Kestrel Group Ltd (KG) 2017 Q2 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Maiden Holdings Second Quarter 2017 Earnings Conference Call. (Operator Instructions) As a reminder, this call will be recorded.

I would now like to introduce your host for today's conference, Mr. Noah Fields, SVP of Investor Relations. You may begin.

Good morning, and thank you for joining us today for Maiden's Second Quarter 2017 Earnings Conference Call. Presenting on the call today, we have Art Raschbaum, Maiden's Chief Executive Officer; along with Karen Schmitt, our Chief Financial Officer. Also in attendance today is Pat Haveron, President of Maiden Reinsurance Ltd.

Before we begin, I would like to note that the information presented here today contains projections or other forward-looking statements regarding future events or the future financial performance of the company. These statements are based on current expectations and future events and are subject to a number of risks and uncertainties that could cause actual results to differ materially from these expectations. We refer you to the documents that the company files from time to time with the Securities and Exchange Commission; specifically, the company's annual report on Form 10-K and our quarterly reports on Form 10-Q. Some of our discussions about the company's performance today will include reference to both non-GAAP financial measures and information that reconciles those measures to GAAP as well as certain operating metrics and may be found in our filings with the SEC and in our news release located on Maiden's Investor Relations website.

Please also note that, unless otherwise stated, all references to common share data in today's discussion are on a diluted share basis and comparative comments will refer to Maiden's results in the second quarter of 2017, relative to corresponding period in 2016.

I will now turn the call over to Art.

Good morning, and thank you for joining us. Results for the quarter were challenged with adverse loss development observed in both of our operating segments. While I'll get into our overall business activities for the quarter shortly, I believe it's important to address this development first. We do not believe that the development observed in the quarter is analogous to the trend observed across our portfolio over many quarters, which specifically emanated from elevated excess commercial auto liability frequency and severity in our treaty portfolio for the 2011 to 2014 underwriting years. That's a phenomenon that has plagued many in the industry for some time.

In contrast, this quarter we responded to specific development, largely related to specific accounts. Adverse development in the quarter totaled $56 million. As we've commented in the past, Maiden maintains a robust process of working with our clients to better understand the underlying dynamics of their loss development. That includes active and ongoing claims management with additional case reserves posted by our claims professionals as necessary and auditing activities, as well as account-specific actuarial analyses that all help to shape our view of ultimate exposure. These activities allow us to respond quickly to adverse trends, and then to work with our clients to effectively react to the development. Our objective is to respond to unanticipated claim development promptly when it's observed and validated. That response can manifest itself in changes in expected ultimate reporting patterns for specific accounts, or as in our past excess of loss commercial auto development across entire classes and types of business when the trend is broader across multiple clients and, in that instance, on an industry-wide basis.

As a result of the adverse development, Maiden produced a non-GAAP operating loss for the quarter of $12 million and generated non-GAAP operating income for the first 6 months of $10 million. Beginning with the AmTrust Reinsurance segment, we experienced $29 million of adverse development in the quarter. We mentioned in our first quarter results that we've been observing changes in loss activity in the small account commercial casualty lines in the last 2 quarters, which were impacting a number of underwriting years. We committed to complete a review of claim activity in advance of the quarter. In our experience, the pattern suggested that they were either fundamental changes in the way losses are developing or there were changes in claim practices that influence the development or a combination of both.

We've completed our preliminary reviews of several large claim offices, and we've observed clear increases in staffing and corresponding reductions in examiner workloads versus prior periods. These types of operational changes can create volatility.

At this point, we're giving some credence to the impact of staffing increases. However, we're also giving way to pattern shift that reflects regulatory changes and elevated severity in specific jurisdictions.

Separately in the quarter, we did adjust our expected loss ratios for the most recent years for the Hospital Liability line to reflect profitable, but more conservative expected loss ratios in view of developing trends and economic factors that could influence future loss settlement values. Please note that to the extent that AmTrust has adverse development, the components of the reserve movements are likely to be different. Each company sets reserve and picks independently from each other. And AmTrust has lines of business for which we don't provide reinsurance. In addition, when AmTrust makes an acquisition, Maiden does not assume any of the unearned premium that comes along from that newly-acquired business or the existing loss reserves, with only one exception in our history, which occurred at the beginning of Maiden's relationship with AmTrust.

In our Diversified Reinsurance segment where we recognized $25 million of adverse development in the quarter, the majority of the impact was largely from 3 accounts: 2 treaties and 1 facultative program. Of the 3 accounts, the largest contributor development was the single facultative commercial auto program, which has been terminated. For this pro rata account, the observed latency was significantly influenced by the clients reserving practices. For the most part, the balance of the development emanated from an individual risk facultative commercial auto segment -- or business, excuse me. It is important to draw a distinction between facultative commercial auto loss development that we observed this quarter, in contrast to the treaty excessive loss contracts that resulted in significant adverse development over many previous quarters, which we believe we've appropriately addressed.

As you may recall, in responding to our treaty-related commercial auto development last year, we mentioned on several occasions that we were not seeing a significant and adverse or systemic trend in our facultative casualty commercial auto business. In this product line, we price and underwrite each specific risk, and we can effectively react to adverse trends quickly in comparison to a treaty contract where we reinsure an entire portfolio of our client's business. The exceptions are the facultative program contracts, which reinsert homogeneous portfolios of risk. In the quarter, as I mentioned, we did react to late develop emanating from a specific commercial auto program. We also observed some elevated loss activity among several accounts in lines of business. But at this point, it's not our assessment that there's a systemic underlying issue across the entire portfolio. We believe that we've responded appropriately to the elevated claims activity and latency in the quarter.

Finally, while not adverse development, we did realize approximately $6 million of non-catastrophe property claims, which also impacted our results. It's important to note that the vast majority of our diversified segment development in the quarter emanates from underwriting years 2011 to 2014. Further, the accounts that drove the majority of adverse development have either been terminated or experienced significant changes in terms and conditions since 2015. And importantly, our post-2014 underwriting years remain within performance expectations.

Notwithstanding the adverse development, we've continued to enjoy growth in invested asset, strong cash flow and overall growth in gross and net premiums written. Despite the adverse development, our balance sheet remains strong with only a modest drop in second quarter book value from $12.19 per share to $11.95 per share. And Karen will provide more details on the balance sheet shortly.

Looking at revenue and business development. Overall gross written premiums were $705 million, which is an increase of 2.5% in the quarter. Year-to-date, gross written premiums have increased by 4.9% from the prior year. Beginning with the diversified segment, the focus on disciplined underwriting continues in the quarter. Gross written premiums were down 4.6% from the prior year second quarter due to the impact of a terminated account in the U.S. that was canceled on a cut-off basis with premium returned to the customer. The U.S. premiums were up year-on-year, absent the impact of this account. As a result, we do not see the quarter revenue trend as run rate.

Our outlook for the U.S. business continues to be positive with a continued expectation for year-on-year growth. In the U.S., during the quarter, we also saw the impact of several other nonrenewed accounts, which was more than offset by the impact of several new accounts in expanding existing relationships. While the market remains competitive, we continue to find opportunities to selectively grow our portfolio as well as to selectively color portfolio of underperforming accounts. The U.S. team is focused on building our highest margin businesses and expanding our existing client relationships. And we continue to roll out additional products and services in response to client needs.

While comparatively smaller than our U.S. business outside the U.S., our auto OEM branded insurance solutions business in Europe has experienced year-over-year growth of 23%. And it reflects the addition of several new programs and the expansion of existing ones. In this business, our team provides original equipment manufacturing, auto manufacturers with distribution and product capabilities in collaboration with insurer partners. Maiden receives fee income as well as reinsurance revenues. The growth we're experiencing is from personal auto extended warranty and, to a lesser extent, GAAP. We're also actively developing several new payment protection insurance programs in both the OEM and the non-OEM marketplace with several programs launching within the next several quarters.

And then finally, our Capital Solution business. We have added several new accounts and are activity entertaining several new prospects. We continue to find a very competitive reinsurance market in Europe, but we are finding great interest in the capital solutions model.

Looking at the AmTrust segment. Gross written premiums in the quarter were $564 million, up 7.8% from the prior year second quarter; that's the same as the year-to-date change. Much of the growth continues to reflect the impact of 2016 acquisition activity versus the prior year. The majority of AmTrust growth is emanating from the workers' comp and small account commercial underwriting portfolios in the U.S. Absent the impact of acquisitions that were first ceded to Maiden in 2016, overall organic growth is relatively modest as they continue to maintain discipline. The warranty and special risk businesses are up year-on-year, while the revenue from Hospital Liability has continued to decline, again, reflecting continued discipline. We expect continued moderation in AmTrust growth as prior year acquisitions are fully absorbed, and disciplined underwriting actions in its program business continue to be implemented.

While we're disappointed with our results in the quarter, we believe that the fundamental strength of Maiden's business model remains unchanged, and we're absolutely committed to stabilizing our underwriting performance and restoring the profitability of Maiden going forward.

I'd like to turn the call over to Karen Schmitt, our Chief Financial Officer, to provide more details on the quarter. Karen?

Thank you, Art, and good morning. As Noah said earlier, unless otherwise stated, all references to common share data are on a diluted common share basis. And comparative comments will refer to Maiden's result in the second quarter of 2017 relative to the corresponding period in 2016.

Last night, Maiden reported a second quarter 2017 net loss attributable to common shareholders of $22 million or $0.26 per share, compared to net income attributable to common shareholders of $31 million or $0.39 per share. The non-GAAP operating loss was $12 million or $0.14 per share, compared with non-GAAP operating earnings of $28 million or $0.37 per share. In general, these results were impacted by adverse prior year development in both of our underwriting segments, which was partially offset by continuing strong investment returns.

On a year-to-date basis, non-GAAP net operating earnings were $10 million or $0.12 per diluted common share compared with $57 million or $0.73 per diluted common share in the first half of last year.

In the second quarter of 2017, gross premiums written increased 2.5% to $705 million from $688 million. Gross premiums written in the diversified segment totaled $141 million, a decrease of 15% versus the second quarter of 2016, which, as Art mentioned, was primarily affected by the termination and commutation of one account.

Our AmTrust Master Quota Share business saw an increase in premiums during the second quarter, while European Hospital Liability business continues to decline and is down more than 20% on a year-to-date basis.

Net premiums written totaled $684 million, an increase of 5%. Net premium written changes were higher than gross written premium changes due to reduced usage of the retrocessional quota share in 2017 versus 2016.

Net premiums earned were $711 million, an increase of 12%. In the Diversified Reinsurance segment, premiums -- net premiums earned increased 7% to $204 million. The AmTrust Reinsurance segment net premiums earned were $507 million, up 13%. The lower utilization of retrocessional support also impacted net premiums earned in both of our operating segments.

Net loss and loss adjustment expenses of $529 million were up 24%. The loss ratio of 74.1% was higher than the 66.8% reported in the second quarter of 2016, largely due to the adverse development recognized in both operating segments.

Commission and other acquisition expenses increased 13% to $210 million. The expense ratio decreased to 31.7% compared with 31.8%. General and administrative expenses for the second quarter totaled $15.3 million, an 11% decrease compared with $17.3 million. The general and administrative expense ratio was 2.2%, compared to 2.7%.

The combined ratio for the second quarter totaled 105.8%, compared with 98.6% in the second quarter of 2016. For the first 6 months of 2017, the combined ratio was 103.4%, compared to 98.7% in the corresponding period last year.

The AmTrust Reinsurance segment combined ratio was 101.5%, compared to 94.9%. In the AmTrust segment, we experienced adverse loss development in the U.S. small account casualty lines. In the quarter, AmTrust announced the purchase of an adverse development cover that would respond to adverse net development over its entire portfolio with outstanding reserves. Maiden does not participate in this contract, and the economic impact does not [in origin] Maiden nor is the cost allocated to Maiden.

As we've commented in the past, Maiden reinsures a specific portion of the AmTrust underwriting portfolio, not the entire book. As an example, Maiden does not assume in-force liabilities of any acquired business with the exception of the single 2008 deal that Art mentioned, and does not reinsure Lloyd's syndicates.

The Diversified Reinsurance segment combined ratio was 112.9%, compared to 103.4%. Diversified segment results were impacted by $25 million of prior period adverse development, primarily from facultative commercial auto as well as a handful of specific accounts across several lines of business, with over half of the development for the quarter emanating from 3 accounts.

In the Other Nonoperating category, we had approximately $1 million of adverse development in our former E&S property book that was the result of expense adjustments to the few remaining open claims from that business, which has been in runoff since 2013. Maiden also experienced elevated non-cat property loss activity in the quarter.

Investment results were strong through the quarter with net investment income of $41 million, an increase of 15%. At the end of the quarter, total investable assets were $5.5 billion and the average yield on the fixed-income portfolio, excluding cash, was 3.23%, with an average duration of 4.83 years. The duration of Maiden's total invested assets, including cash, was 4.42 years as of June 30 compared to the duration of liabilities as of June 30, which was 3.77 years.

In the second quarter, we purchased $259 million of securities consistent with our historical investment philosophy with a rated -- with a weighted average yield of 2.7% and an average duration of 2.97 years. We continued to have strong cash flow in the second quarter with cash flow totaling $178 million compared to $156 million.

Cash and cash equivalents were $439 million at June 30; were $288 million higher than at year-end. Our total cash balance as of June 30 is higher than normal due to the unusually large purchase at the end of the quarter of investments, which settled in July. Additionally, the balance increased due to the net proceeds from the issuance of Preference Shares Series D on June 15, 2017, which was partially used to redeem all of the 2012 senior notes on June 27, 2017. Unrealized gains totaled $42 million as the values of non-U. S. denominated securities increased due to foreign exchange movements.

Shareholder equity was $1.5 billion, up 10.3% compared to December 31. Book value per common share was $11.95 as of June 30, or 1.4% lower than December 31. During the second quarter, we replaced $100 million of 8% coupon senior notes with $150 million of 6.7% coupon preference shares. We have another opportunity to improve our cost of capital later this year as we replaced our $150 million of 8.25% preference shares. We continue to evaluate our options. And we'll alert the market of the timing of the redemption when appropriate; however, lowering our cost of capital remains a priority to improve profitability.

I will now turn the call over to Art for some additional comments.

Thank you, Karen. As we look across the market today, it's fair to say that competitive pressure remains strong. In North America, capacity remains abundant as it does throughout the globe. And other than what has been a noticeable strengthening of terms for personal auto quota share contracts and commercial auto contracts on balance, casualty lines remain in the strong competition.

As I've commented in the past several quarters, we tend to work much harder to find the right business to write, but we still believe that our competitive advantages provide us with the ability to selectively grow our underwriting portfolio. Our operating expense relativities remain among the lowest in the industry. And our unique competitive advantages, such as our collateral trust and our client-focused business platform, continue to serve as well.

As I highlighted earlier, we see a number of areas of profitable growth in our diversified portfolio. And we view this quarter's decline in gross premium as an anomaly that should recover in subsequent quarters. In the AmTrust segment, we believe that recent underwriting actions in the program business and the commercial auto line will benefit Maiden. AmTrust has been less active in acquisitions in the last 6 to 12 months. As a result, we expect revenue trends to continue to moderate.

As we look at the balance of the year, we look to restore and strengthen our underwriting and operating income and see opportunities to continue to properly expand our underwriting portfolio. The adverse development that we're seeing does not invalidate our perspective on the probability of our underwriting portfolio over the last several years. In fact, the long-term combined ratio for Maiden, absent the adverse development emanating from commercial auto, remains solidly profitable. As I mentioned earlier, the diversified segment development in the quarter does not reflect continued deterioration of our treaty excess of loss commercial auto portfolio, which was the source of the majority of our past adverse development. But rather it reflects some specific account deterioration as well as observed latency, primarily in our facultative portfolio.

Within the AmTrust segment, the development that we've observed for the quarter really reflects largely the impact part of our ongoing diligence and active management of this important client relationship. On a relative basis, the development recognized for the AmTrust segment in the quarter is relatively modest. While we can never guarantee that we will not realize future adverse development, what we can and have committed to do, is to respond to underwriting and claims challenges effectively and promptly. Fundamentally, our actions in the quarter are focused on both objectives. And, importantly, we remain confident in our ability to restore profitability in subsequent quarters and continue to build our unique business platform and effectively serve the needs of our clients and our shareholders.

This concludes our prepared remarks. Operator, could you please open the lines for Q&A?

(Operator Instructions) And our first question comes from Randy Binner with FBR.

I kind of came in the call late, so hopefully this isn't too remedial. But I guess, just thinking about some of the reserve items. So there's a latency issue that I think has continued in the Diversified Re area. And so I just wanted to ask -- I mean, is that -- am I hearing it right that, that's mostly commercial auto related? And is that a -- is it just a function of litigated cases? Or is there something going on further from kind of an actual loss perspective?

Okay. One of the things, I want to draw a distinction between the commercial auto adverse development, which really emanated from our excess of loss treaty portfolio significantly that we recognize over many quarters. That phenomenon is obviously the same phenomenon that's been affecting a lot of companies in the P&C space. We reacted, as you know, with a fairly strong reserve movement at the end of the year, and we have seen relatively benign development coming from that segment over the last 2 quarters. So it gives us increasing comfort that we've sort of responded to that.

I think the issues that we're seeing this quarter, Randy, are much more account specific and less, what I would say is, systemic. We talked about program facultative account that generated a significant portion of the facultative sort of adverse that we noticed. We think it's a very client-specific issue in this case. There are certainly operational issues that we think have sort of magnified the latency in this account. We think we've addressed it. We're not on the account anymore. But we do believe that it's as much a client sort of process-driven issue as it is some late phenomenon.

As far as the facultative certificate -- and I mentioned this earlier. And If I'm repeating myself, I apologize. But in facultative -- when we talk about facultative, we're typically talking about individual certificate, individual risk underwriting. So our underwriters will entertain large volume of business. They'll pick the select best risks that they can find out of the incoming submissions. And they'll apply fairly kind of rigid pricing, which is based on pricing benchmarks that we've established out of our own portfolio and also tempered with industry data. We have the ability -- so we write an account for a year. There's no assurance that we'll write it the next year or the year after. And that portfolio turns over fairly readily. But we have -- it's a very quick action kind of portfolio. We can deal with issues fairly quickly.

So our general view is that we're not seeing the same kind of systemic issue relating to our facultative book as we saw in our treaty book. And so we feel a bit better about the development we're seeing there. And I know that when you aggregate all of the commercial lines activity and you add what we saw in this quarter, it just looks like a continuation of the same trend. But quite honestly, we believe that what we're seeing this quarter is a bit different and a bit risk specific.

Karen, would -- anything you would add to that?

Yes. No. I think that's right. And I guess the one thing I would add on the large account, that is a proportional account. So while we do -- we do claims audits as we have for all of our accounts. When we went in and were specifically having a problem with excess of loss treaty, we were able to look at virtually every claim that either had been reported to us or was likely to be reported to us based upon characteristics of the claim. And we were able to do a very deep dive on those customers. It's obviously not possible to look at every claim on a proportional book. So we have looked at the way that the company reserves and changes in their practices. And without getting too specific to this customer, it's -- there were some issues with this one that we chose to recognize this quarter.

Yes. And actually that kind of gets to what the follow-up was. So I heard some comments around claims practices, so that was more kind of these program managers. I think -- just jumping over to AmTrust. I think in the past, you talked a little bit about them kind of staffing up on claims adjusters, and that having some impact on the development. Is that the case there? Or was the AmTrust adverse just more kind of regular review of that book?

Well, obviously, as we do with every account that we write, we do a quarterly complete analysis. We mentioned that in the last quarter -- last couple of diagonals, we'd seen kind of fairly consistent adverse or, let's say, loss development in multiple underwriting areas, which to us either suggest just a coincidence of development or something's operationally going on. And so we committed in the last quarter to go out and visit a select number of offices and just see if we could find any operational changes. And I think what we -- the combination of that effort was that we identified definite sort of increases in staffing, in some cases by 30%. But in addition to that, reduction of individual technician caseloads, which -- certainly, claims operational changes can have a very significant and leveraged effect in loss development. That said, the development that we recognized in the quarter is really more specific to severity we're seeing in specific jurisdictions, New York being one, changes in regulatory trends in places like Florida, but not a broad wholesale movement across the book and still certainly profitable, at least from our perspective.

So the New York and Florida comments, is that like workers comp in those states coming from the AmTrust book?

Yes. It was not -- our movement isn't strictly workers comp. There is some impact from general liability as well. I think commercial auto was fairly stable in the quarter, so we didn't see a lot. And as you know, that was part of our year-end reaction as well to commercial auto there. So I'd say, probably the major lines would be GL workers comp.

And the GLs, the program business, they are program segments, is that right?

No. The GL was not in the program segment. This was all in their small risk commercial business. And we responded pretty significantly at the end of the year to what we saw in the program segment, and we had a bit of additional response in the last quarter. We feel like we reacted effectively there and we'll, obviously, keep analyzing and watching that. But in the quarter, our development was chiefly in the small commercial risk business.

Yes. I guess, I'll conclude. Is this the -- I know there's timing differences, and you have your own processes. But I mean, you know, -- I just kind of want to hear the response, is that -- so they took $73 million of adverse pretax. And it mirrored these segments you're talking about. So, I mean, do you feel like your processes are pretty synced up with what's going on at AmTrust from an adverse development perspective?

Well, Randy, there are timing differences, as you correctly pointed out. So they probably, in a number of lines, have a much more complete quarter than we do because of reporting to us is quarterly (inaudible) typically. And so we get revenue data. And then we take our loss ratio assumptions and bring them up in sync with the earned premium in the quarter. So there can be timing differences there.

I also think it's important to recognize that in the AmTrust portfolio, there are a lot of elements that we don't reinsure, and those have an influence on their total combined ratio. Even within the segments we reinsure, there may be incoming portfolios, let's say, Republic that include unearned premium as well as loss reserves that they could have influences in things that they see. So it's very difficult, if not impossible, to do an apples-and-apples comparison of that. I think in a general directional sense, most of their development I believe in their call yesterday was associated with their small commercial business, and I think they spoke of that. I think they did see, to a smaller extent, an elevation program as well.

And our next question comes from Ken Billingsley with Compass Point.

I wanted to follow up on the retro cover that AmTrust put in that you were just discussing. And just to follow up a little more, does -- in a charge, as the they add to reserves there -- and I understand there's moving pieces and you don't reinsure everything and the timing is going to be a little different. But as they increase reserves and push that through, through the cover, does that predispose them to have an incentive to push more into reserves there to utilize it, which requires you guys to address it and obviously have to respond to any increases that a major partner is increasing reserves?

Ken, I think the most important thing to recognize is we're setting reserves independently. So the fact that they may have an increase in a certain line of business, it sort of depends on their starting point and our starting point, whether we have the same kind of development they have. But whatever they choose to book through that cover really doesn't affect us, because we're still going to be establishing our reserves separately. The base data that we're getting, the paid losses and the case reserves, those aren't going to change as a result of the cover that was purchased. So we will certainly watch with interest to see -- we're -- as we always do, we listen to the call and we find out about their development pretty much the same time you guys did last night. And we're interested in seeing where that is. But it doesn't really affect our process. Our process is independent. We get a lot of detailed actuarial data from AmTrust, but we're setting our own reserves.

And Ken, I would say, just to sort of echo what Karen said, if we see something that they announce in the quarter that relates to a segment of the business that we reinsure that we're not seeing, we delve into that. We want to understand why a client saw something that we didn't see. And so we'll continue to maintain that, and we have been doing that all along. But I don't think the ADC creates any different view in terms of how a company recognize losses. But that would be a question I think more directly addressed by AmTrust themselves.

I understand that you guys set reserves independently. And they've made the same comment between how you guys both look at the reserves. But over time, I would imagine that you both have to come to a similar conclusion that if it's not a quota share agreement, at least on similar lines of businesses, regardless of how you initially set those expectations, now that you've been ensuring some of these businesses for a longer period of time, that eventually you'll have to come to some consensus over time. Would that be correct?

Absolutely. I mean, for the segments that are exact elements of their portfolio. But again, as I say, if you look at just an example, the AmTrust acquisition activity over the last 3 or 4 years, there have been a number of incoming portfolios. There have been a number of portfolios running premium reserves. In the case of Tower, there was also some front-end business that I believe all makes it into the small account universe; that's not in our numbers. So -- but the exact subsets of our business that are exact replicas of those subsets of their portfolio should absolutely be in sync ultimately.

Well, eventually, yes. Eventually, all losses are paid and the answer is the answer. It's just -- at this point, we're both doing separate projections of what that answer is.

But as I said earlier, obviously, if we see something that's very specific to business that we reinsure, that we're not seeing in our diligence, but they're commenting on, we're going to delve into that and understand what the impact is or isn't on us.

So for your reserve, your loss picks to be quite a bit different over time from them. Then we're really talking about that your portfolio of what you're covering is starting to diverge quite a bit from what the AmTrust book looks like, where before what you reinsured was pretty much predominantly all their book. But now it's -- what are we're talking about, 60%, 70% of the type of business that they're writing that's coming on that you're reinsuring? Or is that even less than that and getting smaller?

I think if you actually look at the percentage of the business we reinsure, it's in the 20s now. So yes, it's -- and it's because there may be other programs that are -- that we don't participate that maybe (inaudible) others, the Lloyd's business. And again, all those incoming portfolios as well have an effect on it. But you're absolutely right in saying that in the early years, pretty analogous. And, frankly, most of those years are pretty well developed. And I'm sure if you looked at the 2, you'd see -- if you looked at the relative action here, you'd see a pretty close number. But as you look at -- as you move further out, that kind of diverges.

One thing we do from time to time is we just look at -- just assume that we have 40% of their total portfolio and compare it to -- compare their loss ratio to our loss ratio. And they're within a very small, maybe a percentage difference of each other. And some of that is going to be made up by those differences. So it's not like in the aggregate, we have that big -- that massive difference in our view.

That would also be on an inception-to-date basis.

That's right.

Quarter-to-quarter, we can have variation. We can have variation and mix. There can be all kinds of reasons why --

Year-to-year --

So we do look at -- on an inception-to-date basis, just how close we are to AmTrust. We're actually surprisingly close now, considering how much -- much of the business that we're now participating in. But we certainly could expect that they may not stay close just because of all of that other businesses in there.

And they have -- they mentioned that they're looking to do quota share on that personal lines. Are you planning on bidding on that? Or are you still going to stay away from personal line?

No. We don't necessarily stay away from personal lines. But I think probably at this -- first of all, that business is not included in the quota share to the extent they presented to us. We would evaluate it. But as it stands today, that's not contemplated.

Okay. Let me just -- I want to just clarify what you said. You said if they presented it, you'd look at it, but you're not contemplating it. Does that mean you would --

Yes. I think they've already been -- I think they've been in discussion with other markets and that's certainly within their purview. And it's not concerning to us at all.

Have you changed your view of accepting straight up cat risk on the property side? I mean, you historically have avoided that.

So if that was presented to us, like most of our pro rata auto business, we would have occurrence caps. And then the other markets that are willing without it.

And the other question I have is, you mentioned that on the reserve charge -- I mean, I know you said it was not the same commercial auto exposure losses that you took charges in the last few quarters because this is more on the facultative side. Two parts of the question. One -- I mean, I didn't quite understand some of what you said. But some of the issue that -- you said the processes. Is this the process of the primary company? The way that they were conducting their business didn't allow you to see that you needed to make these adjustments? Or is this a frequency and severity issue that's similar to what the market experienced? It was just delayed in getting to you?

Yes. I mean, I guess -- this one particular account, without getting into too much -- it's a little challenging how specific we can be because we certainly don't want to call out even at a former customer. But in this particular account, we have been doing audits as we do with all of our customers. But we just became aware that the reserving practices were not what they needed to be, I guess. I think we should probably leave it at that.

So as you know, we have a very granular process where we do account-by-account reserve analyses every quarter. That, many times gets preceded by physical audits of the client. And in this instance we completed physical audits. But as Karen mentioned earlier, on a pro rata contract you can just sample. The universe -- you can't look at the census of everything they write. And outside of our sample analysis, we've seen that there's been a bit of a tendency to under-reserve cases, and it's now kind of developing into our quota share.

Doesn't a facultative contract give you the right, though, to like set individually on a case-by-case basis?

Programs are different. Programs are kind of hybrid. They're are sort of like treaties, sort of like individual certificate. In theory, companies have the right to accept and reject within a program. But for the most part, it's a large portfolio of homogenous risk, and you're not looking at every single transaction that gets written. It looks more like a treaty than it does. We don't write a large kind of volume programs today, but that was one program that we did right. And it was a much more like a treaty than it was a facultative certificate.

So it would be safe to say that -- and I know you're not trying to expose who the person is, but -- or the company is, but that -- this isn't that they're experiencing unique or different exposures than what the market had in general? It was just the way that they had been viewing that exposure is what has changed?

Yes. And one thing that I would say to clarify, that was a pro rata account. That was a not an excess account. So most of our issues have been in excess. It's been more of a severity issue for us that has really impacted sort of the many quarters that we reported some adverse. In this account, it was pro rata. And we generally not observed that level of sort of volatility on reserving and adverse and what I would say is historically latent development. Pro ratas, you should be getting the claim in summary data, all the claims the client receives, and their expected reserves. And as I say, we test them. So it is a little different from the rest of the commercial auto we've been seeing. And for the most part, pro rata hasn't been the issue for us. It's been much more of an issue on excess of loss, and this is not an excess of loss contract.

And our next question comes from Christopher Campbell with KBW.

Is it possible to get a breakdown of the accident or the underwriting years contributing to the reserve charge in the AmTrust segment?

For AmTrust, I guess, when I look at the pieces, a lot of it was '14, '15, '16. I don't have the precise numbers in front of me, but that was the majority of it. We had -- I guess, in New York, it actually went back a little bit further. But for the rest of it, it was really in the more recent years. And some of it was -- we concluded that we needed to book a little bit more conservatively, as Art mentioned, for the Hospital Liability. So -- and that was certainly a more current year.

Right. No. That's very helpful. So just kind of -- further question, just in terms of switching over to the premium growth. So if -- so -- all right. So if you're seeing more adverse development in recent years -- you mentioned 2016 -- so why would Maiden be actually increasing its retention on the AmTrust segment, if there are concerns with reserving and pricing?

We're not increasing our retention.

I was just -- so I was looking at the segment. It looks like your retention ratio is about almost 97% this quarter, and that's up from 93% a year ago. I mean, I know it's more in line sequentially.

(inaudible)

Okay. I know what you're talking about now, the corporate retro. Yes. I mean, that -- the corporate retro is really a capital management tool. So as we look at the premium that we expect to write for the year, we had a pretty good idea that this year the growth was not going to be as significant as it's been in prior years, and we didn't need the capital support. When we look at that retro, we're not ceding off volatility, really. I mean, obviously, it passes risk transfer and it can reduce a loss if there is one. But we certainly anticipate going into these underwriting years that we're producing a profit. And so we're ceding off profits in order to do that. So if we need that from a capital management standpoint, we'll do it. But this year, we felt we didn't need it. There wasn't really a specific that we want more of the AmTrust. We were ceding off pretty small amount to begin with, but it's obviously much smaller this year.

And I would also say, Christopher, that despite these movements, most of the underwriting years involved -- relative to AmTrust are profitable for us. So we see movements in every segment that we write at points in time. But there's nothing that gives us such discomfort that we would want lay off that business.

No. That's very helpful. A few questions on diversified reinsurance. Is the gross written premium decline there, is that, that commuted account? Is that commuted account the one -- or the commutated account, is that the one that was causing the reserve development issues?

No. Unfortunately, no. No. It was a terminated account. The one that's -- is what's causing the development is canceled, but it is not terminated.

Okay. So it will just non-renew at some future time period, right?

It did non-renew, yes. It's no longer on the books. What we're dealing with is the run-off as a liability.

Okay. And on the treaty lines, what underlying product lines were you seeing the adverse development? Because you'd mentioned one was facultative, which is definitely -- have more of a one-off. But I'm just wondering if there's any patterns you're saying in the treaty development?

Yes. What's interesting in the treaty development -- I mean, as I mentioned earlier, historically, it's been almost a single-line kind of issue, excess commercial auto. This quarter, there was -- and I say it was specific to accounts -- we had a very small amount of commercial auto. It really was benign. We saw an account that had some workers comp exposure that had seen adverse and then we had sort of the facultative issues. And we talked about 3 accounts that drove a big chunk of it. So it would have been the facultative account, it would have been the comp account and, to a lesser extent, an auto account. Commercial auto.

Great. And then just over on the reserving. Obviously, reinsurance reserving is complicated, right? You're at a level back from the ultimate -- the risk. But Maiden have taken 2 big reserve charges in the past 3 quarters, so there could be some investor concerns about the process. So what steps are you taking to improve this? And do you think Maiden may need to buy its own reserve cover on AmTrust segment, given the relative exposure it has in your book?

Well, let me take those in pieces here. First of all, we're equally concerned. And we want to ensure investors that we don't take lightly what occurred at the end of the year and what occurred in this quarter. I'd still kind of reinforce the fact that what occurred at the end of the year was significantly a function of an adverse industry-wide trend related to commercial auto. I wish, in retrospect, that we hadn't generated as much in those particular underwriting years that created the adverse. It's important to point out that -- and this was in our excess business -- if you exclude the effect of commercial auto and you look at the rest of our excess business, it's been very solidly and quite consistently profitable, even in those same years. So -- and we keep looking to see whether there's any other line that's manifesting the same behavior. And to date, we've not found it. But it's intensified our effort to get in, look at files and get comfortable with preserving and post additional case reserves when appropriate and necessary. So I think that, that issue is an issue that we've been talking about for quite some time. Our action at the end of the year was to try to be more permanently responsive to that issue. And as I say, the first quarters of this year, excess commercial auto development has not been an issue. It's been relatively benign, which gives us some comfort that, that issue is addressed.

The things that occurred in the quarter itself, as I said earlier, were kind of more account-specific issues. And we believe that our reaction to the things that we saw in the quarter are responsive. And we believe that -- our view is that prospectively, we think there's a significant opportunity; obviously, restore our profitability and continue. We can't ever make a guarantee that we'll never see adverse development.

But the other thing I will, I guess, mention is that we have a very granular process. I think it may be one of the most granular processes that any reinsurer uses. And part of it is because we're writing lower-layer business. But we -- and it starts from a very significant technical review process when we evaluate a risk that includes our underwriters, our claim people, our actuaries. But even when the account -- once the account is in the Maiden portfolio, we continue to audit those accounts. And a lot of our effort on the claim side is to develop, particularly on excess to loss business, is to identify what we believe are under-reserved cases, so that we're not caught behind. And I would say that some of the acceleration that we've seen, particularly in respect to excess commercial auto, and even in the development we've seen in the quarter, is a direct result of our very proactive process. But by acting proactively, we can mitigate the effect in future years of an adverse trend we're seeing now. And most importantly, we can help a client identify a situation that may cost them even more than it cost us. So I think in that respect, we are different.

I believe we tend to see things earlier than maybe higher excess of loss writers and even other people that write lower-layer business. So that does definitely accelerate our recognition of adverse trends.

The issues in this quarter are pretty unique to the quarter. I don't believe at this point they're systemic, and we'll keep watching it. But as we look at going forward, I think that the excess commercial auto activity has been addressed. You can never be 1,000% sure, but we believe that we have, and I think the issues in this quarter have been addressed.

In the last quarter, our diversified U.S. business had relatively modest adverse, I think it was about $4 million of movement. So it's -- the adverse that we saw in the last quarter actually was driven auto related, which this quarter looks great. And it was also related to some facultative business that had been written (inaudible).

Now moving to the adverse development cover. From our perspective, we're always investigating retrocessional approaches that we can use to, let's say, improve stability, enhance stability of performance. That said, anything we look at, we'll evaluate in the context of whether we believe it's the right move from a shareholder perspective. So we don't have anything to report at this point with respect to ADCs, but we may look at them in the future. But again, the governing criteria for us is that we believe it's shareholder friendly. I think our issues are a little bit different than AmTrust in that respect.

Right. And you mentioned -- just one quick follow up. And you'd mentioned the diversified ones are kind of like one-off, like client-specific issues, not really a big trend. Would you say that the AmTrust reserve issues are not systemic as well?

At least based on what we've seen -- I mean, we've reacted to what we've seen. It's a very large portfolio business. And if you look at the amount of development that we saw in the quarter. Let's take the Hospital Liability; that's not going to reoccur. That was a judgment decision based on external economic factors that we thought could influence settlement values on the more recent underwriting years. It wasn't heavily driven by development we're seeing, but more the environmental factors that could affect the business. The rest of it, which was less than half of the total development at AmTrust, was related to small commercial. And we'll continue to look at that on a quarter-to-quarter basis. But at this point, we believe we've been responsive to what we've seen.

Karen, anything else?

Yes. I mean, I think the tricky thing with what we're seeing on AmTrust is we know there are a lot of operational changes. And that it does make it a little trickier to figure out just how that should be affecting our actuarial projections. So we continue to evaluate that. We've made the best estimates we can make based upon the information we have. And we just -- and we're continuing to audit and to try to further understand what changes have taken place and how they'll affect our projections going forward.

And our next question comes from Matt Carletti with JMP.

A lot of my questions have been answered, but just one last one. Just around kind of -- can you give us a little color on, I guess, both on the AmTrust side and on the Diversified side, that the answers might be same or different, but kind of the process or the timing around, kind of, how you do the audits? And then where I'm going with that is, does it take a full year for you to kind of get through all the business lines and geographies at AmTrust or your clients on the Diversified side? And so maybe you reacted to some trends you saw in specific areas this quarter. Do those get spread across the broader portfolio and extrapolate it to try to encapsulate the entire relationship with AmTrust or with other diversified clients of similar lines of business? Or is it a full calendar year process? The next quarter you're looking at different pieces of AmTrust and different clients on the diversified side? And if those trends are pervasive and a different geography within the same line, then that would get reacted to separately? I'm trying to just get a feel for -- if you feel that you've kind of covered the whole book with the actions you've taken? Or if there's sort of a cycle to it and it might take a few more audit quarters to feel comfortable with everything?

I think I'm going to have turn over the Diversified discussion to Karen and have Pat talk a little bit about AmTrust. But I will tell you just before turning it over that -- I mean, from our perspective, it's a dynamic ongoing process, and it's somewhat risk-based. Larger accounts may get multiple audits in a year. If we see something developing in a particular line of business -- good example is when we started to see adverse excess commercial auto development in specific accounts, we swept out to all of the customers who we wrote excess commercial auto contracts for. And we had not just one, but multiple on-site audits, and some of them were electronic remote audits, where we tested the case reserves. And it created a lot of additional case reserve activities that we generated from that. Our belief that in some cases, there may have been a bit more optimism on a case than we believe was appropriate, so we raised our reserve. And so that process -- and you may remember how it unfolded -- as we saw it materialize in a couple of clients through audit activity and development we were seeing in the quarters, we then extended out and said, is this something systemic? It's important to recognize, Matt, that our reserve evaluation process is quarter-to-quarter on virtually every account we write. So we don't -- at this point, we don't do bulk reserving by a line of business. We look at each contract and see -- we may try to aggregate some of the pieces. But the bottom line is we're trying to understand how each client's performance develops.

And with that, I think I'll turn it over to Karen to talk about the process in Diversified.

Yes. And I think that's exactly right. We always start the year -- we've got a schedule. Here is how we're going to -- we're going to look at accounts. But then it becomes very risk-based. And part of it is we're looking at every account every quarter, and we're getting together and kind of seeing where some adverse trends may be or some favorable trends. I mean, you want to understand both. And the claims audits are scheduled then to address the risks that we're seeing. And from that basis, we're able to try to incorporate that in the actuarial reserving as quickly as possible. If it's something that is really adverse, as Art said, we'll be back multiple times. And for some accounts, we actually have the ability to look at claims online. But it just depends upon the results of those initial audits and how often the claims folks need to come back.

And I would say, Matt, that one can drive the other. So there can be a reserve -- excuse me, a claim review that proceeds quarterly reserve valuation that drives up the actuarial numbers. And so in that case, we're responding to development that we're seeing that may be under-recognized by the client, and that will trigger sort of an impact. In other cases, actuarial will just see elevation in the diagonal. And they may say -- and they will say what's causing this? And we'll go at the team level. We'll go out and we'll visit the client. And we'll try to understand what's happening. We'll certainly have discussions with them about what may be causing it, but then we'll get into a physical file review.

On excess of losses, it's a much easier process because we know every claim that's been submitted. But we also will then test some of the claims that are below the retention of the contract that we write to make sure that we're not missing anything that maybe should have been reported. But yes, both can happen. But in sort of the shortest answer possible, the answer is we're trying to do it as quickly and urgently as possible. If we go through a quarter, claim audit's clean, no issues, development's fine, there's no reason for us to accelerate audit activity. But if we see something through audit or through reserving, we will accelerate our reviews and get much more active. And to the extent we get a sense that it's line-specific or we see it coming from another account -- another client in the same line, we'll intensify across other clients that have the same line of business.

Pat, on AmTrust?

Yes. I mean, not surprisingly, Matt, I think, you'll hear that the process at AmTrust is -- there's a lot of commonality between what we do in Diversified, what we do in AmTrust. I think it's important over a calendar cycle, we'll touch every aspect of each segment of the business, at least 1x or 2x and often more, but not unlike the process that we go through with Diversified. We respond to -- a risk-adjusted basis is what we see. Art talked in the first quarter about some things that we saw relative to loss (inaudible) that allowed us to adjust our diligence requirements for the second quarter. We took a deeper dive and that will inform what we do in the third and fourth quarter. But at the same time, there really is kind of a parallel path, making sure that we're keeping an eye on all the business, corresponding to the areas that need the most needed diligence at any given point in time. And with the operational changes that we've seen, we think a lot of them are positive, but they require a lot of validation and a deeper dive on our part.

Of course. I mean, in terms of AmTrust, I know it's your largest client and like you said, you have multiple audits on all the pieces, kind of, each year. On a year-to-date basis, sitting here kind of 6 months through the year, is it safe to assume that all the major pieces of that business have been at least looked at once, kind of under -- kind of this year lens? Or are there certain pieces of business, like you mentioned (inaudible) as an example is just completely made up. But you mentioned severity in New York, right, workers comp New York. Has California workers comp not been looked at, and that's something that happens later in the year? I'm just trying to get a feel for kind of 6 months into the year since really we've kind of seen 2 rounds of adverse development, be it on different items, do you feel like you're kind of through that process, during which you've seen the changes in the claims processes and things and like that? Or are there still major pieces of the business to go in the back half of the year that haven't got that kind of routine audit in the first 6 months of the year?

I would think that we broadly touched on every aspect of the business at this point. Certainly, within the first 6 months and certainly the level -- we always continue to increase the amount of diligence that we do, number one. I think that the other thing that we also do is we have regular contact, not just with senior leadership of AmTrust, but business unit leaders as well, whether it's on the underwriting and claims side. I think the -- so as a result, it's not just kind of deep dives or kind of specific audits, but it's also kind of getting regular updates on the business. So I think we've touched each of that. Clearly, areas that are showing volatility, okay, we clearly give kind of a more immediate attention to. And I think our audit diligence activities in the first 6 months of the year have been reflective of that.

And I'm showing no further questions at this time. I would like to turn the call back to Mr. Noah Fields for closing remarks.

Thank you for joining us today. If you have any follow up, please reach out to me. And we look forward to speaking with you in the future. Have a great day.

Ladies and gentlemen, thank you for participating in today's conference. This concludes today's program. You may all disconnect. Everyone, have a great day.