The Travelers Companies, Inc. (TRV) Q3 2013 Earnings Report, Transcript and Summary

Good morning, ladies and gentlemen. Welcome to the third quarter results teleconference for Travelers. [Operator Instructions] As a reminder, this conference is being recorded on Tuesday, October 22, 2013. At this time, I would like to turn the conference over to Ms. Gabriella Nawi, Senior Vice President of Investor Relations. Ms. Nawi, you may now begin.

Thank you, Tina. Good morning, and welcome to Travelers' discussion of our third quarter 2013 results. Hopefully, all of you have seen our press release, financial supplement and webcast presentation released earlier this morning. All of these materials can be found on our website at www.travelers.com under the Investors section. Speaking today will be Jay Fishman, Chairman and CEO; Jay Benet, Vice Chairman and Chief Financial Officer; and Brian MacLean, President and Chief Operating Officer. Other members of senior management are also available for the question-and-answer period. They will discuss the financial results of our business and the current market environment. They will refer to the webcast presentation as they go through prepared remarks, and then we will open it for questions. Before I turn it over to Jay, I'd like to draw your attention to the explanatory note included at the end of the webcast. Our presentation today includes forward-looking statements. The company cautions investors that any forward-looking statement involves risks and uncertainties and is not a guarantee of future performance. Actual results may differ materially from those projected in the forward-looking statements due to a variety of factors. These factors are described in our earnings press release and in our most recent 10-Q and 10-K filed with the SEC. We do not undertake any obligation to update forward-looking statements. Also, in our remarks and responses to questions, we may make mention some non-GAAP financial measures. Reconciliations are included in our recent earnings press release, financial supplement and other materials that are available in the Investors section on our website. And now, Jay Fishman.

Thank you, Gabby. Well, good morning, everyone, and thank you for joining us today. Our financial results this quarter were really strong. We posted quarterly operating income of $883 million, a 15.2% operating return on equity and record quarterly operating income per diluted share of $2.35. Net income was $864 million, or $2.30 per diluted share, producing a 13.9% return on equity. In the quarter, we returned nearly $1 billion of capital to shareholders, continuing to deliver on our long-term financial goal of producing strong profitability and returning excess capital to shareholders. While the current quarter certainly benefited from low weather-related losses, most importantly our underlying underwriting profitability continued to improve across our businesses, largely the result of the successful execution of the pricing and underwriting actions we have been pursuing over the last several years. While we are really pleased with our financial results not just in the current quarter or year-to-date but over now more than 5 years, we continue to expect to be confronted with more volatile weather patterns and historically low interest rates. As it relates to weather, notwithstanding the fact that in this quarter we had very low catastrophe losses, on a year-to-date basis, our pre-tax cat losses now stand at nearly $540 million. To put that in perspective, that number, again a 9-month figure with a fairly benign third quarter, approaches the $600 million of catastrophe losses that we provided at the end of 2009 in connection with our full year 2010 guidance. It is worth noting that over that period of time, we believe that our catastrophe exposure has not changed meaningfully, but losses have risen as a result of increased frequency and severity of weather events. An interesting way of observing this phenomena is through the data on Slide 4 of our webcast where we have displayed the number of extreme weather events with economic losses over $1 billion by year over the last 30-plus years. While this data does adjust losses for inflation, it does not take into account exposure growth over time. Nonetheless, it is hard not to take notice of the trend. We, of course, evaluate weather over the long term and continue to believe that more recent weather patterns are different and cause us to anticipate higher levels of frequency and severity of weather-related losses. As to fixed-income yields, we continue to believe that interest rates will remain at historically low levels, at least until 2016. Although the long end of the yield curve has moved somewhat, it is not yet enough to make a difference in the return profile of our business in any significant way. Total yields remain just off historical lows. Consequently, in our commercial lines businesses, we will continue to pursue improved profitability and product returns by seeking higher rates on an account by account basis so as to continue to minimize disruption to our agents and customers. So far, this strategy has worked exceptionally well, and we expect that at least for some time across our portfolio, we will be able to continue to achieve written rate gains in excess of estimated loss trend. In Personal Insurance, we're excited about the launch of our new product, Quantum 2.0, during the fourth quarter of this year. We are hopeful that this lower cost, lower-priced Auto insurance product will allow us to improve volumes at attractive returns. We will know more as we get into 2014. And with that, let me turn it over to Jay.

Thank you, Jay. Let me begin by saying that our third quarter results, operating income of $2.35 per diluted share, a record level, and operating ROE of 15.2%, were very strong, having benefited from solid investment and underwriting performance. Our combined ratio of 88.9% was also very strong, reflecting earned rate increases that exceeded loss cost trends in each of our business segments, a relatively low level of both cat and non-cat weather-related losses and $158 million of net favorable prior year reserve development, which was net of an increase to our asbestos reserves. As has been the case in recent periods while the underlying asbestos environment has remained essentially unchanged, payment trends have been moderately higher than what we had expected. For the quarter, excluding the asbestos provision, which, in a 10-year table, would be included in the prior to 2004 accident year row, each accident year developed favorably, and each product line developed favorably or was essentially unchanged. We continue to generate capital in excess of what is needed to support our businesses, and, consistent with our strategy, we have returned a very significant amount of capital to our shareholders. Operating cash flows in excess of $1.6 billion were extremely strong this quarter, bringing total operating cash flows to almost $2.9 billion year-to-date. We ended the quarter with holding company liquidity of approximately $1.9 billion after returning almost $1 billion of excess capital to our shareholders through dividends of $185 million and common share repurchases of $800 million. Year-to-date, we returned almost $2 billion of excess capital to our shareholders through dividends of $552 million and common share repurchases of $1.4 billion, leaving us with only $759 million remaining under previous share repurchase authorizations as of the end of the third quarter. Since we may exhaust this amount before the end of the year, as announced today our Board of Directors authorized an additional $5 billion of share repurchases. All of our capital ratios remained at or better than their target levels. Our debt-to-total cap ratio of 21.4% was well within its target range, although up slightly in the quarter due to our successfully issuing $500 million of 30-year, 4.6% senior debt at the end of July. Book value per share increased 2%, and adjusted book value per share, which excludes unrealized investment gains and losses, increased 3% during the quarter. While year-to-date, book value per share increased 1%, and adjusted book value per share increased 8%. The difference between the unadjusted and adjusted year-to-date growth rates in book value per share was driven by the impact that the recent rise in interest rates had on net unrealized investment gains. Net unrealized investment gains were approximately $1.6 billion at the end of the third quarter as compared to $3.1 billion at the beginning of the year. So with that, let me turn the mic over to Brian, who's going to comment further on operating results.

Thanks, Jay. In Business Insurance, we had a great quarter with operating income of $526 million and a combined ratio of 93%. The underlying combined ratio, which excludes the impact of cats and prior year development, improved by about 1 point year-over-year and reflected about 2.5 points of margin expansion due to earned rate increases that exceeded loss cost trends. When comparing the combined ratios quarter-over-quarter, any one quarter can be lumpy. This quarter, the comparison was negatively impacted by an unusually low level of non-cat, weather-related losses in the third quarter 1 year ago and, to a lesser extent, a couple of non-run rate adjustments that helped last year's combined ratio. On a year-to-date basis, the underlying combined ratio improved nearly 3 points due primarily to the earned rate increases that exceeded loss cost trends, and this is a better indicator of how the business is performing. Looking at the production results for the segment on Page 10, retention and renewal premium change remained strong and in line with recent periods at 80% and 9%, respectively, while new business volume was up slightly versus the prior year. The renewal premium change was driven by pure rate increases of 7.1% and exposure and other of about 2%. Rate increases continued to be broad based and range from 6% to 10% across all lines, led by Commercial Auto and Worker's Compensation. So in the aggregate, the production metrics in the quarter were very stable across the segment, and we continue to be pleased with this result. But as we've emphasized in the past, the detail behind how we got to the aggregate result is what's really important, and we remain very pleased with our granular execution as well. The segmented rate and retention results remain largely consistent with what we've discussed in prior quarters. For our best-performing business, retention remained very high with written rate gains exceeding loss trend, while we continue to get significant rate increases on our poor-performing accounts. So overall, a great story in the segment. We saw meaningful underlying underwriting margin expansion due to earned rate gains that continued to significantly exceed the loss trend, which remained consistent at approximately 4%. Going forward, we will continue to execute on our strategy of improving returns by maximizing the rate and retention trade-off at a very granular level. In Financial, Professional & International Insurance, operating income of $160 million was very strong, though down 11% from last year. The prior year quarter benefited from a particularly low level of large losses, which, along with year-over-year lower net favorable prior year reserve development, drove the decrease. Similarly, the underlying combined ratio of 92.5% deteriorated by 1.1 points due to the higher level of large losses in the third quarter of 2013, largely offset by earned rate that exceeded loss cost trend in both our bond and financial products and international businesses. Written premiums in this segment were up 6% compared to the prior year quarter, driven by continued favorable rate in Management Liability, strong construction surety results and higher new business in International. In Personal Insurance, operating income of $262 million was up 27% versus the third quarter of 2012, while the underlying combined ratio for the quarter improved nearly 6 points year-over-year, driven by lower weather-related losses along with nearly 2 points attributable to earned rate increases that exceeded loss cost trends. Looking specifically at Auto production, retention of 81% kicked up from recent periods, renewal premium change was over 7%, while new business volumes and net written premiums were down year-over-year. Turning to Auto profitability, the underlying combined ratio of 97.6% was an improvement of 0.5 point versus the prior year quarter. The underlying loss ratio improved about 1.5 points, primarily reflecting earned rates that exceeded loss cost trend, while the expense ratio was up about 1 point due to some unusual items, including a charge related to the expense action we discussed last quarter. Excluding the impact of these items, the expense ratio would have been about flat quarter-over-quarter. Auto loss cost trends remain stable at about 5% overall, mix-adjusted frequency continue to be benign, while severity trend remain stable at a slightly elevated level. Specifically, bodily injury severity trend this quarter remained in line with what we've seen over the past few quarters. I'd like to take a moment to talk about Quantum 2.0, our new Auto product which we began rolling out this week and will be rolling out in approximately 15 states during the fourth quarter. Our goal is to offer a lower cost, highly sophisticated and segmented auto product that successfully competes in the marketplace and generates an appropriate return. Quantum 2.0 helps us to achieve that goal by leveraging 8 years of data and experience from Quantum 1.0 to create a more granular and highly segmented product. It will also benefit from a significantly lower cost structure, thanks to both the expense reduction that we announced last quarter, along with a base commission rate that is approximately 2 points slower than Quantum 1.0. When fully realized, the combined impact of these savings will be about a 12% reduction in the non-loss cost base for the product. So a significant impact. We're very excited about this product. We believe that Quantum 2.0 will allow us to improve our competitive position, drive increased business volumes and deliver improved returns. Turning to Homeowners, pricing was also very strong in the quarter with renewal premium change of 11%, retention stable at 83% and new business volume that was up slightly from the prior year quarter. The underlying combined ratio for Home was 71.6% in the quarter, an improvement of over 7 points year-over-year. The primary driver of the improvement was lower non-cat, weather-related losses, but we also saw over 1 point of improvement due to earned rate increases that exceeded loss cost trend. So a very strong underwriting result for the segmented in the quarter as we continue to see the positive impact of the underwriting and pricing actions we've taken. With that, let me turn it back over to Gabby.

Thank you. Tina, we're ready for the question-and-answer period. [Operator Instructions] Thank you. Tina?

[Operator Instructions] And our first question comes from Randy Binner of FBR. Randy Binner - FBR Capital Markets & Co., Research Division: I wanted just to go to Page 10. And given that there's so much kind of focus on the red line there, the renewal rate change, I wanted to understand why the previous historical numbers in that chart can kind of adjust down a little bit each time we see this presentation. What drives kind of a revision to those prior quarter numbers each time you update this illustration?

Right, Randy, this is Jay Fishman. I'll start, and then I'll have Brian chime in as well. First, to your -- we enter into a month's production period. Policies renew, policies don't renew. But it doesn't end as of the reporting date. There can be transactions, cancellations, endorsements, changes in pricing even that occur after the reporting date. So we're always going back and making not adjustments but truing up the reported numbers to what the most recent data actually is. Now I'd make a point that the rate change, the rate change actually is far less than the exposure change will be in a given period. Exposure has audit adjustments embedded in it and, as a consequence, as those are reflected, they will have more change. The rate change, if you look back, has actually been relatively modest and generally is a result of cancellations, early cancellations midterms or other endorsements. But I'll ask Brian to chime in a little, too.

Yes. And one of the things I'll just emphasize is that at the top of the page, it says, "Illustrative Business Statistics." So this isn't an accounting number that's getting booked into the financials. This is us looking at our written production results. And it's -- but it's exactly what Jay just said. When we close the books at a given period of time, we try to make our best assessment of what the production is going to be, but there's also always movement in the accounts. And the adjustments tend to be fairly consistent, but they move around. I don't know, Bill, any -- yes. Randy Binner - FBR Capital Markets & Co., Research Division: Just one quick follow-up. And I appreciate that you think that the market's focus on this is too myopic, but I guess relative to the number of 7.2% that you gave at Barclays versus 7.1% now. Is it more just kind of the chatter of those adjustments that Jay mentioned? Or was there actually kind of a lower rate in that last month for the quarter?

Well, the Barclays number was a 2-month number. At least it was as of whatever date that conference was. This is obviously the full quarter. What I don't know is, if we had updated those 2-month numbers, whether they would have reflected any change for midterm cancellations or endorsements. So I honestly don't know, and I agree with your assessment that you're looking at it too myopically. I don't know the answer.

Our next question comes from Mike Zaremski of Crédit Suisse. Michael Zaremski - Crédit Suisse AG, Research Division: In the Business Insurance segment, the underlying loss ratio, not the combined ratio, it looks to have incrementally deteriorated year-over-year. I would have thought it would have improved given pricing continued to trend in the high single digits. I know high or non-cat weather losses were cited. So is that the main culprit? Or are loss cost inflation rates creeping higher as well?

So this is Brian and then I'll let Jay Benet comment. But no, loss cost trends are not creeping. They've remained really flat at right around 4%. Obviously, little pieces move up and down, but the overall loss trend has been very, very consistent. The biggest driver in there is the non-cat weather. Jay, I don't know if you want to go through more of the detail, but that's the biggest change that flows through that.

Yes. No, I would add only a little bit to what Brian said. I mean, if you're looking at quarter-to-quarter comparisons, you're always going to see some variances associated with things like non-cat weather impacting the data, which is why in Brian's comments he really provides the information as to what underlying is -- underlying the underlying loss ratio is taking place with regard to the margin expansion relative to earned pricing versus loss trend. Michael Zaremski - Crédit Suisse AG, Research Division: Okay, great. And then lastly, could you update us on the gap between the new money yields and the existing portfolio yield? And I ask because the fixed-income portfolio yield has actually been steady for 3 sequential quarters now.

Well, the gap is -- there still is a gap between new money yields and...

By the way, this is Bill Heyman speaking.

Oh, hi. Sorry, yes. And yields on securities which are maturing. It's closing. It depends on which asset class you're looking at, corporates and municipals. But in -- let's take long-term municipals. The raw yield, which is an after-tax yield, of new money is still probably 0.4% lower than what is maturing. Michael Zaremski - Crédit Suisse AG, Research Division: Okay, so the fixed-income portfolio stayed steady, though, in recent quarters. Is that just due to some nuances in the portfolio?

I think the -- I mean, we analyze the change in the book yield of the portfolio to the basis point every quarter, and the change in the first 9 months of this year was 11 basis points. Now in terms of the raw returns from the portfolio, that doesn't really amount to much.

Yes, and this is Jay Benet. Just -- let me just add a couple of points. If you look in our -- in the Outlook section of the 10-Q, we are projecting about a $25 million-per-quarter drag on earnings associated with the reinvestment of the portfolio -- pieces of the portfolio that are maturing at the lower rates. And I think part of what you're seeing or not seeing is the impact of just rounding to the nearest tenth as opposed to going out further when you look at those yields.

Our next question comes from Jay Gelb of Barclays.

For the $5 billion share repurchase authorization, if I think about for 2013, the company is essentially running at an annualized retained earnings less shareholder dividend pace of around $2.5 billion annually. So based on that, should we assume that, that buyback could take place of around 2 years?

Well, we've been saying for the last several years, Jay, now that return of capital is ultimately going to be tied to earnings. And so it really is a question of weather and catastrophe losses and all of the various elements. You pick the earnings -- and by the way, Jay always reminds me that we always need to take into account dollars to make pension contributions if necessary or to pay off maturing obligations that we might want to -- might not want to refinance...

Essentially to support the growth of the business.

Support the growth of the business. So there's always an additional call. But you figure the earnings, divide that into 5, whatever the number is, it's probably not an unreasonable assumption. Include dividends when you do that, by the way.

Okay. And then that $5 billion authorization, given the magnitude of that potential buyback, it appears that signals a strong degree of confidence in your overall reserve position?

Our reserve position is always our best estimate of what our reserves should be, and I think you can see how the reserves have been playing out over the last many years.

Right, okay. And then any thoughts you'd like to provide on the proposed accounting changes from FASB on GAAP for insurance companies given that they could be substantial in terms of the -- and also somewhat adverse relative to how insurers present their financial positions?

We'll always do what regulators ask us to do. And however we're asked to report, we'll report. We don't have a view. We follow directions.

Okay, and if I can add a couple of things to that. Certainly, we are always very compliant. This is a draft that's been issued. And like other companies, we are responding to the draft, talking to the FASB about what we think is good about it and suggesting areas to be improved.

Our next question comes from Michael Nannizzi of Goldman Sachs.

Just one question in BI. It looks like property grew the most this quarter. It hasn't happened for a while in terms of year-over-year premium growth. Just trying to understand. Given that, that doesn't appear where pricing is best, is that a deliberate area of growth for you? Are you seeing exposure growth? Or is that just -- or is that new business? Just a little context on that would be great.

Yes. So this is Brian. So if you really get underneath what's driving the growth in premium, it's very strong retention and it's pricing. So we're not growing our risk count there anything dramatically. And in fact, our overall pricing in the property line, our results throughout this year have been strong and actually a little bit ahead of what the total BI average is. So retention is very strong, pricing has been solid and we feel good about it, so.

Okay. And then I guess the year-over-year has been a bit of a focus in BI in terms of the margin. You cited non-cat weather there being a contributor. It sounds like it went the other way in Homeowners. But I guess the question is, is the fourth quarter -- what's the setup for the fourth quarter? Just given that the sequential -- the comp averages improved sequentially in fourth quarter versus third quarter last year. Is it going to be tough comps again because of non-cat weather in 4Q?

Well, let me go back and just clarify the non-cat weather story across the place. So it was, from a meteorological perspective, a light weather quarter. That came through very clearly in Personal lines. In Business Insurance last year, we had an adjustment that we put through in the third quarter to our current year prior quarter property loss pick last year in the third quarter, and it was a light weather quarter in BI. Remembering that Business Insurance is always going to have a little more volatility to the actual losses you could get from an event, you can have large exposures in one year. You have an event, you don't get hit. In the next year, you have one. So that's just to clarify what sounds like a difference in weather patterns between our segments. The bigger story last -- was -- in BI was the last year number was so low. The fourth quarter, I think, we're -- I don't know, Jay. We -- we're not going to get into projecting what we see. Through -- broadly speaking, the thing I would emphasize is the year-to-date, we continue to see something like 2.5, 3 points of margin expansion running through the underlying combined ratio and expect that for the full year, we're going to be somewhere around 2.5 points of margin expansion. So that's, I think, the -- what we're still looking for. And then the results in the quarter will be what they are.

Particularly, weather.

Yes. that's right.

And our next version comes from Greg Locraft of Morgan Stanley.

I wanted to just follow up on the Quantum 2.0 initiative. You mentioned it's being rolled out in the fourth quarter in 15 states. Can you talk about the average price decline? I mean -- or have you filed yet? And can you talk about how that's going to impact pricing in those 15 states?

Sure, Greg. This is Greg Toczydlowski. We look at Quantum on a state-by-state rollout basis. Brian referenced some of the expense reductions across the non-loss component of the business at the minus 12. If you think of that across the country, that's somewhere in the minus 10 to minus 13 range from a pricing perspective or the lower price of Quantum 2.0. Of course, that will vary at a very local territory level. But in aggregate, that's a good sense for you to understand what the pricing disparity will be.

Yes. And specifically, you asked have we filed. We're in the process of doing that. So as we roll out states, obviously -- obviously we've filed in a lot of places. And as we get the approvals, we're rolling out. And our plan is that by the end of June, by the end of the second quarter next year, we will essentially be fully rolled out, which means that there will always be those few states where we have a kind of a unique regulatory environment and work with a different product.

Okay, that's very clear. So just to come back, so the average price decline for this product versus before is 10% -- so this is 10% cheaper to your clients then than what they were getting before, is it? Nationwide by June, that's what you're going to have on the shelf?

That's correct, Greg. And again, that's a generalization. All states have variations, but that's kind of the high-level math.

And Greg, I'd make just 2 points. First, that is literally across the entire book. So you will find we have prices -- we showed you a graph previously, we have prices in 2.0 that are higher than 1.0, and we have prices in 2.0 that are considerably down more than 10%. I would draw your attention back to last quarter, the webcast where we gave you that...

The chart.

That spot diagram...

Yes.

That we have out there...

The Barclays presentation.

The Barclays presentation. But that's -- so it's important to remember that it's a big divergence. It's not 10% for every account. Plus, we really don't know how quickly the renewal book will move over to -- so this is -- it's new only, but at -- available to agents for renewal business where they opt for it. We just don't know yet how quickly that renewal book will roll over either, so we're in a wait-and-see pattern there as well.

Understood, understood. And what are you doing on the Homeowners side? Will Quantum pick up Homeowners business at all? Does it focus on Homeowners? Or what will you be doing there?

Well, the Quantum product is purely an Auto product. The majority of our business is account business, though, however. So we believe we will absolutely see a lift in the Homeowners production side as we roll out a more competitive Auto product.

And I'd make the observation that the decline in policies in force in the Homeowners book is meaningfully tilted towards catastrophe-focused areas where losses have been more challenging in the context of a changing weather environment. It's being driven by higher deductibles, higher pricing, but it's not -- meaning it's not uniquely a pricing issue in Homeowners. It's also a catastrophe management dynamic that sits around it. We were hopeful that it wouldn't have had as much decline as it has had, but it is what it is and we're going to price for catastrophes appropriately and manage our book accordingly.

Our next question comes from Amit Kumar of Macquarie.

So just to follow up on the last question on Quantum 2.0. If the pricing changes approximately 10%, how should we think about sort of a commensurate impact on margins?

Yes, so this is Brian again, and [indiscernible]. But we are targeting the same returns on this product that we do in our overall business. So as we look at the expense actions we're taking and the underwriting and risk selection that we're refining on the product side, we -- that'll be partly driven back to the consumer in lower prices and partly driven to ensuring appropriate margins. We're looking at both -- accomplishing both objectives. But the fundamental target returns on the business have not changed.

And what...

We may experience -- it's entirely possible that we may experience in the early days of Quantum 2.0 a return that is lower. There is -- I don't like the term because it's often confused. There is a new business penalty. It's not a premium penalty. We don't charge less for the product new than renewal, but there is a loss dynamic that takes some time to sort out in the underwriting profile. So it's very possible in the early days that the returns may not be at the level that we'd like them to be. We fully anticipate that, and we'll manage the book accordingly as we make adjustments. But Brian's point about not -- about having it priced to be the same return over time is exactly right.

Okay, that's helpful. The only other question I have is, I guess, going back to Slide 10. Has the source of new business changed? I guess what would be helpful is, is it coming from issues at your competitors? Is it a function of the economy? Or is it new niches?

Yes.

Yes.

Sorry, Amit. Just to clarify it, when you're saying Slide 10, are you asking about the new business dollar amounts? Are you asking about... .

Yes, the source of the new business on Slide 10. Has it changed over the past few quarters? Or has the mix changed? I guess that's what I'm asking.

Yes, I would say the short answer to that is no. There hasn't been any fundamental shifts in where we're getting our new business from and lines of business and industries, et cetera, so.

Over a longer period of time, there has been some shift in lines. But in the context of the last several quarters, no.

Our next question comes from Vinay Misquith of Evercore.

The first question is on the Quantum 2.0 product. I believe you're rolling it out this quarter. What are the indications right now you are getting from the agents on the new business side from that product?

We don't -- it's just -- it -- we don't have any yet, really. It's just too early.

I mean, literally, we just started rolling out this past weekend.

Yes, in 2 states.

So we're out, like, 3 days. And so it's a handful of states so far. And like we said, it'll be 15 or so by the end of the quarter. We will, in the next coming weeks, begin to get a better answer to that.

Okay, that's helpful. The second question was on the Select Accounts in that -- on Slide 11. The new business dropped off a little bit there. Just curious about the pushback you're getting from your clients because it seems that we might be running into headwinds in terms of the amount of rate increases you can get on that business. So I'm just curious about your strategy in this line.

And so Select Accounts -- and I think we've mentioned this on previous calls or at least, I think, the last call actually. Select was an area -- first of all, we think of that business as the Express world of the really small accounts and then Plus, Express being a little bit more than half of our portfolio. And that's an area where, going back a couple of years ago, we felt the returns in that business were very challenged. We were fairly aggressive in what we were doing from a pricing perspective. We feel better about the returns in the Express world now, and that is one where we are starting to shift our strategy to wanting to, first and foremost, retain more of that book. And then we'll see where the new business appetite goes. But...

I'd make just 2 other observations. First, retention in the line has been absolutely consistent. So there's -- notwithstanding renewal rate change, that's been in those -- in that high 6 area. Retention has been absolutely solid. And the difference between $97 million of new business and $89 million in the third quarter of this year versus third quarter of last, those are pretty small numbers. I -- there isn't anything that we've seen that would suggest any fundamental shift any real change going on in Select relative to the pricing environment.

Our next question comes from Josh Shanker of Deutsche Bank.

On personal auto, the nonrenewals compared -- or at least the -- policy counts down 12% from 1 year ago, and you have 9% to 10% rate increases. I would think the margin improvement would be better that you're weeding out the unprofitable drivers, getting great increases on the newer drivers. When do we have to wait for this to come through?

Josh, this is Brian. The first thing I would say is we've talked a good bit over the last probably 1.5 years to 2 years now about loss trend in personal auto and the -- particularly the bodily injury severity side. And that has been running at an elevated level, the aggregate 5% loss trend we've talked about for the last couple of quarters. So as we've earned in the rate increase, the margin expansion hasn't been quite as dramatic as it has in other businesses. So that would be the first point I'd make. The other, on the overall growth, the retention has remained pretty solid and has actually picked up a little bit the last quarter. And the real strategy here is the new product and having a more competitive product in the marketplace and, as Greg went through, the biggest lever in that is what we've done on the expense side. Historically, our loss ratio performance in this business is top quartile, the expense ratio, and that's all in with commission has been the challenge. And so we think the new product will deal with that pretty effectively.

So when I think about the new product and the legacy drivers on the older iterations. Where should we see expense ratio 1 year from today? Will it be higher or will it be lower than it is right now?

Well, expense ratio, we're -- we fully expect will be lower. So the expense actions that we took, the $140 million that we're taking out of our operations, is not product specific. That's coming out of the business. About $100 million of that $140 million, I think, was Auto, and some of it is in the personal lines franchise, so it will affect the other products as well.

I guess I'd look at 2 other things. First, even 1 year from now, the percentage that Quantum 2.0 makes up of our total book is still going to be a relatively small percentage. And so relative to the commission component, there will still be a meaningful portion of Quantum 1.0 that will still be on the books. I don't know if the expense ratio will be lower. I don't know what's going to happen to premiums. I don't know how effective the product is in restoring new business. We know what we've done in terms of expenses. That we've done, and we can be very clear and very certain about that. What happens in the marketplace and what percentage Quantum 2.0 makes of the whole book, that's yet to be determined.

Our next question comes from Meyer Shields of Keefe, Bruyette, & Woods. Meyer Shields - Keefe, Bruyette, & Woods, Inc., Research Division: I'm going to ask this maybe in a more specific way that it can be answered, but if you had the Quantum 2.0 rate decreases in effect now, do you have any sense as to what the year-over-year PIF change would have been?

No. I can tell you what we did in looking at its pricing structure was to do a fair amount of back testing. And we went back with the product, and we went back with the pricing and we went back with the sales to determine at least to be able to have a perspective on competitive profile. Now we feel pretty good about that. The back testing results were encouraging. But that doesn't take into account competitive reaction or what other companies will do or not do. So we don't know. I can tell you that the back testing caused us to be encouraged, but we only control what we control. And so I -- we don't know what the PIF would be if we had rolled it out. Meyer Shields - Keefe, Bruyette, & Woods, Inc., Research Division: Okay, I completely understand. In the 10-Q, you talk about favorable development, even on very recent years, for general liability because, in part, of the judicial environment. Can you talk a little bit about how you can tell, in the near term, whether that's actual benign emergence as opposed to just timing issue?

Yes, this is Jay Benet. When you look at longer-tail lines, you're looking at what's taking place in the current environment. But when we think about the current environment, that's not a snapshot of the last 3 months. That's a snapshot of a longer period of time. So if you see some favorable development in a long-tail line that's taking place and -- relative to recent years, what we're really talking about is going back a few years, looking at the favorable judicial environment, in this particular case that you're asking about, and then saying, okay, now that we know that from the past, what does that mean rolling forward into the loss picks that we had for the more current period? So it's not a look at, okay, what was the expectation for loss in the 2012 accident year, what are the losses there. It's a more fulsome look over periods that preceded that as to what does that mean for our loss pick for 2012. I hope that answers your question. Meyer Shields - Keefe, Bruyette, & Woods, Inc., Research Division: Yes, it does.

Our next question comes from Cliff Gallant of Nomura.

I had another question on Quantum 2.0. You said that you were cutting your commissions, I believe, 2 points. I'm just curious, how does the agent take that? Are they excited about the reduced price? So they get volume? And also, I'm curious, when you think about your total relationship with your agents, how important is that 2-point cut on just that -- on just the Auto product?

Yes, so obviously, it's a question we've thought about a lot. I'd start with, I think broadly, we. And by we, I mean the industry here. We don't give our agents enough credit and that, first and foremost, they fully appreciate the need to have a competitive product in the marketplace. And they are very aware of what's gone on in the private passenger consumer auto business, and they know they need more competitive alternatives. I'd also emphasize that we really built this product with lots of agent input. And hopefully, we've also been able to demonstrate to the agents that we are really committed to the independent agent channel. And this really gets to your last point of when agents look at us and what we do across the spectrum of our businesses. I think they appreciate that we are at the forefront of underwriters that are working aggressively to make that channel successful. So I think all of those things are very positive. With that said, every agent, all things being equal, would rather have higher commission than lower commission. But we feel really good about the dialogue we've had with them and the engagement, and we will see what the results are as they roll out.

And I'd make the observation that the comparative rating process, it's obvious to us, and we've shared that data with you, is trumping everything, that competitive price at the point of sale from a well-respected company and a quality brand name matters a lot more than whether commission is 1 point or 2 higher or 1 point or 2 lower. There are lower-paying commission companies that have grown value, grown volumes significantly, and there are higher paying that have not. It's not -- it's a factor, but at least, as we assess the marketplace, it certainly isn't the most meaningful factor.

[Operator Instructions] And our next question comes from Charles Sebaski of BMO.

Following up on the Quantum 2.0 and the direct market channel. Given this level of an expense cut, is it too -- is this product going to be available in the direct-to-consumer? And do you expect that to grow that channel?

Yes, and we sure hope so.

Okay.

It's certainly -- we -- one of the things that we -- and we've talked about this in the last couple of calls. We've been doing price testing in a couple of states, trying to determine elasticity of demand in a more granular way. And it became apparent to us that we needed a lower-priced product not only to be more successful in a comparative rating agent-driven world but also in the direct-to-consumer world. This will make a meaningful difference. I don't know what happens from the competitive reaction, but this will make a difference in the attractiveness of the product in the direct channel.

Okay. The -- a question on pricing and the competitive marketplace on Management Liability. There seems to be increased focus from some of your competitors. I'm just wondering if you have any color on what you see going on there with pricing and just competition in general.

Sure. It's Alan Schnitzer. There's always competition in these markets. And you really can't look at this as one broad Management Liability market. There are different geographies, there are different products, there are different lines, different segments, and each of them has their own dynamics. So as we've been approaching the market, we price as granularly here, as we do in the other lines that we talk a lot about, seeking to get rate on those lines and those accounts where we need it. And we've been successful doing that, and you see the RPC in the quarter having been the result of that.

Our next question comes from Josh Stirling of Sanford Bernstein. Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division: So I wanted to continue with some comment -- questions around Quantum. And one of the -- I think in one of your presentations, you guys provided a scatter plot, a sort of new versus old price. I'm wondering if you could talk to us generically about the sort of the lift charts you've done where you compare loss ratios for -- against the price changes you're taking, much like the, I think, the slides you guys used a bunch of times now with your commercial lines of business as people how think -- people think about personal, a lot of changes. And I'm trying to just get a better sense of whether Quantum 2.0 is a big deal from the pricing segmentation perspective or whether it's really more about sort of the repackaging, the cost cutting and lower commissions into just a new business offering that you can use to drive volume.

This is Gregory Toczydlowski again. Anytime we bring a new product out, we constantly learn from the prior one. We take that experience, and we get a little more sophisticated and a little more segmented. So Quantum 2.0 will clearly have additional segmentation in addition to some product features that you talked about. But really, when we look at those scattergrams, that really is our understanding of what loss experience is in the marketplace of our quote business and our in-force business. And then we bring the new product out, where there will be, as Jay said, ups and downs across various customers, but it is because of that new segmentation.

And at the Barclays conference, we actually gave a couple of examples of specific underwriting changes that were incorporated into 2.0 as a result of what we had learned in 1.0, areas in which we had been too stringent, where we had set an underwriting standard that not only wasn't competitive but even when we sold the product turned out not to make a difference in the loss performance. So there's some of that as well in this. It's not just a lower price cranked into a new product. Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division: That's helpful. A broader question I'd love to ask is you guys very helpfully gave us guidance on pricing, if I can call it that, sort of your outlook in the Q, which I think I interpret -- can interpret as saying that you're going to continue -- you're going to do what you're going to do. And, I mean, the market will do what it does, but that you're going to continue to raise price for the next -- through the end of the year and well into '14. You guys actually have to plan the business many months ahead, rate revisions, product filings. You're probably doing an annual budgeting process right now. What guidance are you actually giving your staff? And what are you actually planning for in 2014?

Well, first, I'd disagree that we give guidance. We're giving you our outlook. We're required to give one. We do the best we can. We always say that we're going to attempt to do what we're going to attempt to do. We don't know what anyone else will do and, therefore, I don't know whether it'll succeed in the marketplace. We've said that over and over. And also, in the 10-Q, we really speak to renewal premium change in the aggregate, not just rate, but we talk about renewal premium change. What -- our instructions to the field organization in Business Insurance are no different today than they were 1 month ago or 2 months ago or 3 months ago. It's account by account; it's drive business from the right side of the segmentation to the left, meaning from the poor-performing categories to the best-performing categories; as we move left, focus more on retention and less on rate, although it's great if we can get rate in excess of loss trend even in our best-performing business; and at the right side if, and we've shown you this information before, if retention is comparatively low, that's just fine. You're talking about components of the business that are not building value but taking value. And our instructions to them are no different. We don't give them a target. Our experience has been, over a long period of time, that if you give a field force a target, you will get it, but you will not get it in a way that's designed to actually produce improved profitability. You'll get the rate that you asked for on the poorest accounts. You will lose the accounts that are best performing in the minimum rate thresholds. So we give them analytical data. We give them lots of information. And we are very return driven. They try and improve the returns on the accounts, and then we leave it to the underwriters to execute. And that is the way we run the business.

And this will be the last question, please. Thank you.

Our next question comes from Brian Meredith of UBS.

Just a couple of quick ones here. First, you mentioned in the Q that the G&A expenses in the Business Insurance benefited from reduced assessments. I wonder if you can just quantify that for us?

Hold on a minute.

It's not a number we have off the top of our heads. You want to go to your next one while we look for that one?

Sure. Yes. I guess the second one, I'm curious, when you closed the Dominion acquisition, do you plan on implementing the kind of Quantum 2.0-type pricing there as well? And in addition to that, what about -- with respect to kind of commission rates and stuff. Are you going to apply that to the Dominion personal auto book? Because I know they got a pretty big Ontario auto book. And how that's going to be impacted by the 15% rate reduction that Ontario is taking?

It's Alan Schnitzer. We will certainly be applying all of our knowhow in technology and thought process from our U.S. experience into Canada. Whether that exact product, with those specific characteristics, translates or not will be a question for us, but we will certainly have all of our resources and knowhow martialed to bring to bear to that product.

Yes, the commission, the TL&F variance, that you're looking at is about 12 pretax. So it's relatively modest. But when you look at a quarter-over-quarter comparison, it does come into play.

Okay, that concludes today's call. Thank you all for joining us. And as always, Andrew Hersom and myself are available for any additional follow-up questions. Thank you.

Thank you. Ladies and gentlemen, that does conclude the conference call for today. We thank you for your participation and ask that you please disconnect all lines. Thank you, and have a good day.