This a shortened version of Making Risk Flow podcast, episode: “GenAI, Sustainability and the Talent War: Inside BCG’s Value Creation Report”. In this episode, Juan is joined by Nathália Bellizia, Managing Director at Boston Consulting Group (BCG), to analyse the BCG’s upcoming Value Creation Report. The eagerly awaited document shall help identify the patterns of insurers that have created value over the last five years compared to those that have destroyed value. Juan and Nathália also discuss the three key themes which have contributed to value creators and value destroyers: GenAI, sustainability and the talent war.
Listen to the full episode here
Juan de Castro: Today I’ve got the pleasure of talking to Nathália Bellizia, Partner and Managing Director at BCG. Nathália leads Corporate Finance and Strategy in Insurance for BCG globally, and one of her areas of focus is value creation. This conversation is very timely as BCG is about to release the new 2023 Value Creators Report. So in today’s episode, I will discuss with Nathália how the industry as a whole is performing and specifically what are the characteristics of those P&C carriers that outperform their peers. Thank you so much for joining me today. I think we realised a few weeks ago that we’ve got plenty of friends in common and we are alumni of the same business school, but we had never had the chance to meet until probably whatever five, six months ago. Thank you so much again for joining me. It would be fantastic if you start with an overview of your role at BCG and what do you focus on?
Nathália Bellizia: Absolutely. Thank you very much for having me. Very excited to talk to you again. I’m a Managing Director and Partner with BCG in our insurance practice. I spend all my time working with carriers and distributors on a variety of key strategic topics. And in particular, I lead for BCG what we call Corporate Finance and Strategy. The main focus of that particular area is around value creations. So I spend all my time working with insurers, making sure that they are deploying the right strategies. They are thinking about their portfolio, business strategy in order to create value for shareholders. So that’s one of my passion topics at BCG.
Juan de Castro: You said one of the areas where he focuses on value creation specifically with insurers. And I think you mentioned to me before you’re about to release a new report specifically on that. I think it’s focused on the US P&C Value Creation Report. Perhaps you can give us just a sneak preview of themes and key insights from that.
Nathália Bellizia: We’ve been writing the Value CreatIos Report for the better part of the last 20 years. It’s always very interesting to take a pause and look at the industry holistically, see what are we observing globally and then across the different businesses. It’s been a pretty challenging last couple of years for the industry and so when we look at the industry globally, what we’ve noticed is that over the last five years, the annual TSR has been on average very much below cost of equity. So for the last five years, the industry has delivered approximately a 4% annual TSR and if you consider the cost of equity being around 9-10% that’s half of what the cost of capital has been, which I think is a function of a variety of things that we’ve experienced globally, I think pandemic, massive shifts in macroeconomics, inflation and interest rate, supply chain challenges, geopolitics. It has been a very interesting period and challenging period for insurers. And starting last year, what we’ve decided to do was to actually deep dive into particular aspects of the industry. Because what we’ve noticed is that it’s very hard to draw themes around value creation when you are looking across all lines of business and across all regions. Last year we decided to launch what we’re calling a US P&C Value Creators Report and that allowed us to actually get very much deeper into what’s happening in the industry and who are the insurers that are creating the most value in this year’s report. What’s super interesting is that while globally and across lines of businesses, the insurance has not been delivered on shareholders’ return expectations. On the P&C side, they have. The average annual TSR for US P&C players is around 10% per year, which is at or close to above clearing the cost of equity. And despite a very challenging market, I think the challenges are very well known in personal lines with loss caused by inflation, hitting insurers, also a lot of secondary perils and CAT becoming more prominent over the last couple of years as well. On the commercial side, we’ve been experiencing a hard market for a while. And so I think returns were also very close to the 10% mark, but with less disparity across, across players. But it’s very interesting to see some of the themes around those that are actually performing better than others, cause in doing this type of analysis, what we typically find is that the averages don’t tell you the full story. So it’s actually interesting to get deeper into those.
Juan de Castro: To look at the variance within the distribution that leads to that average, right? And have you been able to identify what are the common characteristics of those in P&C insurers that have outperformed?
Nathália Bellizia: Yeah, in terms of different dimensions a couple of themes tend to pop. I think one is around the idea of focus. Players that tend to spend more time on their niches or areas in which they have very strong expertise tend to do very well. A couple of examples that we highlight in the report this year is Progressive which focus on auto insurance and craft that market extremely well. The other one is Kinsale the number one in the ranking this year, targeting small and medium enterprise risks on the E&S market. We typically find that data-driven, and analytics-driven organisations and businesses tend to outperform. Even those two companies are known for a very distinct approach to data that is leveraging underwriting pricing and that discipline that is required for one to create sustained value in the long term.
Juan de Castro: So in those cases, is it less about which niche are they focusing on, but is it more about mastering that niche?
Nathália Bellizia: Yeah, I think you nailed it. It’s not about one niche being particularly profitable in the market, but it’s about issuers being very deliberate in understanding the distribution and the risk better than others. That and being able to leverage data in analytics to be nimble and actually execute on that particular niche that tends to pay out. And it does pay out. If you look at the top quartile players, the average we’re talking about 20% TSR, Visa fees, the 10% of the overall sample. We’re seeing the parity between winners and losers actually getting wider and wider.
Juan de Castro: I think the fact that one of those two, you mentioned Progressive, is focused on auto, is quite revealing. Auto has always been a line of business where most people struggled to make a good return. I think what you’re saying is by focusing on that data-driven approach to really doing superior risk action in those lines, you can still outperform the market.
Nathália Bellizia: Yeah. The reason we look at the return on tangible equity is that we find consistently that’s a very good proxy for what drives the insurer’s price to book value multiple over the years. And so when we look at top quartile players, we see that they achieve 15 percentage points of return from underwriting. And that’s basically the difference. And when we look at the bottom quartile, we find that it’s negative, the underwriting returns. And so that’s the main driver of the overperformance. I think I couldn’t emphasise enough the importance of being underwriting driven and actually focusing on that particular dimension for one to sustain value creation in the long term.
Juan de Castro: Did you also look at whether that superior return is mostly driven by purely risk selection or underwriting excellence, or is it also driven by operational efficiency?
Nathália Bellizia: I think there are two things that are particularly important in P&C. One is the risk selection and pricing, and the other is your expense ratio advantage. I think what is super powerful is when you actually combine the two. It’s your ability to have your cost advantage and be more competitive on the risk that you want to write that allows you to create that flywheel that drives growth at a disproportionate amount in the industry. The number one and number two in the Ranking, Progressive and Kinsale are great examples of that because they actually have both. Progressive with the channel mix that they have and Kinsale with the technology that they deploy. They both can actually grow faster than the rest of the industry by pricing more competitively as well. And so I think what’s super powerful is the combination of both that you alluded to.
Juan de Castro: Yeah. And then because that combination of the two was historic, you go back 15-20 years, there was this false trade-off between risk selection and efficient operations, right? If you would really good risk selection, you want your underwriters to spend a lot of time on each risk. And if I understand you correctly, we’re saying it’s data and analytics. So that skill is what allows you to achieve high performance, both in risk selection and underwriting efficiency.
Nathália Bellizia: Exactly. And I think part of the reason why there is no such a thing as a compromise anymore is that with advances in technology and AI, I think that they bend the curve. You don’t have to prioritise one at the expense of the other. You can actually have both. And those are what the winners are actually trying to do.
Juan de Castro: So so looking at the report, what does it mean for the insurer? What should they get really good at?
Nathália Bellizia: It’s a good question. And I think there’s an argument that top quartile players in the US P&C market have cracked the code. They are creating value in spite of a very challenging market. There is an alternate argument and maybe they’re not necessarily mutually exclusive, which is success. What drove success in the past might not be what will drive success in the future. Because when we look at the market today, we see a lot of challenges and potential headwinds. You look at what’s happening, all the uncertainty around the inflation supply chain and the impact it has on lost cost is pretty severe. I don’t remember seeing that in my time in the industry. And then you look at what’s happening with social inflation. No, it’s not looking like it will decelerate anytime soon. You look at climate change and people saying the secondary perils are becoming primary perils given the activity on a variety of events. You see competition evolving as well. Yes, insurers are always being threatened by disruption, but you see very strong players. You see OEMs as an example, trying to create new value propositions on the auto insurance market. To top everything with now Generative AI on overdrive since the launch of the ChatGPT, there’s a lot for an insurer to tackle right now. What we suggest, and that’s very consistent with conversations I’m having with clients on a regular basis, is to focus on one of four main things to drive success in the future. First one is it’s very easy to get lost. So I cannot emphasise enough the importance of focus on insurance fundamentals and the riding excellence that has been the main driver of results in the past. I don’t believe that this will ever change for P&C Insurers. The second thing we suggest is embracing AI data in analytics in absolutely everything you do. I think there are a lot of people much smarter than I am saying that this is the most transformative technology that they have ever experienced and I tend to agree with that. I think the opportunities are endless. Being able to harness the power of AI will be very important for every insurer. The third one is I cannot talk about AI without touching upon implications for talent, and skill development. I think there are still a lot of questions on whether Generative AI, as an example, is actually more beneficial to high-skilled workers. It’s making your top talent even better, or if it’s actually more impactful on the lower-skilled work. There’s debate. I’ve seen studies on both, but one thing that we know for sure is that it will transform how people work. What are the skills that you will be looking for? An underwriter or an underwriter assistant or a claim adjuster on a before basis. The insurers will have to tackle that. Then last but not least, be ready to energise the fight against climate change. I think this impacts insurers on a variety of ways. Think about physical risk, which is at the core of what insurers do when it comes to property. It impacts insurers as well from a transition risk perspective. You look at what will happen in the renewable market in the next years to come. That’s a massive opportunity for insurers. Also on the litigation risk, I think we’re seeing how fast things are evolving and claims of different nature coming in. I think that those four topics are the ones that insurers will have to be on top of to create value in the years to come.
Juan de Castro: That is a great framework in terms of some of the big themes. Some of your clients, some of the CEOs of this insurance company should be thinking about focusing on fundamentals, embracing AI, Data Analytics, the war for talent and be ready to re-energize the fight against climate change. So it would be really helpful to go one by one. And then if you could share some of the conversations you’re having with the CEOs and how you’re advising them on how to tackle each of them. And perhaps let’s start with the one which is probably more on the news today, which is about AI, Generative AI, Data Analytics.
Nathália Bellizia: Absolutely. We’re having a lot of conversations and the reality is that things are evolving really fast. I think what’s really powerful about Generative AI is actually three things. One is with traditional AI you had to be a data scientist to use it. You have to have very, a lot of data, a lot of clients data you had to learn or know some coding language. And I think what Generative AI does is putting the technology at everyone’s hand, literally everyone’s hand and at no cost. And so I think that democratisation is one thing that is, I don’t think we’ll ever go back and now the coding language is actually English or is Portuguese or Spanish, you name it, right? And so I think that that has such a profound implication for how people will engage with AI. The second thing is the breadth of application. I think before we were talking about developing very specific models to answer particular questions on contained fields of experience. And right now we’re talking about the ability to ingest fact, data, videos, images, and voice, and create things on a much bigger breadth of domains. And so even what the technology is very good at, at summarisation, at creativity, things that we didn’t have before. So this breadth of utilisations is very powerful.
And then the third one is the time to value. I think before you were talking about AI and before you actually build an AI model, you will have to have your data lakes, your data cleanse, all that in place, and then you would train the data off the sample, you would test and test and test. And what I find amazing is how quickly you can get to some, but pretty good results with very limited resources. What makes that possible is large language models that have been already trained. Yes, you will still have to fine-tune for very specific applications, but I get consistently amazed by how well it does with you simply uploading one or two policies and interrogating the data on it. I think those true things lead me to believe there’s no way back. And this will be pretty transformative. There are a few use cases, I think there’s a no regret move, which is to embrace and embed Generative AI in the way you think, and we know there are pretty important responsible AI implications of data protection and how people will actually deploy the technology. But the idea that you can actually have your personal assistant with you at all times is super powerful. The second thing is the what’s around horizontal use cases we call it. So it’s basically, you can think about every single function of an insurer or a company. And then there’s the third category, which is how should insurers think about their core functions. You can think about claims. You can think about underwriting, what should be the role of an underwriter assistant if you had a chatbot that can actually summarise the information you received from a broker, can access your pricing models, can make suggestions for the debits or credits for that particular account, can scalp the news, can listen to your calls and transcripts from interactions with the broker and actually make you a recommendation. How does it change the work of the underwriter assistant? How does it change the work of the underwriter, him or herself as well? So it’s pretty powerful. And I think the same thing is true for claims. How will we be adjudicating claims in the future if we could listen to all the calls and automatically ingest information and compare that with reports from an adjuster with the pictures that have been taken with the aerial image or something and things that for the human brain is very hard to do because you’ll be putting together way too many pieces. I think that’s where insurers should spend most of their energy and time trying to think about the places where their data could give them an edge and a competitive advantage and create those more unicorn use cases.
Juan de Castro: I think Generative AI and large language models have the potential to really accelerate. I‘m using the word accelerate because, in the end, it is about how these new technologies accelerate part of the vision of insurance companies. But given the breadth of applicability, isn’t there a bit also of a challenge, a deer in headlights type of situation where the applicability is so broad in every function, in every way, every in words, that that part of the discussions I’ve been having with some of our clients is like, where do you start? I believe we’re still at that phase where everybody’s testing out ideas, but there’s not a clear strategy, right?
Nathália Bellizia: Yeah, I think you’re absolutely right. We should not ever forget the point around responsible AI and all the challenges and implications it has for how companies think about it. I do find that a lot of insurers are actually shutting down access to it because they’re not feeling in control. And I can understand that up until ChatGPT, you knew exactly where AI was being used in your organisation. And all of a sudden you no longer know if an email was drafted by AI or not. In reality is humans are very bad at identifying if it is or not for drawings or emails and texts, et cetera. And so I think that the trap of thinking “let’s just wait”. And when someone had figured it out I will go and follow because things are happening too fast. And I don’t think that any insurer could afford not to be thinking about it, how we can transform the function. The one other piece of advice we generally give as well is like, sometimes making distinctions between traditional and Generative AI on tackling a problem, it’s not particularly helpful. It’s very easy to run the trap of like, Generative AI is the hammer, let me look for nails. And the use cases we’ve been discussing that are most powerful, they actually combine the two. They actually combine like, for example, you are creating synthetic data on losses to up sample your losses, to actually train your pricing model to identify fraud. So it’s an example, you were actually combining the two, the ability of Generative AI to generate data for the sampling with traditional AI in your pricing models to be trained and better at identify fraud, or have fewer false positives, which typically lead to a lot of costs in the fraud investigation units of insurers. So it’s generally much more powerful if you think about a business problem, and you then look at all the things you have at your disposal, and how can you tackle that problem? Which I think is also tied to one other component of the four things I mentioned, which is talent. I think being very deliberate about the skills you are looking for, and how they will evolve over time. I’ll give the example of the underwriting assistants within every single company, every single commercial insurer I worked with. It’s a stepping stone into becoming and being trained as an underwriter, right? And what if we have fewer of them? Are we going to actually impact our ability to have a pipeline of underwriters in the future? Could we be looking for different skills on those given the things that we can actually delegate to the technology itself? And so I think that talent implications are pretty dramatic as well. We’ve been pretty deliberately in client conversations suggesting that you don’t separate the two because they are so intertwined because the power will be with individuals making use of better technology. You can’t tackle them separately.
Juan de Castro: In that specific example about underwriting, so entry level positions underwriting systems and how AI could potentially replace part of all of those roles, there are two schools of thought. One, which is, okay, that’s going to be a challenge because you are removing the entry point for new talent into the industry and then they get trained, et cetera. The opposite view is actually you can have new talent coming in directly as underwriters and the technology will actually support them much more in their training. You will have general AI that is going to support you. So it’s actually an accelerate of those careers rather than hindering them.
Nathália Bellizia: Yeah, I know. I like your second school of thought. I should also say that I do think in the short term, it’s not about technology replacing the individual. I think it’s the individual using the technology will be so much more productive. So that will be the power as opposed to replacing them completely. And I really maybe like, because the parts of the job that are less enjoyable will be potentially automated or made more efficient. What if we actually make the industry attract a lot more talent overall, it can be great, right? I believe in the glass-half-full perspective more than the other one.
Juan de Castro: So you see the impact on talent, is it both in terms of career paths and roles? Is it also about the skills insurers are looking for in that new talent?
Nathália Bellizia: I do think too, having conversations with clients about what makes it a good underwriter. And so there are a couple of things I don’t think will change. There’s something about intellectual curiosity and the underwriters are typically very curious and they have to truly understand the risk. There’s something about the relationship that they build with brokers and clients that is very important as well. There’s something about the job that is entrepreneurial and makes you want to cultivate your own book of business. I think those things will remain important. I have a few clients asking themselves the question, would I need more of that process management aspect of the job, making sure that everything is on the file that we respond quickly or will those be eventually automated or made better by technology? So you don’t need a person making the first drafts. You need someone revising or you need someone prompting the technology. The way you ask questions has real implications for the responses that you get. So we’ll be spending more time learning how to ask the right questions as opposed to, actually getting to an end. And so it’s pretty interesting. I tell some of my clients that if they haven’t yet hit a wall in terms of their thinking and had trouble sleeping in playing out, what are all the things that technology could do? You probably haven’t yet played with ChatGPT enough.
Juan de Castro: Definitely. We’ve talked a lot about GenAI. One of the things that really resonated was the point you made about the GenAI at the end is a hammer. It’s a tool. I often have these conversations with clients, first thinking about the use case, what are the pain points, the vision, the strategy, and then GenAI is just one more tool. I think you summarised that quite well. We touched on AI, Data Analytics, GenAI, we touched on talent. You mentioned at the very beginning this focus on fundamentals. What are the key points in those regards?
Nathália Bellizia: No, absolutely. So one of the things we are observing in the industry as well is different lines of businesses being at different points on their cycle and conversations I’ve been having with clients is around the importance of you being very deliberate about the book of business that you want to underwrite. And I think reality is that a lot of insurers are simply re-underwriting their books and maybe with two or three points more of rate, but not really taking that hard look, should we be in that line of business? What is our right to win in that particular area? And so having a lot of conversations around portfolio steering, very deliberate portfolio steering and capital locations across the entire portfolio as a way of informing your underwriting guidelines and informing your risk appetite. And at the end of the day, inform what your underwriters should eventually do with their own books of business. It’s not about taking the autonomy away from the underwriters but being very deliberate about return expectations on each individual book of business, in taking a data-driven approach to some of the accounts even. Asking questions like “Are we going to, can we get to our technical price in this particular account?” If not, what is the remediation plan and how should we think about it? It’s a lot more of an active steering of the portfolio towards and better priced risks, that comes also with a good understanding of what are you good at. And where do you have the ability to actually be the lead underwriter in that account and set the terms and conditions and pricing.
Juan de Castro: I’m sure nobody would disagree with that and that’s almost the fundamental of the insurance industry as a whole. How often do you see that done well? That’s much harder to do than to say.
Nathália Bellizia: Absolutely. I like to say that’s no brainer and it’s easy and it’s very straightforward, the concept, right? You define where you want to be and what are you good at. The challenge is execution because it’s first you have to have all the functions within the insurer actually working together. Effectively, you need your finance organisation, you need your underwriting teams, you need the claims so that loop works well, you need your actuaries and your pricing models to be in sync as well. You need to have a be at a reasonable place when it comes to data and technology. I know you won’t be surprised, but sometimes I still get amazed that you ask a couple of questions about the profitability of certain aspects of the business and some insurers cannot cut the data in that particular way. So how can you actually steer if you’re not even having the right visibility into forward-looking metrics? And so I think it’s complicated to execute. The strategy is at times simple, the challenge is how you can have the entire organisation rally behind that strategy and you have the right measurement and performance management in place to actually make sure that that happens.
Juan de Castro: Definitely. And I don’t want to wrap up before talking about the last point you mentioned in your four points, which is about being ready for this, everything around sustainability, climate change, transition to net zero. So give us a couple of bullet points on what’s your advice to the clients and what are you seeing being done well.
Nathália Bellizia: On that point, I would say insurers play a pretty important role in society overall, but in this topic in particular I do think the insurance industry has the ability to be quite influential in a variety of ways. I also think that this is a massive business opportunity. The numbers, depending upon the source, they’re slightly different, but we’re basically talking about 120 to a 150 trillion dollars to be invested in achieving that path to 1.5 degrees Celsius. And that investment will only happen if insurance is in place to mitigate the risk associated with those technologies. So I do think there’s a pretty strong business case for insurers to actually be investing in the transition being heavy on renewable technologies, you have the more developed solar and wind, but also batteries, CCUs, Hydrogen. Those will be technologies that insurers will need to understand. Those will be technologies that will have to be maturing over the years to come. And I think that there’s a pretty compelling business case for insurers to actually be ahead of the game, helping them, their clients transition and taking advantage of that pretty significant opportunity. It will be beyond energy. And it’s not just energy that we need to transform. You can think about what is required from aviation, from marine, from industrial goods, transportation, et cetera. Every sector will have to decarbonize. And I do think insurers will play an important role in supporting their clients and understanding the technology, helping them mitigate risk and transfer some of those risks as well.
Juan de Castro: And do you see a big distribution in how different insurers are supporting that today? Are they, what percentage of them are they thinking about it versus doing something about it?
Nathália Bellizia: Yeah, I think a lot of them, particularly on the commercial insurance side you see more mobilisation. There’s no week that passes by, there is no announcement of a Head of Energy or a Head of Renewables, either in a large commercial insurer or the brokers are also organising around it as well. I think that the ambition level can be very different among insurers, but I think most of them are realising that inaction on that particular topic is not an option. I think some of them are more bullish, want to be the leading player in supporting clients in the transition. Others are more comfortable being, I’m not a fan of the term fast follower, but there are a few that there are more on that camp. Maybe I’m not going to, I don’t want to be the leading one. I just want to make sure that I’m close enough to learn about the developments and start getting the experience and the expertise, the data, so that I can play in the space into the future.
Juan de Castro: Really looking forward to the developments in this space. As you said, I fully agree that the insurance industry has a huge role to play in society in terms of transition. So that will be really, really exciting to see in the next few years. Nathália, it’s been a pleasure having this chat with you, giving us that kind of sneak preview of the report you’re about to release and looking forward to seeing the full report.
Nathália Bellizia: Absolutely. Thank you so much for having me. A lot of fun and also always nice talking to you. Thank you.
Juan de Castro: Cheers. Making Risk Flow is brought to you by Cytora. If you enjoy this podcast, consider subscribing to Making Risk Flow in Apple Podcasts, Spotify or wherever you get your podcasts so you never miss an episode. To find out more about Cytora, visit cytora.com. Thanks for joining me. See you next time.
