American International Group, Inc. (AIG) Earnings Call Transcript & Summary

Great. Thanks, everyone, for joining our virtual fireside chat with AIG at the Deutsche Bank Global Financial Conference. I am Phil Stefano, the insurance analyst here at Deutsche Bank. And we're very excited to have with us early this morning the CFO, Mark Lyons; and Deputy CFO and Treasurer, Sabra Purtill. At least for the insurance and focus of the investors, our 2 esteemed guests likely require a little introduction, so I'm just going to provide a brief intro for both. Mr. Lyons returned to AIG in 2018 after spending the majority of the prior 2 decades holding various roles of increasing responsibility at Arch Capital Group; and Ms. Purtill joined AIG in 2019, previously having been with The Hartford Financial Services Group, where she was Treasurer and Head of IR. So again, thank you both for joining us today. When we start with the Q&A, I'll begin with some questions of my own, but we're also going to leave time for questions from the participants. [Operator Instructions] But first, Mark, I believe that you have a few minutes of prepared remarks, so I will turn it over to you to start us.

Great. Thank you, Phil, and I appreciate the invite very much. It's always good to get out in front of analysts and investors, and this is a great FDE-compliant opportunity. So again, thank you for that. So I guess a few summary comments. I'd say first, and we kind of mentioned it on our first quarter call, but AIG entered this COVID whole situation and environment in a very strong financial position, and that continues to be the case. We enhanced our current liquidity as you probably saw in the trade press in May by raising $4.1 billion in the debt capital markets, and we now have north of $11 billion of parent liquidity at this point in time. Our subs remain in a very strong position, both the life and retirement, and the general insurance fleet RBC ratios were north of 400% at the end of 2019, and they actually improved between then and 1Q 2020. I think you guys know that the general insurance story has been one of massive reunderwriting. Peter Zaffino itemizes a lot of those things on the quarterly calls, or at least he did in '19, and to give people really a flavor for the kinds of effort that had to be expended. And that work continues, of course, but it's more the last -- it will continue, I think, through 2020. But now you've had basically a full renewal of everything and especially in a harder market environment, so I think it bodes very well. And I think the biggest problem child that general insurance has had has been North American commercial. I think that's where, figuratively speaking, the most surgery and attention has been made. There wasn't attention elsewhere, but I'd say proportionately more there. The personalized side of the house is really a pretty strong franchise, especially in Japan. And with the high net worth business, it's one of the industry leaders in that marketplace. We recently launched Syndicate 2019, as we commented on, on the call, in partnership with Lloyd's to really serve that high net worth market and which is a -- continues to be a fast-growing and profitable segment. Life and Retirement is a well balanced portfolio. I'm sure we'll get into that a bit. It's got a strong breadth of products and distribution channels to really let clients be served, I think, in the right way. And as markets change, as pricing conditions change and so forth, there's an array of options for them. Fortitude Re that we've commented on probably since November of 2018 is still on target to close midyear and is going through the regulatory approval process as we speak. Our investment portfolio, I think we commented on this, probably more fully on the last call, has been really derisked over the last 3 to 4 years, really since Doug Dachille got here in late 2015. And I always like to make the analogy that he was faced with similar challenges that general insurance was faced with, with a portfolio that needed major rebalancing, major constructive and construction work associated with it, and I think that's been evident over time as well. And we broke out a lot of information in our financial supplement for investors and analysts. And hopefully, that's been of a value to everyone on this call. Now I think just kind of closing where I started is we continue to have a lot of confidence in our balance sheet and the business portfolio. As it stands, we'll continue to tinker with it, of course, but we feel pretty strong about it, even with all the uncertainty going on at this time. We feel we're in a pretty good place to navigate things, and this current environment is just that, the current environment. And in 3 weeks, it could mean something different. And there could be different governmental policies coming in and programs and so forth that it's really hard to determine to what extent it buffets up the economy or what it does for overall liquidity. But we still feel that this is an earnings event for AIG, not a capital event. And with those intro comments, Phil, I am happy to kick it back to you and begin the chat.

Great. Great. And so one of the questions that we get, especially now that the first quarter earnings is behind us, is trying to parse between the different layers of COVID reserving that people have put out there. So the question that we're going to begin with asking everyone is, to what extent do your COVID first quarter charges represent an ultimate? What's encompassed in them? And if you can just talk about the extent to which there may be legal expenses embedded in there. It feels like these are the 3 kind of primary differences on how people have reported so far.

Okay, okay. It's a fair question. So I guess there's a lot of subparts there. So let me go through and hit some of them. I think, Peter, kind of marched through it, but what we posted, which was $272 million on a net basis, was really composed of travel exposures, accident and health contingency business, did include property, did include credit and workers' comp. We kind of went through, I think, but happy to do it again, some of the rationale for some of that. But to the extent that we could measure it, Phil, it would have contemplated legal expenses and buried in the reserve. The reserves are dominated, as you would expect, to be IBNR reserves, not case reserves. Now as things emerge, could legal expenses be a little different than you originally think? I mean when you look at the papers, and I'm sure everybody is reading the same things, you can have different views coming from different states. You have certain areas trying to get multi-district litigation, MDL litigation, and so forth. So depending on where that goes, it's a different legal spend associated with it. But our reserves contemplated our best guess at the time. So when you -- I think you said, does this represent an ultimate? So the answer is yes. I guess I have to differentiate how an insurer might look at that and what we're responsible for doing and how the question might be phrased. So I guess, first, what's the requirement of any insurer to do at any point in time. So at any balance sheet date, you're required to put up your management's best estimate of loss reserves associated with events that have occurred on or before the statement date. So from that perspective, these are management's best estimates of the exposure from COVID from claims that occurred 3/31 or prior. Therefore, it's not an estimate. If you think of it like on an underwriting year basis or underwriting years basis depending how long this lasts, that will be future loss occurrences. So if there's an event cancellation in September, that's clearly not reflected because it was neither postponed nor canceled as the time of the statement date. So hopefully, that clarified it.

Yes. No, it does. And one of the things that I've been trying to think about just given the uniqueness of this event and you have a reserving process that you put in place, to what extent does the reserving process fluid do you need to start from ground up every quarter? Or how should we think about what the reserving evolution for COVID could look like as we move forward?

Yes, that's a good question. I think what you'll see because I think we talked about kind of a granular bottom-up approach, which involved a lot of different people and functions, and then a top-down approach, which is more exposure. So the bottom-up teaches you about where the exposures are, and the top-down is more of kind of an exposure rating approach, if you will, or an exposure view. So those 2 marry themselves. As time goes on, there's going to be more weight given to the bottom-up than the top-down. I would make the analogy to a reinsurer, right? A lot -- on a normal catastrophe, usually the estimates come out as a function of market share, very top-down. And then as the quarters pass, that kind of gets kicked to the curve, and it's the actual emergence that dominates, and it will be the same thing here in concept. So -- and part of your question, Phil, was, do we have to start from scratch? I'd say it's similar to any other reserve group. So you leverage what you had last time and you look at how things have changed. And you -- depending upon what you have, if it's a frequency line or severity line, you can apply different methods of projection. That's all a function of what comes through.

Understood. Understood. And we had a question coming on the line about COVID. Given that this is -- at least it was described in the opening remarks an earnings event scenario. When you look at stressing the various assumptions that you're making, can you help us understand maybe what the macro circumstances are? What leads this to be a capital event? How does the evolution of losses change to pivot from being an earnings event we should be comfortable with to a capital event that's maybe a bit more of a headwind?

Well, as I said, our view is that it is, with our view of how we've looked at it, it's still an earnings event even with -- I mean, I got to back up a bit because, as you know, we and others have withdrew guidance because there's only limited visibility. I mean I don't know about your crystal ball, but mine is very cloudy on 2020, let alone 2021. It's basically opaque on 2021. So you've got to make assumptions on, for example, where the 10-year is and where that might be over time, what's going on with spreads, what's going on with equity markets and some other aspects you might have. What's going on with other charges that may come through. So what we did on -- I mean it's -- that's the thing with scenario testing versus probabilistic approaches, you really have to pick a set of dynamics and kind of let that run through. So it would have to be far worse than we've assumed in our kind of revised view forward, limited visibility view forward for this to become a capital event for AIG.

Got it. And you had mentioned that the guidance was withdrawn, but at least for the general insurance, the combined ratio improvement expectation was reiterated. To what extent should we contemplate the potential accident year or attritional losses coming through from COVID and serving as a headwind to that improvement target?

I view that as marginal, really, at this point. I think there's the possibility of some upward drift in -- we're talking actually year ex CAT, Phil, I assume.

Right. Yes.

For actually 2020 ex CAT, so yes, there could be some movement. But in AIG's case, because of the kinds of lines affected, it could be driven by our changing mix. I mean when you have your travel business, I mean you know what's going on with travel, it just fell off a cliff. So the travel volume is off tremendously. That affects loss ratios, it affects acquisition ratios and things of that nature. A&H and other lines are off and they'll rebuild as a function of how the economy rebuilds and what governmental policy lets travel really start to occur again. So I give the travel example because it's, I think, a good extreme example, because many people don't have travel books, of how a decrease, a radical decrease in the volume of the book can actually weight and change your loss ratio overall. Even though every line of businesses loss ratio may not have changed appreciably, the weighting may cause some upward pressure.

Got it. And when we think about the P&C commercial lines, I think in the past you had talked about using a machete to help sculpt the business and maybe you're down to a scalpel at this point. I mean can you remind us where in the surgery you are with the changes that you're making?

Well, I think if I talk about one that's still a little bigger than a scalpel would have been the Syndicate 2019 on the high net worth business. So that's a -- it's really a strong performer. The one exposure you have about people in the high net worth segment is they all tend to want to live on the coast, they all want to live near each other. So you end up having risk aggregation exposures that you really can't get away from. But the book of business that AIG has, it's not just auto and homeowners, it's watercraft, it's collections, it's excess liability, there's fine arts. There's a full array, collectible cars, things of that nature. So there's a full array of exposures that really do help you overall. But that is, I think, a good recent example of additional portfolio management that welcomes and allows Lloyd's with its innovative structures of bringing in third-party capital and arranging it in different ways, and it's being complemented with a lot of reinsurance that we have from our third-party -- normal third-party reinsurance indemnity providers. So it's a really good spread, and it allows the profits to be spread as well, but it should, over time, change AIG's profile to that being a smaller proportion of total and also generate fee income because that will be an MGA structure written through Talbot. So it'll still be written. The PCG brand will continue, for example, but it'll be written through AIG through Talbot. So we think it's an elegant spread solution that benefits everyone.

Got it. Okay. Switching gears a little bit. We got a couple of questions that came in through the line about the ceded reinsurance program. I was hoping if you could just discuss the program generally. I know there have been several enhancements over the years, which you've detailed on your earnings calls. As we contemplate the structure of the reinsurance program, how should we think about the potential response? At least it feels like the question is geared towards COVID claims on the property versus the casualty side.

Well, the -- we tend to talk about the property more, I'm happy to do that again. But I think the enhancements -- first off all, it's complicated, I could be here 3 hours, quite frankly. But if I think of the per risk and the CAT, there have been, I think, substantially beneficial tweaks that make the covers more relevant. So on the per risk side, the attachment points have dropped and there were some AADs, annual aggregate deductibles, in some of those, and those have been severely lessened. And because we cut our limits that Peter Zaffino talks about, right, so dramatically, we didn't need some of that extra coverage at the top because our limits nowhere, that didn't have that level of size anymore. And we have made enormous progress towards eliminating those long-term agreements, or LTAs, which were the 3-year deals that carried a lot of that. So you've got that benefit there. And in the CAT program, we not only have lower attachment points, it's better on the aggregate basis. So it will attach sooner irrespective of whether it's a large vertical or a collection of aggregate losses over the course of the year. And both the per risk and the CAT would have coverage for communicable diseases, for example, although on a go-forward basis, we expect and are already seeing the reinsurance market tighten that. And there's generally going to be -- I think it's a general rule, there'll be communicable disease inclusions going forward. But all of our reinsurance contracts have strong follow-the-fortune conditions and/or follow the settlement conditions if their settlements evolve. So that would be my summary, Phil. On the property side -- casualty side, sorry. On the casualty side, there is, again, the gross underwriting changes have to drive what you do. And I think Peter talked about last year, on the casualty side, it was a significant change put in on the growth side by the chief underwriting office, Tom Bolt and his team. So in the past, AIG would compete against itself basically and have different pockets around the world and the brokers knew that, and they knew how to exploit the chinks in the armor. That's no longer happening. So it's more centrally controlled. There are strict ventilation standards between layers where we would operate. Whereas in the past, we may have written $200 million with no ventilation in different pockets of the organization and had a huge net on it. Now there is no more than $100 million that's out and $75 million on new business. There's a $75 million ex $25 million that's 100% ceded, and there's a quota share in the first $25 million that's roughly 55% ceded. So let's call that an $11 million net compared, if we've written the whole thing up to $100 million. So a massively different tail management, number one; and number two, it changes your net mix of business, which is also favorable to us. So that's how I'd summarize that, Phil.

No, that's great, Mark. The pricing momentum that we've seen, at least on the primary side, is something that we've been discussing for a while, and it feels like the hardness or the firming in reinsurance is a bit of a new phenomenon in the industry. Does it feel like the pricing momentum of reinsurance versus primary, are they materially divergent at this point?

Well, some of the things that we see -- because remember, we have Validus now, right, so we're writing reinsurance assumed, not just dealing with markets on a ceded basis. And it's probably worth noting that the amount that Validus rewrites is -- more than offset the sessions that we do. So on a net-net basis in a harder market like this, we still come out winners from a price gain point of view. So with Japan and Florida, there's -- whether it's loss-affected or not, there was clear indications of real strength of reinsurance pricing going up. So AIG's program, you may recall from Peter, we still had a 7% spend reduction compared to the prior year, but the rates have gone up. So the reinsurance market that seemed to lag the primary market and the retro market, they were in the middle, that seems to have been corrected.

Got it. When we think about -- I guess, looking back to the first quarter earnings call, there was a comment that was made, and we had a couple of questions come in, about the exposure to property and the affirmation that -- of coverage for communicable disease was less than 1% of the total limits on property. I think people were trying to just get a better understanding for the context in which that was provided and the extent to which the limits could be a headwind as we think about forward earnings potential. And maybe you could just clarify or talk around the -- if you have any more clarity in that comment and how we should be thinking about it.

Well, I think the intent of that was to kind of give some relativities. A combination of how often it's offered, I guess, is a good way of putting it and the sub-limits which are tiny compared. So it was less than 1%. I think that's what you're referring to, Phil. Less than 1% of the total gross limits in an aggregate sense would be potentially exposed to business interruption if everything went south, including all the things you asked about, coverage, language and legislative efforts and things of that nature. So I think, Peter, in the past, has talked about that depending whether it's North America or international, it amounts to like of $1 million to $1.5 million policy limit gross maybe in North America and less than $1 million internationally. So if you think of it in context from over the year, there was 2 -- these are astonishing numbers, $200 billion in limit reductions, gross limits, $200 billion. So if you take the 1%, that's a $2 billion reduction, assuming they all proportionately had business interruption. That would be a $2 billion reduction in business interruption limits. So the constant resculpting of the portfolio, and I think as Sabra has commented on in the past, that if AIG had been struck, if COVID had happened 3 years ago, it would've been a whole different ball game. But all those efforts that went forth to derisk and restructure the book allowed us to be in this position. So I would summarize that it's less than 1% of the total limits outstanding. The limits have dropped by $200 billion over the course of the year, and that's what puts us in such a great position now.

Understood. Understood. And thinking about Fortitude Re, look, post the sale of Fortitude, are there any parts of AIG that still seem noncore? Or alternatively, is there any area that you see a need to grow inorganically following the sale of Fortitude?

Well, Fortitude dominates but isn't exclusively what legacy is. There's still some pieces that are probably not worth talking about, but Fortitude still dominates that. So yes, that's considered noncore, and that helps the Life and Retirement be comfortable in saying they don't have legacy exposures, right? They don't really have the 2000 vintage variable annuity issues. They don't have LTC issues and things like that. To the extent any of that was there, that's over in the Fortitude book and not a, call it, going concern upon sale. So I would really say that is still our view of what's noncore. If I look at L&R between their group, the individual retirement, the life piece and institutional markets, which is mostly structures and GICs and pension risk transfer deals and things of that nature, I'd say they're all core. Now the Life piece, when we talk about Life and Retirement, it's dominated by R not L. So the Life piece really, compared to others, it's probably not proportionately as big as some of our competitors. But I'd say all of it is still core. And then when I look at GI, it's more -- as Peter would do, we would sit down and look at the portfolio as to what's performing, what's not, what should we grow, what should we coast on, what should we cut back or what should we drop kick. That's an ongoing -- but that's pruning, right? That's not same core, noncore. So I think with the exception of legacy, we don't have things that I would really classify as noncore.

Got it. And when you think about the broader AIG portfolio, does it feel like there are things that are missing that will be easier to solve inorganically than organically?

Well, if we didn't have Validus yet, right, I'd say we need a reinsurer, we need a Lloyd's presence and we need business that we don't have like Crop. And we got all that associated with it. So I don't think we're in a hurry to flood into the Far East, for example, that's super competitive at this point. So there might be some areas. I mean we used to have a Latin American presence, for example, and that got cut back. So I think it's a continual look at the landscape, Phil. So some economics within a geographic region improve, then we'll look in that regard. I mean back to my old company, one of the wavy line tenets was to have more of rule of law where you could depend on it. So whether it's London or England or the U.S. or Australia and some of the others where you had some level of knowledge of how a legal system would work, which is critically important in insurance, so it all goes into the mix.

Okay. And so switching gears a bit. We have conversations with investors about AIG 200. It feels like there is a common pitfall in contemplating AIG 200 as merely an expense improvement initiative. And I was hoping you could just remind us, what are the other goals of this program? How is it helping to position AIG for growth in underwriting in the future?

Okay. Well, I guess a couple of things. I think Peter talked about it and I talked about it a little bit on the call, and I just want to remind the audience here that we -- in our earnings call deck, we purposely put a slide out there for everyone to kind of -- I'll get to the heart of your question in a second, Phil, to see what the costs to achieve are, what we view the year-by-year capitalization to be. And since you can't start amortizing that until it gets put into service, we provided some information on that as well. I think to the tune that of the $400 million being capitalized at the end of the 3-year period, there's still $350 million yet to be amortized, and we gave some views that, that would be $50 million per year for, I don't know, 4, 5 years and then tail off from there. So that's the financial aspect that you asked, but you really asked what other type of views. So I'll tell you where I get excited about it is we wind up being in a position where we're going to have a lot more insight than we have today, the ability to have information commonly defined, no matter where you are in the world, be able to have the same view of the same information, whether underwriting is looking at it, finance is looking at it, actuarial is looking at, ceded re is looking at it, which you can guess is more important given our increased use of reinsurance. So having much more information available in a drag-and-drop sense, so the data strategy is key and a lot of senior people are involved in all of these. So this isn't a measure. It's not an IT project. They're business-led projects with clear tollgates, clear milestones that have to be met. And the data strategy piece is very key to it. So there's a lot of colonels involved on these things, it's not kick the corporals to do. When you kick at the corporals, you get a process-driven result. When you have senior people involved, you get a results orientation on what you're trying to achieve. So I'm really excited about our ability to pivot more quickly in the marketplace. The anticipatory have better information to make the continual pruning that goes on our growth opportunities and looking at things not just from an insurer point of view, but from a customer-centric point of view, from a producer-centric point of view, from a reinsurer point of view. So to me, that's the kind of value and leverage we get that's not obvious from the outside and doesn't show up in dollars and cents.

Got it. We have a couple of questions coming in and probably time for one more, so I'll try to frame as best I can. Social inflation is something that we were beginning to discuss more and more over the past year, and this is pre-COVID. Social inflation was a topic on pressuring attritional loss ratios. As we think about the post-COVID inflection in the broader macroeconomic landscape, is social inflation going to develop or be disrupted in some way by COVID? How can we think about -- how are we going to be talking about this concept of social inflation in the coming quarters, if not years?

Yes. Good question. Hopefully, I have a good answer on that. So I guess you can cut it a couple of ways. First, I'll reaffirm that, yes, before COVID, there was, I think, increasing sense and feeling and evidence of that across the industry. I think there have been many comments about it. The interesting thing is that when you begin looking at whether it's work comp or GL, you are seeing a frequency decrease on the non-COVID claims. Everyone is focusing on COVID understandably. But the non-COVID is really reducing the other way. Will that completely offset? Time will tell on that. But you see that as a potential dampener on social inflation to the extent that there's fewer cases to get to a jury award in some of the counties we could -- around the country we could comment on. So there's a possibility of that. But since there's more jury awards, if -- I mean, conceivably, I'm speculating that jurors read the papers like everyone else, so to the extent that there's like ill feeling because of COVID or something, maybe that could jack things up. It's really hard to know. But we, and how we're looking at things and needed rate changes and things like that, are continuing to reflect that social inflation will not peter out.

Got it. I think we only have 1 minute left, and I'm not sure we have time to get into another question. So we have a few in the queue. We'll try to get to them in our breakouts. But Mark and Sabra, thank you so much for your time again. Hope all is well and continues to be well for you and yours, and we look forward to talking to you again shortly.

Great. Thank you, Phil, again. I appreciate the invite, and I look forward, like you said, to talk to you again shortly. Thank you.

Thanks, Phil.