The Bank of America Financial Services Conference. If you're looking for W. R. Berkley presentation and learning more about it, you're in the right room. We're really pleased to have Robert Berkley, CEO of W. R. Berkley, here with us today. I think since 2015, yes, since 2015, the name is still on the door, and we have the right people here to talk about it. And also we've got Rich Baio and Karen Horvath in the audience. Thank you all for being here. We really appreciate it. And we can get into I have questions. You might have questions. If someone raises a hand, I will happily call on you. But why don't we just get started with a few things I think on people's minds? Thank you for indulging some of my questions, which are sometimes unusual.
I think that back about 20 years ago, in the wake of the 1997-2001 events, it birthed a lot of people talking about accident year results. So people are looking at these core combined ratios. But prior losses are real losses, and CAT events they take out and whatnot. You've generally managed the business to, I think, a more these are results we take on less CAT risk than everybody else. The reserves have been fairly stable for the last few years, despite very, very high ROEs and a lot of people moving around their accident year results. What is the Berkley philosophy on combined ratio and maintaining such consistency over time? Where is that coming in? It's really in the all-in number you're delivering.
So well, thanks for the question, Josh, and thanks for the invitation to participate here. I was pleased to be coming down to Florida. South Beach and insurance usually don't naturally go hand in hand, but it was.
Unless it's from a CAT capacity.
That's exactly right.
That's what one's looking for. But as far as the question goes, I think there are a couple of pieces that I'd like to flag. First off, I think there's a shared appreciation amongst many of us, if not all of us, that this is an unusual industry because we're one of the few industries where you really don't know your costs of goods sold until some number of years after the sale actually occurred. And that is easily said but not so easily managed. We, as an organization, are very much of the view that one has to, on day one, to the best of their ability, take somewhat of a measured or cautious approach in how you come up with your initial loss picks, with the understanding that you may get the porridge too hot or you may get it too cold.
It's very unlikely you're going to get it just right. Our preference is, when possible, to err on the side of caution. As those losses develop out over time, you will tighten up that loss pick. That has been our approach. That's why, more often than not, our reserves have proven to be comfortable, if you will. That having been said, every day after you write the policy, you learn a little bit more about what that ultimate outcome is. One of the differences, perhaps, between the organization that I work for and some others is that every day we're looking at that, and we are trying to refine, A, how we think about our reserves and making sure they're what they need to be, and B, how we think about that instructs our selection in pricing going forward.
So the consistency of our results, or the relatively tight band that you're referring to, is partly as a result of what I'm talking about, that constant refinement every day. But the other piece is, as you flagged, it's also our appetite. You referenced volatility. We have no problem philosophically with volatile product lines, such as CAT-exposed property. That having been said, our view is, when you have to approach the business through a lens that we would define as risk-adjusted return, volatility needs to be a component that is considered when you think about the risk that you're taking on and the return that you should be entitled to. There are moments in time when it comes to more volatile lines, like we have recently seen with property CAT, where you are getting paid for that volatility.
And then the lion's share of the time, most of the cycle, you aren't, which, again, instructs us as to how we will participate. So yes, are we more consistent than many? Yeah, I think that's true. Why is that the case? Because of how we are constantly refining the business every day and also how we think about risk and return and when we choose to participate in certain parts of the market. So I made a joke earlier when you walked in, but something you've said in a lot of calls recently, the expression, the pig through the python. People talk about 2016 to 2019. And that was a very soft market. And reserves have proven to be adverse, offset by COVID and post-COVID favorability on a lot of reserves, workers' comp, and particularly being something that's been very good.
How do we know the pig is going through the python exactly? We spent about maybe two years with suppressed paid claims data refining that consistency, I guess. Courts were fully reopened last year, I guess. Do we think that the timeline getting comfortable with the knowledge on those prior accident years is delayed, or are we getting the information pretty much on schedule to pass?
I think at this stage, first off, if you look at our reserves, the average life of our reserves is approximately four years. So it's pretty simple math to try and figure out how far along are we. Now, that's the total, if you will, or the average of the total. There are some parts of our portfolio that are longer, some parts that are shorter. As far as how far along we are, you have that as a backdrop as to sort of that four years average life. In addition to that, ultimately, you can see the actions that have been taken, and you can see the development for some of the years that you're talking about slowing quarter-over-quarter. And you'll see it in our K, and then you'll see it in our statutory statements.
That information will be made available, and I expect it will come into focus for people. But long story short, as far as how far along are we and what really occurred, what occurred was very simple. The industry, and ourselves included, did not appropriately anticipate social inflation and what that meant for claims costs. It wasn't fully considered when we were writing those policies during that period of time. I think we were earlier than many of our peers in identifying the issue and pushing on rate and talking about it and taking action, which is why I think, to use the analogy, our pig is further through our python than many of our peers. But long story short, I think that we are well on our way to that period being processed, if you will.
Further, I think the more recent years are going to prove to be very attractive. If you look at our paid loss ratio, if you look at our IBNR relative to case, relative to total reserves as a % of net premium, all of those are pointing towards a very encouraging sign. But after what we've been through, we're just not going to declare victory prematurely. And furthermore, there's really no reason to introduce more potential risk to the situation when the business is comfortably generating a high teens, low 20 return.
Another common thing said on Berkley calls is we're slow to recognize good news but quick to recognize bad news. When people look at reserve releases from the recent accident years, what lines of business are being recognized, I assume, more shorter tail lines? How comfortable and prudent is the loss pick on the more longer tail lines?
I think the answer is it is thoughtful and measured, and I think we need to maybe dissect that a little bit. Putting aside property and shorter tail lines, there are some liability lines that, under certain circumstances, you have more visibility. When it comes to workers' compensation, the tail may be there, but to the extent that your frequency trend is proving to be very encouraging, well, you know what? The claim either happened or it didn't. And to that point, it's also on the professional liability front where it's a claims made form as opposed to an occurrence form. If you haven't been noticed on a claim, once that policy period has lapsed, that chapter is closed.
I think that one needs to not just draw the distinction between shorter tail lines versus longer tail lines from the perspective of how long you hold the reserves, but it's also how quickly you get visibility. When it comes to comp, as far as frequency trend, I think that's something that comes into focus more quickly. In addition to that, on the professional liability front, when you're using a claims made form, that can crystallize more quickly as well.
Shifting gears more philosophically to how you run the business.
Yes, sir.
It's a very flat management structure. I think that Berkley creates businesses around very talented individuals to run them like operators themselves and to give capital to those businesses. And maybe there's, I don't know, 60, 65 Berkley businesses today. I'm guessing that maybe that's right around the right number.
60.
60. How does that fit into a cycle management paradigm? When you build a business around someone, but some of their businesses aren't going to be great periods in the cycle for that business, but they're a talented person, how do you allocate capital to that business? Does it get smaller, and can that individual run that business successfully with less capital during a softer market? Let's take D&O, for example, right now. You're not going to get out of the business, but it may not be the most attractive time. From allocating capital into someone who's running a business, how does that work exactly?
So I think it really starts with, no surprise, the people that are running the business, the people that join the organization and their philosophy. You have to start from a place with an understanding, and people that join the organization have to subscribe to this view, that we are not in business to issue insurance policies. We are in business to make good risk-adjusted returns. And that sounds very simple, but it's an important distinction because most people in this industry think they're in business to issue insurance policies, and they lose sight of the fact that, no, that's not really the goal of the exercise. Number two, we are an organization that recognizes that it is a cyclical marketplace. So there are moments in time when the right thing to do is to really lean into it and to try and maximize the opportunity and grow dramatically.
There are moments in time that perhaps the right answer is to recognize where market conditions are and allow the business to shrink. The people that are running the businesses understand what the goal of the exercise is. They understand that their job is not going to be compromised for operating with the right underwriting discipline. If that means their business shrinks and their expense ratio rises, then that's absolutely acceptable. That is in part, certainly, in how we think about the people that join the organization. Do they subscribe to that approach, and are we comfortable them managing capital for shareholders? In addition to that, obviously, it's how people are measured and how both short-term and long-term incentive plans are structured.
In terms of that sort of structure also, I mean, we haven't talked about it lately so much, but talk about incubating new businesses. There always seems to be businesses currently being incubated right now at the firm that might grow into a real business at the company?
Yes. We probably have a handful of businesses at this stage that are notably subscale. And when market conditions present themselves, you will see them scale considerably. And right now, because of their underwriting discipline, they will remain subscale until that market opportunity presents itself. Is it a little bit challenging in the short run? Yeah, because the expense ratio is inflated. But when the day is all done, from our perspective, we think it's a much more controlled way to build the business and create value over time.
Let's talk about specialty markets, excess and surplus markets in general. There's two things. One is, I think there's sort of probably heard it from Berkley first at the earliest, this idea that non-admitted markets will triple in size during a period of time when risk is difficult to place. And then the admitted carriers will say, oh, I could have written that. And their appetite changes. And during the softer market, it gets smaller because business leaves the non-admitted markets for the admitted markets. Marshall, is that an apt characterization of the way the business operates today? And where are we in that sort of cycle of admitted carriers embracing more business that's riskier in the past versus saying, we can't do that?
So the way you characterized it, I generally speaking agree with that. The specialty market in particular, the E&S market, we live off of the crumbs that fall off the standard markets table, if you will. We have gone through a period of time where the standard market, I think, has embraced a greater level of discipline, and they have maintained that discipline. That having been said, the marketplace is different from where it once was, where different product lines, while they remain cyclical, they have decoupled again, as we referenced earlier. So the standard market has great discipline with certain product lines. And perhaps that's most pronounced in the level of submission flow that we're seeing into the E&S space. I think that discipline has remained intact and will likely remain intact for the foreseeable future.
But there are certainly many parts of the market where the discipline is not what it once was. Certainly, workers' compensation, as referenced earlier, would be an example of that, even some of the professional liability lines where some of the standard market appetite has increased dramatically. So I agree with your comment about how things ebb and flow. And generally speaking, it is still a very good moment and is likely to remain a good moment for the specialty marketplace overall.
In thinking about data and artificial intelligence and all these, the ability to get information, it seems like everything could be a specialty risk. Is there changing because we have access to information we didn't have that risks are looking more different than they did in the past? And is that an opportunity for the E&S markets, or does that mean that the admitted markets will be able to also use data and be more comfortable telling the difference between things?
Look, I think the use of data is going to prove to be an ever more valuable tool for the industry. Clearly, AI is going to be a tool that allows you to use data in a more thoughtful, more timely manner than without. That having been said, in my opinion, the opportunity for the specialty and in particular the E&S market tends to be around product lines where there's not a great level of data that is available or homogeneity within that data so you can extrapolate. The world is changing and becoming more complicated every day. Part of the demand that the E&S market is seeing is because, quite frankly, the standard market doesn't have the tools or the appetite to keep up with some of the newfound risks that are coming our way more and more.
Just to be clear, one of my concerns is that technology is cheaper and cheaper as time goes on, that more people will get those capabilities. You're thinking that this is a winners who are ahead, stay ahead type of a situation more?
Look, I think technology has been and will continue to be more and more of potentially a competitive advantage or a competitive disadvantage in this industry. I think the applicability of technology in two broad categories. One is internally and how you process and how you make decisions. The other one is externally, if you will, and how you think about servicing your client, whether that be traditional distribution or insured. Given how technology is moving, I don't think scale is necessarily as much of a competitive advantage today when it comes to insurance technology as it once was in the past. I think there are solutions that are provided by third parties that anyone can have access to. I think the building and development of solutions also is more readily available and can be done more quickly at a pretty local level.
I just want to say, in all these fireside chats we've been doing, no one has asked a question in any of them. So no one has to be shy. I have plenty of questions, but.
Are you looking to break the trend?
I don't know. I mean, Robert's a pretty friendly guy, so I don't know. You don't have to be afraid. All right, so I want to get some numbers right. I'm going to read from these things. I want to be exactly correct here. So over the past four years, Berkley has returned $2.275 billion to common shareholders, including over $1 billion in 2023 alone. In 2021, 2022, there was a special dividend paid, and then there were three special dividends paid in 2023. Also, 2023 had more share repurchases than the prior three years. And the price at which the stock has been repurchased, average multiple in 2023 was 2.3% compared to a higher multiple in the prior years. We don't have to go through all the numbers. Typically, in my mind.
Good.
In my mind, the cheaper the stock gets on a relative basis, the more the appetite would be to buy the stock. The more expensive it would be, the more prone you'd want to be to pay the special dividend. You're doing both.
Yes, that's true.
And so I think they should be in conflict with each other, but they seem to live in harmony. Can we talk about the capital model? What tells you it's the right time to pay a special dividend? What tells you the right time to buy the stock? And how do the two things exist in harmony?
So maybe if I can just take half a step back, Josh, and then we'll do my best at least to address your question. For starters, we start from a place trying to think about what are our capital needs today? How do we see the opportunity over the foreseeable future? And what will our capital be required in order to support those activities? Then we recognize that we don't have a perfect crystal ball, so we don't know what tomorrow will bring. So on top of whatever the first answer is, then we think to ourselves, how much of a risk margin, if you will, for capital need going forward do we want to layer in on top of that?
Then beyond that, when we think about the business, we're going to think about what do we think is putting aside all the CPA, accounting, mumbo jumbo, what do we really believe the true book value of the business is? And then we also think about what do we think the earnings power of the business is over some period of time? So with all of that as a backdrop, then we sort of conclude, OK, so once we do all that, what do we think the business is worth and what's left over or what type of surplus of additional capital? And then it boils down to what do we think the most efficient way is, to the extent we have a surplus that we don't need above and beyond, to get that back to the shareholders?
We look at, to your point, what do we think the stock is trading at? And what do we think it is as a multiple to what our view on book value is? What do we think about the earnings power of the business and how that will be going forward? Some of the periods in time, as you referenced, you talked about what the multiple was for the full year. But there was a moment in time in 2023 where the stock had traded down considerably. And when it did, that instructed us to how we wanted to use excess capital. There are moments in time when the stock had traded up. And as a result of that, to the extent we had excess capital, our conclusion was, you know what? There is a more sensible way to return that excess capital to shareholders.
To the extent that you're looking for a perfect formula, do we have formulas and capital models and all kinds of other stuff? Yes, we absolutely do. But there also is a judgment factor as well. I appreciate why you would say, perhaps, well, which is it? And why are you doing both simultaneously? But two things. One is, when we do the modeling, we think about it that it's not always necessarily one answer or the other. It can be an and. And part of what instructs that is where the stock is trading at any moment in time, not just an average multiple over an extended period of time.
If we go back, oh, please, Mark, go ahead. We got a question. Did break the streak. You see that?
Well done.
Well done.
I have a question about the casualty business, the liability business, and some of the things you talked about. There seems to be this sort of distinction between 2016 and 2019 or 2015 and 2019 and then this magical difference between 2020 and 2023. And so the narrative is, in starting 2020, we took from any company, we took prices up, we underwrote the business, we had more conservative loss picks. They're completely two different books. Can you talk about Berkley specifically? What exactly did you do starting in 2020? How proactive were you in reshaping the book limits, loss picks versus what your initial estimates were from 2016 and 2019?
Yeah, thank you very much for the question. And I very much appreciate the point that myself and other industry participants, oftentimes we speak about there's this bright line that 2019 ended and someone hit a switch, and it became 2020, and the world changed. And the truth is that there are a couple of pieces. One, it was sort of 2017 into 2018 when all of a sudden people started to say, holy cow, the legal environment has changed. Some people, it didn't come into focus quite that early. We started to talk about it during that period of time.
It wasn't as obvious as you might expect, but we started to see it in the claims data where something I'm just picking numbers to try and articulate a point that you would have settled for $500,000 for all of a sudden turned into a claim that you were going to court, and the award was $5 million. It was just this social change that had occurred relatively quickly. When we saw that, we took pause, and we realized that we needed to affect change on the rate front. To your point, that also instructed us about how we thought about limits, how we thought about terms and conditions, and other ways that we were shifting the book. The reality is that takes a little bit of time.
And again, that sort of feeds into this idea of not knowing your cost of goods sold till after you make the sale and that lead time or that mismatch. So what happened was two things. The industry had a, excuse me, but oh shit moment. And people started to take action. And the action started to build a groundswell. In addition to that, as people were taking action on rate and selection, something else happened. It was called COVID-19. And as we all know, it was just shy of the earth stopped spinning on its axis. Everyone went home, and things shut down. So all of a sudden, the exposure that you had been underwriting to changed dramatically. Furthermore, the legal system shut down. COVID proved to be both a challenge and a friend for the industry.
It proved to be a friend because of what happened with frequency during much of that period of time. It proved to be a problem because the way that the industry thinks about loss trend and exposure, it created great confusion. You had all of these problems from the 2016 through 2019 period, and then all of a sudden, claims activity went silent. So long way around the barn, the answer is it was the aha moment where action finally caught up, along with COVID-19, which sort of helped draw that line in the sand that much more deeply. That having been said, we put all kinds of propaganda out on our website. And on one of the pages, you can see, and we publish it, what we were doing with rate increases.
You can see how we, as an organization, started to push, push, push harder, harder on rate. While that doesn't give you visibility as to what some of the other underwriting actions that we were taking and how our book was shifting and how our E&S business was really starting to grow dramatically, it is a good sample leading indicator for how we started to focus on it, again, I think, earlier than many and how that action or momentum built.
More questions?
Probably two, just unrelated. First, I guess, on the social inflation piece, clearly, you've talked about the limits management and the terms and conditions. From what we have seen with the litigation funding industry, the amount of assets that they are gathering and the marketing tools, largely around social media, that they are now starting to leverage, do you think that that can also become a frequency issue down the line? I mean, clearly, limits management helps with severity. But how are you thinking about frequency and claims management around it, especially when that industry can reach the claimant in seconds versus the insurance industry might take days?
Look, the plaintiff bar and related activities is as aggressive as certainly we can remember it in some number of decades, maybe beyond. Some of that fire is clearly fuel, to your point. I share your view by litigation funding. And we can go down that rabbit hole as much as you'd like, as you'd expect. I do have a view on that. But I think what we're talking about is, and I agree with, if I'm understanding you correctly, what you're poking at, it's not just about severity, but it's frequency of severity that's on the rise. And when we're talking about loss cost trend and how that's being driven by this new social environment, as it's been labeled, social inflation, sir, it's exactly what you are highlighting.
We, as an organization, when you look at us thinking about how much rate do we need, how do we think about attachment point, how do we think about terms and conditions, how are we pivoting the organization? The point that you are raising is front and center. The problem is, how quickly do you gain the visibility going back to this mismatch when you make the sale versus when you know the outcome? How many widgets did you manufacture before the new reality has come into focus? That cuts both ways in this industry. Sometimes things pivot and become more benign, and all of a sudden, there's a windfall. Sometimes it proves that things were much more challenging than you anticipated.
It highlights the challenges of the industry and the importance of being as forward-looking as you possibly can be, which is tricky in an industry that is all consumed by the rearview mirror.
Maybe a second unrelated one on medical inflation, another of your favorite topics. You've had a more cautious view versus some in the industry continue to really like the workers' comp line. I think the difference and maybe a question for you there. The difference is expectation on where the medical inflation is. You've had a different view. A lot of others think it's low single digits that they're seeing in the business. Talk about what you saw in 2023 sequentially in medical inflation, because clearly, in more recent earnings calls, there were more companies acknowledging the fact that there's medical severity in the system, but still to the point that it's running well below their expectations.
Well, I don't know what their expectations are. And I don't pretend to have all the answers. I don't even have all the questions. But what I would share with you is this. From our perspective, when you look at a workers' comp claim dollar, over $0.50 on the dollar is medical related. That's a statement of fact. You can call NCCI or whoever you want. They'll give you the information. In addition to that, I think that there is a shared recognition, putting aside the workers' comp space, as to what's going on with health care. And we can certainly spend the balance of the day talking about pharma.
But quite frankly, all you need to do is turn your attention to providers and what has happened with their economic model and the difficulty that they find themselves in because most of their negotiation with payers is at least on a three-year cycle. So we went through a period of time where we saw a lot of inflation, including wage inflation. And those medical providers were not able to reprice with payers. So in my opinion, I think that there is this pinch point that exists that is going to bubble over when it comes to medical costs. And I think workers' compensation, in spite of what has been said by some, well, there are state schedules that they haven't changed yet, that's true.
But ultimately, every payer, other than perhaps the federal government, is going to have to adjust to the new costs and the broken economic model that most providers are struggling through. And that includes workers' compensation. So where do I think it's going to play out? I think that it's going to prove in hindsight that if my colleagues and I are guilty of anything, it's going to be that we were early. I don't think it's likely going to be that we are wrong, at least about this. We're wrong about a lot of stuff, but I don't think it'll be about this one.
We got one more.
Thanks. Can you make some comments on the investment income? Because it seems the yields must have gone up as the rates have gone up. And if I look at your PE multiples before COVID and today, they have gone down a lot. Do you think a lot of that is due to market kind of saying that, OK, we are at a peak investment income yield, and that yield will go down? How do you see it? I mean, and are you changing your investment assets to kind of take care of, OK, maybe we are at the peak? Can you change it? Can you flex it? What are you doing about it?
So I think there were a couple of pieces there. If you don't mind, I want to try and take them one at a time. I think that we went through, as everyone in this room is acutely aware, we went through an extended period of time when interest rates were exceptionally low. And first, it was because of a financial crisis. And just as it looked like rates were going to start to move up a little bit, then this thing called COVID-19 came along. And it was a decade and a half plus of notably low rates relative to history.
As a result of that, I think there were a lot of people that, quite frankly, almost—I don't know if they forgot—but it sort of fell to the back of their mind that a business like the one that I work for, there's really two pieces to the economic model that are noteworthy. One is the underwriting. And the other one, to your point, is the investment income. What has happened pretty quickly is that investment income piece or driver has reawoken. And from our perspective, at some points, will rates perhaps come down a little bit? Yeah. But we don't see them going back to where they were. And we can have a whole conversation around that. As far as I'm going to skip over for a moment how the world thinks about us.
Quite frankly, that's probably a better question for my colleague to my immediate left than me. But just going on to the last piece about how do we think about the portfolio, look, given where interest rates are and when we again, whether it's underwriting or investing, we approach things through a lens that, again, we refer to as risk-adjusted return. There's not a lot of catalyst in our mind for trying to push harder into alternatives. We think that sort of risk, return, liquidity, et cetera, et cetera, fixed income isn't so bad today. So I think what should you expect? You should expect us a couple of things on this front. Number one, I think you should expect us to continue to be very disciplined on the asset quality front, as we have been.
Number two, I think you should expect us to be nudging out the duration of our fixed income portfolio over some period of time. We don't feel a great rush. But directionally, that's where you're going to see it going. Number three, I don't think you should expect us to dramatically shift, though we're not going to abandon alternatives. We think fixed income is a good place to be.
Higher interest income, does it tempt you to take a higher risk on the operations?
No.
No.
So there's a lot of dialogue around this concept of cash flow underwriting. And first off, I think that what led to that was a remarkably different interest rate environment than what we have today. If you go back and if you're having a hard time falling asleep, you can go study insurance industry history. So I don't think that's where we are today at all. Number 2, we tend to certainly look at the business as one, but we also decouple things. And we look at an exposure, a risk, if you will, from an underwriting perspective and say, how much margin do we think we, our shareholders, are entitled to if we're going to accept that risk? And then when we make investments, we will look at, OK, how do we think about the exposure and the risk return there?
It's not that we are so naive to think that they are completely mutually exclusive. But we are also very reluctant to find ourselves in a situation where one is subsidizing another to the extent that you're making short-term foolish decisions. We've seen the movie. We're not looking to be in the movie.
We're almost out of time. One more quick question. A few years back, and just a few minutes ago, you said that our impression of what our book value really is worth is different than what is the GAAP number for what that book value is. And a lot of that was real estate investments that are recorded at cost on the balance sheet but have appreciated dramatically over that period of time. A few years ago, you sold a large share of a landmark building in London. But today, is that still true, that there's a substantial amount of unrealized gains in that portfolio and the book value is?
We believe that the alternative portfolio still has value that the financial statements do not fully recognize. Perhaps the opportunity is not as rosy as it once was in real estate. But we do a variety of other things in the alternative space, including some private equity that we invest in as an organization directly. We have our own team of people. One of the benefits is that, by and large, we never used leverage. And I think that occasionally, if there were some of our colleagues that would scratch their head and say, hey, you're kind of handicapping us because you don't let us use leverage, well, this happens to be one of those moments where the fact that we didn't use leverage has proven not just to be a good thing, but something that's also creating an opportunity.
Well, thank you, Robert, for giving us your time today.
Thank you for the invitation. Thank you for your time.