Selective Insurance Group, Inc. (SIGI) Q4 2019 Earnings Report, Transcript and Summary

Good day everyone, welcome to Selective Insurance Group's Fourth Quarter 2019 Earnings Call. At this time, all participants will be on listen-only mode until the Q&A portion of the call. [Operator Instructions]At this time, for opening remarks and introductions, I would like to turn the call over to Senior Vice President, Investor Relations and Treasurer, Rohan Pai.

Good morning everyone and welcome. We're simulcasting this call on our website selective.com and the replay will be available until March 2, 2020. Our supplemental investor package, which includes GAAP reconciliations of any non-GAAP financial measures referenced today also is available on the Investors page of our website.Today, we will discuss our results and business operations using GAAP financial measures that are also included in our filings with our annual, quarterly and current reports filed with the US Securities and Exchange Commission. Non-GAAP operating income, which we use to analyze trends in operations and believe makes it easier for investors to evaluate our insurance business. Non-GAAP operating income is net income, excluding the after-tax impact of net realized gains or losses on investments, unrealized gains or losses on equity securities and debt retirement costs related to our early redemption of debt security in the first quarter, and statements and projections about of our future performance. These forward-looking statements under the Private Securities Litigation Reform Act of 1995 are not guarantees of future performance and are subject to risks and uncertainties. For a detailed discussion of these risks and uncertainties, please refer to our annual and quarterly reports filed with the US Securities and Exchange Commission. You should be aware that Selective undertakes take no obligation to update or revise any forward-looking statements.On today's call are the following members of Selective's executive management team. Greg Murphy, Chief Executive Officer; John Marchioni, President and Chief Operating Officer; and Mark Wilcox, Chief Financial Officer.And now, I will turn the call over to Greg.

Thank you, Rohan, and good morning. I'll make some introductory remarks and focus on some high level themes and initiatives that enhance our strategy, positioning us well to continue to generate superior performance. Mark will then discuss our financial results and John will review our insurance operations in more detail, providing additional color on key underwriting and strategic initiatives. We're extremely proud of our stand out 2019 fourth quarter and full year results, reflecting continued strong execution across our underwriting and investment functions. For the quarter, we generated record non-GAAP fully diluted operating earnings per share of $1.37, which translated to an annualized operating ROE of 15.2%.Our combined ratio was an exceptional 91.8% and after-tax net investment income was up 6% to $47 million. The fourth quarter capped off an overall stellar year for the company in which we generated strong premium growth of 7% and a solid combined ratio of 93.7%, with underwriting operations contributing 6.5 points of ROE. Our investment results were excellent with after-tax net investment income increasing 13% and contributing 9.1 points of ROE. For 2019, our overall non-GAAP operating ROE was 13.3%, which exceeded our 12% financial target. In addition, the change in unrealized after-tax gains on our available for sale securities of $169 million or $2.83 per share, reduced our operating ROE by 60 basis points. For 2020, these unrealized gains will reduce operating ROE by about 100 basis points and that will reverse as these securities age.2019 was a standout year for us in many respects with some of the operational highlights including: net premium written growth of 7% versus 6% expected for the industry average; delivering our sixth consecutive year of double-digit ROEs, which places us among an elite group of peers that have generated similar results; issuing $300 million of 30-year senior notes in our first ever institutional debt offering, which significantly increases our flexibility and provides access to attractive long-term capital markets; being recognized for our superior operating and financial performance by rating agency AMBest [ph] who put A financial strength ratings on positive outlook in October; continuing to make progress on strategic initiatives such as geo expansion and delivering a superior omnichannel customer experience; making our communities safer through the dissemination of recall notification and alerts, coupled with our efforts to reduce distracted driving with our Selective Drive product offered free of charge to our commercial lines customers; receiving recognition as one of America's best midsize employers by Forbes; and building a solar energy farm at our corporate office, which will annually generate approximately 4 million kilowatts of energy, highlighting our commitment to our communities and society as a whole.Starting off 2020, there are a few specific topics I'd like to comment on. First, we're pleased with the direction of industry-wide standard commercial lines pricing which continues its upward trajectory and should only increase further in 2020. For Selective, fourth quarter overall renewal pure pricing reached 3.8%, which we expect will clearly establish a pricing floor for 2020. Our sophisticated modeling and underwriting tools allow us to administer price increases at an extremely granular level, while balancing retention and growth rates. We've been steadfast in our commitment to maintaining underwriting discipline, having consistently implemented commercial lines renewal pure price increases that have matched or exceeded loss trends. Renewal pure price versus expected loss trend is the primary indicator of future underwriting performance. For the past five-year period, our compounded commercial lines renewal pure price was about 16.5%, more than twice the Towers Watson CLIPS pricing of 8%. As we look forward to the coming year, we are well positioned with expected renewal pure price, in line with loss trend and excellent book of business and the sophisticated tools, technology and best-in-class people to execute on our strategic plans.Second, there's been considerable industry focus on the potential for rising loss trend. We certainly are keeping a close eye on overall loss trends, especially with respect to litigation rates and settlements. Estimated pure premium or loss costs are broken down into two principal measurements: one, estimated ultimate claim counts; and two expected average severity. Over the past two years, our expectation for overall loss trend, which is embedded in our loss picks has increased from approximately 3% to 4%. In addition, we maintain a very disciplined reserve methodology in position with a process that includes quarterly detail ground-up internal actuarial reviews as well as semi-annual reviews conducted by an independent Big 4 accounting firm. We are less susceptible to headline type injury awards as our book of business is focused on smaller account business and lower hazard classes. Our average commercial lines account size is 12,000 and 87% of our casualty policies, excluding workers' compensation, have limits of $1 million or less.We also purchased reinsurance protection that limits per event exposure on casualty policies to $2 million, that reduces our exposure to long-term medical inflation. Inflation, even social inflation is not a new concept. The best way to protect your book's performance is by achieving consistent renewal pure price over time. Selective has consistently done this over many years.Third, while catastrophe losses were relatively moderate in the fourth quarter of the year, the risks of large unexpected losses remains pronounced. Climate-related risks are leading to generally unpredictability of catastrophe events such as hurricanes, convective storms, floods and wildfires. For the year, our total catastrophe loss ratio was 3.1 points, which was below our expectation of 3.5 points. In an effort to limit our exposure, our catastrophe reinsurance program protects us on a 0.4% event probability to only a 5% impact on stockholders equity. Non-catastrophe property losses have also been consistently elevated for the past few years. As an industry, we need to do more to manage property risks appropriately, including through mitigation initiatives, risk sharing, more diligent underwriting and consistent risk-based pricing.Finally, the prolonged low interest rate environment will continue to put downward pressure on industry-wide investment portfolio returns and consequently ROEs in the coming year. This should result in a greater impetus for the industry to improve underwriting results in order to generate adequate returns. With our pricing sophistication and agile execution capabilities, we are well positioned to benefit from commercial lines' pricing tailwinds and feel very confident in our ability to maintain attractive ROEs.Turning to 2020 expectations, our guidance for the year is based on our current view of the marketplace and incorporates the following. One, a GAAP combined ratio, excluding catastrophe losses of 30 -- about 91.5%. This assumes no prior-year development. Two, catastrophe losses of 3.5 points. Three, after-tax investment income of $185 million, which includes $14 million of after-tax investment income from our alternative investments. Four, an overall effective tax rate of approximately 19.5% which includes an effective tax rate of 18.5% for investment income reflecting a tax rate of 5.25% on tax-advantaged municipal products and tax rate of 21% for all other items. Five, weighted average shares of 60.5 million on a diluted basis.Now, I will turn the call to Mark to review the results for the quarter.

Thank you, Greg, and good morning.For the quarter, we reported $1.36 of fully diluted earnings per share and $1.37 of non-GAAP operating earnings per share, both of which are company records. We generated a very strong annualized ROE of 15.1% at a non-GAAP operating ROE of 15.2%. For 2019, our annualized non-GAAP operating ROE of 13.3% was above our 12% target. The full-year operating ROE was reduced by about 60 basis points [ph] due to the significant after-tax net unrealized gains on fixed income portfolio that increased GAAP equity by $169 million or $2.83 per share. These gains reflect the low interest rate environment. Each year, we establish an operating ROE target that is based on at least the 300 basis point spread over our weighted average cost of capital, our outlook for interest rates, and overall P&C insurance market conditions. For 2020, we've established a non-GAAP operating ROE target of 11%. The lower target is principally a function of our lower estimated weighted average cost of capital and the lower interest rate environment, which has put pressure on investment yields, and has also increased GAAP equity.Consolidated net premiums written increased 8% in the quarter with excellent 11% growth in our Standard Commercial Lines segment, driven by strong new business growth and retention and accelerating renewal pure price increases. This was partially offset by premium declines in Personal Lines and E&S. Underwriting profitability remained strong with a fourth quarter combined ratio of 91.8%, driven by low level of catastrophe losses in our footprint and favorable casualty reserve development. On an underlying basis or excluding catastrophe losses and prior-year casualty reserve development, our combined ratio increased to 93.8% driven by an elevated fourth quarter expense ratio and some modest increases to 2019 accident year loss effects, which I'll touch on more in just a minute.For 2019, consolidated net premiums written increased 7% with strong contributions from our Standard Commercial Lines and E&S segments. Our reported combined ratio was highly profitable at 93.7% and our underlying combined ratio was an excellent 92.9%, which reflects 20 basis points of underlying margin improvement in 2019. Despite delivering a very profitable combined ratio in 2019 of 93.7%, which was almost 2 full points ahead of our 95.5% in combined ratio forecast going into the year, the underlying combined ratio ended the year at 90 basis points above expectations. This was in part driven by the excellent calendar year loss ratio that drove the expense ratio up by 30 basis points above expectations, due to profit-based compensation with the remainder largely driven by some modest fourth quarter adjustments to 2019 casualty loss ratio picks. For the year, the impact of these additions to 2019 loss ratio picks was $14.5 million or 60 basis points [ph] and for the quarter, the impact was $11.9 million or 1.8%.Catastrophe losses, which is a reminder related only declines specifically related to catastrophes designated by PCS were modest in the fourth quarter and impacted the combined ratio by 1 point, which is better than expected, while non-cat property losses resulted in a 15.1 point impact and were also slightly lower than expected. For the year, catastrophe losses accounted for 3.1 points on the combined ratio which was better than our annual expectations of 3.5 points while non-cat property losses of 15.8 points came in about 10 basis higher than expected for the year. Our expectations for catastrophe losses remain unchanged for 2020 at 3.5 points. In the fourth quarter, we experienced $20 million of net favorable prior-year casualty reserve development, driven by $35 million of favorable development in workers' compensation line and partially offset by $5 million of development in general liability, $4 million in commercial auto, $4 million in personal auto liability and $2 million in E&S segment. The impact of net favorable prior-year casualty reserve development was 3 points on the combined ratio for the quarter and 2.3 points for the year.Moving to expenses, our expense ratio came in at 34.1% for the quarter and 33.8% for the year. The increase of 60 basis points for the year was principally driven by higher profit-based compensation for our distribution partners and employees, driven by our excellent underwriting results. We expect to pay out a record level of agency supplemental commissions for the 2019 year. We expect some modest expense ratio improvement in 2020, which is reflected in our underlying 91.5% combined ratio forecast. Over the next few years, we believe we can continue to improve our operational efficiency and drive down our expense ratio while also still making significant investments in developing our people, improving our underwriting capabilities, enhancing customer experience, continued product development and geographic expansion and other investments we feel important to manage the Company for the long term.Corporate expenses, which are principally comprised of holding company costs and long-term stock compensation totaled $2.6 million in the quarter compared with $3.4 million in the comparative quarter with the decrease principally driven by a decline in our soft price. For the year, corporate expenses totaled $31 million compared to $25 million in 2018 and included $3 million of one-time items, principally related to severance.Turning to investments, for the quarter, after-tax net investment income of $47 million was up 6% from the comparative quarter. For the year, after-tax investment income was up -- of $181 million was up 13%. The improvements at both periods was driven by active portfolio management and excellent cash flows, which when combined with the net proceeds from our senior note offering in March helped drive our invested asset base higher. This was all partially offset by the lower interest rate environment and a contraction in credit spreads that put pressure on new money purchase yields. The overall after-tax yield on the fixed income portfolio, including high-yield bonds averaged 2.9% for the year. The average new money yield on the fixed income portfolio during the year was 2.7% after-tax, although we've now seen five sequential quarterly declines with the fourth quarter coming in at 2.4% after-tax. In addition, we've managing down our floating rate securities, which now represent approximately 12% of our fixed income portfolio, which is down from a peak allocation of 18%. Despite the decline in LIBOR over the last year, we still find the all-in yield of these securities very attractive on a comparative basis with similar fixed rate securities. All in all, the pre-tax book yield on our core fixed income portfolio decreased 5 basis points in the quarter and was down 14 basis points for the year.On a go-forward basis, we expect continued pressure on our book yield given the low interest rate environment. However, given our strong expected cash flow, our 2020 after-tax net investment income guidance of $185 million reflects the modest growth from 2019, although a lower ROE contribution given the growth in stockholders' equity. Our average fixed income credit rating remained strong at AA minus and the effective duration of our fixed income and short-term investment portfolio is at the low end of the range at 3.6 years. Overall, we continue to have the portfolio conservatively positioned. Risk assets, which principally include high-yield fixed income securities and alternative investments portfolio, accounted for 8% of total invested assets as of the end of the year and remain underway our longer-term target. Our alternative investment portfolio, which includes limited partnerships in private equity, private credit and real asset investment and reports on a one-quarter lag, generated a pre-tax gain of $18 million for the year, which was in line with 2018.Tuning to capital, our balance sheet remained very strong with $2.2 billion of GAAP equity, an increase of 22% for the year. Adjusted capital ratio was 20.1% at year-end, which is well below our target, and provides us with financial flexibility. We continue to operate at the low end of our premiums to surplus target range of 1.4 times to 1.6 times. This flexibility at the operating level, when combined with our $278 million of holding company liquidity, provides us with meaningful capacity to grow if market opportunities present themselves. At our 1.4 times operating leverage, each combine ratio point equates to just under 1 point of ROE. In addition, our 3.05 times investment leverage means that each point of pre-tax book yield on our investment portfolio results in approximately 2.5 points of ROE.With regards to our reinsurance program, we enjoyed a successful renewal of our catastrophe program on January 1. We maintained our existing structure that keeps 1 in 100 or 1% net probable maximum loss of P&L from a major catastrophe risk, US hurricane, at a very manageable 2% of GAAP equity and 1 in 250 nets the amount of 0.4% probability at 5% of GAAP equity. We also renewed our non-footprint catastrophe program that drops our retention from $40 million to $5 million for our five new expansion states as well as our E&S states outside of our original 22 state footprint, which includes states such as Florida, Texas and California. Pricing on cat program reflected the loss free status of our account and our continued efforts to generate strong renewal pricing in our property portfolio and continued efforts to diversify our exposure. As a reminder, our reinsurance program also includes access to loss agreements, which limits the impact to us of individual loan losses to $2 million for both property and losses. Individual occurrences about $2 million are ceded under the effect of the loss agreements.With that, I will turn the call over to John to discuss our insurance operations.

Thanks, Mark, and good morning.Let me begin with a discussion of full-year results of our operations by segment and then provide you with an overview of some of our strategic initiatives. Our Standard Commercial Line segment, which represents approximately 80% of premiums, generated 8% net premiums written growth for the year, continuing a consistent track record of strong and profitable growth. Segment generated new business growth of 8%, stable retention of 83%, renewal pure price increases of 3.4% and an excellent combined ratio of 92.9% or 93.7% on an underlying base. Commercial Lines renewal pure price increases -- increased 3.8% in the fourth quarter, up sequentially from 3.5% in the third quarter. We are happy with the overall direction of market pricing while it is worth noting that the increases so far have been greater for larger accounts and specialty risks than for the small and mid-market standard lines accounts that make up the majority of our book. We pride ourselves on managing our renewal pricing strategy in a highly granular fashion, closely monitoring rate and retention by cohort of expected profitability. Our analysis uses a point of renewal retention measure that removes policies that canceled prior to expiration, as we think that's the best indicator of the effectiveness of our pricing strategy.On our highest quality Standard Commercial Lines accounts, which represented 49% of our commercial lines premiums, we achieved renewal pure rate of 2.1% and point of renewal retention of 91%. On the lower quality accounts, which represented 11% of premiums, we achieved renewal pure rate of 7.8% while retaining 80% at point of renewal. Our ability to analyze the risk and return characteristics of each renewal policy at an extremely granular level, allows us to achieve additional loss ratio improvement through mix of business changes while maximizing overall retention.Drilling down to the results for the year by commercial line of business, our largest line, general liability, achieved an 89.6% combined ratio which included favorable prior year reserve development totaling $5 million or 0.7 points. We will continue to closely monitor this line for frequency and severity trends, including litigation rates, which have been relatively stable in recent quarters. For the year, we achieved renewal pure price increases of 2.2% for this line, excluding umbrella and expect firming pricing environment, heading into 2020. Our workers' comp line generated a 74.1% combined ratio, aided by favorable reserve development totaling $68 million, which accounted for 21.8 points on the combined ratio. This favorable development related primarily to lower-than-expected severities for accident years 2017 and prior. Renewal pure pricing was down 2.8% and we continue to take a cautious approach to underwriting this line as market pricing remains aggressive. Commercial auto continues to produce disappointing results for us and the overall industry as elevated loss trend consumes earned rate increase. Our combined ratio for this line was 107.9%. While we added $4 million of reserves in both the current and prior accident years, loss trends have remained generally in line with expectations. Price increases averaged 7.5% in 2019, on top of similar price increases in each of the prior two years.We've been actively managing new and renewal portfolios in target business segments and improving rating and classification at an individual account level. Our commercial property book generated a 93.9% combined ratio or 7.1 point improvement from 2018. Lower levels of bulk catastrophe and non-catastrophe losses drove the improvement. While market pricing has improved loss trends remain elevated and indicate the need for additional rate level. Our renewal pure price increases averaged 4% excluding in the marine and we are taking steps to address the drivers of the higher loss experience through business mix shifts and safety management efforts.Our Personal Lines segment, which represented 11% of 2019 premiums, reported a 2% decline in net premiums written, mostly reflecting the more competitive market conditions in personal auto. Renewal pure price increases, averaged 5% and retention remained solid at 83%. New business however was down 21% for the year. Market competition in personal lines, particularly for personal auto has become far more pronounced this year. The segment produced a combined ratio of 97.3% or 88.6% on an underlying basis. While maintaining adequate profitability remains a core focus, we will be doing more in the areas of product customization, modeling enhancements and distribution to ensure we are optimizing our position in this segment.In personal auto, net premiums written declined 1% for the year and the combined ratio was 106.4%. Results included $6 million of strengthening for prior accident year casualty reserves, which added 3.5 points on the combined ratio. Renewal pure price increases averaged 8.9% for personal auto liability and 3.8% for physical damage. While benefit to improved profitability of these price increases are clearly putting pressure on new business, with market pricing under pressure and the favorable impact of loss trends diminishing, we would expect deteriorating results for the industry in this line and less pricing picks up again.The homeowners line reported 2% premium decline relative to a year ago and combined ratio of 96.5% that included 15.5 points of catastrophe losses. As the majority of our premium is written on an account basis, our competitive positioning in the auto line has hurt our homeowners' growth. Renewal pure price increases averaged 3% for the year.Our E&S segment, which represented 9% of total premiums, generated 4% net premiums written growth in 2019. The premium decline in the fourth quarter related primarily to our decision to exit the snow removal business, which we discussed on last quarter's call. The segment generated a 95.9% combined ratio for the year compared to 100.3% in 2018. Overall renewal pure price increases averaged 4%. Over the past few years, targeted price increases, business mix changes and exiting specific underperforming classes of business have contributed to the improved combined ratio performance in this segment. Our E&S book consists primarily a small account business with a similar risk profile to our standard lines. While market pricing is increasing, it is more muted for the lower hazard a business we write than for riskier classes that we don't, such as large account casualty for coastal property.Let me now switch to some of our strategic initiatives, which continue to drive our strong operating and financial performance. We are very proud of our accomplishments in 2019 and focused on the following four objectives as we move through 2020. One, using our sophisticated pricing tools to achieve standard commercial lines renewal pure price increases that are in line with expected loss trend. Two, expanding share of wallet with our existing distribution partners and strategically appointing new partners. Three, capitalizing on the investments we've been making to deliver a superior omni-channel customer experience. And four, identifying and addressing opportunities to enhance operational efficiencies and reduce our expense ratio over time, through process redesign and technology enhancements. We remain confident in the overall price adequacy and embedded profitability on our book of business. We continue to achieve price increases where appropriate, targeting accounts on a granular basis that are not meeting our profitability expectations. The level of commercial lines renewal pure price increases we are seeing increase during the course of 2019 and 3.8% achieved in the fourth quarter, are tracking in line with our expectation for loss trend. If industry pricing continues to move higher as we expect, we are well positioned to further improve underwriting margins.Second, our superior distribution relationships and sophisticated underwriting tools are clear competitive advantages. We will seek to capitalize on them as we execute on our strategy to generate profitable growth in the coming years. Our stated long-term objective is to obtain a 3% commercial lines market share. That objective is built around appointing partner relationships that control approximately 25% of their markets and achieving an average share of wallet of 12% across those relationships. We have an additional commercial lines premium opportunity over time in excess of $2.7 billion if we hit our long-term plans and can do so without having to stretch our underwriting appetite or shift our risk profile. During 2019, we appointed 98 new distribution partners, bringing the total to approximately 1,350 or 2,300 store fronts.Third, we have made major strides in recent years to enhance the customer experience. During 2019, we made a number of enhancements to our self-service and digital offerings, including a streamlined activation process and live chat option. We now have a 360 degree view of our customers that allows us to continue to enhance our proactive communication program as we seek to create more customer value. In 2019, we rolled out our new marketing tagline, "Do you legally insure?", which speaks to our differentiated value proposition for distribution partners and customers. Investing in and building out technologies that improve our customers' experience remains a core focus for us.Finally, while we recognize it is essential to continue investing in the initiatives related to our technology platforms, sophisticated underwriting tools and customer experience, we're also committed to balancing these goals with an efficient operating structure. We'll look at a number of areas to improve efficiencies, including workflow and process improvements by better leveraging automation and robotics. We expect our 2020 expense ratio to be slightly lower than 2019 as we balance our investment and expense management initiatives. We are targeting an expense ratio of closer to 32% by the end of next year and we'll seek further improvements in subsequent years.Looking to 2020 and beyond, we are in extremely strong financial and strategic position and are well positioned to continue generating superior results like we have in recent years. We have an attractive book of in-force business and have created a differentiated franchise with our distribution partners and customers. I'd like to close by sincerely thanking Greg for his 40 years of service and 20 years of outstanding stewardship as CEO of Selective. He transformed the Company into a truly unique franchise in the industry. Greg and I have worked in close partnership for much of the past decade. We've put in place a strategy that we believe keeps Selective on a path to be a consistent industry leader and generate sustained outperformance. In Greg's new role as Executive Chairman, I will continue to benefit from his guidance and wisdom as I transition into my new responsibilities.Now, I'll turn the call back over to Greg.

Thanks, John. To wrap up, another excellent year by delivering our sixth consecutive year of double-digit ROEs, which places us among a very elite group of peers that have generated similar results. We have the best teams of employees and distribution partners in the industry. At Selective, we're very proud about what we do as a company by: one, helping our customers put their lives back together by responding promptly and with empathy after experiencing a loss; two, making our communities safer by providing tools and technologies to our insurers such as weather preparation guides as well as vehicle, food and product recall notifications. We are seeking to mitigate the risk of distracted and reckless driving with the offering of Selective Drive, free of charge to our commercial lines customers; and three, providing financial stability to our customers.We are the best positioned in this marketplace with an excellent book of business and the tools, technology and best-in-class people in the industry to execute our plans. I could not be happier to pass on the reigns of the Company to John who will assume the role of Chief Executive Officer effective tomorrow. I have full and complete confidence that he will continue to transform Selective into a truly unique company in the marketplace and an industry leader. As we mentioned on our last conference call, I will stay in the role of Executive Chairman for one-year period.With that operator, we'll open up to questions. Thank you, operator.

Thank you. We will now begin the question-and-answer session. [Operator Instructions] First question is coming from Mike Zaremski from Credit Suisse. Your line is now open.

Hey, good morning. First question, maybe some just clarification on the combined ratio guidance. I've got a couple of questions. So, the 91.5% is excluding any reserve changes and catastrophes, and if that's correct, that compares to the -- for the 2019 number being 92.9%. So, which would imply a good deal of improvements or am I incorrect and there's -- you are baking in some potential for reserve releases in '20?

Yes, Mike, let me start with that. So now, there is no reserve releases there. That number has -- the 2021, the 91.5% is a number with no development in it. And one of the things you have to remember, there's a certain amount of expense ratio improvement in '20 over '19. The favorable development in '19, generated a lot of profit-based compensation to employees and to agents. So, you get a little bit of a mismatch. You see some of the favorable developments rolling through in a line item that you exclude, but the incremental expenses incurred in '19 were in the underwriting expense ratio. So, I just want to make sure you've got that point and that's about 40 basis points approximately between employee and agents overall. And so, you have some of that element. Then, when you look at the improvement rolling to '20, the rest of it comes from, as I mentioned, and John mentioned, the renewal price in line with trend and then there's underwriting and claim improvements outside of that that also add to some of the improvement.

And, Mike, this is John. Let me just add one additional point to the final point that Greg made relative to that, underwriting mix improvement in particular. We cite routinely every quarter the difference by cohort between our best business, based on future profitability and what we believe to be the 10% or 11% of our portfolio that we think has a higher expected combined ratio going forward. By having a lower retention on that business and a higher retention on that business, we expect to produce higher combined -- or lower combined ratios, excuse me. That is where we generate mix improvement that will give you loss ratio benefit in addition to the difference between rating trends. So, just wanted to tie that together with the point Greg was making.

Okay. Yes, that's helpful. And next, I'll -- you might have said this in the prepared remarks, Mark, but the general liability reserve deficiency sounded kind of minimal. Do you have the kind of the full-year 2019 GL reserve development kind of versus '18 and if you don't have it, I can take it offline. And also, remind me, Mark, was there a -- was GL part of the true-up in the current accident year as well in terms of the basis points you mentioned?

Yes, Mike. So, let me walk you through that. We did make some modest adjustments in the fourth quarter related to general liability on the prior accident year. It was an increase in the prior year reserves of $5 million, but we had reflected some favorable development earlier in the year. So, for the full year 2019, there's actually $5 million of favorable reserve development on the TL liability line in 2019 and that compares to just under $10 million, $9.5 million for the full year 2018. So, relatively consistent when you're thinking about prior-year reserves. In the current accident year, specifically related to 2019, we did increase the loss pick for general liability by 2.9 points in the quarter, which was $5 million impact. We had a very modest benefit earlier in the year. So net-net, when you put it all together for 2019, it was a $3 million increase relatively stable, lawsuit was flat. So, I'd just consider that a little bit of fine-tuning into new accident year.

Okay, that's perfect. And I -- lastly, John, in your prepared remarks, I believe you said litigation rates relatively stable past few quarters, is that -- are you referring to the -- what some other firms have called the attorney representation rates and along those lines, when we think about increasing loss cost for the industry and -- I mean, you guys think about it, especially in the GL or commercial auto line, is the attorney rep rate, is that one of the, kind of, biggest unknowns or risk factors or is it predominantly the severity of the lawsuits or both?

So Mike, this is John. I'll start and I'll try to hit both parts of that question. So, there is a difference between litigation rates and attorney rep rates and what we referred to in our prepared comments was the relative stability over the last several quarters in our litigation rates and that's for all of our liability lines. So, litigated file is one that is, as the name indicates, in litigation. Those are easy to track and measure and see a change over time. Our attorney rep rates are going to also include files that are not yet in litigation, but they are -- just are beliefs that there is an attorney representing a client.Those, I will tell you, are not as easy for us or anybody else to measure because it's a little bit more of knowledge on the part of the claims adjusted to understand whether or not there is any -- been any attorney communication. Those numbers will bounce around a little bit more. They're going to be higher than litigation rates, obviously, many of those will not ultimately result in litigation and I think that leads to an important point that we have stressed on this topic in the past. I would say, really two fronts on this. Number one, we do some modeling work and provide our adjusters very early in the claim lifecycle on files that have attributes that are more likely to result in litigation down the road. What that does is, make sure that, hey, we have the right adjuster on that file, but can you make sure that there are early communication with that claimant is going to put them more at ease about how that claim is ultimately going to be resolved and that will, in many cases, help us avoid a litigated file.Now, that sort of ties into your second point, which is, litigated files, generally speaking, and part of it's going to be because of the severity of the injuries claim by the claimant, but litigated files are generally going to carry a higher severity than unlitigated files. Some of that is because of the cost of attorney involvement including your own attorney involvement as the company, but part of it is also going to be because you're generally going to have higher damage levels on both the property side and the bodily injury side when there's lawyers involved.

Yes. So Mike, let me just -- this is Greg. So, obviously, to what you think about loss cost, it's frequency times severity. We closely monitor frequency activity and generally find that that is the leading indicator in terms of the composite of our breakdown, your pure premium claims are coming in. So, we closely monitor that and between that and the severity that John went over is really what drives the overall loss ratio and the amount of loss [indiscernible].

Okay. And just lastly, to be clear on that; by the way your color is always extremely helpful. Are you seeing a little bit, clearly, given the -- I guess, the texts were changed a little bit, are you seeing a little bit of uplift in loss trend?

Mike, this is John. As Greg commented during his prepared comments, we have adjusted our view over the last couple of years of future trend expectations, which had been running around 3%, are now pushing up closer to 4%, just under 4%. So, that is reflective of what we've seen in our own experience. Yes, I think this is probably an important point to just stress what we think is also are critical consideration when you think about changing loss trends in the industry, which is -- you really want to start by understanding how you feel about the initial loss picks that company has established for their causality lines and we've been very transparent about this and very consistent about this.In all of our prior accident years in recent memory, you will note that we talk about having an embedded assumption for future claims trends and that has always been embedded in our loss pick. We've also stressed the importance of achieving pure rate increases or pure price increases that meet or exceed that expected loss trend, which is a great way to think about the strength of the loss picks that sit on every one of those older accident years. And I just think that's an important consideration when you think about that changing environment to the extent it does happen and it happens industry-wide, it will hit the current and also hit the prior accident years, but we have had that embedded assumption for claims inflation in our loss picks, very consistently.

Okay, understood. And Greg, all the best in your new role as Executive Chairman of the Board. Thank you.

Mike, thanks. Operator?

Thank you. The next question is coming from Paul Newsome from Piper Sandler. Your line is open.

Hi, good morning and again, congratulations to John and Greg. So, a little bit of a follow-up. Can we talk about rate versus the claims inflation. I think if I read it right, your true rate increase levels are a little bit below the 4% and does that mean that you'll be essentially pushing for more rate, all things being equal, next year to keep us up to -- the next year you're at rate inflation?

So Paul, this is Greg. Yes, so that's why when we said that the 3.8% will establish the floor, that's what the -- the fourth quarter, we got 3.8% in rate and we view that as the floor going into 2020. So, yes.

Okay. And then, I'd love to hear your thoughts, generally speaking on the commercial auto business, which obviously has struggled for you and everybody else, pretty much. It doesn't seem like re-underwriting or at least pure rate is helping that much. What are we missing there that you think could be the reason why the industry keeps falling behind?

Yes, I think that's a great question and I'll tell you what I think also in the industry, principally was led by way higher frequency counts over expected, that's what started it and then I think it also -- so, I think, that got misstating in '16 and rolled through various accident years was just underestimation and I think that in part reflected some of the societal issues regarding distracted driving, regarding core driving habits, regarding a lot of issues with a lot of congestion on the roads, unemployment dropping, people -- higher driving, less experienced people on the road. All of that added and just the general road deterioration. I mean, I'm avoiding puddles all over the place now it seems like, but it's just the general deterioration of the infrastructure, I think, that's added to that. And I would tell you that one of the products that we offer, the Selective Drive product specifically earmarked at that. We offer that to our customers free of charge. It allows them to manage their fleet, their maintenance records, but not only that, it's the most important that it gives the driver a score at the end of the week in terms of how well they're driving which includes when they're on the phone, what -- hard braking, hard turning, anything. So that's a big -- when you look at the Hawthorne effect, that's a big additive factor that we can offer to our commercial fleets.And then, I would say the other part then is this whole issue around severity that worked its way in. And so, it was two areas that probably took a line that normally ran at about maybe 3.5% to 4% in that neck of the woods from a trend standpoint and pushed it to like 7%. And then -- so everybody was getting rate in the 6%, 7% rate, and that's where everybody was getting more of their commercial lines, but that only started recently. So, all of this build up kind of caught everybody by -- think a little bit behind the 8-ball, but I would tell you that there is a big improvement. Combined ratios in commercial auto dropped from like 150% in the 115% area. We're looking at 106%, pretty much on an accident year basis, at a current year basis. So, we actually feel pretty good about our improvement and when you look at where we're getting rate, that's obviously our lead line from our rate and it's just under 8% in terms of renewal and that's pure rate. So, we feel that we can get ahead of it and we also feel, maybe that the frequency counts relative to vehicles has kind of apexed right now.

Yes, I appreciate the color on the [indiscernible]. Thanks for the answers and good luck in next year.

Thank you.

Thank you. And the next question is coming from Amit Kumar from Buckingham Research. Your line is open.

Thanks and good morning. I had a few follow-ups on the same discussion as well. Just going back to, I guess, Paul's question on pricing versus loss cost and the margin, I think, I'm oversimplifying this a bit, but I would have thought that now would be the time to press on pricing and maybe you can just help me understand that a bit better. Is it a function of the loss trends in your book, the launching of your accounts. But I would have thought that maybe the trajectory of pricing over 2020, the slope of the line would have been a tad different?

Yes. So Amit, this is John. I'll start. So, I guess, first for us, based on the fact that we are achieving our target margins and our focus on commercial lines here, for us, the conversation is more about price adequacy than it is about price maximization. That's what we believe in and that price adequacy is in terms of the overall portfolio and it's also on account-by-account basis. You have seen with that in mind though some sequential improvement and it might appear slight, but when you look at where our margins are, that slight sequential improvement is a positive indication. We don't include in our numbers any exposure, any exposure change. And I think when you look at what's happening in the industry, you are seeing a lot more exposure included in the pricing numbers that are being put out there and while there might be some portion of that price number or exposure change that those act like rate, much of it does not. So, we give you what's probably the most conservative view and insight into pricing.Let's also remember, we don't exclude workers' comp from our number. And in our view, if we would exclude workers' comp, we probably should exclude it from our combined ratios as well to make the analysis complete. So, that's -- it's still a bit of a drag. Now that said, we do think that and we reacted to it by raising our loss -- our future loss trend expectation up closer to 4% and that has caused us to view the rate need in our portfolio as a little bit higher and therefore, we expect to push that as the year goes on.Now, we are going to still do it on a very targeted basis though. This is in our portfolio that doesn't justify a meaningful rate increase won't get one because our view is, we're going to do what we can to protect that portfolio. The flip side to that is the 10% to 15% of the portfolio that we view as having worse than desired combined ratio -- future combined ratios is going to be the focus of our effort and we're going to achieve higher rate levels there. So, I guess, that's how I'd respond overall in terms of how we think about the pricing environment. And the other point I would also reinforces if you look at the Willis Towers Watson CLIPS survey, which we do site frequently in cited earlier and look at it for the small account sub segment. You're going to see a much lower number in there and a lot of the price headline is coming out of the specialty and E&S areas is coming out of the coastal property and very large sharply accounts. We haven't seen it as much in our core small and middle market lower medium hazard risk space that we're predominantly focused on.

That's a fair comment. So, is it possible and I completely agree and appreciate your comment on exposures. So, it's simply not apples-to-apples, looking at your numbers versus others, but do you have the number excluding workers' comp for the quarter or maybe could you even like talk about it, sort of, in generalities how much the delta would be, if we exclude workers' comp from the pricing?

Yes, so -- and we gave you some of the pieces. And I can't calculate on the slide for you what it would be ex-comp, we'd have to wait it out, but -- so GL excluding umbrella was 3.3% for the quarter. Commercial auto was 7.8%, commercial property was 4.1% and BOP for us, which is a smaller line, was 5.3%. And then, you've got workers' comp at a minus 2.7%. So, you're probably looking at something -- probably closer to 5% roughly, if you were to look at an ex-comp number, maybe a little bit above 5% on an ex-comp basis.

Got it. And that's what I thought. Okay, that's good. The other question is also a follow-up on the litigation trends and we've talked about this on several conference calls and we expect to talk about this. I know in the past we've talked about Reviver. This question is not on Reviver. Can you -- when you sort of look at your trend line on the litigation trends and then you look at all the reports out there, we should look at litigation trends, etc., when you look at your own book, do you think the reason why we haven't seen that level of increased activity versus some of the larger companies? Is it more because of the account size, is it the geographical mix, or maybe it's a combination of all of it, but maybe just talk about that a little more so that we can get some confidence that these trends will remain stable versus some of the other larger companies? Thanks.

Yes. So, I think there clearly are differences and you would expect differences in terms of where the plaintiffs' bar is going to target in terms of individual companies and the type of company we do right is going to be less of a target, but I think we do look at inflationary trends or claims inflation as not just entirely focused on the headline litigation or headline settlements amount that you see. To the extent it's happening, it will ultimately start to make its way up and down the marketplace. I don't think there is a geographic difference when you look at our footprint. Our footprint is concentrated in the Eastern half of the US with some of the Southwest expansion, but for us, you also want to look at the hazard levels and we do still right a predominantly low and medium hazard class of business with a low limits profile.And I know a lot of companies cite their limits profile and ours is certainly on the lower end for most and that will provide some protection, if in fact, we do see an acceleration of claims. I'll also say that while our own litigation rates haven't really moved as we've mentioned and we focus in terms of our view of our reserves on our own experience, we certainly pay attention to what's being stated and what others are saying in the broader industry and have a keen eye on that relative to what might potentially emerge, but to sit here and say with clarity that frequency or severity trends are going to be X or Y in future years is just tough to get comfortable with.

Got it. The last question I have is on the GL development. I think you were responding to Mike and I'll go back and look at the transcript, but my understanding was, there was adverse of $5 million, and did you -- was that from AY 2018, '17. I didn't get the details in terms of the moving parts of that $5 million? Thanks.

Yes. Amit, it's Mark. It was $5 million of adverse development in the quarter, but I think it's best to look at the full year, which was $5 million of favorable development in the general liability lines. So, there's obviously always some quarterly volatility. It's a relatively small number on a big book of business. This reflected a little bit of fine-tuning in some respects is going into the new year and then as it related to the current accident year and I mentioned this earlier, but just for the sake of clarification, it was also $5 million on the 2019 accident year, so we had a modest benefit earlier in the year. So, for the full year, it was $3 million. So, if you net the $3 million in the $5 million, it's a bit of a wash against the current and the prior related to GL line, it reflects a line of business that we've had reasonably stable trends.

Got it. And then then, was that related to like an account or is that just more fine-tuning across the book?

It's more across the book. And when you think about that small dollar number across multiple of prior accident years, there's not even going to be an individual year you would look at, but I would say, it would definitely have been more -- in the more recent accident year '17, '18.

Got it. That's all I have. Thanks for the answers and good luck in your new roles, Greg and John. Thanks.

Operator, next question?

Thank you. The next question is coming from Mark Dwelle from RBC Capital Markets. Your line is now open.

Yes, good morning. Let me just start by giving my well-wishes to Greg, I very much appreciated your valuable insights and you will always be among the Ivy League of CEOs in my book, So good luck to you going forward.

Thank you. Thank you, Mark.

First question just something fairly simple, the corporate expense line item is much lower than the other quarters. I think this was the same as a year ago. Just, is there something -- is there a credit or some seasonality or something, I would have thought the comp expense alone would have been more than $2.6 million in the quarter.

Yes Mark, it's Mark Wilcox, let me jump in and Greg and John can follow on as well. The corporate expense line item, does have an element of volatility and a little bit of seasonality to it. So it was running in the $35 million range a few years back at $25 million per year. So we've gone this year mainly include -- so mainly includes stock-based compensation and a portion of that, call it 25% is what we call liability based and so there is some volatility. It is driven really by a couple of factors, one that the peer group factor; two there's also a total shareholder return factor. And the reason it was down in the quarter was soft kind of pretty big seller, as you're well aware in Q4, as that drove a benefit or a credit in that line item in the quarter.

And just as a reminder, if you look ahead to 2020, there is some seasonality. You typically see a heavier front end load, I mean Q1 of the year related to the corporate expense line item, given the retirement eligibility and the full expensing our stock-based compensation for those individuals that are requirement eligible and receive an award.

And I would say, just the changes that we made to the plan overall, now that you have all the years that are now eligible to be expensed in theory that 3-year period, under the new plan it will reduce the overall amount of volatility in the line.

Okay. That's helpful, thanks. Second question, John, you had provided some pretty good pricing segmentation information with respect to the commercial lines book. Do you have similar information you can share related to the E&S book?

Well, we haven't traditionally, I can tell you that we do a similar segmentation, but it's more based on industry classification and our targets pricing levels to achieve our target ROE we do certainly manage it that way internally. It's just not something we reported about externally. But it's the same very granular approach that we take in managing both new and renewal business in the E&S segment.

Okay. Staying on the E&S book, any additional insight you can share related to the reserve addition in that business. I mean I know it's kind of gone back and forth over the years. But is there anything different about loss trend or litigation or anything there?

Mark, it's so small, it's not even that large. So I would say it's just minor year-end adjustments that come through based on our slight modifications, so it's not material, really not significant at all.

Yes, was $2 million in the current quarter of the prior accident year and $2 million in the current quarter on the current accident year. It's relatively modest. Again a little bit of fine-tuning going as we roll over to new accident year making sure we feel confident in our reserve inventory but nothing, I think we'd point to in terms of trend. I think the good news obviously with the E&S segment highlighted is a strong level of profitability, its best year yet, from a profitability perspective in E&S. Still some work to be done to drive towards the targeted level of profitability in E&S. But the reserve, obviously we had a $12 million addition. So trending in the right direction. And just some fourth quarter minor adjustments.

Yes, I would just add to that, Mark, we've come a long way US about target surcharges that joining [ph], we track that we need to tell you exactly where the new normal is and we are been in this market we -- for the past several years. I have never been totally satisfied with what we have been getting or the amount of increases that we're getting on the renewal side. We have seen change in that. We have done a lot of work on our renewal inventory to get them closer and closer to where our target surcharges need to be and we have been very happy relative to the new business that we put on the books and where that is relative to target surcharge.So that's come a long way and again, you know us, we put in a lot of discipline in it and we want to make sure that that business is profitable. We've always told you that it will -- the topline will go up or down, more based on market conditions and that I think will happen overtime.

Okay, that's helpful. And then one last question, and this is I suppose it's more of -- almost a thought question as anything else, but is -- I'm trying to gauge kind of a sense of where the market tone is -- the -- from the customers' perspective. As you go to customers with renewal rate increase requests, whatever level they happen to be, 5%, 2%, 8% whatever, I mean what kind of push back are you getting? Are people understanding of the need for rate? Is it more than unusually adversarial? How does that conversation go at this stage?

Yes, Mark, it's John. So obviously, the answer is going to be intense. It depends on the size of the rate increase and it depends on the individual producer who owns that relationship at our independent agency and their approach to managing that process. I will say for us, we have really prided ourselves on maintaining that routine balance between price and trend over a long period of time. So we're not in a situation where you've got a 10% or 15% rate need on the overall portfolio. So that minimizes the impact overall, when you think about what we're asking of our agents in terms of selling increases.Now that said, we do have very specific discussions and expect our agency partners to be able to have that specific discussion with that smaller percentage of the inventory that is going to get a higher than average increase based on risk characteristics or experience of that account. And I will say that based on our overall retention and our retention by cohorts, the customers understand that and our distribution partners do a good job of explaining that. So I don't necessarily feel like that's a huge impediment at this point, executing on our pricing strategy, but I will say if we just took an across the board approach and we're trying to fix a loss ratio problem, which were not, it would be a lot harder to execute in this kind of an environment because there are enough companies like we are, that are well positioned with their pricing sophistication, to go out and take those opportunities that look attractive when a company is taking an across the board rate increase.

And Mark, this is Greg. So the other side to that too really comes down to more than the general economic conditions. So with GDP growing businesses growing, top line revenue growing, price increases are little bit easier to take it as a customer versus years ago when GDP was not growing. And you're trying to raise rate and everybody's looking through their balance -- looking through their P&L to figure out what expense lines they cut. That does make it a little bit easier relative to that.And the other point that John made earlier relative to what we're doing customer experience, relative to everything that we're doing to build more customer connectivity, whether it's our security mentor product that we recently rolled out, our drive product, the product recalls that we're offering. All of the additional work that we do to build more and more contact ability to the end customer, we want to make sure that -- when we do need to sell larger price increase someone would sit there and say okay this is all the value that I'm getting from Selective and it's not just the insurance policy, it's way more than that. And we believe that over time, that should help our retention, help our net promoter scores help all of our OSAT scores everything else that we track internally that will make it make different long-term.

Thanks very much for that. I appreciate the additional insight. Thanks.

Thank you. And the next question is coming from Matt Carletti from JMP. Your line is now open.

All right, thanks. Just have a quick one, I wanted to ask a question on workers' comp, growth was up 5% in the quarter. That's the strongest number we've seen in a while, actually first positive we've seen in a while. Just wanted to ask, what's going on there. I mean, I think you gave us the rates we know it's not that, is it the new business production underlying kind of economic work activity, was there some audit premium catch-up, just what kind of gave us that growth in the quarter?

Yes Matt, this is John. I would say, and again remember we write predominantly account business and very little monoline workers' comp. So you want to think about workers' comp growth in the overall context of the very strong quarter we had in commercial lines. Commercial lines growth was at 11% and that was driven by both strong retention and a solid new business quarter. And I would say when you look at the growth in comp relative to the other lines, it's still the lowest growth line at that percent, but stronger in the quarter certainly. I don't -- I wouldn't necessarily as you said pricing wasn't that materially different than we saw in the first three quarters. I just think we had a stronger growth quarter overall and that helped a little bit the growth in the workers' comp segment.We have not seen meaningful change in the competitive landscape in comp. I will tell you that continues to be -- especially for small lower hazard workers' comp probably the most aggressive market in any segment that we play in. You've seen a lot of companies continue to be very aggressive and we love our results, we love our book of business. We feel good about the book of business we have, but we also recognize that loss trends in that line can continue the downward trajectory they've been on for both frequency and severity, and we're just being cautious about growing that relative to the overall growth of the Commercial Lines operation.

Got you. Just kind of tying that last comment there together, I mean am I right to infer that as you go through the component pieces of -- obviously gave us guidance that implies kind of accident year margin improvement, '20 versus '19 for the overall book that this might be a component where it might tweak the other way a little bit that rates down a few percent. And I mean less trend continues in that direction in which it sounds like it's not an assumption that you'd make off the that -- that might be a component that goes a little bit the other way as we go forward.

I -- we give you overall combined ratio guidance, and I will give it to you by line, but based on how we talk about rate and trends, I think the way you described our view of comp would be a fairly accurate description.

Great. Well, Greg and John congrats and best of luck going forward.

Thanks, Matt.

Thank you. And the next question is coming from Ron Bobman from Capital Research [ph]. Your line is open.

Hi, thanks a lot. And Greg you've have been awesome. I sure have learned a lot and you've always pointed us in the smart direction. So thanks a ton.

Thank you.

Greg, a little while ago in the call, you mentioned the commercial auto combined was at 115 and now it's at 106. But I'm wondering what sort of date -- what window what quarter, or what year you were sort of referencing that, if I have those numbers right, the movement occurred over.

Yes, so let me just pull that out in a second, hold on a second, let me spread the lines here. So for 2000 and -- so year-to-date commercial auto in '18, the cap rate, the combined ratio call it $1.16 was the calendar year trend and the accident year was $1.08 [ph], pretty close to that in the underlying combined ratio for that year. And now we're running, call it the -- for this 2019 year, commercial auto, the combined ratio was a $1.08 and the underlying combined ratio was like a 100 -- call it 107m that could work. So that gives you a little bit idea as to the movement and what was happening and that's what I was referring to earlier.

Got you. And would it be -- given how rates have been -- should I think that Q4 '19 numbers are marginally better than the 108 and 107 that was booked for the whole year or the rate increase has been pretty consistent?

Ron, I think you're -- the way we build our plan is, let's put it this way, the way our planning process works, the base year is a 4-year average and that we're projecting that forward. So we look at any liability line, we're estimating break it down ultimately through a frequency severity number about and then that ratio is pretty consistent in liability throughout the year. We're not market weighing, or price adjusting quarter to quarter. So generally speaking on any liability line, you'll see unless we're making modifications to it, it would be the same ratio every quarter. What you do see as volatility in commercial auto though could be the property side and that could be weather specific. So that could come from -- whether it comes from property damage frequency on the liability side, but it also could come through physical damage on this side. And that's what could create a little bit more up and down on that number versus a straight liability line like GL or workers' compensation, where you see virtually the same kind of ratio throughout the quarters.

Okay, thanks. You mentioned, I think just qualitatively but I could have missed it, retention. I'm curious to know how Commercial Lines retention was in Q4, as compared to Q3 sequentially, please.

Yes. So retention for our commercial lines was 84% in the quarter, and I give you the full year number, which was 83%. I don't have the individual first three quarters. But that would suggest that it was relatively stable. A year ago in Q4 '18, it was a little bit lower at 83% and then throughout the year 84%, 83%, 84%. So in the full year, 83% so relatively stable retention throughout the year.

Got you. That's great elasticity as far as that support you to push more rate I guess, or like you said a baseline. I think what Greg said going into 2020.

Thank you for picking up on that.

Last question. The putting aside Selective and you highlighted really small E&S book as compared to your admitted book, from a more general perspective, this challenge of increased loss cost, a portion of which is driven by attorney representation. And the distinction between that and litigated claims, do you think that an E&S book would be more resilient to that challenge than an admitted book or not -- it's both, you're going to sort of face the same delta from loss cost inflation being driven by litigation or attorney representation in one quarter or another.

Ron, this is John. I would say generally speaking, I don't think it's just going to be that different. But it also depends on the segment of the E&S market, we're talking about. I do think if you're writing in the space of the E&S market that is really high exposure products, higher hazard classes of business, you might see a little bit more of an acceleration in terms of claims trends, but I would say for us. Our book is predominantly small contractors, small habitation, small restaurants and [indiscernible] and service businesses, small limits profile not high profile exposures. It would be more akin to what we would expect to see in our standard lines book.

Yes. Greg quoted a number earlier on that limits profile, casualty with the stated commercial lines of $1 million less than 87% of the book and E&S casualty for the $1 million and below is 98%. So a very similar limits profile even perhaps a little bit more conservative.

And then again, everybody focuses on the issues that could rise -- raise loss cost, we do as well. And I just want to make sure we get out. There is a tremendous amount of effort internally in the organization. John touched on the litigation propensity models, the escalation models that we have, what we're doing in the area of complex claims, some of the other work that I know we're starting on relative to robotics in certain areas. What we're also doing to find out when we look at losses and sit there and say what is it that we can do to help mitigate some of the losses, whether it's notification, whether it's education and we've got a lot of efforts on that and those are things that will pay off on the longer term and normally would -- we do not see any credit come through in the near term, relative to the loss ratios, but they are part of analyzing the loss costs, analyzing, how you we get back to strength, analyzing how are some of the things you can do ensure tech to help the mitigate that, are all things that we need to be focused on to reduce loss costs. So those efforts are underway, but will pay off over time.

Thanks. And Greg, thanks for time again, been great and you're leaving us I think well protected and well led with John. Take care.

Thank you. And we show no further questions in queue at this time.

All right, everybody, and I appreciate the comments from everyone. Thank you. If you have any follow-up, Rohan and Mark are available. Thank you very much for all your questions today, very interactive call. Thank you.

And that concludes today's conference. Thank you all for your participation. You may now disconnect.