U.S. Markets close in 5 hrs 17 mins

Edited Transcript of AFH earnings conference call or presentation 7-Aug-18 12:30pm GMT

Q2 2018 Atlas Financial Holdings Inc Earnings Call

ELK GROVE VILLAGE Aug 22, 2018 (Thomson StreetEvents) -- Edited Transcript of Atlas Financial Holdings Inc earnings conference call or presentation Tuesday, August 7, 2018 at 12:30:00pm GMT

TEXT version of Transcript

================================================================================

Corporate Participants

================================================================================

* Paul Anthony Romano

Atlas Financial Holdings, Inc. - VP & CFO

* Scott David Wollney

Atlas Financial Holdings, Inc. - CEO, President & Director

================================================================================

Conference Call Participants

================================================================================

* Frederick Timothy Shepard

Capital Returns Management, LLC - Analyst

* Jon Paul Newsome

Sandler O'Neill + Partners, L.P., Research Division - MD of Equity Research & Senior Insurance Analyst

* Matthew John Carletti

JMP Securities LLC, Research Division - MD and Senior Analyst

* Robert Edward Farnam

Boenning and Scattergood, Inc., Research Division - Senior Research Analyst of Property and Casualty Insurance

* William J. Dezellem

Tieton Capital Management, LLC - President, CIO and Chief Compliance Officer

================================================================================

Presentation

--------------------------------------------------------------------------------

Operator [1]

--------------------------------------------------------------------------------

Greetings, and welcome to Atlas Financial Holdings 2018 Q2 Earnings Results. (Operator Instructions) As a reminder, this conference is being recorded.

I would now like to turn the conference over to your host, Scott Wollney, Chief Executive Officer of Atlas Financial.

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [2]

--------------------------------------------------------------------------------

Thank you very much, Jeremy, and good morning, everyone. With me today is Paul Romano, our Vice President and CFO. We're very pleased to report continued favorable underwriting performance in the second quarter highlighted by a strong combined ratio, greater than 20% return on equity and book value appreciation.

I'll now turn it over to Paul to provide details about our quarterly materials and review our policy regarding forward-looking statements.

--------------------------------------------------------------------------------

Paul Anthony Romano, Atlas Financial Holdings, Inc. - VP & CFO [3]

--------------------------------------------------------------------------------

Thank you, Scott, and good morning, everyone. Yesterday, after market close, Atlas issued its second quarter 2018 financial results. Copies of this press release are available at the Investor Relations section at the company's website at www.atlas-fin.com.

On this call, Atlas may make forward-looking statements regarding the company, its subsidiaries and businesses. Such statements are based on the current expectations of the management of each entity. The words anticipate, expect, believe, may, should, estimate, project, outlook, forecast or similar words are used to identify such forward-looking information.

The forward-looking events and circumstances discussed on this call may not occur and it could differ materially as a result of known and unknown risk factors and uncertainties affecting the companies, including risks regarding the insurance industry, economic factors in the equity markets generally and the risk factors discussed in the Risk Factors section of its Form 10-K for the year ended December 31, 2017.

No forward-looking statements can be guaranteed. Except as required by applicable securities laws, forward-looking statements speak only as of the date on which they are made, and the company and its subsidiaries undertake no obligation to publicly update or revise any forward-looking statement, whether as a result of new information, future events or otherwise.

When discussing our business operations, we may use certain terms of art which are not defined under U.S. GAAP. In the event of any unintentional difference between the presentation materials and our GAAP results, investor should rely on the financial information in our public filings. All amounts discussed on this call are in U.S. dollars unless otherwise indicated.

We will be utilizing a slide show presentation in conjunction with this call. Though we may address a few slides specifically and generally -- in general, we will use this as an accompaniment. Feel free to follow along as we will follow the basic structure of the document. This presentation is available on our website's Investor Relations section and then under the Earnings Release Info selection. For those of you following along with our presentation, we'll begin on Slide 3.

With that, I'll turn the call back over to Scott.

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [4]

--------------------------------------------------------------------------------

Thanks, Paul. We continue to be pleased with our results in 2018. Our earnings per share are on track to achieve the level to which we previously guided, and we're taking advantage of the favorable competitive environment in commercial auto and our niche in particular from a rate perspective. As always, we're focused on margin expansion as opposed to top line growth and want to emphasize the importance of that dynamic. The significant investment we've made in predictive analytics is helping to optimize the value we're able to deliver as a specialist. Results from our in-vehicle technology initiatives continue to show promise and are beginning to create opportunities to generate meaningful amounts of new business at above-average level of expected profitability.

As highlighted on Slide 3, our results in the quarter include a strong combined ratio based on our measured approach to new business writings, improved net income per share and an annualized return on equity of 23.8%. We're confident that continued progress in these areas will create significant shareholder value and book value growth going forward. Paul will provide more detail on our financial results in a few minutes.

Atlas has a long-standing focus on delivering a strong value proposition as a highly specialized writer of commercial auto insurance products to the evolving niche market of people using light vehicles to move people around. Within the $2.25 billion sub-segment on which we focus, we've continued to see a shift from more traditional taxi business over to livery, limo and para-transit. We've detailed the changing landscape on past calls, and many of these same trends have continued. In short, taxi premium is down but stable, nonemergency para-transit is growing and livery/limo continue to expand with the most growth attributed to transportation network company-related activities.

Slide 4 illustrates the way in which our business mix has transitioned along with these trends. We're proud of the fact that our value proposition continues to benefit traditional operators and has also been evolved to enable us to capitalize on the disruption scene in our niche in recent years. We've maintained our focus on owner operators and small fleets, which we believe has been the most profitable sub-segment of commercial auto, but are increasingly seeing opportunities to engage larger accounts through the use of in-vehicle technologies and other ensured tech-driven aspects of our business model. To the extent that part-time transportation network and other gig economy drivers increasingly purchase commercial insurance, the size of our market could increase by as much as $500 million to a $1 billion. It takes time to cultivate these relationships but we're enthusiastic about our pipeline.

Atlas' competitive advantages remain relevant through the evolution of our existing infrastructure that provides localized and more effective customer care with the benefit of a larger integrated national platform. As shown on Slide 5, our business is geographically distributed across the country with premium generated in 42 states plus the District of Columbia. Our overall market share is approximately 12% to 13% nationwide. We have long maintained that proportionate share would be 20% of the market, which provides a long runway for potential growth and is achievable given the current hard market for commercial auto.

While our year-over-year gross written premium growth in the quarter was relatively flat, we captured market share in key areas like California and New York. We increased retention for accounts predicted to be better than average and reduced our exposure to those predicted to be worse than average from a loss perspective. Paul and I will both share more detail in this regard later in the call.

Moving on to Slide 6. Demand as measured by new business applications is trending up. Our competitive environment is currently populated primarily by smaller, local, nonstandard companies. Our comparative advantages are driven by niche expertise, customer-centric service and claims handling and other differentiators that stack up favorably to these more generalist companies. These are the things that our customers prioritize when they decide where to purchase their insurance and how much of a premium they may be willing to pay. Furthermore, there are no apparent signs of new entrants.

We have seen our hit ratios decline slightly as increased rate has been introduced during the first half of the year. In the past, when we've taken significant rate, we've also seen competitors follow price leadership and expect that to be the case in the second half of this year as well. In the meantime, we're generating a level of new business consistent with last year as illustrated by bound policy count shown in red.

The levels of price change I referenced are detailed on Slide 7. The ISO rate changes shown in blue are a good proxy for an average competitor. You can see that in all cases where ISO recommended rate changes, our increases outpaced the competition. In addition to an overall increase in average rate levels, our use of predictive analytics is helping to shift our overall business written to a larger percentage of accounts expected to generate below average losses as denoted by the graph on the right side of this page. Our goal is to maximize hit ratios for these favorable accounts and reduce the relative amount of less profitable accounts.

Slide 8 shows our in-force business, which remains near record levels, and we expect continued profitable growth going forward.

Paul and I will now discuss our financial results, and I will then share details regarding the impact of the pricing dynamic as well as a more fulsome discussion regarding market conditions and strategic opportunities.

With that, let me turn it over to Paul.

--------------------------------------------------------------------------------

Paul Anthony Romano, Atlas Financial Holdings, Inc. - VP & CFO [5]

--------------------------------------------------------------------------------

Thanks, Scott. As always, I encourage you to review our filings, our slide presentation and to reach out to Scott or myself with any questions.

Slide 10 provides some key financial highlights. For the second quarter 2018, gross premium written was $57.4 million for both the 3-month periods ended June 30, 2018 and 2017. We continue to push our average rates higher and are focusing on underwriting profit and risk selection, which may pressure top line growth in the short run. At June 30, 2018, in-force premiums increased to $270.9 million from $268.5 million at December 31, 2017.

Net income for the second quarter 2018 was $5.6 million or $0.47 per -- earnings per common share diluted compared to $5.5 million or $0.45 per diluted common share in the prior year quarter. Annualized return on equity was 23.8% in the second quarter 2018.

At June 30, 2018, our operating leverage, as measured by net written premium to combined statutory surplus, was roughly 2.2:1 as indicated in the top chart on Slide 11. Managing our operating leverage is important for both regulatory and rating agency perspectives as we continue to manage to an approximate 2:1 relationship between net written premium and statutory surplus. During the second quarter of 2018, our quota share session rates were 30% and 25% of subject written premium for the ASI Pool companies and Global Liberty, respectively. Net written premium for the 3-month period ended June 30, 2018, was $34.8 million compared to $46.3 million for the prior year period. Through the balance of this year, we will continue to see deleveraging as a result of the increase to the ASI Pool session rate during Q2 of 2018.

Slide 12 breaks out the key components of our combined ratio after accounting for the effects of reinsurance. Atlas' combined ratio was 88.9% in the 3-month period ended June 30, 2018 as compared to 86.2% in the prior year period. The loss ratio relating to claims incurred for the 3-month period ended June 30, 2018 was 61.1% compared to 60.1% for the 3-month period ended June 30, 2017. The loss ratio increased over the prior year period primarily due to the company's continued review of underwriting profitability of our product and state and higher-than-expected claim costs associated with Atlas' participation in nonvoluntary assigned risk pools.

The underwriting expense ratios for the 3-month ended June 30, 2018 was 27.3% compared to 25.6% for the 3-month period ended June 30, 2017. The ratio included $750,000 of nonrecurring expenses related to file year-end work and business system changes. Actual payment timing of expenses versus 2017 were also impacted.

Slide 13 provides both quarterly and annual combined ratio trend information in terms of its components. As always, we encourage users to look at these ratios on an annualized basis rather than focus on changes from quarter-to-quarter as there are a number of factors that can impact the shorter-term view.

Slides 14 and 15 provide an overview of our overall balance sheet strength and our cash in invested assets. We have always maintained the view of a conservative investment portfolio was important. Our primary investment objective is to protect capital to support underwriting operations. As seen on Slide 15, while the majority of our $226.4 million of cash in invested assets are invested in publicly traded high-quality fixed income securities, we do have approximately $30 million of other than fixed income investments, or OTFI.

Half of these OTFI investments are private commercial real estate-backed investments. The remainder are a combination of investments deemed appropriately conservative by our investment committee. All of these investments are audited annually and qualify as admitted assets from an insurance regulatory standpoint. We do not have concerns about the value of any of these investments. However, based on the current book value, we have made a decision to reduce the percentage of OTFI investments during the year.

In the second quarter, these investments were reduced by $1.4 million. We are continuing our effort to reduce our allocation in the OTFI investment category. In some cases, withdrawals or sales that were initiated in the first half of this year [were announced] during the first half of 2018.

With that, let me turn the call back to Scott for his concluding remarks.

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [6]

--------------------------------------------------------------------------------

Thanks, Paul. Before I continue discussing opportunities ahead of us, it's important to provide an update regarding reserves. This information is presented in the same format as last quarter's call with some additional detail based on input received during and after our Investor Day in May, which took place shortly after our last conference call.

On Slide 18, we highlight key changes from last quarter in connection with the ASI Pool companies which represent the largest portion of our reserves and have been affected most by the implementation of analytics. In particular, you can see that overall inventories have declined and more challenging areas, like Michigan, continue to represent a lesser amount of our exposure. Since year-end 2017, overall accident year 2017 and prior reserves have decreased by 33.2% while claim counts have been reduced by 58.8%.

Slide 19 identifies our overall gross loss and loss adjustment expense reserves for accident year 2017 and prior as broken out into categories that have distinct characteristics. These include older nonmodeled claims, claims in inventory which have been modeled, Michigan as a separate category, newer physical damage and other claims that have not yet been modeled and the remaining reserves available for incurred but not reported losses. We are committed to providing both transparency as well as updates relative to the benchmark expectations on a quarterly basis. This detail is laid out in the next 4 slides by reserve category.

Predictive model based case reserves, which are set using the methodology we introduced in early 2016 and expect to represent the bulk of our reserves as older claims are adjudicated, are addressed on Slide 20. In total, claims in this category continue to close with redundancy relative to case, returning reserves into IBNR. Aggregate reserves of $62.7 million were set by the model since inception and the amount paid on closure, including closures without payment, totaled $55.2 million.

Predictive model based case reserves represent the majority of our pending third-party liability claims for accident year 2017 and prior at this point. Our expectation is that overall claims in this category will close at or slightly below the case reserves predicted by our models over time. These models are refreshed regularly to refine predictive ability as well as to capture underlying loss trend information.

Claims that were too old to be modeled when our current protocol was introduced carry a factor reserve. Activity related to these claims is provided on Slide 21. As part of our year-end 2017 work, experienced members of our claims team conducted a detailed review of these well-developed claim files and created a database to benchmark expectations regarding ultimate loss costs earlier this year.

At that time, there were approximately 1,300 claims in this category. Year-to-date in 2018, 420 older clams, which were not scored by the predictive model, were closed with an aggregate paid amount of $12.7 million compared to an expected benchmark range of $7.8 million to $18.3 million. Given the age of the remaining claims where we have taken case-by-case review, they are expected to have a higher tendency for severity than our average calendar year paid severity across our book of business.

The remaining count of nonscored clams is 638 with an average allocated case and IBNR per feature of approximately $35,500 apiece. The gray bucket represents a combination of case reserve and an allocation of IBNR equal to the expected ultimate loss costs related to these claims. We are pleased to see that thus far in 2018, claims in this group are being paid well within our estimated range, and we'll continue to monitor and share with investors as we provide -- or as we make progress through the remainder of 2018.

Slide 22 outlines activities related to nonmodeled Michigan claims. This has been a challenging state for Atlas as well as other commercial auto insurers in recent years. Our overall exposure to Michigan was significantly reduced since it was identified as trending in the wrong direction and now represents less than 1% of our business and is shrinking.

Open claim inventories have been reduced to 351, of which 68.7% have been scored. All of the remaining older 160 claims were file reviewed, and as of June 30, 2018, we expect the total loss cost related to these claims to be $10.1 million and are allocating an average of $63,000 per claim in case an IBNR for the runoff of the open nonmodeled inventory, which represents a higher payout over historical averages in the state. The remaining Michigan claims that have case reserves set by our predictive model are included in the light blue section.

On Slide 23, the remaining IBNR for accident year 2017 and prior claims that have not yet been reported to us or are less than a month old, so have not yet been scored is shown at $42.4 million. Based on historic frequency trends, we should pay approximately 750 of new claims for accident year 2017 and prior going forward. We'll continue to update this information quarterly to provide clarity regarding progress.

While there is a significant focus on reserves, we also think it's important to share information about actual claim closure and severity trends. Slide 24 confirms that we continue to close claims faster than in the past and for lower calendar year paid severity amounts. Ultimately, the goal of introducing analytics in our claims process was to accomplish this, and we're pleased to see data supporting this expectation.

As claim activity appears to be moving in the right direction, market data from the Council of Insurance Agents & Brokers, which is graphed on Slide 25, demonstrates that commercial auto rates continue to be increased in general. Pricing levels are higher than they have been seen in more than a decade at this point and appear to continue to strengthen. This is expected to create buoyancy in our niche.

As shown on Slide 26, our insured vehicles in-force have been relatively flat year-over-year. We're comfortable allowing hit ratios and persistency to fall in the near term in exchange for greater underwriting margin in the longer term. For those of you who've been following the Atlas story for some time, we have seen opportunities to push rates when the variables in our niche market support doing so. We had similar instances in both 2014 and 2016. Our expectation is that this approach will create the most shareholder value in a favorable market environment. We also expect these to return to levels similar prior years as our competitors once again follow Atlas' price leadership.

Slide 27 shows that our risk selection appears to be working with claims presented continuing to be marginally lower than growth in insured vehicles.

On Slide 28, we provide additional detail with respect to price elasticity and the relative shift and expected profitability of our book. The graph on the left side illustrates the relationship between persistency in rate changes. This relatively linear relationship suggests that we are on the front-end of the supply/demand curve and truly leading our niche market.

In the upper left, you can see that where rate increases were limited to less than 20%, our persistency was between 80% and 100%. In the lower right, those are accounts receiving the biggest increases, and these are generally more challenged from a loss ratio standpoint, were much less likely to renew with us. As competitors raise rates, we would expect a more concave slope. We'll be tracking this information carefully and take a thoughtful view on price change with the objective of optimizing return on capital especially in an environment where significant [premium] Growth could necessitate greater use of reinsurance.

In the chart on the right half of the page, we show the relative persistency for renewal of below average, average and above average accounts as scored by our underwriting predictive models. The first row, labeled 1 to 400, are the bottom 1/3, while the third row, level 701 to 999, are the top 1/3. As you can see, in the past year, we retained 67% of the bottom 1/3, 87% of the middle and 91% of the best categories, respectively. As a result, a disproportionately large amount of our business is in the category predicted to be the most profitable, which should have a favorable impact on combined ratio in the future.

In terms of expectation, we always put a priority on margin over top line growth. As I mentioned previously, we expect local competitors to follow the price leadership Atlas exhibited in the first half of the year, and we do not expect the change to our A.M. Best rating to have a material impact on the quality or quantity of business we write. We have already begun the process of expanding our fronting relationship in states where there may be more risk of rating sensitivity. But other than normal seasonality, we have not seen any significant slowdown of application activity in recent weeks.

We continue to identify significant opportunities in the transportation network company-related space as well. Last year, we highlighted some of our successes in this area, and I look forward to confirming future successes as opportunity in our pipeline are realized. From an operating margin perspective, we will continue to focus on increasing rate versus exposure while maintaining operating efficiency. We will leverage the value of our core assets through partnerships of reinsurers, analytics and technology companies and other strategic resources.

Innovation is a critical aspect of our strategy, and we will continue to expand the utilization of in-vehicle technologies and our ongoing commitment to analytics across the enterprise. We are beta testing our digital usage based product designed for part-time transportation network drivers and are excited about the next-generation aspect of our business which builds on the foundation we have in our traditional niche. Provided this pilot validates the digital platform, we expect to begin distributing the insurance product itself later this year. Once that's underway, we'll begin to share more details.

We continue to expect to generate full year net income per share in excess of $2, which equates to a greater than 20% return on equity. Provided loss activities continue to behave as we've been observing, the full year loss ratio should trend in a positive direction as well. As always, we will continue to manage our business in an operationally efficient manner, and as a result maintain an expense ratio in the range we've seen in the past.

Now let's open it up for questions.

================================================================================

Questions and Answers

--------------------------------------------------------------------------------

Operator [1]

--------------------------------------------------------------------------------

(Operator Instructions) Our first question comes from the line of Bob Farnam from Boenning & Scattergood.

--------------------------------------------------------------------------------

Robert Edward Farnam, Boenning and Scattergood, Inc., Research Division - Senior Research Analyst of Property and Casualty Insurance [2]

--------------------------------------------------------------------------------

I had a question -- the interpretation on Slide 20, I just want to make sure I heard you right. So the actual paid amounts have come down. Well, the actual paid amounts have been $55 million, and the reserve you have for that was $62 million. So that extra $7.5 million is -- was redundant, and that -- you didn't recognize but you're putting that into IBNR. Is that how you dealt with that?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [3]

--------------------------------------------------------------------------------

That's correct.

--------------------------------------------------------------------------------

Robert Edward Farnam, Boenning and Scattergood, Inc., Research Division - Senior Research Analyst of Property and Casualty Insurance [4]

--------------------------------------------------------------------------------

Okay. Just remind us of the flexibility of the Swiss Re quota share. I mean, how much can you cede? And how often can you make changes to that?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [5]

--------------------------------------------------------------------------------

So the maximum we can cede is 50%, and the minimum is 5%, and we are able to elect to change that on a quarterly basis. As part of that agreement, there is a structure designed to maintain an annual relationship between net written premium and surplus at 2:1 or better. As Paul pointed out, we're slightly above that right now, but the 30% session for the ASI Pool and the 25% session for Global Liberty based on forecasts for the year will bring us below to 2:1. So at this point, we are not changing the session rates for the third quarter. But we do have the option if we wanted to increase it in future quarters.

--------------------------------------------------------------------------------

Robert Edward Farnam, Boenning and Scattergood, Inc., Research Division - Senior Research Analyst of Property and Casualty Insurance [6]

--------------------------------------------------------------------------------

It'll bring you down below the 2:1 by year-end or within what time frame?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [7]

--------------------------------------------------------------------------------

By year-end.

--------------------------------------------------------------------------------

Operator [8]

--------------------------------------------------------------------------------

(Operator Instructions) Our next question comes from the line of Bill Dezellem from Tieton Capital Management.

--------------------------------------------------------------------------------

William J. Dezellem, Tieton Capital Management, LLC - President, CIO and Chief Compliance Officer [9]

--------------------------------------------------------------------------------

A couple of questions. The first one relative to your part-time TNC offering that you're beta testing now, what can you tell us about that beta? And did we hear correctly that you are planning on rolling that out in full force later this year?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [10]

--------------------------------------------------------------------------------

Well, for -- the test currently is focusing on the technology platform itself. It is a mobile phone-based distribution, where the app, which is an Android as well as an iOS app, is effectively serving not only as the distribution platform but also the underwriting tool. And so first, we are testing that technology itself before insurance coverage is provided. So we have a beta group that was assembled, and we're confident that within a relatively short period of time, we'll be able to identify any challenges with the technology, resolve them. Once we've done that, we then plan to introduce the insurance aspect. So rather than acting as a beta group, initial users will actually be purchasing the insurance over the platform. So our intention is to launch that in Illinois only. And as we've commented on earlier calls, our goal is to limit the premium to a range of probably $3 million to $20 million in the first year. And while that seems like a pretty wide range, obviously, it's a new product. We're going to be the first to market with that. We are very comfortable that $3 million is sufficient to validate all the underwriting algorithms, and we want to limit the premium levels to below $20 million with the idea that we don't want to create too much exposure on a brand new product until we have an opportunity to validate that all the assumptions that went into underwriting and loss ratio expectations will bear out. Once we have that confidence, which will be in 2019 probably towards the end of the year, that's when we would anticipate beginning to roll it out in other states. So I wouldn't describe what we're going to do at the end of this year as a full-fledged launch. It will be a full launch of the product, but only in 1 state and with some pretty specific parameters around what we expect to do. And then again, once we validated all the assumptions that were made in terms of the underwriting components and the algorithms behind it, that's when we'd expect to start rolling it out incrementally in other states.

--------------------------------------------------------------------------------

William J. Dezellem, Tieton Capital Management, LLC - President, CIO and Chief Compliance Officer [11]

--------------------------------------------------------------------------------

And are you finding that there are -- or is there anything happening to increase driver awareness of the need for insurance? And I posed the question in the spirit of the last time that I was in Uber. The driver said his personal insurance was exactly what he needed.

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [12]

--------------------------------------------------------------------------------

I think there is more awareness. I still think there is a lack of awareness in general. However, we have partnered with a number of different resources that focus on the part-time transportation network drivers in terms of providing them with information to help them be successful. And one of the things that we're finding that they are increasingly emphasizing is the insurance aspect and the risk that is created for the drivers if they don't buy primary commercial auto insurance. The other thing that we're seeing is an increased perspective among at least some of the transportation network companies to encourage drivers to find primary insurance. And so while now I wouldn't describe that it as a push at this point, I do think that there is a bit of a nudge going on in terms of how information is shared. So our goal is to first make sure that we have a product that satisfies that need in a way that's appropriate for the drivers, is commercially viable and will be profitable for us. And at that point, we will then want to try to work with some of these different partners to try to market and promote that product. So I think the awareness is increasing, but like I said, it's certainly not across the board. But we do think that's going to continue to trend in the right direction over time.

--------------------------------------------------------------------------------

William J. Dezellem, Tieton Capital Management, LLC - President, CIO and Chief Compliance Officer [13]

--------------------------------------------------------------------------------

And then on a completely different note, you'd referenced a number of actions that you think will benefit the combined ratio. How long do you believe it'll be before that's apparent to us with the financial statements?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [14]

--------------------------------------------------------------------------------

Well, from a loss ratio perspective, as we said earlier this year, especially given the strengthening at the end of last year, we want to be conservative about our current accident year loss ratio, and we are going to want to see the prior year loss ratios validated at the end of this year before we really take any significant credit. We are carrying a loss ratio this year that's above what it was last year even after reserve strengthening, although we have been putting a significant amount of rate in the market in excess what we believe loss trend is. So all of those things point to a better loss ratio for accident year 2018, but I don't think we would choose to recognize that until some point in 2019. And again, only after going through a fulsome actuarial review at year-end which is -- has always been our practice and is common as you know in the industry, so I think that's longer-term expectation. In terms of expense ratio, there were some higher than average and nonrecurring expenses in the quarter that Paul had commented on, so we do expect that on a full year basis to run about 27%, so better than we recognized in the quarter but more consistent with kind of our trailing 12-month expense ratio for the past couple years. So that should be recognized earlier. And overall, we're comfortable with the EPS forecast that we made based on those 2 expectations.

--------------------------------------------------------------------------------

Operator [15]

--------------------------------------------------------------------------------

Our next question comes from the line of Frederick Shepard from Capital Returns Management.

--------------------------------------------------------------------------------

Frederick Timothy Shepard, Capital Returns Management, LLC - Analyst [16]

--------------------------------------------------------------------------------

Just on the investment side, you mentioned that you had started, and I think, planned to continue reducing your exposure to the OTFI investments over the course of 2018. Can you maybe provide us with some order of magnitude whether it be dollars or percentage that you hope to reduce the investment by over the course of the year?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [17]

--------------------------------------------------------------------------------

So the steps we've already taken should reduce that exposure by about 30% from where it was at year-end 2017. We may reduce it further depending on the timing and liquidation of some of the real estate investments that make up half of that OTFI investment. Although those are castling positively, and as Paul mentioned, we're very comfortable with the value we're carrying them at, they are relatively illiquid in the short run. All of those were investments expected to have a 3- to 5-year sort of time horizon, and so it will take some time for those to be liquidated. And we have definitely made the decision we do not intend to liquidate OTFI assets at a loss simply to reduce exposure to them, if we're otherwise comfortable with the investment, especially those real estate investments that are cash flowing positively and appear to be appreciating as expected. So we'll provide an update every quarter on that. But I think by year-end, the 30% magnitude should certainly be achievable and it could be more depending on the timing of some of the real estate sales.

--------------------------------------------------------------------------------

Frederick Timothy Shepard, Capital Returns Management, LLC - Analyst [18]

--------------------------------------------------------------------------------

Okay. And then, I guess, my second question just on the assigned risk business. I think it had a 1.4 percentage point impact on the loss ratio in the quarter, which you mentioned is elevated but is in line with last quarter. I was just wondering how we should think about this going forward? And are there any states in specific that might be causing more of an impact than others?

--------------------------------------------------------------------------------

Paul Anthony Romano, Atlas Financial Holdings, Inc. - VP & CFO [19]

--------------------------------------------------------------------------------

Yes, it's very hard to peg an estimate on that, Fred. I would put an expectation kind of on a run rate perspective. I think if you utilize what we've done already this year and just kind of forecast that out for the balance of the year, I think you'd be safe.

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [20]

--------------------------------------------------------------------------------

Yes, I mean part of what's driving that, of course, is the fact that commercial auto, in general, has seen a lot of development, and so you're seeing 2 things. As rate increase, more risks are going into these assigned risk pools. And of course, all commercial auto writers are required to take a proportion amount of that exposure. I think the good news is some of the more challenging states that we've highlighted like Michigan are also challenged in terms of the assigned risk result. And as we write less business in those states, the assignments we receive will also be less. So that should trend it in a better direction, but as Paul said, it's hard to forecast because you effectively get a bill at the end of the quarter from the various assigned risk pools. So there's not much that anybody can do to manage exposure there other than reducing primary exposure in those more challenged states, which should then ultimately reduce the level of assignments that you receive as well.

--------------------------------------------------------------------------------

Operator [21]

--------------------------------------------------------------------------------

(Operator Instructions) Our next question comes from line of Matthew Carletti from JMP Securities.

--------------------------------------------------------------------------------

Matthew John Carletti, JMP Securities LLC, Research Division - MD and Senior Analyst [22]

--------------------------------------------------------------------------------

Just going back to that Page 20, question is -- I believe that's closed claim counts that we're looking at there. And so my question is kind of the open inventory, by those 3 categories that you show kind of with varying sizes, how would that population look? Is it more heavily in the -- between 25 and 50 and above 50 buckets because the less than 25s tend to close out first much more quickly, they tend to be easier? Or how did that mix look on kind of what remains in inventory?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [23]

--------------------------------------------------------------------------------

So keeping in mind that we are only 6 months out from the end of 2017. There is still a reasonably good mix given that most of our claims get reported after -- within 12 months following an accident year. So it has been the case that we've closed more claims under the $25,000 amount as what you see there. But in general we'll get a lot more those claims as well. We will see about 75% of our claims close for an average pay severity of somewhere between $1,000 and $1,500, so there's a big percentage of small claims that are going to get paid. And then obviously, a smaller percentage of bigger claims which was really at the heart of why we focused on using analytics to try to identify those claims in the range of $10,000 to $50,000 in the first place. It's a relatively small percentage of our overall claims, but claims in that range will be about half the dollars. So from a distribution standpoint, I think we'll see a slight increase in the percentage of claims that are predicted to be more severe as a run rate because the inventory -- as the inventory grows, we will be settling out the smaller claims generally earlier and be carrying an inventory older claims that will tend to have a higher predicted severity and probably a higher actual severity. And so that is also part of why we had [backed] that as we go forward, what we pay is going to be closer to the overall predicted severity, where early on there was a lot of redundancy in those predictive model based case reserves. But the process is designed to predict ultimate. It's not designed to have embedded redundancy. So while there has been embedded redundancy in the first couple of years in those case reserves, we do expect that to migrate to effectively sufficient. And so that's the expectation. Part of that is the balance of newer claims versus older claims and the related severities you highlight. And part of it is as we continue to refresh and refine the model, it's going to get tighter and tighter. And so that's the -- that will be by design. So that's the thing that we're monitoring and making sure that we're comfortable with. Did that get to the heart of your question?

--------------------------------------------------------------------------------

Matthew John Carletti, JMP Securities LLC, Research Division - MD and Senior Analyst [24]

--------------------------------------------------------------------------------

Yes, that was helpful. And then my other question just kind of looking at accident year loss ratios this year, can you help us through -- keeping in mind that kind of 2 -- north of $2 or $2 or better guidance, it seems like accident year loss ratio is the biggest lever and that to get there, there's probably going to have to be some improvement on that metric beyond kind of just maybe some volatility with the assigned risk and things like that. And so can you walk us through how you're thinking about that? I mean, you've been getting really good pricing. Is that really what you expect to start showing through and bringing the ratios down? Or am I thinking about that the wrong way?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [25]

--------------------------------------------------------------------------------

Well, certainly, pricing is a big part of it. And obviously, we started increasing rate in the first half of the year. The relative risk selection that we highlighted that we're seeing in our renewal book should also have a favorable impact. So the fact that we have shifted proportionately, as I highlighted earlier in the call, where we renewed 91% of accounts predicted to be better than average loss ratios, and we did not renew almost 40% of accounts predicted to be worse should also provide a positive shift in that respect, so that should help. And again, we have taken a position that we elected a carried loss ratio that was slightly higher than what our outside actuaries believe last year's fully developed loss ratio will be, and so when we feel like the actual loss information reinforces that, especially at the end of the year, we are going to be more likely to want to take credit for that, where right now we want to maintain as conservative of an approach as we think is appropriate. The other thing I would suggest is we are seeing what looks like a 10% to 25% reduction in paid severity for those claims in the $10,000 to $50,000 range, which we believe is the actual outcome of using the predictive analytics to route those claims and triage them earlier and effect better outcomes. Settle the meritorious claims ideally for less early on and then defend more vigorously, more effectively the claims that should be because there's either a lack of causality or the claim amount is inflated. And so if we continue to see that track through, that is also going to benefit because right now, we're not effectively taking credit for that improvement in severity in our carried results. So there is a couple of different moving parts. But fortunately, the data that we're observing and that we shared on the call today all support the fact that those things seem to be moving in the right direction. Well, the other comment I would just make is even with the relatively flat gross written premium in the second quarter, we do expect earned premium to continue to grow based on the fact that we have a significant unearned premium reserve, and we do anticipate that we'll see some growth in the second half of the year as well. So obviously, that earned premium growth at the same combined ratio is also going to increase EPS.

--------------------------------------------------------------------------------

Operator [26]

--------------------------------------------------------------------------------

Our next question comes on line of Paul Newsome from Sandler O'Neill.

--------------------------------------------------------------------------------

Jon Paul Newsome, Sandler O'Neill + Partners, L.P., Research Division - MD of Equity Research & Senior Insurance Analyst [27]

--------------------------------------------------------------------------------

The only question I have left is about the fronting agreements and both the possibility of them being used more widely as well as is what the incremental costs might be if they are used more widely.

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [28]

--------------------------------------------------------------------------------

Sure. So there is a handful of states where we've been actively using that, in particular states where there are airports, for example, that had an A- rating requirement, like some of the smaller airports in California. And so we identified 3 additional states where we think there could be some potential to capture more profitable business if we're writing it on A- paper or A paper as opposed to our current ratings. And so we want to be able to do that. And in terms of cost, there is obviously a fronting fee that will be incorporated into the pricing that we introduce, so we would not expect it to reduce operating margin. It would cause the product to be higher priced where an A rating is going to be important or necessary for the insureds, but that is the way we would structure the price. If we're not able to absorb the fronting fee, we would not write the business. And in the current market environment, we are comfortable we would be able to do that because the market again has been very hard, and there are very few, if any, options to obtain primary coverage for our niche customer with A paper right now.

--------------------------------------------------------------------------------

Jon Paul Newsome, Sandler O'Neill + Partners, L.P., Research Division - MD of Equity Research & Senior Insurance Analyst [29]

--------------------------------------------------------------------------------

If you found the ratings more important broadly speaking, how flexible is the use of that fronting fee beyond just this initial 3 that you've identified? Is it pretty easy to ramp up the fronting agreement in your opinion?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [30]

--------------------------------------------------------------------------------

Well, I think the fact that we have an existing agreement with the a partner, who are comfortable and committed to working with us, I think will make it very easy for us as we go forward. That arrangement took a number of months to establish, and then it took almost a year to ultimately implement in terms of rate filings and systems changes and everything else. So I think given that it is an existing strategic arrangement that we have, we have a good amount of flexibility in terms of speed to market if we did feel that, that was important or valuable or if we could expand that in a way that would be margin-expanding for us. I think if we were starting from scratch, it would be much more challenging from a timing perspective at least.

--------------------------------------------------------------------------------

Operator [31]

--------------------------------------------------------------------------------

Our next question comes in the line of Bill Dezellem from Tieton Capital Management.

--------------------------------------------------------------------------------

William J. Dezellem, Tieton Capital Management, LLC - President, CIO and Chief Compliance Officer [32]

--------------------------------------------------------------------------------

I just want to do a quick follow-up relative to your in-vehicle technology that you're implementing. Would you give us a more detailed update than you did in your opening remarks, please?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [33]

--------------------------------------------------------------------------------

Sure, I'd be happy to, Bill. I mean, we've highlighted a couple of different partnerships that we have with in-vehicle technology companies. They played incremental roles in terms of our securing, especially some of the larger accounts that are transportation network company related last year. And so we focused on some of those programs as sort of a pilot in 2017 to make sure we really understood the expected impact on loss cost. What we're finding is that we've been able to achieve as much as a 35% reduction in loss costs through the active use of in-vehicle technologies, especially those that help us identify distraction before there are accidents where we can then work with those accounts to actively manage and train the drivers. And so with that information, we've been able to begin marketing, to our agents, a program where we can allow credits for the use of those technologies for accounts that have the right risk profile and have a commitment to receive and use the information that helps us to manage distraction. So obviously, we're -- we are limiting the amount of credit we give to the minimum benefit, we think, the account could get and then if a greater benefit or more expected benefit is realized, obviously, both we and the account would benefit for future renewals. And so we have been much more actively promoting that this year where last year we had a fairly limited sort of pilot by design. And so the result has been very favorable, especially in this hard market environment. Insureds are much more open than they would be in a softer market to making a commitment to manage their driver pools or to utilize in-vehicle technologies, even invest in those technologies with the expectation that it will enable them to achieve insurance at a price that they think is more reasonable. And so it is a scenario where everyone can win. We can achieve a better loss ratio. They can pay less premium. And now that we've really been able to validate certain technologies as being key and effective, we're really wanting to promote that. So today, approximately 20% of our overall insured customers have some in-vehicle technology that we believe is effective. And where we are comfortable, we can either help to -- help them prevent losses or ultimately pay less when there are crashes because of the data and potentially video that we get from that equipment. Our goal is obviously to increase that percentage, and we think that the marketing strategy we launched earlier this year is ultimately going to help us to do that. And as I highlighted in my comments where we think we might have the most opportunity to leverage those partnerships are larger fleets or leasing companies, especially those that are growing significantly or have grown as a result of transportation network company expansion across the U.S.

--------------------------------------------------------------------------------

William J. Dezellem, Tieton Capital Management, LLC - President, CIO and Chief Compliance Officer [34]

--------------------------------------------------------------------------------

You had mentioned that approximately 20% of your fleet today has some of this technology. What was that number at the end of 2017? And given the trend line that you're experiencing, where would you anticipate it to be at the end of '18?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [35]

--------------------------------------------------------------------------------

So it -- end of 2017, I would estimate that number to be -- because I don't have this exact number in front of me, probably around 15%. So it has grown in the first half of the year, again, I think a lot because of the active promotion that we've undertaken. I don't have a good forecast for where that will be at the end of this year. But again, it is moving in the right direction, and it's something that we think is going to have a dramatic impact in terms of being able to mitigate losses going forward, especially preventing losses as a result of distracted driving, which obviously was a key element we focused on -- at our recent Investor Day. And I think really we're able to demonstrate the impact that these technologies can have.

--------------------------------------------------------------------------------

William J. Dezellem, Tieton Capital Management, LLC - President, CIO and Chief Compliance Officer [36]

--------------------------------------------------------------------------------

Given that you are now promoting the in-vehicle technology this year whereas you weren't last year, would it be a fair characterization that it is gaining momentum now with your insureds? And as a result, if you were to have gained the 5 percentage points going from 15%to 20%, that it would be more than that in the second half of the year? Or should we be thinking about it in some other way?

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [37]

--------------------------------------------------------------------------------

Well, I think it's certainly gaining momentum in terms of interest, but it is important to keep in mind it takes some time for the technology we rolled out in a particular fleet just because of the physical nature of it. And of course, an interested account is also generally going to wait till renewal to make a commitment. Because although we are getting credit for the use of the technology, we are also expecting them to make a financial commitment to invest in the technology as well because we found that, that is where you get the most benefit. So there is definitely more interest in terms of whether you will see that pick up at the same level or more. By the end of this calendar year, it's harder for me to say because of those 2 components, but we are absolutely getting a positive response from our agency base and also from, especially larger potential customers. So I do think the -- we will see increased adoption in the near future. I'm just hesitant to commit to a number for the end of this year because there are a number of underlying variables that are really not something we can control. But the interest is there. I think the way we've been able to use the technology to differentiate relative to competitors is compelling, and we are definitely seeing momentum in terms of activity. So it will translate into increased penetration of the technology certainly over the next, call it, 18 to 24 months.

--------------------------------------------------------------------------------

Operator [38]

--------------------------------------------------------------------------------

Ladies and gentlemen, we have reached the end of the question-and-answer session. And I would like to turn the call back to management for closing remarks.

--------------------------------------------------------------------------------

Scott David Wollney, Atlas Financial Holdings, Inc. - CEO, President & Director [39]

--------------------------------------------------------------------------------

Great. Thank you very much, Jeremy, and thanks, everyone, for joining our call today. We look forward to talking with you again in the future.

--------------------------------------------------------------------------------

Operator [40]

--------------------------------------------------------------------------------

This concludes today's conference. You may disconnect your lines at this time. Thank you for your participation.