide as a business what you want to do about that. That presents new issues and new challenges for clients. Do you pay a ransom? Do you not pay a ransom? But whilst this is going on, there is the business interruption risk that clients are going through. Clients are experiencing a reputation risk and business interruption. What this deal does is it brings a lot of new clients to the market looking for support with their cyber insurance. Previously, it was just companies that held data that were worried about cyber insurance. And now any organization around the world has an exposure to ransomware and some of the newest cyber threats that we've seen. Breach Response is still very, very critical for these ransomware events, and I want to underline that. What we've seen is that clients are now seeing the value of that breach response team again. So the investments we've made, that we've continued to make, are bringing back those benefits that we saw as we did with data breach. But as the threats evolve and as we see ransomware, we've also been looking at how do we help clients with resilience. Cyber threats change and evolve over time. And what we need to do is help clients become more resilient so that they're more protected and better protected from whatever cyber threats exist. During 2020, as we saw ransomware impact on our clients, we looked at what underwriting actions we were going to take, and now I want to give you a quick overview of some of that. We really divided it down into 5 key pillars, and we worked on this in second quarter of 2020, and it was fully deployed in Q3 2020. This is a huge investment that we've made, both in terms of money, people and a huge project to get off the ground. Firstly, we made a big investment into data analytics. We're working both with our own data, with third-party data, lots of other organizations. We've brought in specialist data scientists that are now part of the cyber team to look at. Based on our data, our clients less or more likely to have a cyber event. What can we tell about them from the data we have, how can that help us drive both clients to increase their resilience and also help with our underwriting. Secondly, we developed a ransom application. This has been a critical piece of information that we've been using as part of the process. When we're underwriting an account, every client tells us about the ransomware exposures and controls and mitigation that they have in place. We've been scanning organizations. This is done with third parties, and it's an outside-in view and it's becoming quite common in the marketplace. We scan clients not only when we insure them, but we also scan them through the life cycle of the policy. If we see new vulnerabilities appear, these are the same kinds of vulnerabilities that bad actors are looking for. And so if we can see them, they can see them. So what we're doing there is notifying our clients about these vulnerabilities and helping them to remediate them, helping them to shut them down. 9 times out of 10, these are reasonably simplistic things to fix and things that should be addressed quite quickly and quite easily. We don't know how many ransomwares we may or may not have helped clients avoid, but we know that we've shot a lot of vulnerabilities down. These 3 things are really focused on frequency. We've been targeting frequency very hard. If we can help our clients be more resilient, we will therefore see less claims, and that has been our approach to addressing the ransomware issue. However, we have also done some limits management. We've looked at our book and we've targeted specific areas where we think we need to look at the limit we have deployed, and we've been reducing some limits in certain sections. This will also help with severity. And then the last thing is, we were talking about the rating environment. Absolutely, the answer to cyber insurance is to address the frequency and focus on frequency. And as I've just explained, we've been making massive efforts with that, but we knew that rates needed to increase. So during 2020, we started to build the rate runway and increased rates. And as Sally said, we're seeing some great rate changes currently, and I'll talk more about those shortly. The other bullet point I want to make is we provide great support to our underwriters globally. One of the things we did is we created a global dedicated cyber underwriting management team that reports to me. This has been invaluable, because as we're taking action, as we're navigating a very fast changing and volatile market, having a team focused on this and profitability and supporting the team through that process has been absolutely fantastic. And then lastly and more -- very excitingly, we're making a huge investment into risk management. Every client has their own view of risk. We scan them. They talk about their ransomware application. We look at our data. Every client is different. Everyone's got their own view of risk. Everyone gets their own terms. We treat every client individually. And I think that's really important, and we want to continue to do that, but we want to continue to help with resilience. And so with that, we're making big investments into our cyber ecosystem. We've been to the exec during the course of the year, and we've had approved significant investments to look at the Beazley cyber ecosystem and think about how can we help make risk management and resilience and support our clients and brokers much better with the cyber threats, the ones that we have now and ones that may exist in the future. So this is a diagram that shows you our cyber ecosystem and the investments we're making. And what I'd like to do is briefly touch on each of these areas. The first is the data and threat intelligence. I spoke about that. We've been doing some of that already, but we're now working with more providers, more data, more threat intelligence. We're overlaying that, as you'll see there, Lodestone, this IT security company. They are now providing us with threat intelligence. We've established a threat intelligence network that involves people not just from companies that work in cybersecurity, but also we're working closely with people that are ex government, ex law enforcement. And we really are trying to make sure we ensure we stay on the front foot of the emerging threats for our clients. Secondly, risk management services. This is absolutely critical. It's one thing to have the threat information. It's another thing to figure out how you can help prevent these things happening. And so what we're doing is we're creating a lot of new risk management services for our clients. We're currently hiring a head of cyber services that will be focused on building out this risk management services that we do even further. Incident response. We've obviously been doing that since the very early years in 2009. We have a fantastic team that work on that around the globe. We are now launching that further. One example would be, internationally, we've established Lodestone in the U.K. to help us with response. We're looking at how we build out the model further afield. We have it in many countries, but we're making sure we have that response and that unit and that team everywhere where we need them. And then we're looking at the U.S. team and constantly thinking about how do we evolve, how do we make sure we're on the very front edge. We're looking at different services, different technology and instant rooms that we can work with our clients on. And then most importantly, underwriting and claims. How do we build what we're learning, what we're seeing, the risk management, into our underwriting and claims processes going forward so that we continue to manage the frequency, continue to manage the book and continue to help our clients and maintain profitability? I think it's really important to say there's a lot of things on this page that I could talk about for a lot of time. It's a huge project. It's an undertaking that's involving 80 to 100 people across the company that are coming together to collaborate on this, which is fantastic, and I really appreciate everyone's support with that. But it's not about doing any one of these things well. It's about doing a lot of them well and putting them all together. That's where the magic is in this. And that's what we're committed to, and that's what we're committed to continue investing in. So what's the impact of the underwriting and all the action and the risk management that we've been doing? What's it having so far? We wanted to give you an early view. And I would say it's still very early. We've been looking at accounts that have been bound since October that have been through our full process. And although it's early days, we are seeing some positive signs. We've seen a 20% ransomware frequency reduction on a policy basis, and we've seen a 50% reduction on a per premium basis. And as Sally said, we've seen increase in rates. In the last 3 months, we've seen the cyber rates doubling in all parts of our cyber books. We show this graph quite a bit during our results presentations, and it's been slightly updated this year with a new line. The bottom chart show the premium growth of the cyber business. And as you can see, the premiums are up in 2021. We've actually reforecast within slightly more than planned. This is driven predominantly by rate. This is shown by the pink line at the top. The pink line at the top shows the policy count. And as you can see, due to the underwriting actions that we're taking, we've shed more policies than planned due to our continued focus on risk selection. So the overarching message of this, is that we have less exposure than planned this year with significantly more rate. The other benefit of that is that we are now seeing that our systemic event exposure is reducing. And I wanted just to take a minute just to talk about systemic risk and cyber just for a second because again, it's another common question that comes up quite a bit. Certain covering cyber policies are more systemic in nature. Business interruption, dependent business interruption. When we talk about systemic risk in cyber, we're thinking about things that can impact more than one client. When we're thinking about attritional cyber risk, it's something attacks that are focused and just really impact one client. And I've given some examples on the slides. A client being targeted for a data breach, that's typically one client, one data breach, it's an attritional loss. Things that are more systemic in nature, for example, cloud outages that can impact lots of clients at once, are really what we're thinking about when we think through systemic risk in Beazley. Managing the systemic risk, it's really no different than any other line of business at Beazley. We manage systemic risk in the same way. We have a group of people that work on the systemic cyber risk that are focused on it and are constantly meeting. Firstly, we have a set of scenarios. We spend a lot of time putting together our realistic disaster scenarios that we work on. This involves not just the internal expertise we have, but we work with a lot of third parties around the world, a lot of external experts, and we're constantly updating them every 6 months, 12 months and they're constantly under review. As part of that, we're not just looking at what happened. We spend a lot of time thinking about what could happen? What are the black swans? What could they look like? How would we manage those? What would be the reality of those happening? What's the likelihood? And we model a lot of scenarios based on that basis. We have a suite of RDS scenarios that we run. We can run them many time at the push of a button, and we're constantly reviewing them on a weekly, monthly basis. The other thing is diversification is key. The one thing I've learned through the years of coming through the cyber business here at Beazley, is that the more diversified book you have, the better you can manage systemic risk. And the thing I'm very proud of, and I know the team is very proud of it at Beazley is, we have a diversified book not just by revenue, but now by geography, by client size, by industry. Managing diversification is absolutely critical. We set a risk appetite for cyber, the same as we do for any other line of business at Beazley, and we manage that risk appetite. And if we think we could go outside of that risk appetite, we purchase additional reinsurance to keep us within that appetite. Exposure to systemic events has been coming down. This is due to 2 things. Number one, as you saw, the policy count is down, because we've shed more business than planned. But, we're already working since 2019 on reducing our exposure to these events by looking at the limits we have exposed, by the types of coverages that cause the main aggregation with our clients, things like business interruption and dependent business interruption. So for my final slide and in closing, I've given a good overview, I think, of how we manage our exposure, the work we've been doing, how we're evolving our cyber ecosystem, what we're seeing from an underwriting action standpoint, what we're seeing from the market and the rates. But what does this mean for us looking forward? As we continue to get evidence we're profitable and frequency is reducing and that the risk management is really helping our clients with resilience, there is -- there could be a rare opportunity for profitable growth. This is actually further supported by the market conditions that we're seeing. However, it's still very early, and this is going to be a continually evolving process. We're going to keep learning as we do this. We're going to keep looking at the threats, looking at the data, looking at the underwriting action, thinking through how we're going to navigate these things. At Beazley, we remain agile, and we want to evolve. We saw a market opportunity in 2019, and that was to establish a leading position through the creation of BBR services. We think there's an opportunity now through risk management and resilience to create a seamless kind of opportunity. And with that, I will pass it back to Adrian.

Thank you, Paul. Okay. So I hope you found that insightful. On then to the outlook. So I expect growth to continue at the current levels for the rest of the year but also for that level of opportunity to persist into 2022. We have a particular set of opportunities and a particular sort of products at Beazley that are more than just transitory. And that's why we think there is opportunity for continued good growth at the similar levels of this year into next year. Yields currently on our fixed income is 0.5%. And when we think about that altogether, our combined ratio guidance for the year remains unchanged. The year-to-date natural catastrophe appetite, including the floods and the wildfires that are currently occurring remain so far within our -- the margin that we hold. So there's no need to update our guidance at this stage. And once again, we intend to pay a dividend at the year-end, all things being equal and share the strategy at that point. Talking of strategy, I thought this was an opportune moment to spend a little bit of time outlining our strategy, which we have refreshed this year and also reaffirming some of our attributes. So to start with strategy then. We're a specialty insurer. What does that mean? Well, most of the products that we underwrite are not written by the majority of the general insurance market as they are, by definition, specialists and so require certain specific tools or skills to do well. These products are less commoditized which allows for an insurer to apply IP and differentiate their offering. Given that, there can be more margin and better cross cycle returns than in general insurance. And lastly, demand is growing faster in specialty than in general insurance markets. Why? Because products are generally at an earlier stage in their life cycle and new products are created here to address new and emerging risks, which occurs now at an increasing pace as the world continues to evolve faster and faster. So a successful specialty insurer is in a market that has better cross cycle returns and persistently high levels of growth, and where do we fit into that? Well, our strategy is to employ people who have genuine subject matter expertise and who take pride in continually refining that knowledge and then to deploy them at the point of sale and empower them to use that expertise. We then find niches where deploying that expertise drive some competitive advantage. And I think there are 2 ways we do this. Firstly, as we out underwrite out risk select, out-claims manage our peers on a risk-by-risk basis. And secondly, our customers, brokers and insurers value having access to that expertise, to people who know what they're talking about, understand their industry and have the authority to make decisions. In most of the insurance and reinsurance world, even some of the specialty areas, those capabilities don't really drive competitive advantage. But in some, they do. And that tends to be in areas where risk is more complex or volatile or changing or new. And our job then is to find those areas where that applies and whereas there is margin, if you do it well, to properly reward us for our efforts. So our product strategy is relatively simple, refine our product set to maximize that competitive advantage by ensuring that the businesses we are currently in, are fit for purpose and constantly looking for new ones. Our footprint strategy is also relatively simple. As a specialist and relatively small, personally, I don't think scale is of itself important. I think our relevance comes from having to say something, having something to say, rather than by size. But given that, it makes no sense for us to have an ambition and have a presence in every geography across the world. Rather, we have a preference to locate where there are large pools of demand for the specialist products that we sell. And they are in the large nonlife markets in North America and Europe and in the big wholesale markets for us, London, Singapore and Miami. The business that we build in those large, mostly retail markets in North America and Europe is sticky, more mid-market and SME and painstakingly built over decades rather than years. The business that we build in the wholesale markets is more larger risk, more volatile and more actively managed through the market cycle reflecting the fact that opportunities in the wholesale world do wax and wane considerably. When you combine that then, you have a business that can create and maintain differentiated products where demand growth is higher. That gives us long-term opportunity for compound growth with levels of cross-cycle margin significantly higher than the cost of capital. It's an appealing model, I think. But it's hard by very definition, if the areas of risk you concentrate on are new changing volatile complex. And we have to be good at it to get it right, and not only good but relentlessly looking for ways to do it better. And that's the challenge that we set ourselves every day. And that's why we invest heavily in improving our underwriting technology, seeking new data, which can generate insight, working with third parties to harness different technologies, building out our risk management services, all to drive doing specialty business while designing those differentiated products and building barriers to entry. And that's our strategy in a nutshell. It's not new, but I think it's good to articulate it every now and then. The second topic I thought I would address is what are the attributes of Beazley as a company. We've long been associated with a number of them. And I think it seems appropriate now to affirm to the ones that we wish to commit to afresh. And the first is underwriting discipline. This is particularly important to me. We're an underwriting company. It's what we hang our hat on. And as I hope you heard, it's central to our strategy and our competitive advantage. And that means both capitalizing on opportunities while also knowing when to pull back or to stop when needed. And you will have seen over the last few years that we do reassess our product set, and we do stock when what we have doesn't really drive competitive advantage. Engineering, trucking, SME marine are examples of that over the last couple of years. Secondly, capital discipline. We've always put great steadiness, returning capital to shareholders when we don't have attractive opportunities to deploy them ourselves, and that commitment remains. Thirdly, financial discipline. I do believe in the discipline of regular dividends and the prudential financial management that, that drives. Next, reserve discipline. We commit, as Sally said, to maintaining a consistent and prudent philosophy of setting reserves which should, other things being equal, provide for a consistent flow of reserve releases over time. Next, diversification. We've long valued the benefits that diversification brings. One of our favorite charts has been that, which shows rate change across all teams, which is significantly less volatile than each team of itself. That portfolio management philosophy remains. And lastly, responsible business, which we will discuss, I think, more at the year-end. But we are building out strategies to address the E, the S and the G of being a responsible business. I want us to be a company that embeds these values and those goals at its heart. Our vision is to be and be recognized as the highest performing specialty insurer, and I believe that highest performing includes a commitment to becoming a better business that contributes to those goals. So with that, I hope you found that useful, and I open up to Q&A.

[Operator Instructions] Our first question today comes from Kamran Hossain from RBC.

So the first 2 questions were on cyber, because I guess it would be rude not to ask Paul some questions given the presentation. The first one is just about market behavior in cyber. Obviously, you've been in this market for the long haul. There have been some people that have gone into the market more recently and probably have been burned a little bit in the last year. So I'm interested in whether people have exited or the behavior is changing. And maybe are kind of barriers to entry now a little bit higher than they were a year ago? The second question on cyber is I guess there's the risk changes all the time. Do you think we'll get to a world where, I guess, given risk management capabilities, we can get to more real-time pricing? Would you think an annual contract is just the thing that we will see in the market for the next few years? So kind of a very big picture question. And then the third question, I guess, the growth outlook for next year sounds really exciting. In terms of the areas of growth for next year, should we expect them to be similar to those that you've seen kind of year-to-date? Or are there any other areas that you think we should highlight?

Right. Paul, do you want to tackle the first 2 cyber questions? And I'll think about how we're going to grow.

Thanks, Kamran. Yes, absolutely you're right about market behavior. I mean we are definitely seeing a situation where capacity is restricted in the market, and that's driven prominently because of the losses, but -- and some of the ransomware activity. And definitely, as you've seen from the market conditions and the rate changes we're seeing that demand is outweighing supply. There have been some entrants in the market. I haven't seen as many in recent times. In terms of barriers to entry in investment, I think the word investment is critical. I do think the barriers to entry to be able to do this well for clients and be able to manage clients and help them with their resilience and manage their own frequency and severity. I think the investment is colossal, so it's not just a monetary investment, it's a people and IP investment. There's just a huge investment, so I think the barriers to entry are definitely, definitely growing. In terms of real-time pricing, I think it could get there one day. I think we're a long way away from it. I think it's early doors. I think we've always thought about that, and that's been something that's been in the back of our minds. I think it will become more realistic in the coming years as clients start to increase their resilience and as we can get a better understanding from clients constantly around what are they -- how are they constantly evolving. How are they constantly seeing their controls evolve? But I do see that one day, that could be something that potentially the cyber market starts to look at.

Thanks, Paul. So on to growth, I think what we've been talking for the last 18 months or so about the sort of 85-15 split of how we're feeling about our portfolio with the 15% really being the areas that we think are exposed to this COVID aftershock claims environment as well as social inflation. I think that underlying assumption still remains, Kamran. So we're looking to grow across most of our book. We still remain quite defensive on things like employment practices liability, certain parts of our health care book and certain parts of our E&O/PI book. But I think overlaying that, there is some stuff which is a bit more live than that. We are beginning to open up as an economy. So there will be more evidence generated, I think, in the second half of this year about what's happening to those COVID aftershocks. Is social inflation going to reemerge as the quarter open? So we'll be able to update our assumptions, I think as the years go on, which as the year goes on, which will influence what we want to do next year. Likewise, I think we're continuing to build evidence about how successful our cyber strategy is, and that will influence how much exposure we want to put on the books or not. And of course, we're in a very live natural catastrophe environment, not only in the middle of a windstorm season, but there's other stuff going on as well. And I think that will impact what we think is going to happen to the market as well as what we think we need to do to our own underwriting strategy. So the sort of short answer is our 85-15 remains, but we're in a very live claims environment at the moment. I think that will influence both what the market does and what our view of risk is. And so that will influence what we do.

Our next question comes from Will Hardcastle from UBS.

First question on cyber, second on capital and dividends, okay. So the first one, cyber, we're seeing these enormous price increases. There sounds to be potential opportunities. Of course, it's got to weigh the risk up. And you're sort of suggesting that there could be a great opportunity. I guess when will we know if that opportunity truly exists, now we're sort of 7, 8 months through starting the remedial action there. And then in that context, that level of growth, what should we really see as the maximum level on a group level that you'd consider to be comfortable to be cyber exposed? Perhaps, when there will be premium exposure or however you look at that, that would be helpful. And now on capital and dividend. I guess we've seen a really big earnings bit today, [ per expectation ] but still no dividend. The capital leaving after assuming what feels now a higher premium growth expectation for next year is robust. Can you just help us to understand the logic in your thinking perhaps of the no interim at this stage? And what's the greatest risk in your mind that perhaps prevented you from distributing that despite these financial metrics, suggesting that you have been comfortable to do so?

Okay. Thank you, Will. I'll have a crack at those. And please feel free to weight in if I miss anything. So on to cyber, as we've been very careful to say that the data that we have is very early, because as you say, we're only 7, 8 months into something. And in insurance terms, that's just a blink of an eye, right? So having said that, sort of first-party side, the ransomware side, predominantly of our cyber business is very short tail. So I think we gather more evidence month-by-month. We'll have a reasonable idea in time for the business plan for next year as to whether we think it's successful or what else we need to do. And so we should be able to give more guidance towards the end of the year about what we want to do as the evidence builds. And I think because it's so short tail, that's faster than normal, I think. Going back to the little speech I gave about the attributes of Beazley. We do still believe in the benefits of diversification. You definitely can have too much of a good thing, at the same time, as you say, there may be an opportunity here. And we will look to capitalize on that if we think there is one. Nonetheless, we want to make sure that we remain a balanced business. Our rule of thumb is around 15. We may flex it up or down a little bit depending upon how big of an opportunity it is, and we may well use some reinsurance partners to build the business with us. So I think as a share of the overall, we'll flex around a little bit but we remain committed to a diversified business. But we may grow a little bit more gross in use of reinsurance. And likewise, as you sort of intimate well, there are sort of 2 ways of measuring this, by premium and by exposure. And because rates are going up so fast, premium may not be an adequate way to do this. And so we'll look at both. But we remain committed to diversification, because I think it's important to a consistency of results. In terms of sort of capital and dividends, we are pleased with where our capital is in the sort of top half of the range. And we did consider whether or not we should pay an interim dividend. I think the underlying driver of why we've decided to postpone it is because paying a dividend is not just that. It's paying a dividend and explaining what your dividend strategy is going forward. And to us, really, that sort of feels like a year-end decision when we can take a look at the whole thing and then set out something new for next year, and that didn't seem to be the sort of thing you want to do halfway through a year. It's a sort of thing that you do when you finish, and you can have a look back and set out something that's going to last for the foreseeable future. So it's just a timing thing for us, really not much to read into it apart from that.

Okay. Okay. That's really clear on the dividend. I guess one thing I just wanted to clarify, it's not a new point. You mentioned something about at the full year, obviously, you sound committed but obviously, things can change. But just thinking about that more, if that comes, I'm not sure if you delivered this year, but will that be the sort of final sort of equivalent to maybe roughly 2/3 of the year? Or will it be covering the whole year? How would you look at that?

Yes. And so you're right to ask that question, because we've worded it quite carefully. So the intention for equal is that the dividend will reflect the full year's earnings rather than just second half. So thank you for asking.

Our next question comes from Andrew Ritchie from Autonomous.

Thanks for the detailed presentation on cyber today. That was very useful and timely. Questions. First of all, I noticed in the first triangles, there continues to be recoveries on various accident years on your specialty/CyEx aggregate excess of loss. I mean you just recapped what comfort there is on remaining room on those recoveries because at the minute, there's a helpful spread that gross deterioration and net improvement. Should we expect that to continue basically? Second question, we are in wind season. What should I think of as your property exposure or property reinsurance exposure? You have -- obviously, you don't -- you give a start through your realistic results scenario annually. But should I think of it has gone up from last year or about the same? Because you -- I can't tell to what degree your exposure has grown in those classes. And linked to that, is there any sense, can you give us that you will, or what your sort of remaining cap budget would be for the year? Sorry, I had 2 other quick questions. One on cyber, on Lodestone, that's an intriguing business that I didn't realize you had set up. Is that a business that can generate meaningful fee income, I suppose, from noninsurance clients? I think I just checked out the website and it appears to be able to noninsurance clients. So I'm just curious what the sort of strategic thinking behind that business, longer term is. My only other question, to Sally, I'm a bit confused by your EPS calculations. I don't get why the average share count is the same year-on-year, the half year. And also the translation from cents to cents, appears to assume last year's average exchange rate. Is that -- maybe just clarify what's going on there.

I'll start with -- I'll take that one away to look in detail at that, because I probably can't do that live.

The earnings per share one?

Yes.

Okay. Right. We'll get back to you on that one, Andrew. If we go through the first 3 questions then. In terms of reinsurance recoveries available, I think we may have an echo. If that's the case, I do apologize. Yes, we have plenty of reinsurance left available across our specialty lines and CyEx. You're right to notice the difference in the gross of the net, but nothing is exhausted and the scenarios that we run don't have us exhausting it either. In terms of exposure growth, I think there's sort of 2 things to look at. Yes, our property business and our reinsurance business in aggregate has grown a little bit. And I think that has driven up a little bit our [ 1 in 250 ], sort of on a pro rata sort of basis. But there's more to it than that, I think. And one of the things we've been talking about the last few years is looking at the near end of the curve as well as the end of the curve, and we've been quite careful to make sure that we're very strict about growing exposures on the near end of the curve. And so it's not all pro rata. I would love to be able to share the net CAP budget with you, but I'm not allowed. In fact, I almost asked you to say that I thought I can say that myself. And Paul, why don't I pass over the Lodestone question to you?

Yes. Thanks, Andrew. Lodestone, as you rightly pointed out, is doing fantastically well. Yes, it can and it does do work with non-Beazley clients. It's something that we're seeing a lot of people, particularly in the middle market, middle market clients having a huge amount of value from. So if you take a step back for a second, Lodestone does 2 things. Really, when you want to think about Lodestone, they do incident response work. So they're a forensics provider and can do incident response work when someone is having some kind of incident like a data breach or ransomware. And secondly, they do a lot of proactive risk management services, so helping clients become more resilient, understand their risk before they have some kind of cyber event. We work with them closely on both across our client base. And it's extremely exciting. We now have cyber security firm that can benefit our insurance greatly. And as we're evolving our cyber strategy and our ecosystem, we're actually -- we've been sitting down with Lodestone every step of the way to think about what does this mean for Lodestone and Beazley and how do we partner on this, and what is the opportunity for both organizations. And it's very much done in partnership and very exciting that they've just launched in the U.K., and there's hopefully a lot more to come.

But yes, Andrew. I think when we first conceived Lodestone a few years ago, it was -- we started it because of that concept that you suggested, right? If the world is getting more digitized and that equals more risk, and there's an opportunity in insurance, but there's also an opportunity in risk management, there could be a business for us to build there. And so we started it quietly, used it to help our clients only, but with optionality to pivot according to where we see the opportunities. And that's exactly what we're doing now is we're assessing what can we do with Lodestone, both to help our own proposition but also in and of itself because cyber risk management -- cybersecurity risk management is a very fast-growing opportunity and then maybe it's something we can capitalize on there.

And owned by Beazley today?

Sorry?

It's 100% owned by you today? Is it -- to clarify that?

Yes, it is. Well, that and the employee, but yes.

Our next question comes from Iain Pearce from Credit Suisse.

All cyber-related, please. So just -- I was hoping you could touch on sort of what's going on more widely in terms of claims trends in the cyber market. Obviously, you're sort of flagging some improvement in claims experience. Can you just touch on what the marketing, because I don't think that's sort of being replicated across the market? Secondly, just in terms of cyber rates, I mean, you're sounding a bit more bullish on how long this sort of hardening can continue. So just sort of if you could flag what your expectations might be for sort of cyber rates in 2022. And sort of if you think hardening can continue beyond that potentially as well. And then on the loss ratio that you booked in cyber. Could you just talk a little bit around sort of how you thought about booking the loss ratio at the start of this year for cyber? And whether you're reflecting some of that positive claims experience in the cyber loss ratio because the CyEx loss ratio at start of this year was still quite high. And then finally, just in terms of the investments that are being made, could you just put some numbers around that and sort of how you're treating that from an accounting perspective, whether that's all been expensed.

Okay. I think there are 4 questions there.

I'll do the third one.

No, that's perfect. Just making sure that we address them. I'll pass the fourth one about the investments to Sally. I'll have a go at the loss ratios. And if you want to talk about the trends and the rates into next year.

Sure. So I'll do 1 and 2 first. Okay. So yes, I mean you're right in, we're seeing right now, as I said, rates doubling. When you look at the wider cyber market and you think about trends, I think there are several things. Definitely, all the cyber carriers that certainly we see are certainly taking action. Everyone's got their own approach and got their own action and their own course that they're taking, but we're definitely seeing a varied set of actions across the market and everyone is starting to really address it. Some focused on rate, some focused through other things. So you've seen a mixture. It's definitely picking up as we move through the second half of the year. The rates that we're seeing for the last month or 2 were not the rates that we saw in the first 3 months of the year. So I think there's certainly a lot more to come in terms of runway. I think you're going to see the rate environment with us for a little while, certainly through the rest of the year and into next year. We're in a very fast-changing market. We're in a very, very fast-changing market. And as capacity restricts, you could well be that rates continue to extend into and through next year. That's one of the reasons why we're keeping our plan very agile and why we plan to reassess later in the year, because had we looked at rates 3 months ago and made assumptions that have been very different 3 months ago than they are today, they want a -- definitely a very steep curve and upward trajectory. And so I think once we get to the end of the year, we'll think about what is the most likely scenario for 2022 rates. But right now, yes, I certainly see and from talking to brokers and our partners and others, we certainly see the rate environment and rates continuing to increase through next year.

Claims trends?

Claims trends. When we look at the claims trends, obviously, we've got some positive claims trends on frequency. It will be interesting to see. There's also a number of other factors that are starting to happen around the world that are going to impact claims trends we're seeing. Certainly, different governments start to take action and law enforcement start to take action around some of the cyber threats, and we'll be interested to see what they do to claims trends. Obviously, we've seen the reduction in frequency. It's very early to tell on severity. But obviously, the action that we're taking through limit management and the targeted limit management approach now that we've been doing that for 7, 8 months, we'll start to see that in the coming months, start to really pull through as well, because that limit management, having less aggregate exposed to these things will have an impact on the severity. The claims trends again is something that we watch on a monthly basis. It's something that is, as Adrian said, it's a short-tail class of business. And so once we get to the end of the year, we'll have another 3, 4, 5 months' worth of claims trends data, we'll have a much better handle on where the market is at that point in time from a rating standpoint and then we'll be able to finalize the plan.

So I think cyber frequency is down generally or just?

I think cyber frequency was not just -- may have been down slightly recently, but I don't think generally. I think cyber frequency is down slightly. And it can be a bit seasonal, cyber frequency. You see certain hotspots. So July 4 in the U.S., that holiday weekend, Christmas time, Black Monday, the shopping -- the busy shopping period. You do see some seasonality in the cyber events over the years. But from what we're hearing and what we're seeing and from talking to a lot of the experts that we work with, they're not seeing any dramatic drop in frequency, although there are things underway that might help along with the underwriting actions.

Thank you, Paul. Loss ratios, so we have not taken credit in our loss ratios for the reunderwriting that we've done. From a GAAP perspective, any impact that, that has is likely to be felt at the tail end of next year. And it's not something that we'll be booking this year. It's, as we say, it's early from a business line perspective. It's even earlier from a booking and profit perspective. So we'll wait a little bit longer before that flows through to our P&L or not. Sally, do you want to talk about the investment dollars?

Yes. So no, it's not a dissimilar answer actually. So the cost of the ecosystem that Paul described is minimal within the first half. It's going to -- I expect it to increase in the second half and then probably increase further in the fullness of next year as well. It will come through our admin costs, I would expect. We may look depending on what we do to look if we want to capitalize of that, but that decision hasn't been made yet. So I can update in the future on that. But in terms of the financials we're seeing today, it's not a significant amount within those.

Our next question comes from Freya Kong from Bank of America.

I've got 3 questions if that's all right. Firstly, on your dividend strategy and what you will be giving us at the end of the year. Has anything changed in the business for you to step away from your previous progressive dividend policy? And what sort of rebasing should we expect? My second question is on COVID. How are you tracking versus the $50 million of extra exposure that you flagged for 2021? I would think by now at this stage, we would have decent visibility for the rest of the year. And of the $340 million that you've booked for reserves, could you give us a sense of the IBNR? And last question is on net premium growth. So you previously guided to slightly below gross premium growth. But now there is a 10-point difference in the guidance that you're giving us. Has there been any change in the reinsurance that you expect to take on this year?

Great stuff. Okay. So has anything changed on our dividend strategy? We -- yes and no. So we haven't decided what it is next year or going forward yet. So it may be a resumption of what we've been doing before or maybe something different, and we haven't thought through that. So I can't really give any guidance as to what that's going to be, because we haven't had those discussions yet. I do think it's a good opportunity for us to set a path for the next 5 to 10 years. I want to make sure we take the proper amount of time to think that through. And once we have, we'll share it. So I'm not being evasive, we just haven't done the work yet. Moving on to COVID. We haven't used the $50 million yet. So as I said, the sort of behavior in terms of what events are happening and how they're happening. It's relatively in line with what we had expected to happen. And so if that continues as is, we don't expect to have to use the $50 million of extra that we had flagged. But -- and by the time we reached the end of the year, most of the events that we're concerned about that will have either happened or not happen. So we'll get clarity on month by the end of the year. But as we stand here, we're not planning to use the $50 million. I'm not able to share with you, unfortunately, how much of the $340 million is in IBNR. There is some, but we're not going to give any detail beyond that. And in terms of the difference between net and gross premium, you're right, we had flagged last year some of the growing areas. We were growing in partnership with some reinsurers, and so we were sharing some of that risk on a proportional basis. That is coming through now. I think the net premium growth of 14 is sort of in line with the mid-teens that we were guiding to and which we're still guiding to and which we're thinking it will persist into next year. Tell me if I missed anything?

No.

Does that help, Freya?

Yes. I mean the net premium growth guidance hasn't really changed. Your gross premium guidance has increased. Where -- in which lines are you sharing more with the reinsurers?

So the biggest extra reinsurance we've been buying is a proportional reinsurance on our specialty lines and our CyEx books, which are the most capital intensive that we have, and that's why we chose to share that and also the ones that are growing the fastest.

And then you've got the one-off purchases...

And then you've got the loss portfolio transfer, that's right, in property, which does have a one-off impact. So the difference between gross and net shouldn't be as much at the end of the year. Yes, you're right.

Our next question comes from Ashik Musaddi from JPMorgan.

Just a couple of questions again on cyber and one on the overall as well. On cyber, I mean, prices are going up quite a lot. I mean 50%, 100%, as you mentioned. Would you say that most of those price increases that you are seeing is expected to be absorbed by like just increase in the frequency or severity? Or would you say that it's just too early to say, and you've been trying to be a bit more conservative with the loss picks that you are putting. I mean this question was asked earlier as well, but I'm just trying to get a bit more clarity on how do we think about the price increases and the reserves that you're booking in at the moment. So that's one. Secondly is, I mean, clearly, you are doing a bit of a portfolio pruning as well. But is it possible to get some views on how do we think about new customers coming in and old customers going out? How do we think about things like whether the new customers are coming in are at that 100 -- like double premiums versus same exposure or they are coming with a lower premium, i.e., most of the price increases are with old customers? So that would be second. And just related to that, I mean, would you say that going forward, you still expect more new customers? Or would you say that this portfolio pruning will just continue for a year or 2 just because you're trying to watch out what's going on? And the third one is pricing ex cyber. I mean, clearly, pricing in second quarter ex cyber has been more or less stable versus the first quarter, whereas there has been a general noise in the market that going into second half pricing would most likely started taping off in specialty lines, reinsurance, et cetera. So I mean it's early days, I agree, but any thoughts on where the pricing is heading towards, ex cyber.

Okay. Right. So let's try and tackle that bit by bit shall we. So the -- yes, we're also seeing significant price increase, and we're pushing price increase. And at the same time, we're improving our risk selection and we're seeing some early impact of that. The expectation is, if all that plays out, as we're hoping it would that, that will bring the margin back, but we have not reflected that in our loss picks or reserves so far. We're going to wait for further evidence of that. The portfolio pruning is an interesting way of putting it. So what we've been doing is with all the analytics support we're referring to, when we scan our clients and we do our due diligence, our clients are either green, or amber or red. And what we've been doing is having a pursuing a different strategy with clients according to which of those buckets they fall into. And we are treating new and existing customers exactly the same, because that's the right thing to do. But essentially, we're doing our due diligence to see how well protected they are and how resilient they are. One of the interesting things is that itself is also changing, because not only are we being more active in what we're looking for, gradually, the corporate world is realizing that risk management for network security is an increasingly important thing to do. Our clients and companies are behaving now differently to where they were a year ago, because they're realizing they need to improve their own risk management. And so our own renewal clients will be different now than they were a year ago, because they will have done a lot of that work themselves. And so part of what we're in now is a transition from a low risk managed environment to a higher risk managed environment. And that's essentially what we're trying to do is to help them through that. And so we're not really pruning necessarily, we're just making sure that the clients we have, new and existing, are appropriately risk managed, and that is essentially up to the clients.

I would agree with that. And I would add that we are writing new business where we see, we're putting new business through the same process where we see new business that, as Adrian said, is green, that we think when we do our underwriting due diligence and we run them through the technology and the processes we have, we are writing that new business. But the rates we're seeing on that new business are significantly higher than the rates we saw in new business last year. So we're seeing for the amount of new business that we write, it's actually a lot less policies that we need to add to do that because of the rating environment. So we are seeing a lot of new business, both in the U.S. and internationally.

And when we look at our rate change ex cyber, your -- it's a good observation. It's sort of consistent with where it was, and it is slightly different from what we're hearing in the market. And that's due to some of the impacts that we were talking about in Q1, right? So we, like everybody else, have noticed that there is a little bit of tapering on some of the shorter tail lines of business, which continues. But some of the other longer tail lines of business we've got in specialty lines and CyEx continue to exhibit strong rate changes. And so overall, that amounts to the same sort of number, but the drivers are slightly different. And because I think we do have a higher weighting of that business compared to some of our peers, that generates a higher average rate change that we are demonstrating today.

Okay. And just any visibility on like going forward, you would expect it to continue? Or would you say that for your book as well, probably we are getting towards a bit of tapering now?

No, I think -- so I don't think we have anything new to say on the short tail lines of business. What you're hearing from others is consistent with what we are seeing. I think the longer tail lines of business were -- took longer to turn. So, whereas the property market, for example, really started to turn at the end of 2017. It was 2 years later on some of the long-tail classes. So it's slightly behind the market cycle, and our expectation is that it will persist slightly longer because of that. So that's why we're guiding to a similar level of rate change for the second half of the year.

Our next question comes from Faizan Lakhani from HSBC.

Congratulations on the very good set of results. I have 3 questions. First is on reserve releases -- on the reserve margin, sorry. So your margin at 6.6%, I remember at the full year, you mentioned that the 6.3% was depressed due to catastrophe and COVID. So can we assume that the buffer is flat or it has to be down relative to half year or full year? Or is this actually a genuine improvement? And question 2 is on cyber. Could you just give some indication of how severity has developed since October? And also how cyber is coming relative to 2019? And final question is on capital. Could you just give us a breakdown in terms of how the capital developed from implications of the change in investment portfolio, rates and volumes?

Okay. So I think you want to do questions 1 and 3, Sally?

Yes. Sure. So yes, so just to remind everyone, at the year-end, we spoke about the fact that there was a depression caused by a set of claims within the reserve surplus being, COVID in particular, but also the fact that we had some caps, which we have had again in Q1 and to a slightly lesser extent in Texas. And then the other thing that was causing the depression was the fact that the excessive loss we've been speaking about on specializes and CyEx was leading to the net position on both the actuarial number and on our balance sheet being very similar. All of those things still remain. They're not necessarily exactly the same, so that effect still remains in the reserve margin calculation, and it will do to some extent, whilst we're still recovering on the excessive loss insurance, hopefully.

Do you remain at the same extent? It remains to the same extent at this half year?

It probably hasn't changed significantly. It's what I would probably say because the -- if I look at how the gross versus net has moved, as we spoke about earlier, we're still in a similar position. So I would say, it may have moved, but not to anything significant.

And then capital?

And then on capital, short answer, kind of no, so we don't give that breakdown. I guess what you're seeing is an overall increase in the ECR around 5%. So you could kind of work out, well, if we're growing net premium by 14, and we're getting 10% offsets from other things, both -- and that would include both rate change and claims, we don't give that breakdown. That's just a high level...

Did the investment portfolio have any impact at all in terms of how the capital developed? Or is that something modest in the grand scheme of things?

Not particularly. The 2 things that would impact it, is if we changed our risk appetite, which we haven't. And secondly, if there's some volatility change in the market, which there has been up and down over the last couple of years, but it's not a significant driver of our capital. It will do in the rounding, but our peak capital isn't driven by our investment portfolio because relatively, the risk is lower compared to other risks that we take.

So cyber severity?

So cyber severity, it's still very early when we're looking at the business that's bound since October, to talk about severity. I mean we'll take a little bit more time for us to be able to look at that and take a view on that. But with the action we've been taking, both in the limit management, both with program changes where clients -- some larger clients take much larger retentions now than perhaps they did a year ago, we will see, I think, severity start to pull through in the coming quarters, and we will be able to take a look and take a view on that. Performance relative to 2019, well, when we show you the 20% reduction that we've seen so far in the 50 based on premium, that is very much comparing it to 2019 and 2020, early 2020, which is when we saw the most of the ransomware activity. So those numbers are looking at not just how is it down, but how is it down compared to what we saw in 2019, 2020, because that's predominantly where the ransomware threat existed.

May I ask a couple of follow-up questions if that's okay? In terms of cyber, your frequency is down 20%. You're giving a loss pick -- your loss isn't allowing for the improvement in cyber. Can we very [ rudimentary ] assume that the loss pick and CyEx be sort of bolted down 10% to current year? That would be an additional buffer to your reserves? That will be fair?

Say that again, sorry?

Yes. So I think the question was given that we haven't amended our loss pick given the positivity that we're seeing in the data, can we imply that there's more reserve coming reserve margin coming out of that. I would say if the trend continues, then that would be the outcome, but I would not rush to that conclusion at this stage, because I'll say again, as Paul and Adrian have, it's a very early indicator of the data. So I think we need to wait for a year's data at least, probably until the year-end to start drawing those conclusions. But you're completely correct that if we continue with this trend, then that would be an outcome.

Perfect. And just in terms of the second half of the year, if rates go up, do you think that will think just has the worsening trends? Or is that -- can we assume improving cyber rates from here would lead into better experience in the second half of the year?

So rate action is a backward-looking thing, right? So rates are going up because of the claims that the insurance market has paid. And so I don't think rates will continue to go up because of -- what happens in the second half of 2021, rates will continue to go up or not because of what's happened in the first half of 2021 because the insurance market reacts to what it has seen. And I think the rating environment will persist until the insurance industry, in general, has reassured itself that it's on top of this problem, and it is a long way from that, yes.

Our next question comes from Ivan Bokhmat from Barclays.

I've got a couple of questions, both in cyber. The first one would be related to the capital intensity of the business. I'm just wondering if given the panel that you've got on the call right now, if you can help us understand for cyber, what is the -- what cost of business would you compare the capital intensity to? Is it -- would be akin to specialty or akin to property? Is clearly some of the aspects like the accumulation risk would make cyber quite close or similar, let's say, with Property CAT? But I presume that the risk charge under your model, under the most rating agency models is far lower. Maybe if you can give some general thoughts on that, maybe you're aware of any discussions happening broadly of perhaps increasing risk weights in the Lloyd's or Solvency II framework, that would be very interesting. And the second one, I suppose you spoke about the attritional versus systemic risk. And I'm just wondering, could you explain a bit how you can distinguish between your customers? Are you able to offer more narrow policies, limiting that systemic risk to some others outside than just the risk limits or typically the policies you offer are quite broad. And therefore, you'd still be exposed to let's say, cloud outages even for the reach response customers?

On capital intensity, why don't you...

You have a crack. You quite like it.

Yes, it's a good observation, right? So it does have attributes or elements of both property and liability in it. And if cyber have been born in a different way, it would have been in our property team. But as Paul said, because it emerged as a regulatory and legal obligation, it emerged in our specialty lines team. I think -- and the -- so it has its own place in the internal model. It is modeled separately. I would say if it was like anything, it's probably like D&O in terms of how capital intensive it is. And I think a significant part of that is still the fact that although we've been doing this since 2009, it's still a new business with a lot of uncertainty to it, and that generates an extra load. So part of that is because of the volatility that it exhibits. But actually, a large chunk of it is because we still know the capital ratios because it's new and there's uncertainty around that newness. Attritional versus systemic is an interesting one. We have, haven't we discovered as we've done more modeling, which bits of cover create more systemic issues than others. And the cyber coverage has broadened over the years. And as we thought about that more and more, we have realized which bits do. And as Paul, I think, alluded to earlier, the real thing that joins all businesses together is business interruption and it's cousins, dependent business interruption and service interruption. And part of the management that we've been doing these last couple of years is to be much more stringent and careful as to how we give that coverage out because of the fact that we do know if it accumulates. Paul, is there anything else you'd like to add on top of that?

I think we've been doing that, but we also do a lot of work to understand what these scenarios would look like because they don't mean the same for large accounts, for example, as they do middle market. It doesn't mean the same for international as it does in U.S. It doesn't mean the same for different industries. So when we're thinking about the cover we give to clients, we're thinking through those issues, and we have a good understanding of when these scenarios happens, who is that going to impact? And therefore, what coverage are we giving them? And how much are we giving them? Are we -- it's a combination of all those things put together.

Does that help, Ivan?

Yes. Maybe if I can follow up. I mean, considering that a lot of your portfolio would be market risks, I'm just wondering whether you perceive the wordings of the contracts in your book different than from broader market exposures. There's some differentiation that you can help illustrate.

The differentiator between our cyber wordings and the market cyber wordings? Is that what you're after?

Yes. As in, if we look at the largest cyber writers like [ Mid-Free ] would presumably take more larger clients. And it doesn't have to be many, of course. But whether you think there is a difference in the wording between you and the largest players, let's say, in the U.S.

Yes, I think the biggest difference is around the fact is in our Beazley Breach Response product, which gives us control over the claim of -- why don't you...

So when you look at the cyber market and the wordings, firstly, agents spot on the main point of differentiation in the wording is the services on services. Look at our wording. Our BBR product is designed -- it's typically used in the middle market. The middle market clients are the ones that gravitate to the Beazley Breach Response product, because they really crave a lot of those services and the support that we give with the policy. When you line up wordings more generally, the coverages there are all very similar in terms of the actual coverage within the policy. But that said, we also write a lot of cyber that isn't on BBR. We said we're striving for a diversified book, so we write large accounts where perhaps BBR may not be the best fit. We have a middle market book, and we also have an SME portfolio as well. And so we really work hard to keep the balance of diversification between those. But right now, if you were to compare the policy wordings, I think the services and the risk management services and the way we structure those and the policy, is the main difference that you would see.

And that manifests itself in how the claim presents and that's the important thing for us, right? So the value to the client of Beazley Breach Response is if they have an issue, they can call us up and we will handle it for them and take a lot of the stress out of that, because it's a very stressful thing to do. And we have a very experienced team that can do all that. The advantage it gives to us is that we're controlling how the claim goes. We're doing a lot of it in-house. We're controlling the costs, and the result of that is that not only is the claim safely handled, but it's handled by people that we know and trust and therefore, the costs are contained. And that control over how a breach is handled and how the answering claims is handled, which you don't always get an insurance because oftentimes, it's an indemnity policy so the client does it and then we reimburse them. The control that gives us is a significant advantage for us.

The other significant advantage that gives us is speed. We see data far quicker, because clients are with their first call, some things happening, they call us first. So we will see trends. We will see costs. We should see movements far quicker, because we have it in-house rather than waiting for third parties to tell us about what they're seeing.

Great. By the way, the focus session on cyber is brilliant.

Thank you.

Our next question comes from Ming Zhu from Panmure Gordon.

Just 2 questions, please. First is U.S. liability line. There's a comment that you said you're expecting the return of social inflation. I think that's quite in line with the market expectation. But could you just give a little bit of color in terms of what you've assumed? And what actions have you taken in that line? And my second question is that you're quite bullish in terms of your growth for full year 2022? And how much of that is actually embedded with rates momentum, outlook? And how much of that will be organic growth?

Okay. So to take the first question, we're assuming that social inflation goes back to the sort of levels that it was and the sort of drivers that it had before the pandemic struck. And so we were talking quite a lot about social inflation and where that was having an impact and the sort of impact that was having back in 2018 and 2019. And our central assumption is that that's what we revert to. And we took underwriting action around that back in '17 and '18, and we have superimposed on top of that additional action reflecting the fact that we do think there's a danger of COVID-related claims activity on top of that, and that's what we've done. And those actions remain in place. When we look at growth next year, it is partly rate and partly organic growth. I think our overall level of rate change is probably slightly down from this year into next year with slightly more organic growth, noting that overall that the rate environment should be more positive, but we'll be able to provide a little bit more guidance on that at the end of the year when we've got more certainty about what that mix looks like. But at the moment, those are just our first and I'll first pass at it.

Our next question comes from Andreas van Embden from Peel Hunt.

I just really had one question. The -- we've seen sort of 3 years of rate-on-rate increases across your portfolio and your peers as well, and I appreciate you're seeing an acceleration in cyber now. But I just want to see how this sort of filters through into your underwriting margins. Are there rate increases across your whole portfolio? Are they well above loss cost trends now? And are you seeing this translate into an improvement in your attritional loss ratio by your loss ratio, excluding the reserve releases and the large losses and the cap losses you incurred during the first half?

Okay. I'll have a go then. If I get it wrong, please, please. So I think -- so yes, I mean what was Sally saying earlier, she was saying that one of the goals she had was to go over the couple of years ago, was to show more consistence and better prior year reserve releases. And we've gradually been getting closer to where we used to be as an organization, which I think is evidence of both the reunderwriting that we've been doing and the rates that we've managed to charge. So we are hopeful that the underwriting margin is back in the business, and we will start to see that flow through to our financials, which is one of the reasons why I think we're guiding to the low 90s. Having said all that, we live in risky times, right? So we're talking about European floods. We talk about wildfires. We talk about uprising in South Africa, we're talking about cyber changes. The world is a very volatile place. And so the game is never finished, right? So we consistently have to react and make sure we're on top of what's new, as well as, fixing what we know about from the past, which is sort of a gentle hedge. I said before, that all other things being equal, yes, it should begin to flow back in.

And how long do you think that will take for those rate increases to really sustainably take you back to the sort of low 90s or perhaps even better combined ratios?

Well, we're guiding to low 90s this year, Andreas. So we're hoping that, that we'll start to know then. When you look at our combined ratios in the past, when we had combined ratios under 90, those were generally years with significantly lower than average CAT activity. So we had a sort of a run rate of 92-ish, low 90s and then the bonus, if you will, of a low CAT or CAP for a year.

Understood. So the rate increases will bring you back to where you were on average before?

That's what a low 90s projection would imply, yes. Absolutely.

And our final question comes from Nick Johnson from Numis Securities.

Just 2 questions, please. Firstly, on capital. The required capital growth for next year estimate of 5% is a lot lower than recent years. It's been around about 15%, I think, you mentioned the benefit rate in that calculation. Just wondering to clarify, does the estimate for next year anticipate any change in the percentage of premiums ceded? Is it possible that the required capital number for next year could go up once you finalize your cyber growth plan for next year? And then the second question is on SPAC D&O growth. Perhaps you could just talk a bit about the risk management in that category, sure to be filled by quite a risky segment.

Right. Why don't you take the first one while I bring my D&O underwriter and ask -- rather answer the second one.

Yes. Nick, so the -- so yes, you're completely right. What that 5% increase allows for is what we're planning to -- currently planning to grow in the business plan early days on it, net of reinsurance, which, as we've said, is similar to what we're looking at this year. So that does allow for the reinsurance -- the growth in every reinsurance to continue into next year. If definitely, that is an early estimate of the number, and as in every year, that will evolve between now and the end of the year as we continue to evolve the business plan. And then also depending on what happens on cyber and depending what happens in cyber in terms of rate and exposure growth and what our decisions are within the business plan, which we continue to work through, they will all have effects on the capital requirement. They can go up and down because, as I said earlier, one of the reasons the capital is down compared to the overall premium growth is the exposure is down on cyber in particular. So we have to work through all of that to see where we end up in the year, again, following on from the comment Paul made earlier that we've got some encouraging data, but there's more to learn between now and when we finalize the business plan. But yes, it will definitely evolve in the coming months. Over to Adrian on SPAC.

Thanks. Thanks, Nick. That's a great question to finish the morning with. So I guess to start with, SPACs have been around for a long time. I first had SPACs explained to me when I was running SL back when, must have been 5 or 7 years ago, and they've always been slightly curious sentiments, but they've been around for a while. And they've really come into vogue as we all know, the last couple of years or so. And they're quite esoteric in terms of the particularities of how the different transactions work and what the liabilities are around those different transactions. And so we designed a policy form, so that it was very clear what cover was being provided when because it wasn't always obvious, just using a normal D&O policy. And that's sort of what we're trying to do, right? When things are a bit different, we customize and we bespoke to us so that it's very clear that what is being done where, and there was a chance to do that here, I think. We are -- although they've been around for a while and although we've managed to produce policy form, I think is a very good one that produces that clarity of what's being covered. They always have been, and they remain quite high-risk things just like IPOs are high-risk things, and that is reflected in our underwriting due diligence and how much limit we're willing to pay out and the sort of premiums that we charge. And we're very careful about what SPACs we do, avoiding the celebrity endorsed ones and so on and so forth. We'd like to do the ones that are proper businesses and so on and so forth. But they're a valid part of our D&O book, and we have some history with them, but we're very careful because they are high risk, just as we've been very careful with IPOs over the years.

We currently have no further questions registered. So I'll now hand back to Adrian, Sally and Paul to conclude today's call.

Great stuff. Thank you very much, indeed. Well, thank you very much for calling in this morning. I hope that was useful. If you have any follow-up questions, please get in touch with Sarah, and we'll do our best to answer them, and we look forward to seeing you very soon. Thank you very much.

Thanks, everyone. Have a good day.