Deutsche Bank Global Financial Services Investor Conference

Jun 1, 2016

Thank you for coming to the Arch Capital presentation. Everyone can come on in. We are really pleased. We have Mark Lyons here who is the CFO of Arch Capital. He is one that has been a lifer from Berkshire Hathaway, AIG, virtual encyclopedic knowledge of the industry. I think that is where I want to sort of start. I managed to throw away my questions. Luckily, I found them on my BlackBerry. Yeah. All works out pretty well. Looking at the market right now, if we think about 2016 to 2018, what is the ROE of the industry? People talk about in the dry spell, I guess. Great question. I would say the industry, that comes down somewhat to mix, but if you have an average cat mix, it is probably 10 at best. My guess would be. In terms of out selecting or your approach to it, some companies try to deal with the cycle and bad pockets by having massive diversification around the world. ACE, I think, is a great example. Others, you might try to do some out selection. I think you got to be careful to not fool yourself to the degree by which you can out select. From the beginning, we always said we wanted to be the best, not the biggest. The bigger you are, the more you converge to any index or any industry average. We kind of like to continue to be nimble. We think between our pockets with growing mortgage insurance as a greater percentage of total with, I will call it an even keel on insurance and more difficult views of reinsurance. That is what we do. We manage mix, and we manage the cycles pretty effectively. I think part of the secret sauce, and you can push back, I think you won't, is incentives. People at Arch from the beginning, you design an incentive model early on that for the cash rate, there's really a 10-year payout if they do a good job and whatnot. Obviously, the 2003 to 2007 vintages in the past for those years were wonderful for Arch underwriters. Can you retain the talent in the dry spells to the same extent that you can in the good times? How does an incentive system sort of maintain a core cadre of underwriters as things are less optimistic? Well, one thing is the messaging internally is very consistent. You cut through all the mechanics of it's effectively a quota share between the company and the eligible employees on the after-tax profits that emerge, firstly. Secondly, since it's an underwriting year basis, which you kind of alluded to, there's an initial payout, then you have to earn the rest. It gets rid of all the arguments of IBNR out of it. Actuaries overdoing it or underdoing it. Eventually, over 10 years, you're totally or predominantly paid and everything's self-evident. You get incremental payments over the period of time as justified. It always balances into the balance sheet reserve, so we're not construction guys. We don't have three sets of books. There's one set of books. The core of your question is in this difficult marketplace, and it is, you can still make it all work because for the 2016 year, you're still getting payouts back from 2006, and in some cases, we kept the years open. It might be 2005, 2004, still producing payouts in calendar 2016. The employees who wrote that business favorably are "locked in" to get their maximum payout. They're still getting paid for the old business. If they're thinking rationally, yes. If they're thinking. Makes sense. I think I'm always sort of nickel and diming you a little bit on share repurchases, trying to figure out how much you paid. Okay. Generally, the firm has said, can we get paid? Are we buying back our stock at a discount to book value 3 years compounded out? Rule of thumb. Rule of thumb. I think I would argue might have gone over that in the most recent publicly disclosed pace of share purchase. Maybe it's a rounding error you've decided that you went for. First of all, to what extent are you sort of confined by that system? Two, if you're pushing the limits on what is an acceptable share repurchase for you, what does it say about every company? Are you confident the single best thing you can do with capital right now versus growing mortgage insurance, making an acquisition? I know you don't love dividends. No. How confident are we that a share repurchase maybe three and a half years of compounded book out might be a better use of capital than something else? A lot of questions here, Josh. I'm making it tough. You're earning it. Your first point I disagree with. We bought back at about 135% of average book value for the quarter. Okay. We've been talking about 10%-12%, right, depending on our mix on a forward look, and that would be 10.5%. Okay. It's steeped over in there, into that territory. It doesn't mean it can't be increasingly tough. It does mean we have a rich man's problem. What we view, because we're in the mortgage space in a lot of different ways. We're in the U.S. primary MI, as you know. We're in it on a reinsurance context. We're not just in U.S., we're international, and we're innovators and highly participative in the GSE structured credit transactions that are dominated, but not exclusively 60-80 LTV. It's increasingly capital consumptive. Some of the capital associated with those is still partially unknown from some of the rating agencies, especially on the GSE transaction side. We keep some of that in the bank. Our major ROE growth engine is consumptive, and to some extent, there's volatility around how much that needs, so we're maintaining that. On the second part of capital management would be a dividend or a special dividend. We've never had a dividend on the side. M&A or redeployment, I guess. There's a lot of others. As you know, the second there's going to be signals of a P&C hardening, we press our foot on the accelerator probably harder and faster than most, and we also back off coast and apply brakes differently. Part of being successful isn't just maximizing good times, it's minimizing bad times and playing defense and giving cultural awards inside for Gold Glove Award, not just the high batting averages, which is what we do. On the special dividend side, if I could for a minute, the difficulty with that is we have a lot of long-term investors, meaning decade or more, with very low cost bases. The last thing a lot of them want is to have a not forecastable taxable event jammed down their throat. When it's share repurchases, they choose when to the extent of participation. Special dividends, a little different story. That doesn't drive our decisions, but it's a component. Let's shift to Watford a little bit. As some of you know, Watford is a business that Arch founded in 2012 or 2013. It was really 2014 that it really started to move. 2014, the alternative assets allow the company a little bit more leeway in terms of the underwriting margin. Has Watford performed in line, outperformed, underperformed expectations? Obviously, when you founded, I don't think you knew what the 2016 market would look like. That might be market-driven to some extent. Is this a great business given the climate today? Let me give you a two views on that, and there will be a definitive answer. Okay. I'll take them. We invest $100 million into the common, there's an investor perspective. We're also a service provider by providing the underwriting and sourcing services. Those are different views. On the service provider side, I think it's done exceedingly well. It's basically a run rate of half a billion dollars of premium, and the sourcing's gotten broader. The market acceptance has gotten deeper. Compared to a year and a half to two years ago, there's more sources and different types of business coming in. It's a softening market, so it's tougher, but you're getting a broader net. There are more submissions that come in that Watford will want to write. It's increasingly difficult for them as it would be on a normal PC insurer or reinsurer. From an investor point of view, I'd say the jury is still out. Not I'd say on Watford, but on the model, the aggregate model, because when, I'll call it reinsurer two or hedge reinsurer two, started to come out is when the market started to really go south, it's kind of timing. From every index I have seen of the performance of the sectors on the asset side that Highbridge is doing, they've outperformed the indexes. That's a relative comment. That's like saying you lost money at a slower rate than others on the investment side. Overall, I think we're satisfied. We're happy with the underwriting side of it, and the investment side has yet to be proven. I was talking to someone that you and I both know who is a underwriter for some might call competitor of Watford. He said, "Look." We were talking a recent deal that was done by two very large, a primary and a reinsurer. He said, "Look, I don't really care," he said, "what they say." The going rate for a lot of reinsurance right now all in is about 105% combined ratio excluding catastrophes. One of the problems is we have an alternative strategy. We were willing to write at 105 before. Now we're finding at this point in time, there are a lot of traditional reinsurance models that are also willing to write at 105, we're finding it harder and harder to find good business because we were finding more competition from people in the past who weren't competing with us for that same level of margin. You've broadened out the Watford distribution. Is it harder or easier for Watford to find that sweet spot margin today than it was a couple of years ago? I would say it's not easier. I think people are earning their salary a lot more. There's more choice now just because of broadening of because of where we started, being so recent. The jumpstart, though, compared to some of the other analogous firms like you talked about, is they had to hire people and reestablish distribution from scratch. Whereas with us, they piggybacked on 14 years of relationships and risk appetites and things like that, and they could hit the ground running. That's one of the reasons they got the $500 million so quickly. Let's be serious. In this environment, if you're targeting 105, it'll be 108 to begin with. Returns will not satisfy you. Any company who is thinking as their business model that they're going to write 105 is not going to have good results. I have more questions about the mortgage insurance business, about investments. I want to make sure that people have a chance to ask questions. Can I, in the back, Tony, there, right back there. Thank you. Just a quick one. Can you hear me? Okay. Yeah. Just a quick one in terms of June 1 renewals, which I assume closed today. Just if you could give us a quick update in terms of what you saw for pricing and just terms and conditions. In general, or are you talking cat in particular? For property cat. For property cat. Property cat averaged about, I think, or averaging, it's still in process, about five down, I believe. A softer landing than it's been on some of the prior. I think we have to be close to hitting bottom. I don't think we're going to have a negative return interest rate environment for us. I was a little bit pleased by that actually, that it was only to that extent of a drop. More challenging elsewhere. For example, the property business. Here's one thing to differentiate, that we take a minute. Everybody talks about property cat. Property cat, I would use the word loosely, is a derivative product on the insurance companies that they're insuring. Early on, when property cat rates were plummeting, there was strength still or at least stability in the underlying business. Now the property cat seems to be stabilizing more, there's deterioration in the underlying business. Whether you're sitting on top as a per risk or a catastrophe XOL, or you're a quota share participant taking severity and attritional losses, your economics are changing even if you're flat on renewal and reinsurance terms. Those things have to be taken into account, the underlying primary and how you're risk managing it, and then secondly, the reinsurance on top of it. On other lines of business, it's tougher. Shorter tail businesses are under, I think, more competition. You still are seeing minor fall-offs in some of the third-party lines. I think D&O is an exception. I think D&O continues to fall off a little faster than it should, especially in the high capacity lines. Virtually any high capacity line is under pressure. By that I mean you're putting up $25 million limits, gross limits, something like that. The capacity has to be built through reinsurance purchases. It's becoming overly commoditized. That high capacity business is tougher risks on average. It's exceedingly well built and they generally have risk managers. It's very hard to make money in a difficult market like that in those lines, which is why we have cut it back dramatically and to the level that we still retain it, we've reinsured it more. I got a little off track there, so. The way you account for your mortgage insurance business, there's a lot of upfront costs and then the business earns out over five or six or seven or eight years. Just wondering what the pattern looks like in terms of, is it going to take a while for the earnings to start kicking in, and when are they going to start kicking in sooner than later, and how's the slope going to be? Yeah, good question. It's earnings certainly, but I would even say it's steeper on writing recognition. On written premium, not just, when you write a bunch of singles, you get writing recognition and then you get earnings over time. When you're on the monthly business, which dominates the industry, written and earned come in simultaneously. You don't front-end load it and have massive persistency assumptions and front-end load it. It's as it comes in. Where we are, at least in the USMI piece of the overall mortgage business, that's still in incline mode. That's still growing, I think, fairly steeply. I wouldn't expect steady state for a few years, if that's the kind of question you're asking. How often does earnings recognition really account for? Earnings recognition, first you got to recognize, again, the USMI accounting model, which I view as broken. It overestimates profits early and underestimates them later. You're only allowed to put reserves up on loans that actually have a notice of default. You cannot project an IBNR, if you will, for PC parlance, for currently performing loans that will become unperforming loans, which is insane from a property casualty perspective. We're cognizant of that. Now, on the reinsurance business, that restriction is a little more lax in that regard. Make a long story short, the keeping combined ratios, I've got to be careful what I say because every time we talk combined, I always talk duration at the same time or combined ratios mean nothing. Assuming the combined ratios don't change too materially, you're going to see a pretty big increase in the dollar underwriting profits because it's just proportional to the recognition of the earning curve as it comes in. Thanks. Just following on the mortgage business with the soft rate environment here in the U.S., what's the appetite to keep building that offshore mortgage business? We still believe its returns are north of our PC operations at this point. There's different cycles. Mortgage business is cat business, at the end of the day. The underwriting years are all autocorrelated. You go through a bunch of attritionals and you get an explosion. Your market cycle management is key on that, and your risk management and your acceptance of tolerance, measured it and managing it is pretty critical. We still like the returns. The returns differ on primary versus international versus reinsurance deals. You can set the terms differently. There's still fewer participants in the reinsurance market, your terms are a little better. The GSE transactions, we still feel, although thinning, are still profitable. We still have appetite. Our marketplace approach is likely to be driven by our internal risk tolerance more than it is market forces. That potential describing MI as a CAT business. Given underwriting standards at the mortgage origination level today, is it possible that you could have a huge spike in losses suddenly? Right now, I don't want to say something and spoil forever, but it seems like you're printing money a little bit at this point in time, given how strong and hard it is to get a mortgage right now. Here's how you have to look at it. A good analogy would be workers' comp. Workers' comp, there's a lot of medical components, like 60% medical now. When they make big indemnity changes, whether it's a big law amendment change, it cuts across all underwriting years. It's imposed on a calendar year basis, and prior and current underwriting years are affected, varying degrees depending upon the age. The older it is, the less impacted it is, and so forth. Analogously in mortgage, it is, we believe, highly profitable now. I think the risks of a meltdown in the foreseeable future are very low. If it's, let's say, six or seven years from today, that's going to cut across the board. Just like surety business. Let's put it that way, too. In surety business, when a contractor goes down, it affects the entire credit exposure you have with that contractor. It's not just bonds that you recently wrote, it's bonds that are still outstanding and written over many years, and therefore, it cuts across the underwriting years. That's what I really mean by a CAT business, because it's not isolated to an underwriting year. It cuts across underwriting years. When we talk about, and maybe it'd be worthwhile, I feel like early on, this was 15%-20% ROE business was sort of how I think it was described to me. Does that include the meltdown year, or is that a modal-type return? That was more of a return that we felt was for business being put on the books then, which was probably our 2015. Given one of your competitors allegedly has put pricing pressure on that business over the past year, what is new business being written today? Is there sort of a loose range of what you think the level of ROE is for new business going forward, say? Given the increasing capital commitment, I think it's south of that. I think it also depends on what your mix of monthlies and singles are, and whether it's borrower or lender-paid, and how it's done in bulk versus not. I think it's more challenging. I believe, and I hope that not incorrectly, that the introduction of our risk-based pricing, I mean, UGC has it and we have it, where think of it as, for any stat guys out there, a big discriminant function that takes many variables into account to isolate characteristics unique to identifying risk and matching risk with price. The marketing side, having it exceed price and possibly pricing yourself out of certain areas because you have a much better zeroed-in view of relative risk versus market price and what you can achieve. We really think that for those companies, so it's more of a first-mover advantage. I think others will catch up to it, ultimately. In the short term, it allows you to isolate the better risk reward trade-offs. I think that could help in a health selection point of view, but I think it's transitory. I think with the RateStar, I think it was about 5 quarters ago, I think that it sort of became the top du jour that everyone was really wondering. Your competitors, if they so chose, I assume, could've caught up or could have started. Most of them haven't. Most of them are still saying the customer doesn't want this product. To what extent do you find that that objection is true? The receptivity, we're here 1 year later. Are people saying, "Okay, we'll do this. Well, it's really not the year. I mean, we've been talking about it for 1 year, but the traction is probably closer to half a year of real introduction. I think it was December that we really got some introduction. We've been very happy with the receptivity. I think even Dinos mentioned on the last call that it was roughly 50%-55% of new, I'll call it submission flow or applicant flow, coming through RateStar. In the last month of the quarter, it was closer to two-thirds. That trend has continued. I don't see broad disapproval of that. There are pockets of a bank here or there that might get not complete acceptance yet, but I think eventually that will fall, too. I'll just give a chance for there being questions. Good. Okay. Can we talk a little about a 0% interest rate environment and interest-bearing investments versus non-interest-bearing investments, and what a 3-year plan, how the investment portfolio might need to change or if it changes at all? Well, if I feel like you, I don't even open my retail bank statements at home anymore. Because the fee I pay is more than the interest I've gained. As you know, we're a total return shop. It's not just because of the lower interest rate environment. We've been talking about that for quite a while. It's increasingly hard to know where to place your money. We have allotments to equities and alternatives, as you know. We're kind of adhering to that. We have about $1 billion in the ground of investments on the alternative side. We have roughly, I think, $1 billion of committed but not yet funded. I don't think we'll ever really have $2 billion at the same time because you get return of capital, and it floats through. Some of those are unique investment opportunities, some of those are just a combination and mixtures of other more standard assets, perhaps done in a different way. The one thing we will not do, back to your interest bearing and so forth, we will not have our fixed income go further out on the curve and exceed what we think the duration of the liabilities are. That's first and foremost. It's really the asset supporting shareholders' equity where we take the more risky investments, alternatives, and so forth. We're not going to have a 4-year duration of outflow and do 6 years on our investments, and have that open risk in case there's a spike in interest rates and get hit proportionally more than the average or our peer set. I don't know if I've addressed all your questions. I mean, there's a lot in there. I think one thing, I guess I started doing this in 2002, over the course of various times in the last 14 years, there's been duration shortening exercises as people were anticipating interest rates going up. We've been consistently wrong. Yeah. This is a par for the course, whatever. The extent to which the idea of obviously, there's no liquidity risk in being shorter than liability duration, whereas one might present itself as being a little longer. I guess there are some companies who are willing to go a little bit longer on the asset duration and liability to get a little more yield. In terms of past cycles, is there evidence that that is a very real risk, that somebody could have a liquidity moment by doing so? Based on past cycles? Probably not. There's been different reasons underlying cycles. You go back to the '80s with that idiotic cash flow underwriting when everybody thought 14% interest rates were going to continue forever, and hence matching zeros that go forever out there to match their investments. I don't think there's been a lot of empirical evidence on it. I think there's common sense that that happens. Back to your point, we've had an opportunity loss, effectively, by purposely having a lot of the assets associated with shareholders' equity being under a year. I remember a couple points in the last few years, we were like four tenths of a year as our duration on those assets. In anticipation of being hurt less, that it's not symmetric, and that at some point they have to go the other way. We didn't want to get caught if it was violent. We sacrificed net investment income associated with that. We're still happy with it. We note it, we acknowledge it, we thought about it. We wouldn't have done it differently. Is there a house view right now? Are you particularly short right now? It doesn't feel like you're making one of those- No, we're still under a year on the shareholder assets supporting it. Our duration on the outflow as cat has lightened. We still think that's the right decision. Others can grow and gobble it up, and God bless them. When a couple events occur and we see whether the new capacity has any stomach, then we'll see what happens. I'm happy to let them gobble that up in the meantime. Even if we wrote nothing else and didn't increase our claims-made in the current base businesses, our duration would increase because the cat side is falling away. You've made a conscious effort to radically shrink your exposure in cat, more so than most anybody in the industry. You still have a finger on the wind. This is going to be a reasonable sized cat quarter, I guess, between Texas, Ecuador, Japan, and Canada. Does- This is a cornucopia. Yeah. All of them. There is still some debate on whether or not these are reinsurance events. Do you have any market intelligence on any of these various events at this point? Well, as you might guess, we're still going through it ourselves. Let alone we quarter's not over. That's right. You just had Bonnie, which is on the lower-end scale of things, but others could emerge. It varies by event. I think it's less so an event probably in Canada. Well, it's all over the map. Some have pretty low attachments in Canada, and some are gobbling a lot of net. I think it depends on which of the companies we're talking about. See, we selfishly, when these events occur, none of these, other than the Canadian event, are overly large, from our perspective, and that's what risk management's about. You understand the trade-off, you take the actions, there's demonstrative actions, and when events like this occur, you should have no worse than an average hit if you've done it right. I mean, at the upper end. We view that as just an active part of cycle management. At this point in time, I'm not a spokesperson for Arch, the Canadian event, there's a great deal of debate, I guess, on whether it's going to be a reinsurance event or not still. I guess maybe even today's the first day they allowed- I know it's been difficult claims to be managed. We did it. Take a look at that. All right. There's time for one more question. All right. How are you? Just on that point, you guys mentioned on scaling back. What comments would you have about a couple of your peers that have been growing pretty vigorously? I'm not going to have any directed comments. If we're cutting back dramatically, I think the conclusion towards someone who's growing, you probably wouldn't agree with them, but there's a lot of variability and volatility in the business, and it's not always as evident as you think. We could take reinsurer A, reinsurer B, look at their books, assume they're on identical programs, right? They're on the same cedents, and you could have radically different loss ratio results based upon where they play and what they retro out. On the surface, they may be growing, but you don't know the extent. There's some complicated retros out there. It's possible that that growth is ameliorated by the triggers involved in getting retro protection. It'd be pretty naive of me to be too generalized, but I'll leave it at that. Well, thank you everyone for attending, have a great day the rest of the conference. Thank you to you. Thank you, Josh. Appreciate it. Thank you.