Gabriella Nawi - Senior Vice President of Investor Relations and Member of Operating Committee Jay S. Fishman - Chairman, Chief Executive Officer and Chairman of Executive Committee Jay S. Benet - Vice Chairman and Chief Financial Officer Brian W. MacLean - President and Chief Operating Officer Alan D. Schnitzer - Vice Chairman, Chief Legal Officer and Chief Executive Officer of Financial, Professional & International Insurance Doreen Spadorcia - Chief Executive Officer of Personal Insurance and Executive Vice President of Claim Services

Michael Zaremski - Crédit Suisse AG, Research Division Amit Kumar - Macquarie Research Gregory Locraft - Morgan Stanley, Research Division Keith F. Walsh - Citigroup Inc, Research Division Jay Gelb - Barclays Capital, Research Division Michael Nannizzi - Goldman Sachs Group Inc., Research Division Matthew G. Heimermann - JP Morgan Chase & Co, Research Division Larry Greenberg - Langen McAlenney

Good morning, ladies and gentlemen, and welcome to the First Quarter Results Teleconference for Travelers. [Operator Instructions] As a reminder, this conference is being recorded on Thursday, April 19, 2012. At this time, I would like to turn the call over to Ms. Gabriella Nawi, Senior Vice President of Investor Relations. Ms. Nawi, you may begin.

Thank you, Tina. Good morning and welcome to Travelers' discussion of our first quarter 2012 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, Chief Financial Officer; and Brian MacLean, President and Chief Operating Officer. Other members of senior management are also in the room, 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 up for questions. Before I turn it over to Jay, I would 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 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. Jay S. Fishman: Thank you, Gabby. Good morning, everyone, and thank you for joining us today. We're pleased to report a very strong start to 2012 driven by solid underwriting as well as investment results. In the quarter, we produced net income of $806 million or $2.02 per share and generated a 13.1% return on equity and a 14.7% operating return on equity. Even excluding the impact of net favorable prior-year development, which states the returns on an absolute year basis, the corresponding percentages are 9.9% and 11.1%. Net income of $806 million compares to $839 million last year, but last year included a $104 million benefit from the favorable resolution of prior-year tax matters. As we've shared with you previously, our current operating focus is simple and on course. Given the potential for continued severe weather in the United States as well as the outlook for continued low interest rates, we remain committed to improving our profitability by actively but selectively seeking price increases and improved terms and conditions on the products we sell. In Business Insurance, we continued to achieve improved pricing. Renewal rate change in the quarter was a positive 8%, up from 6% in the fourth quarter of 2011, while retention remained stable. Importantly, we believe that we differentiate ourselves from many in the marketplace by not taking a "one size fits all" approach to pricing. In our individual underwriting businesses, each customer's needs, characteristics, geographies and risk and loss profile is different, and as such, each transaction is analyzed and priced separately. In Select Accounts, analyzing the returns we earn from our insured’s industry segment in conjunction with the individual insured characteristics is critical to success. We continue to leverage what we believe is industry-leading data and analytics to help us optimize rate and retention, that is, achieving higher retentions on our best customer relationships while simultaneously identifying the walk away point for our least-profitable relationships. This differentiation is best evidenced by an analysis of rate and retention by historical loss ratio performance that Brian will share with you later. We're very pleased how we're executing in the field, and this analysis summarize and demonstrates the effectiveness of our analytics and the thoughtful way we're approaching business pricing. In our Personal Insurance segment, we continue to take the necessary actions to increase returns, which include improved pricing, terms and conditions. Renewal price change in the quarter improved in automobile and Homeowners to 4% and 10%, respectively. In many states, we have implemented higher deductibles and changes to underwriting terms and conditions predominantly for new business. And while Personal Insurance new business in these states has been and may continue to be impacted for some time following these changes, we are convinced we are on the right course. I want to take just a moment to acknowledge the work our financial professional and international groups have done in recent years. Within FP&II, there are numerous businesses and product lines, each with its own individual story, but in the aggregate, we recorded a 24% improvement in profitability quarter-over-quarter and are more comfortable with our risk profile and better positioned for opportunity in our international business than we have been in the past. Additionally, the fact that our surety and Management Liability businesses have performed as well as they have is a real testament to risk management scale, especially given the economic environment over the last several years. The solid work these teams have done and continue to do is evidenced in their strong results. Lastly, I'm also pleased to report that our Board of Directors approved a 12% increase in the company's quarterly dividend, raising it to $0.46 per share. This brings our annual yield, based on last night's closing price, up to 3.1% from 2.8%. Since the merger of St. Paul and Travelers in 2004, the dividend has increased each year and, over this time period, has grown at a compound annual growth rate of 10%. As always, we remain committed to returning excess capital to our shareholders in the form of dividends and share repurchases. And with that, let me turn it over to Jay. Jay S. Benet: Thanks, Jay. I would characterize our first quarter results as being very strong and as having the following attributes: fixed income investment returns that were in line with our expectations; non-fixed income investment returns that benefited from strong equity market performance to the end of the quarter; underlying underwriting results that benefited from the significant price increases we have been achieving in recent months as well as growth in exposures and favorable audit premium adjustments; weather-related losses that were more or less within our expectations; and a relatively high level of net favorable prior-year reserve development. Brian is going to provide comments on our underwriting results, but before he does, let me say a few words about our reserves and their development. The strength of our reserves is evidenced by the net favorable prior-year reserve development we continue to experience: $200 million after tax in the first quarter, up from $155 million in last year's first quarter. While each segment once again reported net favorable prior-year reserve development, the Business Insurance segment was responsible for $162 million or approximately 80% of the total. BI's net favorable prior-year reserve development was driven by better-than-expected loss experience in general liability for the 2004 through 2009 accident years and commercial property for the 2010 accident year. Partially offsetting these favorable adjustments was the strengthening of our commercial auto reserves for the 2011 accident year due to higher-than-expected frequency and severity. Financial, Professional & International Insurance contributed net favorable prior-year reserve development of $31 million after tax in the quarter, a little higher than last year's first quarter primarily due to the better-than-expected loss experience in our surety business within Bond & Financial Products for the 2006 through 2008 accident years as well as in several lines of business in Canada and Lloyd's. Turning to financial strength indicators. All remain in excellent shape. First quarter operating cash flows were very strong at $814 million, a holding company liquidity of $2.2 billion and all of our capital ratios ended the quarter at or better than their target levels. Net unrealized investment gains, which were $4.35 billion at the end of the quarter or $2.8 billion after tax, decreased only slightly from year end, entirely due to movements in interest rates and credit spreads, while book value per share rose to $63.81, 2% higher than the beginning of the year and 7% higher than a year ago. And finally, as always, we remain fully committed to identifying and returning excess capital to our shareholders. During the quarter, we returned a little over $500 million to our shareholders through common stock repurchases of $350 million and dividends of $162 million. Brian is now going to provide further commentary on our underwriting results. Brian W. MacLean: Thanks, Jay. A great quarter on many levels. And from an underwriting perspective, our GAAP combined ratio of 92.2% was 2.5 points better than 2011, and adjusting for cats and prior-year reserve development, the combined was 94.6% or 1.1 points better than 2011. So a really good start to the year. In Business Insurance, we are extremely pleased with our results this quarter. Operating income of $612 million was up slightly from a very good quarter last year, which included a $76 million tax benefit. Pricing continued to be very strong, with renewal premium change of 9% that included pure rate increases of 8%. Importantly, the improved rate that we achieved was broad based, with rate increases in all product lines between 6 and 9 points, with the workers comp and auto lines at the top end of that range. Net written premiums were up year-over-year driven by these pricing gains as well as solid exposure growth and audit premiums. By line, the growth was driven by workers compensation, again resulting from price and exposure increases. For your reference, Page 9 of the webcast shows the reconciliation of net written premiums quarter-over-quarter, and we believe this speaks to a picture of improved profitability on the business. Retentions remained solid, up slightly from the fourth quarter of 2011 driven by improvements in commercial accounts. As you can see on Slide 11, Select Accounts retention has declined from 81% in the first quarter of 2011 to 75% in the current quarter. As we have discussed many times, our select business includes Select Express, the smaller-account-sized business; and Select Plus, the larger individually underwritten accounts. In Express, retention has come down modestly and is currently at 78%. In Plus, retention is down much more significantly to 71%. In the Plus business, based on our view of account and segment profitability, the competition has continued to price business very aggressively and at returns we do not consider acceptable. Accordingly, our retention is the result of our active and very deliberate strategy to seek the appropriate price for each account, and we are very comfortable with this volume and profitability trade-off. So stepping back to rate and retention for all of Business Insurance, we're very pleased with the results, but importantly, we're also very focused on how we got to those results. By that, I mean, did we get the rate increases where needed? And did we retain our best business? An example of how we manage this approach is on Slide 14, which shows our rate change and retention data for commercial accounts segmented by the individual accounts' long-term loss ratio. So the bar on the left represents our best-performing business, accounts with long-term loss ratios of less than 60%, and the bar on the right represents our worst-performing accounts with long-term loss ratios exceeding 90%. You can see that the retention ratio was stronger for the better accounts and the rate change is dramatically higher on the poorer-performing accounts. I want to emphasize 2 points. First, our actual pricing and underwriting decisions are made on individual accounts. Secondly, that this is -- this exhibit presents summarized data and we actually measure and manage our performance on a much more granular level: by business, by product line, by industry and by geography. And the conclusion is we feel great about how we're executing at those levels. Turning to loss trend. Across Business Insurance, we are seeing loss inflation at about 4% overall, which is consistent with our expectations. As with the rate story, by line, there's a pretty tight range around that number, with workers comp and auto slightly above the average. Looking specifically at workers comp, the frequency increase we talked to you about in prior quarters has stabilized. And in commercial auto, we continue to see a modest pressure in bodily injury severity and in physical damage frequency and severity. So in closing on Business Insurance, a critical takeaway: In the quarter, the earned impact of our pricing actions exceeded our current estimate of the quarter's loss trend by about 50 basis points. And this is obviously a step in the right direction. In the Financial, Professional & International Insurance segment, we had another very strong quarter, with operating income of $149 million, up 24% from the prior year. The improvement is due to lower level of cat losses as well as improved underlying margins in our international business where we believe underwriting initiatives have resulted in a better balance of risk and reward. Looking at production, Management Liability net written premiums were up 7% year-over-year as retention levels and renewal premium gains were at their highest levels in several years and new business volume was also strong. Management Liability is an area where we've made significant investments in technology and analytics, and the increased volume we are seeing shows that those investments are paying off. In addition, we are seeing the most growth in our more profitable lines. In surety, volumes continued to be impacted by decreased construction spending, while in international, net written premiums were down $8 million or 3%, reflecting the impact of our decision to exit the personal lines business in Ireland, as well as an increase in ceded premium year-over-year. So all in, a very strong quarter for the segment. In Personal Insurance, first quarter operating income, excluding cats and prior-year development, was $176 million, up 5% from the first quarter of 2011. The GAAP combined ratio, also excluding cats and prior year, was 92.3% and, excluding the impact of the direct-to-consumer initiative, 90.3%. So the underlying profitability, the agency business, was very solid. Looking at production. In Agency Auto, renewal premium change was 4%, up 1 point from the fourth quarter, and retention remained strong at 83%. In agency property, we also saw improved pricing, with renewal premium change of 10%, up from 8% in the fourth quarter, along with continued strong retention. In both lines, new business volumes have been impacted by our pricing and underwriting actions. Loss trends in home have remained within our expectations, while in auto, we continued to see severity pressures that we disclosed last quarter with a slight uptick in physical damage lines. Given the desire to improve auto margins that we discussed last quarter and the possibly of continued unusual weather patterns, we will continue to take the actions necessary to improve profitability in both auto and Homeowners. We've filed for additional rate increases in both products in a number of states. And in Homeowners, we've also continued to implement underwriting changes along with changes in terms and conditions, such as higher deductibles. As with Business Insurance, we continually measure the balance between profitability and growth in a very granular way, and we remain very comfortable with our actions. With that, I'll open it for questions.

[Operator Instructions] And our first question comes from the line of Mike Zaremski of Crédit Suisse. Michael Zaremski - Crédit Suisse AG, Research Division: I'd be curious to hear your thoughts on insurance pricing levels. I know no one said that they've peaked, but they clearly are in excess of loss cost trends in most lines. Are you guys still seeing momentum there? And can you comment on Management Liability? Rates increased 4% there, do you think there's momentum in that specific line of business? Brian W. MacLean: Yes, Mike, so this is Brian MacLean. I'll start off. And speaking specifically to Business Insurance, I'd start by emphasizing and repeating what we just said, that we feel good about where the market was this past quarter. And the 8% rate, 9% total price and the retention levels we're getting and, specifically, where we're getting it by account and by type of business, we felt good about. Our position in the marketplace right now as we move forward is, we're not changing our strategy at all. We're looking to continue to prove -- improve our returns. You are absolutely right, that we're now at a point where we're earning in something in excess of our current view of loss trends. And we don't make long-term predictions on the market, but we do have pretty good visibility into what's going to happen 60 to 90 days out. We're obviously quoting business out there and we have a decent view into April. And so we feel good about what we're seeing in the second quarter and, from a written basis, feel that the momentum is somewhat continuing. So we feel very good about where we are. Michael Zaremski - Crédit Suisse AG, Research Division: Okay. And then lastly, in regards to workers comp frequency stabilizing, could you comment on what level it's -- has stabilized at? Brian W. MacLean: I don't know if we've specifically disclosed it, but it's... Michael Zaremski - Crédit Suisse AG, Research Division: Would you say it's at a level that's going down in terms of the trend, the absolute level of frequency? Jay S. Benet: This is Jay Benet. I think, in the fourth quarter, we've seen it level out. And we really haven't seen a change in the first quarter relative to what we have been seeing, so it's running within our expectations. Nothing has really changed. Brian W. MacLean: Yes, I mean, it's pretty close to 0, is what we're seeing and it's very stable. Michael Zaremski - Crédit Suisse AG, Research Division: And have rates, pricing, has that continued to improve on that line? Brian W. MacLean: Yes. I mean, we've -- workers comp is at the highest end of the price increases we're getting, and we feel good about what we've been able to do there. On the Management Liability side, Alan... Alan D. Schnitzer: Mike, it's Alan Schnitzer. You asked about Management Liability pricing, as well. And you mentioned before, the pricing is actually up 2. Overall, RPC is up 4, and 1/2 of that's rate and 1/2 of it's exposure. We are -- it is our strategy to continue to try to get rate in those lines where we think we need it. And just based on our feel for the market and what we're doing 30, 60, 90 days out, we think there's some room for that.

Our next question comes from Amit Kumar of Macquarie Group. Amit Kumar - Macquarie Research: The first question is on the discussion on retentions. Obviously, the retentions have trended down successively every quarter. And I'm just wondering, if you keep on pushing rates, do we get to somewhat of a low-70s retention number as we approach end of 2012? Or do you think that the retentions numbers are exhibiting some signs of stabilizing, going forward? Brian W. MacLean: Yes, so a couple responses to that. This is Brian. I'll start, and then -- let me start by saying that, historically, when we look at retentions in our core Business Insurance, at an 80% level, with the right execution and the right analytics, we feel that -- as a matter of fact, we don't have any doubt that we can perform very well at an 80% retention. We had said consistently, going back a year or so ago when we were running at 84, 85s, that those were at historical high levels and we did not expect to continue to have the business there. So we feel pretty good about where we are today from a retention perspective. As you pointed out, quarter-over-quarter, it went up a blip so fundamentally stable where it is. The other thing I would emphasize back to that exhibit we showed in the webcast: We look really hard at, as retentions move up or down, why they are moving up or down. And to state the obvious, but it's -- the execution on this is critical. If we're retaining our best business and retention is softening in the weaker business because the market -- because we can't get the price increases that we think we need, we're going to be perfectly fine with that trade-off. So we don't sit here with a magic number and say, if it gets to 75, we're at a panic level. If -- whatever. It's really about the composition of what's going on inside of it. Jay S. Fishman: And this is Jay Fishman. I'd just add 2 quick comments. First, just to clarify, in Business Insurance, the retention was actually stable, dead on, first quarter to fourth quarter. So sequentially there, that was actually a tiny bit up. It rounds to a flat number, but it was actually the tiniest bit up. And so sequentially, it is right on. And I'd point also to Brian's comments about the Plus business. There's no question that, if we had a different strategy in Select Plus, the retention could be 5, 6, 7 points higher, but our analysis suggests that, that's not the way to create long-term value. And a lot of -- in many of these market segments, in fact in most, it's an active decision, not a passive reaction. We make the decision of the price-retention trade-off all the time. And the slide that Brian showed that ties loss ratios to retention in rate is absolutely critical in the analysis particularly as you get into a granular level. And we actually -- if you do some quick arithmetic on that slide, what you'll see is the summary of that slide shows an 18-point rate increase against accounts with a historical loss ratio or long-term loss ratio in excess of 90%. That -- the result, the arithmetic result of that is a 14% decline, not a 14-point decline but a 14% decline in the actual loss ratio. I mean, it's meaningful. So it's -- as you have the analytics and the ability to look at your book of business just as granularly as we do, we tend to view the retention as a result of thoughtful, well-executed strategies, not a target or a byproduct that's in effect for us -- on us. Amit Kumar - Macquarie Research: Got it. That's helpful. Just one other question and this is a follow-up to the prior question. In terms of California comp, you mentioned that -- I'm sorry, the workers comp line, you said that the rates were at the highest end of the range. And I couldn't quite discern what exactly the number was, what you were alluding to, what exactly that number is. I think you mentioned the loss costs were 4%. Just trying to understand the delta between pricing and loss cost. And if I understand correctly, what you were suggesting in the prior answer was that, probably, the way you look at it, 2010 was the peak in terms of loss cost. And I know you've discussed that in the past, too. Brian W. MacLean: So let me do a couple things and I'll repeat some of the stuff, just to make it clear. We said 8 points of rate increase between 6 and 9 by line, and comp was at the high end of that. So -- and then on the loss costs, I said 4% across all of BI, with a relatively tight band by product line around that. And comp is at the high end of that. I mean, comp and auto were the 2 lines where we're seeing the most loss pressure and we're seeing the highest rate increases. So we think there's rational behavior there. Jay S. Fishman: And what Brian gave you in rate just a moment ago was the written rate increase. And importantly, he shared with you that the earned effect of the pricing strategy, not just rate but that portion of exposure that acts like rate, actually had an earned impact in the quarter of approximately 50 basis points higher than the -- than our current estimate of loss trend. Brian W. MacLean: Right. Now you made a comment and I wasn't quite sure where you were going but that we implied that the fourth quarter was kind of the peak of... Amit Kumar - Macquarie Research: No, the 2010. 2010 was the peak in terms of loss trend.

[indiscernible] Amit, this is Gabby. When Brian said the question we had been asked was workers comp frequency, which we were saying was essentially flat. But certainly, that is never implying that workers comp loss costs are flat. Obviously, we've always talked about the severity dynamics to it. Amit Kumar - Macquarie Research: So is it still going up, overall? Or you think that it's stabilized at this level? Jay S. Fishman: Well, the question is what is -- that you're asking is, we're not exactly sure what you mean by it leveling. Our -- when you add in frequency and severity and you look at workers compensation -- and I'll say this without checking all the facts, but Gabby can. I think, in virtually every year, that it's a net positive inflation rate on loss costs that, even if severity -- even if frequency has been down, our assumption about the combination of medical inflation and indemnity inflation, both, more than offset whatever the underlying assumption has been even in a declining frequency arena. So our analysis always is that loss costs are always rising year-to-year. Brian W. MacLean: Right. And then so the pieces we're giving... Jay S. Fishman: As we've already [ph] said. Sometimes, we get favorable reserve development years later, and lots of you comment on that, where our initial loss estimates turned out to be way too high where, in one year, too low, and it obviously changes. But as we initially set them, we're always contemplating an increase in the inflation dynamic on a combined basis. Brian W. MacLean: And I'll repeat the pieces we had given you, and I think they can get you pretty close. I mean, we just said frequency was about 0 across all of BI. Frequency Plus severity is about 4%, and comp's a little bit above that. So that, I think, gives you the pieces, yes.

And our next question comes from Greg Locraft of Morgan Stanley. Gregory Locraft - Morgan Stanley, Research Division: Wanted to just ask about the buyback and the dividends. A very nice increase in the dividend in the quarter, but it looks to me like you -- the combined $500 million in buybacks and dividends was less than operating income. And I just wanted to -- and perhaps you said it in your commentary, but I just want to just reassess, what is the guidance for capital management, going forward? Is it 100% of operating income to be returned to shareholders? Jay S. Benet: Yes, this is Jay Benet. We try to be as clear as we can about this in our 10-Q. So let me just reiterate a couple of points. First of all, just in terms of the magnitude in the first quarter, we did return, as I said, a little over $500 million, $350 million in the form of repurchases, the rest in terms of dividends. Obviously, we don't know the actual amount of earnings in a quarter until the quarter is over, so we are buying back our shares during the quarter based on some presumption of where we'll be. I'll remind you that, and this is what is written in the Q, the combination of dividends and share repurchases isn't likely to exceed the income for the year. And the share repurchases are also going to depend on a variety of factors. There are capital needs that we have as we grow our business. Maintaining capital levels commensurate with ratings. We do buybacks on the basis of having excess cash available. So we take into account pension funding needs and other corporate needs. So there's lots of things that go into it. And I wouldn't read from any particular quarter the earnings of the quarter versus the amount of the share repurchases as indicative of much of anything at this point. Gregory Locraft - Morgan Stanley, Research Division: That's very helpful. And then shifting gears, on the core margin side, excellent trends. Everything you're saying is sort of playing out in the reported numbers. The one thing I'm trying to tie, though, Brian, is you mentioned earned premium is ahead of losses -- loss trend by 50 basis points in Business Insurance. And yet, if we're doing our math right, it looks to me like the core margin for that specific unit actually deteriorated by 30 bps year-over-year, which is -- points to a very near-term inflection point for that division importantly. And so I'm trying to tie the "earned premium to loss" commentary with the reported number. And I guess the question will be, is there a lag? Or how should we be thinking about that in the models? Brian W. MacLean: And I assume, when you say you're looking at the reported -- if I look at the x cats, x prior-year development combined ratio for Business Insurance, it was 96.4%. A year ago in the first quarter, it was 96%, but importantly, for the full year of 2011, it was 97.9%. So we've got some pretty significant improvement: It's come down 4 consecutive quarters. With all of that said, in any given quarter, there's going to be a lot of stuff that's going to go into the combined ratio that, when I talk about that core margin expansion, I'm not talking about unusual weather, I would -- so I'd be taking out favorable or unfavorable weather, large losses, mix of business changes. We could be moving something in the accident year that's going to distort that given quarter's GAAP combined ratio. So I think, in a specific quarter, there can be a lot going on. Obviously, as we go forward, we're going to speak to those things. But the core underlying margins showed some improvement in the quarter and bodes well for the future. Jay S. Fishman: And while we -- Greg, this is Jay. Thanks for the question. It -- while we don't -- certainly don't want to make predictions about rates or anything else, the one thing that is quite just arithmetic and quite predictable is the effect of written rate because you can go back, you take 1/12 of the previous month, plus 1/12 of the month before that and 1/12 of the month before that and you calculate the rate dynamic, and you can actually calculate the rate impact on premium, the earned rate impact. Now there's a portion that's not nearly as visible to outsiders. There's a portion of exposure that really looks like rate. It doesn't have loss cost associated with it, a change in valuation on a property. So there are elements, so we very granularly go through that. And so on the presumption that, that exposure piece stays the same quarter to quarter, what we can say with reasonable assurance, because things would have to change so dramatically is to be unlikely, is that the earned rate impact in the second quarter, earned rate, will be higher than the first. That's pretty ascertainable at this point. We don't know obviously what our second quarter view of loss trend is, and so it's very difficult to make predictions and feel like you're doing it in any substantive way, but the earned rate impact is really quite discernible.

Our next question comes from Keith Walsh of Citi. Keith F. Walsh - Citigroup Inc, Research Division: Thanks for the slides on retention. Very helpful in the pack. Just getting back to Jay Fishman's comments earlier, on Slide 14. Can you talk a little bit about the elasticity of retention and the trade-off between raising rates on the really good loss ratio business that looks like it's sub-60 versus keeping more of it, retaining more and maybe not raising rate as much? Jay S. Fishman: Yes, that's -- it's -- actually, it's a very good question and it really gets to the core of how we execute in the marketplace. The -- first, in its most simplistic form -- and you all are knowledgeable enough about our industry to understand it: If you get 1 point of rate, 85% of that, I'm taking 15 points of commissions and taxes, direct costs, if you will, fall right to the pretax line. So 1 point of rate is really quite powerful from a profitably standpoint. Now 1 point of retention has dramatically lower impact on profitability because you have all the loss costs associated with it. So you have the variable cost that goes up, but you also have the direct loss cost that goes with it. So it's quite easy to say that 1 point of rate is worth a multiple, 5x. And I'm guessing at that number, sitting here this morning, but 5x 1 point of retention. Now the -- that theory is good until your last account. At some point, it breaks down. But as we sit here looking at 1 point of rate on the margin versus 1 point of retention, that's a really easy call as long as we're being thoughtful about how it's executed. Now I'm going to get back to the slide you referenced. One of the reasons why we don't at all set a corporate mandate for pricing -- and there are some, we believe, companies that do. They say, "You've got to get x percent on a renewal or you can't renew it." Well, what happens is you'll lose your best accounts because they don't need the x percent of rate. And in your least-profitable accounts, you get the x, and it needed twice x. And so while it looks initially like you've done good work, the fact is you've got underlying real adverse selection working against you. So first on the margin, when you're in an improving profitably dynamic, that point of rate versus the point of retention is just such an easy call. On the marginal dynamic, that's an easy one. And -- but importantly, the ability to execute and understand where one draws the line on that chart is just critical to the actual improvement in profitability. Otherwise, you're going to have a whole series of operating dynamics that are going to look great, but unfortunately, it's never really going to translate into improved profitability. So it's the combination of recognizing the trade-off between rate and retention, which again, on the margin, is an easy one, but executing it based upon the history of the account. I'd make one other point with respect to that chart because it's really an important one and it can get lost. This is the long-term loss ratio. So if we -- this is the actual experience of an account up to as much as 7 years, if we have the account that long, because we also understand the premise that an account can be terrific performing over 2, 3, 4, 5 years and then have an unusual loss, an unusual fire. And you certainly don't want to react dramatically to one insured loss. We're in the insurance business, we anticipate losses. So it's not -- we're not looking at a 12-month loss performance here because, again, that would give you a potentially very adverse effect as well. This -- analytics here enable us to look at the long-term profitability of an individual account. Brian W. MacLean: I just want to emphasize. This is Brian again. One other thing I made in my comments generally, but I think it's really important to understand who we are and why we think we can perform the way we do in the industry. The exhibit that we've got is a very simplified view of how we actually manage this. So what we actually use in our operation doesn't just have 3 segments, it has at least 5 by each of our business. So we've got our business separated by quintiles and it's got the long-term combined loss ratio plus a bunch of other factors. So we're not just looking at the loss experience of the account. We're looking at a lot of other things. And we're looking at that by individual business, by industry, by geography. And believe me, when we see the numbers come out, we go, "Okay, they look okay, they don't look okay, but let's really understand why." So your question, again, Jay said it is great, that keeping the best business for the right reasons and taking the right actions on the more challenged business is what we have to do. And at the end of the day, that's going to result in numbers. And if we did the right things, we're going to be happy with what the numbers are. Jay S. Fishman: And I -- while I can't prove this to you because I'm giving you back the reflection of conversations with agents and brokers -- we tend to forget about them in these conversations. There's an agent or broker who represents every single account that we're dealing with here. And that agent or broker understands the market deeply, understands alternatives and choices and flexibility and in the end makes a recommendation to his client based upon the actions that we're taking. One of the reasons, and this just comes from conversation, that agents continue to be supportive of the strategy here is that it's not being applied as a "one size fits all" dynamic. When we sit down and have a conversation about an account, it becomes apparent why we're having that conversation. And we're not -- we talked earlier about never -- we're trying hard not to be disruptive to an agent's business or to our insurance business, and that means getting granular enough to understand that the pricing actions fit the individual account and account strategy that you have. And that's one of the reasons why agents can embrace it because they understand what we're doing and they understand the economics behind it. Keith F. Walsh - Citigroup Inc, Research Division: Okay, that's very helpful. And then the second question I just had on Slide 9, the new business number and the revenue roll-forward or the premium roll-forward of $423 million; retention, really good, keeping it stable. It seems like this number is a little more challenged. Does this speak to a more competitive market with lots of capacity for new business? And just how do you -- how does that intertwine with your view on expanding rates, going forward? Jay S. Fishman: Is your question -- I'm sorry, just to clarify, is your question that the new business amount seems low to you? Keith F. Walsh - Citigroup Inc, Research Division: Yes. Jay S. Fishman: I think we would agree with that, actually. It's really very difficult to -- we remind folks all the time that the profitability of our business and its long-term value-creation is a lot more dependent on how we treat our renewal book than what happens in our new. And the last thing that we really want is for agents to be confused, to find that we're being thoughtful and disciplined in our renewal book but we're being aggressive from a pricing standpoint from new business. I -- and so we recognize, going into this, that it was entirely possible as we became somewhat more urgent about profitability improvement, that the new business levels could very well fall off. Now this is a summary of lots of areas. There are some areas where I think we've overdone it a little bit where I think we could -- I think we've been almost too conservative with respect to new business and we've got to dial up. It gets -- it's not an overall strategy. There are regions where we've got it just right when we look at the numbers and it's working terrifically. And there are others where we've gotten a little too conservative and you want to dial that back just the littlest bit. But I don't perceive it as a comment on the competitive market. I'm sure there's some truth to that, but I think the more substantive dynamic is the approach that we bring to our enthusiasm, our aggressiveness or our conservatism at the point of new business.

And our next question comes from Jay Gelb of Barclays. Jay Gelb - Barclays Capital, Research Division: Jay, looking at the overall x cat and x prior-year development combined ratio of 94.6%, it's meaningfully improved from any of the 4 quarters of 2011. And I just want to get your sense as to whether we should think of that as the right run rate, going forward? Jay S. Benet: Jay, this is Jay Benet. I think it's -- what Jay and Brian were talking about is the earning end of rate increases that we've been getting over the past 12 months. And when you think about whatever estimates you're making for a loss cost trend, you have to look at the amortization effectively of these written rate increases. And you've essentially seen an upward sloping line of rate increases in that period of time. So just arithmetically, if nothing else changed, if the weather was exactly the same, if loss cost trends were exactly the same, you would just see the combined ratio move in a favorable way based upon that amortization. So we'll see that happens with all the things I just mentioned. But it's not that it's reached a steady state, it's reached a different state than it's been in. Jay S. Fishman: And Jay, I would add that one of the things we're taking for granted here as we're responding is that you understand the impact not only of catastrophic weather, which you immediately pulled out, but small weather, such as we call it around here, has had a significant impact quarter-to-quarter and even has some seasonality to it. It's not just a property loss. We've seen -- traditionally, fourth quarter, there are more auto accidents. We can have a conversation about why that is. And even the fact when there wasn't bad weather in the fourth quarter, there still seems to be some seasonality, so it's not quite as smooth as we'd all like to think. It would sure make it easier for us to manage if it was, but there is some seasonality and movement up and down, so I'd resist the temptation to just say, that's the number and it gets better from here. I have no idea what the next 3 quarters will bring, but I'm sure we're going to be sitting in front of you with some quarter where small weather was either much better than we thought it was going to be or much worse than we thought it was going to be. And that's going to evidence itself in the numbers that quarter. Jay Gelb - Barclays Capital, Research Division: I see. And my follow-up question is, was there any seasonal influence on the pace of reserve releases in the first quarter? Is there a certain book of business that gets analyzed because that picked up as well versus any of the prior 4 quarters? Jay S. Benet: No. Our reserving processes are very uniform quarter-to-quarter. So this is just getting really the fourth quarter data from last year, analyzing in conjunction with everything else we've known about the reserves and seeing some very favorable movements. As you know, when you're looking at reserve development, you're seeing trends, you have to always determine whether those trends are real or whether there's some anomaly that's taking place. So we do want to make sure that there's enough data that has manifested itself as being real before we make adjustments, but that's something that's going to happen on a quarter-to-quarter basis. There's really no magic about the fourth quarter versus any other quarter.

Our next question comes from the line of Michael Nannizzi of Goldman Sachs. Michael Nannizzi - Goldman Sachs Group Inc., Research Division: I just have actually one question about -- I know there was a case, a John Mansville (sic) [Johns Manville] case, that was talked about in the first quarter just in the media in an asbestos settlement there. Did that have any impact on the reserve development in the first quarter? Jay S. Benet: This is Jay Benet. No, we had no change whatsoever in our asbestos reserves in the first quarter or our environmental reserves. Actually, you can see that in all the discussion associated with the litigation and asbestos information. Jay S. Fishman: There's no provision. I'm sure we have paids. Alan D. Schnitzer: Right. Jay S. Benet: I’m sorry. We had no provision. No incurreds. We did have paids... Michael Nannizzi - Goldman Sachs Group Inc., Research Division: And then you -- I assume having asked this before, is -- on the 2010 book, you had some latent frequency, some late-reported claims. And I just wanted to see if we can get an update in terms of, has the late-reported claim activity abated for the 2010 comp accident year? And is it too early to talk about whether or not any of that activity has manifested in the '11 accident year? Brian W. MacLean: So this is Brian. And again, you're talking specifically about comp year. So no, I think that we've seen the frequency dynamic relative to that business level off, and so we did not see any further activity in the first quarter on the 2010 year and are not seeing a similar dynamic through 2011. Jay S. Benet: There's no change of any magnitude of -- in workers’ comp reserves for either 2010 or 2011. Michael Nannizzi - Goldman Sachs Group Inc., Research Division: And then just last one. It was on the Slide 14, you mentioned -- what is the retention? What has been that sort of trend of retention over time, if you can? And I realize there's a lot of information in that slide, but are you seeing an improved retention of the business you want or stable retention of the business you want as you've taken rate over the past several quarters relative to kind of the retention of the business that you would rather not have that's on the higher side of the loss ratio spectrum? Brian W. MacLean: Yes, this is Brian. I would say, directionally, from a retention perspective, this is pretty consistent what we would say. It's the strongest retentions to the left here, and this one obviously, the middle bar, is slightly higher. But very solid retentions, in the plus 80% and the weaker activity on the far right. So I think, directionally, it's going to be similar. Jay S. Benet: Yes, what we do from time to time is we look in our management information system for things that we use to manage the business and share them with you. So this is a slide that is prepared for many of our businesses on a monthly basis that we have just compiled for you to see the granularity in which we actually manage it. But it's not something that's showing a different view than what we would previously have had, had we been putting this together for you. Jay S. Fishman: The analytics are not new. The analytics have actually been around for a long time. This is -- particularly just given the rate environment and some of the focus that some of you have had on rate and retention, we thought it would be helpful for you to actually see it and understand how granular you need to be for the strategy to be successful. Brian W. MacLean: Yes.

And our next question comes from Matthew Heimermann of JPMorgan. Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: A couple of questions. First, just on the Page 14 slide. The over-simplistic view of the world might be that you'd expect lower retentions as you go up in terms of the loss ratio bands, but when we look to 60% to 90%, it's a little bit higher than the less-than-60% band. I'm assuming a lot of that just has to do with mix, but would love a little color there. Brian W. MacLean: Yes, it's Brian. I wouldn't overly analyze the precision of the -- which bars are higher. It was meant to be pretty much a representation of something, as we've said over and over, that we actually manage on a much more granular basis. We also don't look at just the loss ratio and the loss performance, so it's -- you've got a lot of other things going on there. As Jay said, it's a long-term loss ratio, so... Jay S. Fishman: I'd actually remind you, because this is I think a pretty good example. We have shared with you in previous quarters, this goes back perhaps a year, that when we bring on a block of new business, what that loss ratio performance looks like and largely through the implementation of our own loss management practice, we're sitting with a group of 750 professionals who are really dedicated to helping our middle-market customers lower their losses. So if we've come into a situation that's not been managed, that we've not had before, it's new and they are engaging with us in loss management practices so that we actually believe that we can help them lower their loss experience in a reasonably short period of time, that will have us have one view of the account versus another kind of account that simply doesn't do it. That's another example of an important variable that goes into what our thinking about price and return acceptability is. We're -- we -- the reason we give you a long-term loss ratio here, how we look at it, is because that's what we're trying to achieve: long-term loss ratio performance that creates real value for everyone, by the way, not just for us but also for the customer. So that's an example of a pretty good variable, an important variable that we consider as we go through the pricing analysis. Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: Yes, that's fair. I just wanted to make sure I was thinking about it right in terms of -- to the point you just made, there might be business in there you think shifts to a better cohort in terms of loss ratio over time to the actions you think you could take. There could be business in there that is Homeowners in the state of Oklahoma, which probably is suffering from last year, but you're making changes to try to -- I just want to get around some of -- I'm just trying to make sure I'm thinking about some of those things correctly. Jay S. Fishman: You are, although, I'll point out this is just commercial accounts you're looking at. [indiscernible] This is just commercial accounts. So Homeowners Oklahoma, not part of this, but your concept is right. Brian W. MacLean: But your concept is right. Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: Yes, okay. I could have said small business Alabama instead. Okay, no, that's fair. And then just on the dividend, I guess one of my questions I have is, if cycles have changed, and we've talked about this a little bit in the past, and we're now troughing out at higher levels of -- higher earnings return levels than we would have in the past, how do you change -- you just raised the dividend, but if I kind of look at the new dividend, it's only -- if 2011 accident year earnings with kind of a 3.5% cat load is the right way to think about kind of the earnings base, that would still only represent kind of a high-30s payout ratio, which when you think about historical trough payout ratios or peak payout ratios for all the trough earnings seems pretty low. So just curious how you -- how -- whether or not that impacts how you think about the dividend over time and potentially kind of what a dividend payout could look like. Jay S. Fishman: First, I'll tell you that we analyze it here independent of any cycle dynamics. We just, as I -- we shared with you before, we're not big believers in the concept of some mythical cycle that carries everyone up and then carries everyone down. But the nice thing about our whole capital strategy is that our philosophy is to return it all, is to return all the excess capital, and it's really in a combination of share buybacks as well as dividends. And so we don't perseverate a lot about the payout ratio. We also recognize that you can have a 1 in 500 event tomorrow and then you could have another 1 in 500 event the following week, and suddenly our payout ratio would look real high. So we always stand ready to pay claims, have liquidity, be available to make the payments we need to make and yet have a philosophy of we're not keeping any capital here that we don't think we can use productively. And we could certainly have an interesting debate about whether dividends and share buybacks have preference. And I've talked to many shareholders and I would tell you that it's split about 50-50 as I talk to them. So it's not something that we spend a lot of time, really, dwelling on. As long as we have a competitive yield, an attractive yield for owning our stock and we're committed to returning the rest of the capital, we think we're in pretty good position.

Our next question comes from Larry Greenberg of Langen McAlenney. Larry Greenberg - Langen McAlenney: On the commercial auto reserve strengthening, you had strengthening last year for that line. It's continuing now. You mentioned frequency and severity. I'm wondering if, given how long this is persisting, whether or not it's become more of a severe -- severity issue on the margin for you guys. Jay S. Benet: Yes, Larry, this is Jay Benet. The trends that we had been seeing, we factor into our loss specs. And what we saw in the first quarter, based upon the analysis of the fourth quarter results, was a higher level of frequency and a higher level of severity. In the physical damage area, we're just seeing higher costs associated with car repairs. And every time you get more data, you're trying to evaluate what the proper pick is. And in some respects, we're chasing it a little bit. The dollar magnitude is not very much. When you look at what we're seeing in commercial auto as part of the reserve development, it's like $50 million. So don't think -- on a pretax basis. Don't think of this as some very, very large item for us, but like everything else we're doing, we react quickly and decisively to the kind of information we see. And as Brian said, it does impact our pricing as we move forward. Larry Greenberg - Langen McAlenney: Great, that's helpful. And then can you just give us a little color, if there's any to be had, on your lower advertising spend for your direct consumer? Clearly, the underwriting loss associated with debs [ph] also came down in the quarter. Has -- is anything changing on the margin with the strategy there? Jay S. Fishman: Larry, no. Our strategy remains the same, and I think this is worth just 1 minute or 2. It's -- we made this investment as a long-term investment, and I'd just start off by saying that none of us feel differently about how we believe our independent agents' personal lines business is going to act. We know the customers we have. We know their socioeconomic dynamics. We know their risk management needs. It's hard for us to believe that, that class of customer is ultimately one day going to really want to take it on himself without someone advocating for choice and alternatives, but we also recognize that we could be wrong. And so we're going to -- we began to underwrite this with 2 concepts. First, a defensive one, which was, if we were wrong about the longer term and we wanted to be positioned to be able to respond. And then offensively, we also saw the possibility. And we do see who the customer is who buys direct, that's one of the real learnings of this. We understand now with real clarity who it is who's buying, that to the extent that change continued could we create a model that would build a successful business around it? And we started with nothing. We started without the technology or people in seats or an understanding or a marketing program. And over all these periods, I can tell you the things that have gone just terrifically. We have a technology platform that's solid and reliable. We have developed a marketing approach that actually does generate response, importantly, things that we learn. We figured, once someone would begin a quote process, they'd finish it. And we learned in the early days, not so. That in our early days, lots of the customers would start and then drop off. That required changes in processes and people and the rest, and ultimately, we now have solved that problem. We're pretty good at getting people to respond. We're pretty good at getting responders to get to a quote. And the next step is to get the quotes converted to sales. And so now, we're talking about not only process but pricing and competitive positioning and how the product is sold and the advice that's given. And we're going to gear our advertising spend now to when there's something to be learned when we're making a change, implementing a process and we've got to see what happens, so we'll go out and we'll be more active and we'll drive activity and we'll be able -- we'll be able to see what comes back at us. If we -- the great risk in all this, and 20 years ago, this happened to us once before, we had a great marketing success but a financial failure. And we're going to guard very carefully not to have that happen. And before we begin to really step down, we're going to make sure that there's a real -- a high likelihood of success financially that we can make the requisite investments, start to do the spending that's necessary to drive more activity and ultimately convert it, not only to sales but to sales that are priced and underwritten that produce an appropriate return. And we're just not there yet. We're not at that level. But we have a talented group of people who are driving the effort. It gets the attention of the senior leadership here. It's 1 of the 2 most strategic initiatives that we have. It's not a backwater activity, it's front and center. I'm proud of what we're accomplishing, but we're not going to make the real bad misstep of being convinced that we're ready to go before we are. So it's just going to be cautious, and we're going to keep learning and we're going to keep going. But I wouldn't read anything into the timing of our spending other than what's tied to the underlying process changes from which we're trying to learn.

Jay, could I add one thing? This is Doreen Spadorcia. And I just wanted to make one other comment. In addition to that, we will have a lumpy spend. It's not going to be a set amount of dollars that you'll see every month. For example, we're looking at local heavy-up in the next few quarters, so you may not even see us on national TV. But in particular states where we know we've got a broad product appetite and match, we'll be in those places. In addition, we are spending quite a bit of time on digital marketing. Again, things you might not see and the cost structure might be different but very targeted, again, to the kind of customer that we match with so, again to Jay's point, that we can push that conversion rate up. So you just won't see probably big national spends, but we are doing marketing. Thank you.

At this time, I'd like to turn the call back over to Gabriella Nawi for any closing remarks.

Very good. Thank you, Tina. Thank you for joining us all today. As always, Andrew Hersom and myself are available in the Investor Relations department for any follow-up. Thanks, and have a great day.

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.