Hartford Insurance Group Inc (HIG) 2018 Q2 法說會逐字稿

Good morning. My name is Dan, and I will be your conference operator today. At this time, I would like to welcome everyone to The Hartford Second Quarter 2018 Financial Results Call. (Operator Instructions)

I would now like to turn the conference over to Ms. Sabra Purtill, Head of Investor Relations. You may begin your conference.

Thank you. Good morning, and thank you all for joining us today. Today's webcast will cover second quarter 2018 financial results, which we announced last night. The news release, investor financial supplement and second quarter financial results slides and 10-Q are available on our website. Our speakers today include Chris Swift, Chairman and CEO of The Hartford; Doug Elliot, President; and Beth Bombara, CFO. Following their prepared remarks, we will have time for Q&A.

Just a few comments before Chris begins. Today's call includes forward-looking statements as defined under the Private Securities Litigation Reform Act of 1995. These statements are not guarantees of future performance, and actual results could be materially different. We do not assume any obligation to update information or forward-looking statements provided on this call. Investors should also consider the risks and uncertainties that could cause actual results to differ from these statements. A detailed description of those risks and uncertainties can be found in our SEC filings, which are available on our website.

Our commentary today also includes non-GAAP financial measures. Explanations and reconciliations of these measures to the comparable GAAP measure are included in our SEC filings as well as in the news release and financial supplement, which are also available on our website.

Finally, please note that no portion of this conference call may be reproduced or rebroadcast in any form without The Hartford's prior written consent. Replays of this webcast and an official transcript will be available on The Hartford's website for at least 1 year.

I'll now turn the call over to Chris.

Good morning, and thank you for joining us today. Second quarter marked another strong performance for The Hartford. Core earnings were up 36%, core earnings per share rose 40%, and book value per share, excluding AOCI, was up 8% since year-end 2017.

Our annualized core earnings ROE for the first half of 2018 was 13%. Earnings growth came from Commercial Lines, Group Benefits and Mutual Funds, including the lower effective tax rate. Personal Lines underlying margins continued to improve but were outweighed on the bottom line by higher catastrophe losses. Despite higher CAT losses in the first half, we remain on track to achieve our underlying margin and profitability outlook for the full year.

During the quarter, we achieved progress on several other important goals. First, we closed the sale of Talcott on May 31, just 8 weeks ago. Net cash consideration to the holding company was about $1.5 billion. In addition, we retained a 9.7% equity stake in Talcott, which is carried on our books at $164 million at June 30. Second, the Aetna Group Benefits integration, about 8 months underway, is proceeding well. Given the complexity of the numerous activities involved, I am very pleased and impressed with the pace and overall progress. We are optimistic that we can reduce expenses by more than our original $100 million target.

Aside from the integration, Group Benefits sales are off to a strong start in 2018. First half sales totaled $539 million, almost double from last year. This includes $7 million of life and disability product sales to Aetna's group medical customers. These sales are the result of our agreement with their major medical sales force to work with us and continue to market group life/disability products to their customers. These results speak to the success of our cross-sell partnership.

We are also executing on our projects across the company that are making The Hartford a customer-centric and easier company to do business with. It's a long list, with an emphasis on technology and digital tools working in an agile environment. One example is in Small Commercial. On our ICON platform, we have increased the percent of accounts that can be quoted on the glass. We have streamlined the underwriting process and reduced the time it takes to get a quote, which is an important competitive advantage in this market. Another example is the expansion of our underwriting capabilities and specialty product orientation to achieve our goal of being a broader and deeper risk player.

In Middle Market, we have grown and added industry verticals, responding to increasing demands from agents for deeper industry underwriting expertise with a broader product suite. This aligns with our strategy to expand account rounding with our workers' compensation policyholders who value our claims and customer service capabilities. Examples of our focus on industry verticals include technology, a traditional strength at The Hartford, which grew premium 4.5% over the last 2 years. In addition, our construction practice has grown premiums 35% over the past 4 years. Lastly, we entered the energy market late 2016 and, over the past 4 quarters, wrote $24 million in gross written premium. These are just a few examples of our approach to organic business development.

Taken all together, we sustained very good momentum this quarter in all our businesses. I am thrilled with our progress and our future potential.

Last week, we announced a 20% increase in our quarterly dividend. This decision was based on the strong performance of our businesses, the sale of Talcott and lower tax rates, consistent with our long-standing dividend philosophy. Last week's dividend declaration was the sixth consecutive annual increase, and it will increase our dividend yield and payout ratio to be more in line with peers.

With regard to other uses of excess capital, our philosophy has not changed. Investing in our company remains the cornerstone of our strategy. We want to achieve profitable organic growth, particularly where we have attractive returns and strong competitive advantages.

Along those lines, acquisitions that are aligned with our long-term strategic and financial goals are a compelling use of capital. Acquisitions can help build greater competitive advantages, add operational capabilities and accelerate earnings growth compared with building a business from the ground up. For example, in 2016, we acquired Maxum, which allowed Small Commercial to expand into the E&S market. In 2017, we purchased the Aetna's U.S. group life and disability book, which was a unique opportunity to expand our market position while acquiring an industry-leading claims and leave management platform.

As I shared last quarter, when we assess areas of our business that could benefit from the accelerants of an acquisition, we have a focus on specialty lines and industry verticals in the commercial insurance segment that would expand our product sets and underwriting expertise. Since the recent sale of Talcott, there has been speculation and questions about future share repurchase plans. This is understandable as our prior capital management actions have included a large amount of debt reduction and share repurchases. While capital management remains a valid alternative, it is not our primary focus at this time. Therefore, we have not authorized a new share repurchase plan. We will continue to evaluate options that will generate long-term earnings growth at good returns. And if we conclude that there is not an alternative option to support growth, a share buyback plan could be put into place relatively quickly. But as I have stated previously, we will be thoughtful and patient regarding capital deployment, with our focus on creating long-term sustainable shareholder value, which is why we are maintaining the option of investing capital to expand the business.

To wrap up my comments, The Hartford had very strong performance for the first half of the year, both financially and operationally. We are intently focused on continuing to execute on our goals and sustain our momentum through the second half of the year. I look forward to updating you on our progress over the balance of the year.

Now I'll turn the call over to Doug.

Thank you, Chris, and good morning, everyone. This was another strong quarter in Property & Casualty and Group Benefits as we advance our key business initiatives and address evolving market conditions.

Commercial Lines had a strong quarter as we continued to balance growth with competitive market conditions. In Personal Lines, improved auto trends continued, although overall results were hampered by catastrophe losses. And in Group Benefits, we had another excellent quarter, with strong favorable trends in group disability.

Before I touch on second quarter results for each business segment, let me cover current and prior year catastrophe losses for Property & Casualty. In the quarter, we had $188 million of current year CAT losses, $33 million higher than a year ago, driven by significant wind and hailstorms in various regions across the country. Included in prior year development is the reduction of our estimates for prior year catastrophes, largely attributable to hurricanes in the third quarter of 2017.

As a result of lowering these estimates, we no longer expect to receive a recovery against our aggregate catastrophe reinsurance treaty for the 2017 accident year. The benefit from the aggregate treaty was allocated to each business unit based on our estimate of ultimate losses for the full year. This quarter, as you can see on Page 13 of the slides, 2017 gross loss estimates decreased in both Personal Lines and Commercial Lines, causing total P&C losses to fall below the attachment point for the treaty. However, the decrease in gross losses was greater in Commercial Lines than Personal Lines, and therefore, unwinding the aggregate cover resulted in net adverse development for Personal Lines.

Let me now shift into the results for our business segments. In Commercial Lines, the combined ratio improved 4.5 points from prior year to 90.1, driven by favorable prior year development, partially offset by slightly higher current year catastrophe losses and expenses. The prior year development was primarily due to continued favorable trends in workers' compensation and lower estimates on catastrophe reserves, as I just covered, partially offset by an increase in reserves for higher-hazard general liability exposures. This portion of our general liability book remains profitable, but we are responding to an increase in loss trends in accident years 2015 through 2017. Trends in the remainder of our general liability book have been slightly better than our expectations.

The underlying combined ratio for Commercial Lines, which excludes catastrophes and prior year development, was 90, improving 0.9 point from last year. The improvement was largely driven by favorable loss trends, particularly non-CAT property as well as general liability. This was offset by higher expenses and slight margin compression in workers' compensation.

Let me provide a few additional thoughts on workers' compensation, touching on 3 important factors: frequency, severity and rates.

First, on frequency. In both Small Commercial and Middle Market, our frequency has been trending slightly higher during 2018. We see this as a broad-based economic-driven trend across many states and industry classes. Keep in mind that absolute frequency is still at historically low levels and margins remain attractive. However, expanding participation in the labor market often means less experienced workers are on the job, a driver of frequency.

Second, on severity. We continue to closely watch inflation trends for upward pressure on both wages and medical costs, drivers of severity. Although accident year 2018 data is immature at this point in the year, severity remains within our expectations.

And third, on rates. Loss costs in workers' compensation have been quite benign for several years. And as a result, the NCCI and other bureaus have been filing negative loss cost changes in many jurisdictions. Industry rate filings are based on these loss cost trends and have put downward pressure on rates in 2018 and will likely do so in 2019.

Combining these 3 factors, we expect -- we continue to expect some margin compression in our Small Commercial and Middle Market books of business. Since each of these businesses has unique market dynamics for pricing and growth, our approach to successfully managing these rate changes is also dynamic. In Middle Market and National Accounts, greater weight is placed on individual risk characteristics when pricing the account. In Small Commercial, pricing is more heavily weighted to class factors, making book management a critical tool.

Over the last 5 years, we have significantly advanced our actuarial analytics and data science capabilities to improve our ability to identify these trends and respond confidently with pricing, underwriting and claims management initiatives. Given our strong margins and organizational capabilities, we will continue to execute a disciplined pricing and book of business management strategy in the months ahead.

Moving to pricing in the quarter. I'm very pleased that our renewal written pricing in Standard Commercial Lines was 3.1%, up from 2.4% in the first quarter of the year. As I suggested in my comments last quarter, we saw a positive movement in Middle Market across property, general liability and workers' compensation, with strong increases continuing to come from commercial auto. Our pricing in Standard Commercial Lines, excluding workers' compensation, was 5.2 in the quarter.

Looking at our Commercial Lines business units. Small Commercial had another strong quarter, with an underlying combined ratio of 85.6. Written premium was off slightly, the result of downward pressure on workers' compensation rates and competitive market conditions for both new business and renewals. The margin improvement versus last year was driven by non-CAT property, general liability and auto, partially offset by higher expenses.

In Middle Market, the underlying combined ratio of 94.1 improved 0.8 point from 2017, mainly due to non-CAT property and general liability. This was partially offset by modest margin compression in workers' compensation and higher expenses. Written premium was up 3% over last year, with new business production of $138 million.

And in Specialty Commercial, the underlying combined ratio of 98.5 deteriorated 2.6 points. This was driven mainly by higher expenses and margin compression in National Accounts and Financial Products. Written premium was up 9% for the quarter, reflecting strong growth in Bond and Financial Products.

Moving to Personal Lines. The underlying combined ratio of 90.4 improved 2.2 points from a year ago, driven primarily by improvement in auto and, to a lesser degree, better homeowners performance. The underlying combined ratio in personal auto improved 2.6 points to 96.5. We continue to experience relatively stable loss cost trends, and we are satisfied with our bottom line performance.

Personal Lines premium was down 7%. Consistent with my comments in recent quarters, our marketing spend continues to accelerate, and we're focused on driving new business growth. In AARP Direct auto, our bellwether line, new business was up 19% in the quarter. This is an encouraging result as we focus on growth through new business and improved retention. We're implementing additional product and process changes throughout the remainder of the year that I expect will continue to improve our top line trends.

Turning to Group Benefits. We had another excellent quarter, with core earnings of $104 million and a margin of 6.9%. Drivers were similar to last quarter, including favorable disability results, lower tax rates and the contribution from our 2017 acquisition, partially offset by slightly higher life mortality. Persistency on our employer group block of business remained strong at approximately 90%, and fully insured ongoing sales of $85 million were up from prior year. This was a solid sales quarter, and we're pleased with our traction in the market.

Our integration of the Aetna group life and disability business remains on track and is picking up momentum. We began converting small case business to our current platform this month, and we will ramp up those efforts throughout the year. The new disability claim platform, now branded The Hartford Ability Advantage, is expected to come online for new cases effective in 2019.

As Chris noted in his comments, our expense reduction efforts are on track. And as our business conversion process continues in '19 and '20, we are confident that we will be able to exceed our original savings target.

In summary, this was another strong quarter for our Property & Casualty and Group Benefits businesses. At the halfway point for 2018, I'm pleased with our execution and our performance. We're responding to loss cost trends with appropriate pricing action and disciplined underwriting. We continue to innovate in all areas of our business to deliver a superior customer experience. And we're committed to becoming a more relevant partner for our agents and brokers, with an expanded product portfolio to better meet customer needs and deliver profitable growth to The Hartford.

Let me now turn the call over to Beth.

Thank you, Doug. My comments today will cover second quarter results for the investment portfolio, Mutual Funds and Corporate, impacts on the quarter from the sale of Talcott and June 30 book value and debt leverage, before taking your questions.

Starting with investments. Net investment income performance and yields remained steady. Pretax limited partnership investment income was flat with the prior year at $39 million, for an annualized return of 9.5% in second quarter 2018. Excluding LPs, the annualized portfolio yield before tax was 3.7%, down slightly from 3.8% in second quarter 2017 due to the impact of the Group Benefits acquisition.

The P&C yield was flat year-over-year at 3.8%. Consistent with the first quarter, the yield on the consolidated Group Benefits investment portfolio is lower than last year because acquisition accounting rules require that the acquired portfolio is marked to market. As a result, the annualized yield before tax, excluding LPs, in Group Benefits was 4.3% in third quarter 2017 before the acquisition but dropped to 3.7% in the fourth quarter. This quarter, the Group Benefits annualized portfolio yield was 3.9%, up slightly from first quarter 2018. This is a result of our reinvestment of the acquired portfolio towards our target sector allocation, with less municipal bond exposure, resulting in a higher pretax yield.

Investment credit performance remained excellent this quarter, with no net impairments due to a generally benign credit environment and the credit strength and diversification of our portfolio.

Turning to Mutual Funds. Second quarter core earnings were $38 million, up 58% from last year due to combination of lower tax rate and higher AUM from positive net flows and higher market values. Income before taxes rose 21% as a result of the operating leverage in our operations, with a 6% increase in revenues but only a 3% increase in expenses. AUM growth is due to net flows, which totaled $1.9 billion over the last 4 quarters, and changes in market value, which totaled $8.2 billion.

Our ETP business drove $500 million of net flows over the last 4 quarters. Hartford Funds' strong performance is driving the positive net flows, with 61% and 66% of funds beating their peers on a 3-year and 5-year basis, respectively.

Corporate core losses totaled $76 million, higher than first quarter 2018 due principally to a tax true-up for the reallocation from the business segments to Corporate for the impact of nondeductible executive compensation.

Income from discontinued operations was $148 million, up from $112 million in second quarter 2017. In second quarter 2017, this represented Talcott's net income for the quarter. In contrast, in this quarter, about 90% of the income reflected an increase in our estimate of the retained tax benefits from Talcott due to a change in our estimate of Talcott's tax basis. This change increases our estimate of the value of retained tax benefits to about $830 million.

Going forward, other revenue will include the return on our investment in Talcott, which will be included in core earnings and in the Corporate segment along with other Talcott impacts. The other impacts include, beginning this quarter, income from investment management and transition services as well as the related operating expenses. The amounts recorded this quarter reflect 1 month of activity and did not have a significant impact on the corporate bottom line. We expect transition services and stranded costs to decrease over the next 12 to 18 months.

In total, The Hartford second quarter core earnings were $412 million, up from $303 million in second quarter 2017 due to higher Commercial Lines, Group Benefits and Mutual Funds earnings before tax as well as a lower federal income tax rate.

The effective tax rate on income from continuing operations was 19% in the quarter compared with 18% in first quarter of 2018 due to a slightly higher proportion of income from underwriting results and taxable investment income as all income other than municipal bond interest is taxed at 21%.

Book value per diluted share at June 30, 2018, was $34.44, a 7% decrease from December 31, 2017. The after-tax unrealized gain on our fixed maturity portfolio at June 30, 2018, was $211 million, down from $1.8 billion at December 31. This decrease was due to the Talcott sale and the reduction in the market value of our fixed maturity portfolio due to higher rates and wider credit spreads.

Book value per share, excluding AOCI, was $38.15, an increase of 8% from December 31, 2017, reflecting the increase in retained earnings as a result of net income in excess of dividends for the first 6 months of the year.

The core earnings ROE was 8.4%, calculated using rolling 4-quarter earnings and average stockholders' equity since June 30, 2017. On an annualized year-to-date basis, core earnings ROE was 13%, which is above our outlook for the year.

Turning to our debt leverage. In June, we called at par $500 million in junior subordinated debt, reducing our total debt outstanding by $323 million since year-end 2017. Despite the repayment of debt and strong earnings, our rating agency debt-to-total capital ratio of about 30% was flat with March 31 due to the inclusion of AOCI in this calculation. Over the long term, our focus is to reduce the rating agency adjusted debt-to-total capital ratio to the low to mid-20s. However, since it includes the favorable or unfavorable impact of AOCI, this ratio can be volatile during periods of changing interest rates.

So we also focus on the total leverage ratio, which is calculated by dividing total debt, including hybrids and preferreds at par, by total debt and capital, excluding AOCI. At June 30, our total leverage ratio was 25%, which is still at the high end of our long-term goal. We continue to focus on decreasing this ratio over time.

To conclude, we remain on a very good path in 2018, with strong underwriting and investment results and success in both closing the sale of Talcott and the continued integration of the Group Benefits acquisition.

I will now turn the call over to Sabra, so we can begin the Q&A session.

Thank you, Beth. We have about 30 minutes for questions. Dan, could you please repeat the Q&A instructions?

(Operator Instructions) Your first question today will come from the line of Elyse Greenspan with Wells Fargo.

My first question, in terms of the Commercial Lines margin, pretty strong improvement this quarter. I was hoping we can get a little bit more color, if you can break down how much you saw coming from favorable non-CAT property, also from liability and then also from the auto, the 3 things you called out in the press release, if you can just give us a sense of the contribution from each. And then is the 90 basis points of improvement just based off of how you see the rating environment and what's going on in comp? Is that the right level we should think about in Commercial going forward?

This is Doug. Let me try to tackle both of them separately. When I look across our markets, as I mentioned, both Small and Middle had excellent non-CAT property quarters. And Small, about 1 point better quarter-to-quarter, and Middle, a couple of points, 2.5 points better. So significant drivers of positive performance. Small, on the workers' comp side, about our expectations. And I mentioned that just slightly, we made an adjustment to our workers' comp in Middle by a small amount. So when I look across, I feel good about all the non-comp lines in Middle. We made an adjustment in comp. And all in, a very, very solid quarter for Commercial Lines relative to combined ratios. Can you repeat the other question a little bit more?

Yes. Like I said, I was just trying to think about, going forward, is kind of the 90 basis points of underlying margin improvement you saw in Commercial Lines this quarter, is that the right level we should be thinking about or maybe adjust a little bit just for the favorable non-CAT weather in the quarter?

Well, first off, I think the improvement of 90 basis points was a terrific quarter. And so I'd love to think we could outperform like that going forward. But that is -- that will take experience in the next couple of months for us to be able to determine that. I am suggesting that we're seeing a little bit of turn in our frequency in the workers' comp line. So we expect to see some compression there because rates are moving in one direction and frequency is moving in the other. So we're watching carefully what that means to our book of business, and we'll take appropriate actions going forward. But as you know, there are headwinds on the pricing side in workers' comp because those loss cost trends, over the last several years, are so favorable that they're dropping in state by state to the pricing algorithms. And so we're making adjustments as appropriate there. So feel great about our improvement in the quarter. I'd love to think we could continue, but that would -- that's going to take some time for us to show that in -- through the P&L.

Okay, great. And then on the capital side, so I appreciate the disclosure about how you guys are now evaluating M&A as well as capital return. I guess I also want to -- was wondering how you guys also think about managing down your leverage. Pro forma for the debt maturity that comes due early next year, your leverage is probably still running in the high 20s. So how does managing down your leverage balanced against if you end up with excess capital and how you're thinking about potential for buybacks?

Yes, great question. So as we talked about before, we do have maturing debt, as you pointed out, at the beginning of 2019, which our current intention would be to pay that down. And when we look at that, combined with just the earnings power of our businesses, we believe that puts us on a very good track as we think about managing that ratio down. So I think we've done a good job in the past of using maturing debt -- the opportunity for that without having to pay a large premium to reduce our debt outstanding. And I'd say just we will continue on that path.

Your next question comes from the line of Tom Gallagher with Evercore.

First question, Chris, just on the capital management commentary you made. Does the priority of M&A over buybacks suggest that you'll -- you see current attractive opportunities in M&A? Or is that more of a medium-term comment, that you'll patiently look for opportunities and like to build a bigger capital position while you wait for those opportunities?

Yes. Thanks, Tom. Again, we're just talking about things that we've talked about before. So I would say, again, from the building the business, investing in our business, looking at acquisition opportunities, I mean, that's been our consistent philosophy for a number of years now. So I'm not signaling any change one way or other. It's just sort of our playbook of priorities that we would go down and explore. And as I said in my prepared remarks, if we can't find a good use for our excess capital, we're more than comfortable in returning it to shareholders.

Got you. So really no change, just consistent with getting maybe to a bigger level of excess before you'd consider doing -- using it for alternatives like buybacks?

Yes. I'm not going to try to sort of size the level of capital here. I mean, we just closed Talcott. We've gotten excess capital. You've heard Beth just comment about what we want to do with debt. We're comfortable where we are right now in giving ourselves a little time. I mean, it's not -- we're not looking for years and years and years here. We just want the option and the flexibility to explore, using our capital to invest in businesses and/or new revenue streams.

Got it. And then just a question on Group Benefits. From the disclosure in your Q, it looks like you had another good quarter of very favorable prior year development. By our estimate, it's more than half of the earnings for that segment. Now you've been having that favorable development for several years now, so it doesn't appear to be a onetimer, seems pretty sustainable. So my question is, should we think about most of that development being recent in accident year releases? Or is a lot of that coming from recoveries of older accident years? Can you provide some perspective on how that -- where that's coming from and then maybe the sustainability of it?

Yes, sure. It's Beth. I'll take that. So again, on the group side and specifically on the disability side, as we look at our trends, we have been seeing favorable trends, more so in the more recent years. As those exposures develop, I think, in the disability block, it's important to remember that there's a lag in the timing of when someone goes on disability and then when they actually end up on long-term disability. So we peg those lines and then look at how the development comes in over time. And our incident rates and our recoveries on all those fronts have been very favorable. It's hard to predict obviously going forward, but we're very happy with the trends that we've been seeing as we kind of go into the second half of this year.

Your next question comes from the line of Josh Shanker with Deutsche Bank.

I don't know if I'm going to get a better answer than the prepared remarks, but you talked about being more confident about the cost savings associated with the Aetna transaction. You were formerly at $100 million. I look at about $400 million of run rate expenses coming with Aetna. Can you sort of break out what's in that $400 million and why we couldn't expect half of it to go down or how we should think about that?

Josh, let me just provide an opening commentary, and then we could add additional color. I mean, we're, as I said, in essence 8, 9 months into the Aetna integration. We feel very good about the integration, both from an operational side, a go-to-market sales force side, customer retention side. And all we're signaling right now is that our 9-month indication is that we most likely will outperform our $100 million target over a longer period of time. And if you remember, why we say a longer period of time is we're timing the conversion of a lot of these policies and books of business from Aetna paper to our paper over a 2-year period of time so that we don't disrupt that customer base. We just completed installing their claims system on our technology and our hardware here. So the real conversion process just begins. So it is a little longer term than maybe a typical integration activity given we're dealing with 3-year rate policy -- 3-year rate guarantees and moving the entire administration platform into our network and our capabilities right now. But I'd like to -- yes, Doug, if you want to provide any additional commentary, Doug.

No, I think, Chris, that's a really good baseline. Josh, the number I have in my head for the Aetna baseline on costs is roughly $330 million. So when I think about $100 million or $100-plus million, pretty significant change. In addition, what Chris said relative to us moving accounts, we are being very careful trying not to disturb relationships of account managers and account executives on key Aetna renewals. So this is a multiyear process that I think we are very pleased about the initial 9 months, but it is a several-year effort. And as we move through that period of time, I know we're going to find opportunities to be more effective and efficient in operations, and we'll capitalize on those.

And look, there's always the numbers. You gave us good detail on what's going on in workers' comp. You took down reserves there, but you took up your accident year pick, and then on GL, you added to reserves and then took down your accident year pick. I'm just sort of wondering how past information is different from what you're doing on the current accident year?

Yes, really good question. So let me attack both comp and GL separately. When we think about our comp book of business and our reserves that we currently carry, we feel very good about the adequacy of our reserves. And in fact, this quarter, we did release some of those prior year reserves, primarily accident years '15 -- '14 and '15. But our position on the balance sheet is very solid, and feel good about that. When we're talking about accident year '18, we're starting to see some headwinds, and we're looking at frequency, which is a leading indicator, and so we've made an adjustment in Middle Market. But we're connecting the dots through the accident years. And I just want to point out, there's a difference between what we're carrying and what we're seeing today. So that's really what's happening in the comp world. In the general liability world, our normal GL book is performing according to our expectations. What we did in the quarter, we have a specialized high-hazard, heavy products group in Middle Market, and that is the book that we saw some increase in both frequency and severity in older accident years. So we took action to strengthen those years. That was our high-hazard book, and I want to differentiate that from what we're seeing in our normal go-forward GL book. And there, we continue to watch but feel pretty good about current conditions.

Your next question comes from the line of Jay Cohen with Bank of America Merrill Lynch.

Jay Cohen here. A couple of questions. First is, the commercial insurance expense ratio, that did pick up. I understand, on the Personal Lines side, you're spending more from a marketing standpoint. What's driving the higher expense ratio on the Commercial side?

Jay, I'm looking at Doug. He will add his commentary, but I would just say, we continue to invest in our infrastructure technology and digital experience for the customer. So yes, it is elevated from trends over the years. But again, it's a conscious part of our strategy. I would say, particularly from the Commercial Lines capabilities side, we're probably 60%, 70% through of some of the core systems that we want to replace. And I would say, there are always true-ups from quarter to quarter on commissions, whether it in essence be profit sharing or anything else commission-wise. So Doug, that's what I would say, but what would you add?

Yes. It's a large piece, and there's a little bit of compensation there, Chris, as well. So as we look at plans and we look at performance through 6 months, just some true-ups that we normally do, but IT is a driver, and continue to invest inside our businesses.

Got it, makes sense. The other question, on M&A, without issuing stock and given your leverage, can you give us a sense of how big a deal in dollars you could do at this point?

I mean, let's just say, yes, at this point -- and really, Jay, I'm going to sort of probably disappoint you a little bit and refrain really from a lot of speculation here because it just doesn't do anyone any good. But I think the metrics that we've talked about in the past really are, in the premium range, sort of a $2 billion premium company is still accurate as a target as a bolt-on. So that's what we define. So that's what I would just say right now, is that we're still in that bolt-on category, and that's what I would leave you with.

Your next question comes from the line of Brian Meredith with UBS.

So a couple of questions here. Chris, just quickly, back on the whole M&A thing, can you remind us how you think about kind of GAAP earnings accretion when you kind of balance M&A versus share buyback?

Yes, happy to, Brian. I mean, again, we tried to address it in an early-on question. I mean, it's part of the equation on the metrics, on time frames, payback periods, IRRs, returns on tangible, intangible capital. So we look at it all. And I'm not going to tell you, there's a hard-and-fast rule, but we do -- we want to earn acceptable returns. As we always defined, it's above our cost of equity capital in a relatively near term. And we define that term somewhere in the 3- to 4-year period of time. So that's what I would say right now.

Great. And then, Doug, I'm just curious, looking at your Small Commercial business, policy account retention just continued to kind of slip for the last, call it, 1.5 years or 2 years. What's going on there? And what are you guys doing to maybe improve that policy account retention?

Brian, we are -- we're certainly focused on both the retention and the new. There's a little bit of pressure in the micro space. I think we've seen more entrants in the micro end of Small. But there isn't anything material that I would point out that is worthy of spending a lot more time on this morning. Overall, we've been very steady growers of this business organically over the past 5 years. I know in the quarter, we were a little off. There's a workers' comp dynamic to it. There's competition to it. But we're being thoughtful about this business. We continue to innovate. Some of the innovations that we're dropping into our platform, I do expect will show progress and growth in future quarters. So I don't look at this as the full trend for the next couple of quarters, but I do remind you that workers' comp is an important part of our Small Commercial platform, and therefore, we're going to be in a different pricing environment over the next couple of quarters than what we've seen in the prior probably 2 to 3 years.

And Brian, I would also add, too, some of that trend, you're talking about policy accounts and retentions, influenced by the commercial auto environment broadly. So I mean, there's been a little bit of -- my words would be pruning of monoline commercial auto, so that's affecting those trends a little bit. And as you know, that market is still not at adequate returns, so we've been very thoughtful about putting additional premiums on -- in that line of business.

Your next question comes from the line of Gary Ransom with Dowling & Partners.

I was going to ask for a little more detail on underlying loss trends. You did give us a little bit of a picture on workers' comp and high-hazard liability lines. But I'm wondering if you could just give us a sense of the entire array of what you're seeing. I mean, is workers' comp the good end of the loss trends, still relatively benign? And then what's at the worse end, from your perspective? And -- is it commercial auto? Or is it some other liability line?

I would say that the reason I made mention of workers' comp is we do have full attention on this frequency dynamic. As I think across the rest of the lines in Commercial, yes, we have a little spike in our high-hazard GL in our prior book. But generally, all of our other lines are still in a relatively benign low single-digit loss trend environment. So I don't think a lot has changed as we looked at the quarter in our loss performance relative to trends. And I spoke about the line that we saw some degree of change in workers' comp.

Yes. So there's a lot of talk about inflation, not just you but other companies as well. And yet it doesn't really seem to be showing up in the numbers. Everyone's concerned about it, but it's not quite there yet. Is that something you're able to respond to realistically? If you have the concern of inflation, can you raise rates in Small Commercial? Or is that part of what's competitive activity is not allowing you to do that?

Well, I shared our ex comp pricing in the quarter at 5 plus. And feel very good about that, and I feel good about the Small Commercial component of that and the Middle Market. So we've been working at not only comp but the other elements of our book of business from a pricing perspective over time. And I think our performance demonstrates the progress of that pricing, yet it is a competitive marketplace. But the quarter we just punched, I feel really good about. I think in terms of absolute performance, a very strong quarter, Gary. And I would say we've been able to price for what we've seen in the marketplace relative to trends successfully over the last couple of years.

No doubt today was a great day, but I'm always thinking about tomorrow and what comes next.

I know, so are we.

So are we.

Your next question comes from the line of Ryan Tunis with Autonomous Research.

Just a couple of, I guess, follow-ups on workers' comp for Doug. I guess the first one, just thinking about the commentary on accident year margin and pricing in workers' comp being a function of a few things. Is it safe to say, though, it's still mostly the fact that pricing isn't as good, that's probably the biggest reason why you're talking about the margin headwinds there?

What I would say is that in '18, the biggest reason is really twofold. One is yes, we're watching these loss costs drop in, and we're dealing with the state-by-state dynamics of where our loss experience is and what to do about our multiplier in these various states. So that's point A. Point B is that we've got a bit of an inflection on frequency that we're watching very carefully. So 2 quarters don't make a full trend, but we've had a couple of quarters now of positive frequency, and that's the first time we've seen that in several years. So there's full attention on what our own book of business is telling us relative to signals in frequency. As I mentioned, the severity signals, well within our expectations, so we're watching severity, both medical and indemnity, but feel pretty good about that. Just looking at the combination of both pricing and frequency, we are very focused on choices and options in front of us relative to workers' compensation.

Got it. And so you've seen a couple of quarters of positive frequency. Should I take that to mean that -- I guess you're assuming in your new loss picks that there's positive frequency in workers' comp?

So I don't want to spend too much time talking about reserving process, but in general, we're looking at earn patterns. We use historical, and we bring in current year both severity and frequency as appropriate. So the reason we adjusted Middle is that we're trying to make sure we're recognizing what we're seeing in our patterns in the first 2 accident quarters of 2018. And in the Middle, we're seeing a pickup greater than we expected, which is why we adjusted our reserves. We'll have to continue to assess what third quarter and fourth quarter bring. But at the moment, we made adjustments based on everything we could see in our book of business to make sure we closed out second quarter where we should have been from a loss ratio perspective.

Got it. And then I guess the last one I had was just thinking about some of the favorable results in the other casualty lines this quarter like general liability, commercial auto. And I remember you guys adding the reserves in those areas in, I think, what, '14, '15, at one point now as sort of a headwind. Is the reason those results are getting better, because you're finding out now that some of the things that you saw back then didn't end up being quite as negative? Or is there something else driving that?

There are probably lots of things driving that. I think our behavior and our discipline in the marketplace starts that discussion. I think we've become a very solid, thoughtful underwriter using both skill sets at desk level and also data and analytics. So I'd start there. Secondly, I think, with Beth and Chris over the last 7, 8 years, we've worked hard to be disciplined on the reserving to bring forward, to be more current as we're looking at data. So I think our entire reserving process is much stronger today than it was over the last 10 years. Putting all of that together, we're also trying to be very consistent on our approach quarter-to-quarter. And the reflections of all of those behaviors, I think, lead us to feel much better about our balance sheet today than we probably did 7 years ago.

Your next question comes from the line of Ian Gutterman with Balyasny.

Beth, can we start out just -- if I recall from the Q, we have $2.3 billion at the holdco now. Can we just walk through what's left to come as far as the activities in the second half versus the first? I don't know, I know it's separate for the year, but I don't know what's been taken already versus what's left to come.

Sure. So again, I'll remind you that with some of the actions that we did at the end of last year with the Aetna acquisition, that really decreased our dividend capacity for 2018. So we are not expecting dividends from P&C or Group Benefits in the second half of this year. We would expect to see a modest amount from Mutual Funds. But for the most part, there is not additional dividends coming in from the subsidiaries in the second half of this year. And for P&C, because some of the Talcott proceeds actually went through various legal entities to get to the holding company, our dividend capacity for ordinary dividends in P&C really probably for the most part won't be there until the second half of 2019.

Got it. And then do we have any tax sharing payments coming in second half? Or did those all come in first half?

Yes, so a little bit. I mean, obviously, that will be dependent upon actual taxable income forecast for the second half of the year. But based on our current estimates, I'd anticipate probably another $150 million maybe will come in, in the second half of '18 to the holding company. Again, that can bounce around just based on actual results. And then we've also, in the past, highlighted the fact that we will -- we do anticipate a refund coming in, in '19 for our AMT credits. And that will come in when we file our tax return, which could be as late as September of 2019.

Okay. I was going to ask about that one, too. So if I take the tax sharing minus the corporate dividends, it would suggest you end the year somewhere around the $2.3 billion you're at right now at the holdco?

Yes, maybe a little bit less than that depending on other corporate actions. I'm sorry, hit my mic off by mistake. Yes, maybe a little bit less than that depending on other things that we might do relative to contributions to our pension plan that we make usually in the third quarter. But I'd expect to be roughly around $2 billion.

Got it, okay. And that's without the AMT because that doesn't hit cash until '19?

Right, that would not be in the 2018 numbers.

Perfect, okay. So Chris, I guess I was hoping you could talk a little bit more. I know you've already commented, but on the capital side, I guess I don't understand the harm in putting an authorization out there even -- you don't have to necessarily commit to using it tomorrow. But let's just say that what happened with Facebook yesterday or the president does something or whatever, and the market's down 20% within the next 3 months, wouldn't you want to have an authorization out there? And why not have it out there as an option? Just because you put it out there doesn't say you have to use the whole thing.

Thanks. Ian, I understand your point of view, I do, and you've communicated it clearly. I guess the simplest way I can explain it is, given our real intention -- and I understand different scenarios that you just pointed out, but I wanted to be as crystal clear as possible, if we weren't going to be buying shares and we wanted time to continue to deploy that capital into revenue streams if possible, we didn't want to create any confusion. So that's the simplest way as I could say it. And I didn't really want to signal that we were going to be in the market. And you did the math on the holdco. We're not sitting on a lot of excess capital today. It does build over time. So as we sit here and now and project in the near term, I just didn't want to confuse anyone.

No, that's fair. I guess the one thing I'd push back on a little bit is that you don't have a lot of excess capital today. I mean, it's a significant part of your market cap, right? And you could do another Aetna and still be fine and then have healthy dividends for '19 plus AMT plus tax sharing, right? I mean, it's generally -- when we project to the end of '19, you get to some pretty significant numbers. So I mean, can you give us a sense of the timetable? Is this something every quarter, we should expect another update on whether there'd be a change or we're going to -- maybe you won't address it again until you give guidance for '19? Or sort of when should we expect another update is maybe the best way to ask it.

Well, Ian, as you pointed out, I mean, the end of '19 is 6 quarters away. We'll have a lot of opportunities to communicate and interact and keep you posted. As you know, I think we're very transparent, so all I would ask you to be is patient, and we'll keep you posted.

And your next question comes from the line of Jay Gelb with Barclays.

I will not ask about buybacks. How about that? First, on pace of reserve releases, it's 5 consecutive quarters that Hartford has been able to put up overall reserve releases. Is this something you think we should start baking in on a go-forward basis?

So Jay, it's Beth. I mean, we've said this before. I mean, we evaluate our reserves every quarter, and we make adjustments accordingly. We don't predict whether or not there'll be future reserve releases. We've been very pleased with the underlying trends that we've seen. And we'll just continue to evaluate it every quarter line by line, which is what we do, and give you the transparency as to where we're seeing either improvements or areas that we need to add.

I see, okay. And then broadly, on asbestos, with the new talc-related exposure for J&J, just trying to think about how that might affect Hartford if there were some old-occurrence liability policies out there from decades ago. And I know the company typically does its annual review in the fourth quarter, and it does have the adverse development cover in place with Berkshire. But just wanted to get your -- any broad thoughts you might have on talc-related exposure.

Sure. So a couple of things. I'm sure it won't surprise you to know that we have a team that is constantly looking at emerging tort issues. And the alleged connection between talc and ovarian cancer has been on our radar for quite some time, and we take into consideration all the facts that we know as we evaluate our reserves and overall feel very good about where our reserves stand. As it relates to our adverse coverages, one thing that I will point out, we do have the adverse cover with Berkshire. But specifically, alleged connections between talc and ovarian cancer and exposure there is specifically excluded from that cover. And we obviously take that into -- we take -- pardon me?

I'm sorry to cut you off. Why is that? Why would it have been excluded?

That was part of the contracts and what we negotiated, so that was specifically excluded.

All right. So I guess I can imply it was known about at that time when the deal was...

Known that we excluded it?

Known that it was a potential exposure.

Yes...

Jay, as Beth said, we've been following this. I mean, we have a world-class claims team and particularly a mass tort team. So a lot of these things aren't new to us, and we've been on it for a while.

Your final question today will come from the line of Randy Binner with B. Riley FBR.

Yes. I just had a couple of follow-ups real quick. Did you cover specifically on commercial auto where you think price versus loss cost is now? And then the second one, on the frequency in workers' comp, I'm not sure I actually heard what it is. Is it people getting in car accidents while they're at the job? Or is there some kind of slip-and-fall thing happening out there? You alluded to less experienced workers, but if there's a thing that's actually happening that causes the frequency at work, I'd be interested in that.

Okay. Randy, let me take them each separately. On the commercial auto front, we're still getting strong single-digit pricing in auto, and I expect that to be on top of trends. So good news there from commercial auto. Still more work to be done, but good news in terms of where we are in the current quarter. Relative to frequency, I suggested that we believe there's some macro factors across the industry relative to employment. And inexperienced workers that is driving the trend, it's going to take time for us to mature those observations, and we're spending a lot of time looking at SIC class, geographies, size of risk, et cetera, et cetera. But know, we're cross-cutting the data very hard, and at the moment, it looks a little more broad-based than just a couple of classes. And we see this inexperienced worker dynamic where they tend to be injured in a more frequent basis than more experienced workers. I don't think that's a surprise to anybody on the call. It's just the fact of something, as underwriters, we have to deal with every day in our risk business.

Okay, but you can't point to it being -- I mean, so we've seen auto accidents creep into some other workers' comp companies. So I'm just -- I'm trying to just isolate that, and I guess the answer is you don't know. And then back on commercial auto, you said you're over trying to buy it. Like how many -- roughly how many basis points do you think you're over -- pricing over loss cost in commercial auto?

Yes. Let me just go back to your first point. At this point, I don't see the full connection. I don't think auto is driving our workers' comp frequency increase. And we'll continue to study it, and if it changes, I'll share that going forward. And secondly, we don't share specifically exactly those points. But it's a couple of points over top of loss trend at this point relative to pricing versus trend.

But Doug, I'd make the point, too, that it's still not anywhere near where we want to be from a long-term return point of view. So the current year and maybe the last 18 months, we were outearning loss trends. But there's still a ways to go to get to an overall acceptable combined ratio.

I agree with that, Chris.

Thank you all for joining us today. I appreciate the attention to our earnings results and all the questions that you have. Sean and I will follow up with those of you who were still in the queue after the call. Thank you very much, and have -- hope you have a great summer weekend.

Thank you to everyone for attending today. This will conclude today's call, and you may now disconnect.