HCI Group, Inc. Q4 FY2023 Earnings Call

Good afternoon and welcome to HCI Group's Fourth Quarter 2023 Earnings Call. My name is John, and I will be your conference operator. At this time, all participants will be in a listen-only mode. Before we begin today's call, I would like to remind everyone that this conference call is being recorded and will be available for replay through April 6, 2024, starting later today. The call is also being broadcast live via webcast and available via webcast replay until March 7, 2024, on the Investor Information section of HCI Group's website at www.hcigroup.com. I would now like to turn the call over to Matt Glover, Gateway Investor Relations. Matt, please proceed.

Thank you, John, and good afternoon, everyone. Welcome to HCI Group's fourth quarter 2023 earnings call. On today's call is Karin Coleman, HCI's Chief Operating Officer; Mark Harmsworth, HCI's Chief Financial Officer; and Paresh Patel, HCI's Chairman and Chief Executive Officer. Following Karin's operational update, Mark will review our financial performance for the fourth quarter of 2023, and then Paresh will provide a strategic update. To access today's webcast, please visit the Investor Information section of our corporate website at www.hcigroup.com. Before we begin, I would like to take the opportunity to remind our listeners that today's presentation and responses to questions may contain forward-looking statements made pursuant to the Private Securities Litigation Reform Act of 1995. Words such as anticipate, estimate, expect, intend, plan and project and other similar words and expressions are intended to signify forward-looking statements. Forward-looking statements are not guarantees of future results and conditions but rather are subject to various risks and uncertainties. Some of these risks and uncertainties are identified in the company's filings with the Securities and Exchange Commission. Should any risks or uncertainties develop into actual events, these developments could have material adverse effects on the company's business, financial conditions and results of operations. HCI Group disclaims all the obligations to update any forward-looking statements. Now, with that, I would like to turn the call over to Karin Coleman, Chief Operating Officer. Karin?

Thank you, Matt and welcome everyone. HCI Group wrapped up 2023 by reporting another excellent quarter, with pre-tax income of $54.2 million and diluted earnings per share of $3.40. In-force premiums increased 30% in the quarter to approximately $1 billion. Similar to prior quarters, each of our business segments had a positive contribution to our results. At our insurance division, Homeowners Choice generated another quarter of consistent earnings, and TypTap Insurance Group reported its fourth straight quarter of GAAP profitability. HCI continued to deliver on its commitment to shareholders, paying a dividend of $0.40 per share, our 53rd consecutive quarterly dividend. On our last earnings call in November, we discussed several initiatives underway. They included both of our existing carriers being in the process of assuming policies from Citizens. We said we expected those to total between $150 million to $250 million of in-force premium. I'm pleased to announce we completed three depopulations from Citizens, totaling $273 million of in-force premium. We also spoke about plans to form a new reciprocal carrier to be named Condo Owners Reciprocal Exchange, or CORE for short, with the intention of writing commercial residential insurance in Florida. In five months, CORE went from being a concept to now being a fully licensed carrier with a Demotech rating, has fully placed its reinsurance, and last week completed its first Citizens assumption of $38 million of in-force premium. This is in addition to the $273 million I just mentioned. CORE also has been approved for an April assumption and has applied for a June assumption as well. I look forward to providing updates on these in the future. In addition to these significant accomplishments, we took the opportunity to retire the Centerbridge preferred shares and TypTap a year ahead of schedule. We also successfully raised $85 million through a common stock offering. To summarize, in a few short months, we were able to add a significant amount of premium, improve our gross loss ratio, launch a new carrier, and simplify our balance sheet. Now I'll turn it over to Mark to provide more details on our financial results.

Thanks. So as Karin mentioned, this was another really good quarter for the company. Pre-tax income was over $54 million for the quarter and $117 million for the year. Diluted earnings per share were $3.40 for the quarter and $7.62 for the year. These results reflect the continuing positive direction we've been discussing for a while now: growing premiums, higher investment income, better loss trends, and expense management. Gross premiums earned were 18% higher than the same quarter last year, driven by higher average premium per policy and enhanced by the takeouts we've done with Citizens. While the full impact of these takeouts is not yet reflected in earned premium, they did add $23 million of earned premium in the quarter. Investment income was about 50% higher in the fourth quarter than in the same quarter last year. When interest rates were low, we held onto our cash, and now we are seeing the benefits of careful duration management. Cash and fixed-term investments now total over $1 billion, and we have positioned the portfolio to generate impressive yields with minimum capital risk. Another positive trend you can see in our results is the continued improvement in the gross loss ratio, which was 30.4% in the fourth quarter, down from 39.4% in the same quarter last year. When the Florida legislative changes were announced, we said we expected the consolidated gross loss ratios to drop from 40% to 30%, and that's exactly what's happened. We got here through careful underwriting and rate actions, along with lower claims and litigation frequency. We're not getting to these lower loss ratios by reducing reserves. In fact, net reserves at the end of 2023 are higher than at the end of 2022. If you looked at the balance sheet, you might notice that total reserves are lower than a year ago, but those are gross reserves. They're down for two reasons. First, because of the payments made on storms like Ian and Irma, and second, because we have significantly reduced the ultimate expected loss for Hurricane Ian. Originally, back in September of 2022, we set the ultimate expected loss for Ian based on the models at $960 million. At the end of 2022, we lowered that to $845 million, and at the end of 2023, we lowered it again to $740 million. While we have lowered the ultimate by more than $200 million to date, we are still at the top of the actuaries' range for this storm. The last trend I wanted to mention relates to expenses. If you look at the combination of labor, policy acquisition, and operating expenses, they're flat quarter-over-quarter, and as a percentage of premiums, they're down as we continue to manage expenses. The company is growing, but our expenses are not. Along with the declining loss ratios, expense management is driving significant improvement in the combined ratio, which was 85% for the full year. Now that I've talked about improvements to the income statement, I should also talk about improvements to the balance sheet and liquidity driven by profitability, debt management, and capital management. As you know, we raised $85 million of new equity during the fourth quarter, issuing 1,150,000 new shares. We also expanded the capacity of our credit facility with Fifth Third Bank from $50 million to $75 million during the quarter. As of December 31, 2023, we had just over $215 million of cash and financial investments at the holding company level, and when combined with the credit facility, about $290 million of total holding company liquidity. This is about $100 million higher than it was a year ago. As Karin mentioned, there are a couple of other capital transactions that have happened since the year-end. First, we redeemed the preferred shares owned by Centerbridge, and second, we began the process of converting the balance of our 4.25% convertible notes into common to be completed by the end of the first quarter. A couple of other numbers to mention, book value per share continues to grow. During 2023, book value per share increased from $18.91 to $33.36. Our debt to capital ratio has also improved considerably. At the end of 2022, it was just over 65%, and at the end of 2023, it was down to 48%. With the transactions happening in the first quarter of 2024, this should reduce further. By the end of the first quarter of 2024, we expect the debt to cap ratio to be under 40%. To summarize, the fourth quarter was a great ending to a really strong year. Revenue is growing, expenses are not, the balance sheet is improving, and so is our holding company liquidity. We've positioned ourselves well, and we look forward to the coming year. With that, I'll hand it over to Paresh.

Thank you, Mark. Sometimes the numbers just speak for themselves. Karin talked about the multiple operational achievements over the last five months. Mark provided an update on the financial impact of these achievements, both to the income statement as well as to the balance sheet. Each of these items completed successfully is great just by itself. The fact that we managed to do all of them at the same time is really something. This is possible because of our people and the technology that we have developed. The net result of this is that we crossed the billion dollars in-force premium, which is a major milestone, and with record earnings. Stepping back from the numbers for a moment, our actions have impacted not only our shareholders, but our policyholders as well. We now provide coverage to the most policyholders in our history. The steps we took helped to improve market conditions and reduce insurance anxiety in Florida. Through our depopulation efforts, as well as forming new carriers, we have helped the situation. Citizens is still too large with over 1.1 million policies, but it is smaller today than at any time in 2023. We look forward to helping shrink it further. And the events of the last five months were not impossible. When prudent preparation and planning meet the right opportunity, and if you know how to execute successfully, that is exactly what we have done. In closing, while Q4 2023 was our best quarter so far, it is only our best quarter so far. With that, we will open for questions. Operator, please provide the instructions.

Thank you. We will now begin the question-and-answer session. The first question is from Matt Carletti with Citizens JMP. Please go ahead.

Hey, thanks. Good afternoon.

Hey, Matt. Good afternoon.

Good afternoon. Paresh, your last comment there, you gave us a little peek into your continued appetite for business and to grow in Florida. Can you give us a broader picture of what the competitive landscape looks like and maybe how that might have changed, if at all, over the past three, six months?

Sure, Matt. So, what has happened is we started carriers last year. We've done depopulations. We were very successful at it in the sense that when we made offers, over 70% of people accepted the offers. All of this was very successful. And we watched the competitive landscape as a whole. We noticed that there are a lot more depopulations occurring now. I think there are depopulations happening throughout the year, all the way to June at the moment. So, all of these are very positive things. There are other carriers stepping up to depopulate for the first time ever. So, these are all positive signs, building on the things that Mark talked about last year. That litigation reform that was passed will have an effect, and we're seeing everybody else vote with us that that is occurring. Having said all of that, I also want to make sure that we understand where we sit. Citizens, which had swelled to those expectations, they would swell to almost 2 million policies by the end of 2023, has started to shrink. It is now about the same size as it was at the end of 2022. Right? So, the depopulation impacts are having an effect on the size of Citizens. It's still very big, at 1.1 million. So, there's plenty of opportunity. But these are the first signs of a healthy market returning to Florida. And we're glad we participated in it.

Great. Can you provide us with more details on the $34 million in-force assumption that Karin mentioned last week, specifically regarding the filing approved in April for May? Could you give us an indication of the potential size of those assumptions?

Hey, Matt, let me take that one as well. When we started CORE, given its capital funding, etc., our initial objective was to grow it to about $75 million of in-force premium. So, 38 is a very good step towards that. The April and June depopulations will sort of top us up to the number that we want to see. And we've already bought reinsurance for all of that. So, CORE is up and running and healthy.

Great. And then just a couple of numbers questions, probably for Mark, but for anybody. TypTap surplus at year-end, as well as what was gross written premium for TypTap for the year?

So, the surplus for TypTap is $92.5 million. And what was the other question, Matt?

It was gross written premiums at TypTap for full year '23.

For the full year, hang on a second, TypTap was $363 million.

Awesome. Thank you very much. Congrats on a really nice quarter of a year. Thanks for the answers.

Yeah. Thanks, Matt.

Thanks.

The next question comes from Michael Phillips with Oppenheimer. Please proceed.

Thank you. Good afternoon, everyone. I have a general question. Paresh, regarding your growth loss ratio, you mentioned you expect it to be around 30% considering the situation in Florida. Why aim for such a low ratio, which is significantly better than anyone else in the state or the country? Do you think you might miss out on growth opportunities if you are targeting such a low loss ratio? I also assume there's some caution behind this, given your actions regarding reserves.

Good question. I think the growth loss ratio, and just for the sake of it, so it follows along, we define that as our total losses that net loss is divided by gross premium. That low is partly because some of it's because of Florida. But the reason we also do that is it's just how well our technology works, right? We charge a market competitive premium. It's just that our technology is so great. Sorry, I'm bragging about it at this point. It picks the right policies and we've curated a set of policy orders now who file a claim when they have to, but they don't file a claim when they don't have to, and it creates a better stable outcome, right? It's just been one of those things. The other side of this also, it's a moment in time because when Mark a year ago was talking about the loss ratio going from 40 to 30, right, it seemed like a very big lift that that would occur. And here we are, done and dusted. So that's the item of where we sit, right? It's something we aspire to, and once in a while we achieve it, which is a great thing.

So, is it harder to get those policies that you talked about that have such great loss ratios? Is it harder to get that same kind of client base the more you expand that technology outside of Florida?

Actually, let me try to explain in a slightly different way. And we actually talked about this, what's going on inside Citizens, right? So Citizens has 1.1 million policies in it, yeah? When you look at those policies, you just see 1.1 million policies. What our technology allows us to do is very quickly look at all of those homes, and we think of them as red houses and green houses. We've talked about this before, right? Green houses are ones that if you take at current prices and keep them for let's say 100 years, it will be a profitable outcome. Red houses are ones that won't be profitable, and it could be a combination of things: reinsurance cost is too high or losses will be too high or premiums too low, any number of reasons. The computers decide what's green and what's red. But if you can separate them out, you can out of the same book separate into red house and green house and have a much better outcome than the pool you're selecting from, right? And things that we can see, we can see that in Citizens, there's still probably about 400,000 green homes, but we can tell them apart from the 700,000 red homes in there. A lot of other people can't, and that's what gives you that item, right? We are not getting these numbers because we charge more or less. We're charging a very market competitive rate. It's just better policy selection, and we have an advantage because we can tell red from green as opposed to let's just say somebody who's colorblind.

Okay. Thank you for those analogies. I guess, sort of related, what can you say about the margins of the books that you're getting from Citizens? I know there's a couple of anomalies like for now, reinsurance costs and commissions and things, but on a normal run rate, how does that profitability, the margins in that book look compared to your normal book? And I guess part of the reason why I asked is because what you just talked about is presumably you're not using your technology for those new policies that you're getting from Citizens, but you will eventually maybe when you renew. So, is there a different margin profile in that book compared to your normal underlying book?

Our technology is integrated throughout the Citizens book and helps us identify which policies are performing well and which are not. The policies we acquired are largely processed based on our expectations, which highlights the efficiency of our technology. On the finance side, we adopt a conservative approach. Until we have thoroughly analyzed the new book and allowed it to mature for about six months to a year, we reserve for a higher loss ratio than might normally be required. This cautious strategy is appropriate. Once the book gains sufficient history under our management, we can adjust the loss ratio accordingly. Initially, the assumed Citizens policies have a higher loss ratio compared to our existing TypTap and Homeowners Choice policies. Does that clarify things?

Yeah, it does. Thank you. Last one, kind of a numbers question, but it sort of relates to a real-world type of a question. In the gross written premium number that you give us at $320 million, you split between TypTap and Homeowners Choice. I think there's some accounting of the depopulation in that, and I don't know if you can just wipe that out for us. The reason I ask is because what's the underlying growth of the two businesses without that number in it?

Good question. It's Mark. Looking at the consolidated figures we provided in the press release for Q4, the $320.5 million reported does not impact any other quarters. Out of that amount, $143 million comes from the Citizens depopulation. Regarding your inquiry about the breakdown by underwriter, for Homeowners Choice in Q4, there is a total of 182, but I apologize as I don't have the specific figures in front of me. What I do know is that $143 million of the $320 is the consolidated number.

Yeah. And Michael, a new thing is, presumably you're asking this to update your model, etc. You should be aware that our two underwriters don't like to renew policies evenly throughout the year. TypTap is very heavily skewed in Q4 and early Q1. It has a very high GWP rate in Q4 and early Q1. Homeowners Choice, on the other hand, tends to do most of its renewals in April through August timeframe. So it creates a different dynamic from a gross written premium perspective. Obviously on our own premium, it just evens out.

And hey, Michael, it's Mark again. Of the $143 million in Q4, the consolidated total includes $19 million from TypTap. So to break it down, approximately $20 million comes from TypTap and $123 million from Homeowners Choice.

Okay. Perfect. Perfect. Thank you, guys. I appreciate it.

The next question comes from Mark Hughes with Truist. Please proceed.

Thank you. Good afternoon. The loss rate reached 30%. Do you believe this fully reflects the progress you've made in your underwriting reforms, or is this just a step in the process?

I believe that 30% is a solid estimate of our current position. If you're looking to project into 2024, I think 30% accurately reflects the situation. This figure is consolidated, and although we cannot control the weather, we typically experience a bit more weather impact in the first and second quarters, which could lead to slightly higher numbers during those periods. Overall, 30% is where we stand. As mentioned in my prepared remarks, the effects of the legislation have aligned closely with our expectations. We anticipated a significant decline in claims frequency due to a large percentage of claims being Assignment of Benefits claims, and that's exactly what happened. We also expected a reduction in litigation incidence, which did occur as well. Both metrics have decreased in line with our predictions; however, we are only a year to a year and a half into this change, so we're monitoring the situation closely and applying a level of caution to these figures. We had anticipated reaching 30%, and I believe that's where we are now.

Would you say the improvement in the Ian losses is due to AOB and litigated claims as well? Or is it just more information driving it? Do you think the same factors are affecting those storm claims?

Yeah. So, Mark, don't forget Ian was after the first round of reform, but before the second round of reforms. The bigger thing, I think, was driving the numbers there. It's a totally different set of things. Mark's initial number, I think 960 he said, was entirely driven from the models, right? RMSA or all the models that are covered to estimate the losses right after the storm makes landfall. That's what drove that number. We were already thinking that that number was way overstating for us. As time has gone on, the actuary and we've got more comfortable because of actual developments and actual claims and everything else to reduce the number, because keeping that original model number is just not justified. And as Mark did say in his comments, we're still at the top end of the range. Putting it differently also is that the book that we are curating seems to outperform model losses when an actual cat event happens, right? This is pretty big. And we didn't just miss it by a little bit. We seem to be improving in the models by a huge amount. And we think this is a good thing for future events, yeah?

What is at this point your appetite for growth and voluntary policies in Florida? You're obviously doing well with depop. Are you interested in the green policies out there from a voluntary standpoint?

Yes, we are. Throughout most of last Q4 and Q1, we continued to write voluntary policies. However, the volume of those policies was more than offset by the depopulations. This is just the nature of the situation. I believe we will continue to write new policies, but the numbers are small compared to what we experienced with the depopulations. We had communicated this because people were looking for ways to reduce Citizens. Many individuals want to leave Citizens and just needed a better option to move to, which we provided. In all three takeouts, we were very selective about whom we extended offers to. Over 70% of those we made offers to accepted them. I mention this because, during the November takeout, we had a 70% acceptance rate, while others participating in that takeout had a combined acceptance rate of around 30%.

Yeah. And could you expand on that? Why do you think that is? Is your premium any different than the others or just the size, the brand? What's driving that?

I think it's all of the above, right? You have to bring multiple things to the table. Track record, make sure you make a compelling proposition to the policyholder that they should come with us. We have long-corded agents, and we should give a shout-out to them. Almost universally, all the policies we selected, the agents who are actively telling the policyholders, this is a better place for you to go to. If I could have put you there in the first place, I would have. So take the offer. So these are all individual little items that all come together that work well. The other side of things also, in all three takeouts, we actually made fewer offers than we were approved for by the OIR. We did some of that because unless we think the policyholder is going to be happy with us long-term, we tend not to make the offer, right? We want people to join the HCI family who want to be with the HCI family. You're seeing all the combinations of all of these things coming together in the right way, yeah?

Yeah. Okay. Mark, given the timing on the takeouts in the first quarter and the magnitude, any thoughts about kind of earned premium contribution in Q1? I think when we get into Q2, it might be a little more straightforward exercise, but just given the pacing of things, I wonder if you can provide any guidance on that.

Yeah. So I think Karin and I both mentioned $273 million of premium in-force and that's three takeouts. One in November, one in December, one in January. So I'll just focus on the ones in 2023 first. We booked $23 million of earned premium in Q4. If you look at the timing of the assumptions that were done in '23 and in '24, earned premium in Q1 would be closer to $60 million to $62 million instead of the $23 million that we have. And then a little bit higher than that in Q2.

Yeah. And then the January takeout.

I kind of changed my approach halfway through, Mark. I apologize for including both points. In Q1, there should be approximately $40 million more in earned premium from those takeouts compared to Q4. Additionally, there will be a bit more in Q2 because one of those assumptions was made towards the end of January. Does that make sense? Does that answer your question?

Yeah, that's perfect. Thank you.

It's about $14 million. And of course, that number will be significantly higher in Q1.

Okay. Does that margin flow through? Is that 14 out of 23? And then if we just did a little simple math, would the other things be equal?

So keep in mind from now until May 31st, you've got no reinsurance in there, right? So the margin, initially, the margin is about 65%. We're reserving 35% on that book. So you've got 65% initially. There's also no policy acquisition expense initially. Now as policies start to renew in March, some policy acquisition expense will start to creep in. So that'll erode that margin a little bit. But for the first five months of the year, the margins are obviously very significant. And then when June 1 comes along and reinsurance kicks in, of course, the margins will deviate toward the norm of the rest of the book. But for Q4, Q1, and Q2, you've got a very significant amount of premium that's coming in at a very high margin.

Yeah. Okay. Super helpful. Congratulations. Thank you.

Thanks, Mark.

Okay.

We will now open the floor for a question-and-answer session. The next question is from Casey Alexander with Compass Point. Please go ahead.

Yeah. Hi. Good afternoon. I have a couple questions here. You're doing the CORE assumptions in April and June and just thinking about the company's timing; normally, they don't do assumptions right ahead of the busiest storm season. Can you discuss the timing of those assumptions and why not wait until later in the year when you're past the storm season before making those assumptions?

Casey, that's a great question. The reason for our approach is related to the mechanics of depopulation and how it aligns with the startup phase. CORE initially had a surplus of $25 million, which we need to manage alongside our reinsurance when we conduct our first depopulation on February 27th. We arranged for reinsurance that fits the size of our book. Furthermore, due to the schedules and blackout periods related to Citizens, it's challenging to reach the peak depopulation amount all at once because of renewal cycles. Therefore, we're planning additional depopulation efforts in April and June to supplement what we achieve in February. While theoretically, everything could be done in one round, it’s actually more practical to stagger the process across three takeouts instead of one. So essentially, the April and June efforts can be considered as delayed takeouts from February.

Understood. If you already have the reinsurance in place, then you're covered for the upcoming season. So that makes perfect sense. Thank you for that. Secondly, your discussion about the declining reserves against Ian, the change over $200 million from the models, I'm just curious: the next time there's a storm, would you again just go off the models first and then work it down, or is your experience with the book knowing that it tends to outperform the models? Would you reserve it differently next time or would you do the same thing and just go by the models and then work it down?

So, Casey, after a storm, everyone relies on the models and uses those figures. This is similar to what happened recently. If you remember the days following Ian, the industry was busy estimating costs—initially saying it would be $30 million, then $40 million, and eventually over $70 billion. Given our technology and market share, we believed that spending $70 billion on claims was virtually impossible. However, we still need to account for actuaries and industry models. That's why Mark is nearly required to adhere to the model's estimates. It typically takes five to six months for us to transition to our own claims experience model, which is the current phase we're in. Mark mentioned needing to revise Ian's estimates, and he indicated he was at the upper limit of the range. The actuary isn't suggesting he can allocate more than that amount for Ian at this time. We still have adequate reinsurance, but this process is ongoing. As I mentioned in a previous answer, this outcome is not coincidental; it's a result of our technology, which curates a strong portfolio and demonstrates solid performance. These results are clearly showing how effectively our systems are working.

That again makes perfect sense. Thank you. My last question is, TypTap has now generated four straight quarters of profit. You've kind of removed Centerbridge from the equation. So, what else does TypTap have to do before you guys would be willing to create a capital transaction for the company?

Okay, Casey. I don't think TypTap has been held back by the necessary steps for a capital transaction. I believe we can manage that reasonably well and in a relatively short time frame. Remember, we went as far as filing an S1 three years ago. However, this situation is influenced by two factors. First, there are the prevailing macro market conditions regarding IPOs, which we continue to observe. We're being informed that the market conditions are favorable for follow-on offerings, but not as much for IPOs, which gives us some hesitation. The second factor is whether TypTap actually needs new capital. If not, should we focus on increasing its value and demonstrating its strengths until the market fully recognizes it? Only then would we proceed with a capital event. These considerations are why we're not moving as quickly as some might expect. By being patient, I believe we are actually enhancing value for our shareholders, which is ultimately our objective.

Okay. Thank you for that. And my last question is, there's significant indication that interest rates are likely to be declining at some point in time here in the next several months. Just look at Chairman Powell's testimony earlier today. And you guys have a significant cash hoard. At what point in time do you start to extend duration a little bit and try to capture some of that yield curve for a little longer compared to where short-term rates are likely to go over the next 12 to 24 months?

Funny you bring that up. We were just having that discussion internally, right? I think in the coming months we will be extending duration and you're going more towards fixing the yield on our cash hoard. So yes, we are exactly on the same mind and we are starting to move in that direction.

Okay. Well, thank you for taking my questions. Appreciate it.

Thank you.

Okay.

At this time, this concludes the question-and-answer session. I would now like to turn the call back to Paresh Patel who has a few closing remarks.

Thank you. On behalf of the entire management team, I would like to thank our shareholders, employees, agents, and most importantly our policyholders for their continued support. Thank you.

At this time, this concludes our question-and-answer session. This concludes today's call. You may now disconnect.