Greg Toczydlowski - President of Personal Insurance Jay S. Fishman - Chairman, Chief Executive Officer and Chairman of Executive Committee Bill Malugen - Jay S. Benet - Vice Chairman and Chief Financial Officer Doreen Spadorcia - Chief Executive Officer of Personal Insurance and Executive Vice President of Claim Services Gabriella Nawi - Senior Vice President of Investor Relations and Member of Operating Committee William E. Cunningham - Executive Vice President of Business Insurance William H. Heyman - Vice Chairman and Chief Investment Officer Brian W. MacLean - President and Chief Operating Officer

Matthew G. Heimermann - JP Morgan Chase & Co, Research Division Michael Nannizzi - Goldman Sachs Group Inc., Research Division Jay Gelb - Barclays Capital, Research Division Cliff Gallant - Keefe, Bruyette, & Woods, Inc., Research Division Joshua D. Shanker - Deutsche Bank AG, Research Division Gregory Locraft - Morgan Stanley, Research Division Vinay Misquith - Evercore Partners Inc., Research Division Brian Meredith - UBS Investment Bank, Research Division Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division Jay A. Cohen - BofA Merrill Lynch, Research Division Keith F. Walsh - Citigroup Inc, Research Division

Good morning, ladies and gentlemen, and welcome to the fourth quarter and full year results conference for Travelers. [Operator Instructions] As a reminder, this conference call is being recorded on Tuesday, January 24, 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, Carlos. Good morning, and welcome to Travelers discussion of our fourth quarter and 2011 results. Hopefully, all of you have seen our press release, financial supplements and webcast presentation released earlier this morning. All of the these materials can be found on our website at www.travelers.com under the Investor 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 or 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 on the Investor section on our website. And now Jay Fishman. Jay S. Fishman: Thank you, Gabby. Good morning, everyone, and thank you for joining us today. I'm pleased that we posted solid results in the fourth quarter with net income per share at $1.51 and return on equity of 10%. The story for the full year, of course, was the remarkable weather. Weather losses were far and away the worse in our history, and as a consequence, net earnings per share of $3.36 for the year were considerably lower than what we've become accustomed to. Nevertheless, we price our product for the long term, and of course, we don't control the weather. There will be years such as 2006 and '07 in which weather losses are surprisingly low, and of course, there will be years, but hopefully not too many, in which losses will be high. Also a less financially significant, but nonetheless important piece of the 2011 story was the persistent low fixed-income investment environment. A 1.5 years ago, in the middle of 2010, we made the important decision that we were going to actively take steps in terms of rate terms and conditions to improve future returns in light of the insurance pricing environment and our view that it was increasingly likely that we would be facing low fixed-income yields for some time to come. We increased those efforts in the middle of 2011 in recognition of the possibility that weather patterns might be changing prospectively for the worst, particularly given the active weather we experienced in 2010 and into the first half of 2011. We've shared with you before that we are just not big believers in magic market cycles. We understand that our business is about one underwriter and one agent discussing one account at a time. But we believe that our franchise were strong enough that we could seek improved profitability through improved rate terms and conditions without adverse effect to our business. And 18 months later, we have proven that to be the case in most of our commercial businesses. We take these actions selectively where analytics suggest they are warranted, and we try very hard never to be disruptive to our agents or insureds. In our judgment, the success we've experienced speaks to the value that we bring to agents and customers and the success we've had positions the company very well going into 2012. In particular, Business Insurance, excluding National Accounts, had a pure renewal rate gain of more than 6% for the quarter and 8% for December. As a result, given our current view of loss trends, we anticipate widening of the underlying underwriting margins in Business Insurance in the first half of 2012. In Personal Insurance, the regulatory environment is such that while the progress we've made is important, the real evidence of success remains ahead of us as we continue to roll out the pricing and underwriting changes that we have decided to pursue. But I want to emphasize that the steps we are taking are not limited to rate. There are a number of levers available to us in addition to rate as we pursue improved returns. It include underwriting actions, particularly with respect to risk selection and location, as well as changing underwriting standards with respect to roofs and loss history in Personal Insurance. We've always been attentive to our expense base and we continue to be very thoughtful in that regard. Our commitment to improving perspective returns in Personal Insurance is no less than that which we had in Business Insurance when we began this program 18 months ago. Every personalized franchise is unique. Every company's agent group is different, and importantly, the value proposition to the insureds is quite different company to company. We believe that we will be successful given our agents, our insureds and the value that we bring. So while the bottom line financial results this year were less than we would hoped, we couldn't be more pleased and more proud of our people when we look at the progress we've made to achieve improved returns, particularly given the difficult environment that continues to confront the U.S. economy. We have really demonstrate the value of the Travelers franchise to agents and insureds as well as the strength and diversity of the business that has allowed us to achieve all that we have over the past 18 months. We ran the business for the long term and focus on returns over time. We're also pleased and proud that since 2005, we've generated an average annual upgrade and return on equity of 13%. And with that, let me turn it over to Jay. Jay S. Benet: Thanks, Jay. Let me begin by stating that we have maintained our strong cash position, ending the year with holding company liquidity of $2.4 billion. Operating cash flow is $351 million for the quarter, which included unusually high claim payments that resulted from this year's severe weather and $150 million payment to our qualified pension plan that even in this very low interest environment kept our funding level close to 90%. And consistent with an ongoing capital management strategy, we continue to return excess capital to our shareholders. During the quarter, we repurchased $1.2 billion of our common stock, which was at the top of our previously announced range and paid $166 million in dividends, bringing year-to-date common stock repurchases to $2.9 billion and year-to-date dividends to $669 million. As was the case this quarter, there have been many periods during the past several years in which our share repurchase and dividends significantly exceeded our earnings. We achieved this by diligently and systematically identifying opportunities to free up capital, returning that freed-up capital to our shareholders. After several years of working at it, the process of freeing up capital is now largely complete so we now expect the future common share repurchases will be driven more by future earnings levels, taking into account capital needed for business growth and corporate needs such as death service and pension obligations. Having said that, we remain fully committed to identifying and returning excess capital shareholders, whatever the source. I'd also like to point out that all of our capital ratios remain at or better than target levels at the end of the year. Net unrealized investment gains increased by $337 million during the quarter to $4.4 billion pretax or almost $2.9 billion after tax and book value per share rose to $62.32, a 7% increase since the beginning of the year. So now let me turn the microphone over to Brian to go through the results in more detail. Brian W. MacLean: Thanks, Jay. I'll start with Business Insurance, and the real news in this segment is that the rate on the business we are writing is up significantly in the last 6 months and is now meaningfully exceeding our current view of loss trend. Renewal premium change of 8% for the quarter included pure rate increases of over 6%. This marks the fourth consecutive quarter that Business Insurance has seen positive rate change on renewed accounts, and as with recent quarters, rate increase across all product lines, with the largest increases in workers' comp and auto. We are very pleased that the months within the quarter continue to show sequential improvement in pricing with December renewal premium change of 10% and pure rate increase of 8% for the segment. And based on our preliminary look at January business, pure rate gain looks pretty similar to what we achieved in December. Net written premiums were up year-over-year, driven by pricing gains and positive audit premiums. Retention and new business levels were solid, though down somewhat from recent quarters. As we mentioned last quarter, execution on our pricing strategy may impact business volumes going forward, and so far we continue to remain very comfortable with a balance between improved pricing and volume. Through 2011, this has been a very easy call. I want to emphasize here that when we speak about pricing strategy, our goal is not simply to increase the price on every account. It's a very granular approach based on analytics where we're looking for the appropriate price for the individual risk or the class of business. It's never been a one-size-fits-all approach. Turning to losses. Overall loss trend was relatively benign. In workers’ comp, we saw some higher-than-expected frequency from 2010. In commercial auto, 2009 and 2010 bodily injury and 2011 physical damage losses were above our initial expectation. So some movement in loss trend concentrated in the 2010 accident year, but more than offset by increased prices. In summary, we feel extremely positive about the results in Business Insurance. And in our current pricing levels, we believe that we will be generating underlying underwriting margin expansion beginning in the first half of 2012. In the Financial, Professional & international Insurance segment, operating income of $152 million for the quarter was very strong and consistent with the prior year quarter. This caps a very good year for the segment with full year earnings up more than 4% over what was a good 2010 result. In surety, we continue to see strong margins and slightly less volume driven by sluggishness in construction spending. We could not be more pleased that we've been able to maintain our current position as the market leader in this business during a very challenging economic environment without a meaningful increase in losses. That we have done so speaks volumes about our ability to select risks, the credit quality of this business and our approach of running this business for the long term. Looking at management liability. This remains a line of business where market pricing continues to be challenged, but less so at the smaller end, which is the source of most of our premium. While our pricing here has lagged most of our other businesses, we were able to achieve positive renewal premium change for the second quarter in a row. This is reflective of favorable production results in our private and non-profit business and is due in part to benefits from recent technology investments. In international, net written premiums were down quarter-over-quarter due primarily to the timing of a few small reinsurance transactions and our exit from the personal lines of business in Ireland. Over the past several years, we have increased our focus on our international operations, and we are pleased with the progress to date. The quarter-over-quarter improvement in the underlying loss ratio more than offset the increase in the expense ratio, which is a good sign that the investment is paying off. In Personal Insurance. Looking at auto profitability, the story is essentially the same as discussed for commercial auto. We've seen a moderate increase from expectations in bodily injury severity for the most recent accident years and the higher-than-anticipated level of claim frequency in severity in the physical damage coverages for the 2011 accident year. The quarter-over-quarter impact of these items both from prior year and 2011 prior quarters was almost 6 combined ratio points. Because of this volatility and the normal seasonality in the quarter, we think a better way to focus on the run rate of the business is to look at the full year combined ratio, which after adjusting for cats and prior-year development, was 99.5. This includes 1 point directly attributable to higher non-cat weather-related losses. And so 98.5 is our view of the current run rate, which is only 2 to 3 points higher than the combined ratio required to get us back to our historical target return. As Jay mentioned, we are committed to improving returns in this product in 2012 by aggressively pursuing rate and implementing local underwriting strategies. In Homeowners and Other, profits were once again down in the quarter due primarily to adverse weather. Although, the impact of the weather in the fourth quarter was dramatically less than in recent quarters, it was still slightly higher than expectations, driven by the unusual northeast October snowstorm. In addition, we did have some increased claim activity in the non-cat weather from earlier in the year. As we have said many times, we obviously don't know what the weather will be in 2012, but given the experience of the last several years, we are managing the business with the expectation that weather losses may continue to run above longer term historical averages. Accordingly, we are taking a broad-based assertive approach to managing the property product, which includes not only executing local pricing strategies but also seeking to tighten underwriting standards and changes in terms and conditions. As examples, we are underwriting for age and quality of roof, re-assessing how we factor in the number and type of previous claims and implementing higher minimum deductibles for weather and non-weather claims. Additionally, we are changing a small portion of our re-insurance program to create a structure that better responds to higher frequency, lower severity events. So we are taking a lot of actions in homeowners and are confident that these actions will further enhance the long-term profitability of our industry-leading franchise. Turning to production. For both Agency Auto and Property, renewal premium changed and retention remains strong while new business continues to trend down slightly as we accelerate our pricing, underwriting and terms and conditions action. With that, I'll open it up to.

Carlos, you can open it up for Q&A now.

Yes, ma'am. Ma'am would you like me to read the Q&A instructions now?

No. It's okay. You can open -- you can open it up to questions now.

[Operator Instructions] Ma'am, would you like me to go to the first question?

Yes, please.

And the first question comes from the line of Keith Walsh with Citi. Keith F. Walsh - Citigroup Inc, Research Division: Couple of questions. First, Business Insurance. You've alluded to in the press release the underwriting margin, it looked like it would potentially improve in the first half of '12. Is this more optimistic than what you guys talked about last quarter where you seem to be a lot more adamant that this would take time to earn in over the next 12 months? And then I've got a follow-up. Jay S. Fishman: Well, Keith, this is Jay. The rate that we earned -- the rate that we wrote -- I'm sorry, the rate we wrote in the fourth quarter was obviously at a higher level than that which we wrote in the third quarter. So going into 2012, as we look out into 2012 as it earns, we are not more optimistic because that implies an attitude. This is fact. The fact is, the rate in the fourth quarter was higher than the rate in the third. So wherever we were in the third quarter with respect to margins, we're better off today, and it will simply take the time to earn that written premium into earn. So I'd say that it's not a matter of optimism, meaning feeling, it's really a matter of arithmetic and fact. We're in a better position today than we were at the end of the third quarter with respect to the speed with which those margins will expand, and assuming that we continue to earn rate at this level, the effect that it will have in the margin arena. Keith F. Walsh - Citigroup Inc, Research Division: Okay. And then just sticking with Business Insurance. Obviously, the rates are good, but let's focus on the other part of the game here. I guess, the retention in new business which are both pretty -- down pretty sharply on the new business side, especially. So what are the dynamics of not just the accounts you're keeping, which we can see in the rate, but what about the accounts you are losing? What's really going on there? If you could talk about new business hit rate. Thanks. Jay S. Fishman: First, I would disagree categorically with your characterization of this down sharply. The retention in Business Insurance for the fourth quarter was at 81%... Jay S. Benet: 79%. Jay S. Fishman: I'm sorry. 79%. 79% was commercial. 79%. In the context of thinking about business long term, that's still an exceptionally high retention rate. I mean, we've gotten used to rates in those mid-80s that, frankly, we've never ever seen before. If you go back to the period in the late '90s and early 2000s and middle market retention, we struggled to get it above a number that started with a 7. So what's happening on the Business Insurance side is really -- it's just fine. This is an easy trade-off. The one segment of the business where retention has been lower than others has been in the larger end of Select, right? You may recall that we define Select as the smallest end, which is highly technological. And then there's Express Plus, which is the larger end of Select. That category remains a more challenging pricing environment. It has to do with the way we think that regionals compete and how they view middle market pricing. They don't have the technology to really compete effectively at the low end of small commercial. They don't have the infrastructure to compete effectively at the true middle market business that we define, so they tend to land in that larger end segment. We've established our own thresholds for return expectations, and to the extent that we can capture business that meets those return thresholds, we'll take it, and to the extent somebody else decides they want to take it away from us at a lower price, that's just fine. We're -- so far, this has gone better than I think any of us would have expected 18 months ago. And the trade-off at this point between return gains and retentions, as we described it, is losses. It's just fine. What was the 81%? The 81% was... Brian W. MacLean: 81% was Commercial Accounts... Jay S. Fishman: Commercial accounts, yes... Brian W. MacLean: Overall -- this is Brian. Overall in retention, x the high end of Select, we are right around 80%, a little bit north of 80%, and we feel great about that. On the new business front, we're going to look at business in the marketplace and if we can find stuff that meets our return thresholds, we're going to write it. And in this environment, we're not going to search below that. Jay S. Fishman: I'd make other observation -- and it's more challenging for us in quantify it and demonstrate it, but we have indicators that will support what I'm about to say. New business pricing is also better than it was. It's part of the measure because you simply don't have the prior-year expiring premium to compare it, but we are comparing new business pricing against the manual rate that we construct. And when you look at that pattern, there are 2 things that are really encouraging. One is that there's clearly lift going on. And two, that the gap between -- and I'm not speaking necessarily about this particular quarter, but looking at it over a longer period of time here, 6 , 8 months, the gap between renewal and new has narrowed. And so both of those things give us a pretty good sense of confidence that the pricing for new business is lifting as well. Can't quite quantify it with the level of precision that we can renewal, but we feel awfully good about it. Brian W. MacLean: The one comment last comment I'd make on new is that our flow is still very solid. So if this was a matter of we weren't getting opportunity to quote on business and we weren't quoting, we'd be a lot more concerned with that. So we feel good about the flow, we feel good about our activity in the marketplace and profitably levels will be where they will be.

Our next question comes from the line of Jay Cohen with Bank of America Merrill Lynch. Jay A. Cohen - BofA Merrill Lynch, Research Division: Two questions, both related to pricing. Can you talk about your desire in 2012 to continue to push for rate on the commercial side? Because obviously, you've achieved a certain amount of rate already. Is that desire still very much in place? Or will you be satisfied now where the priced ROEs are and you don't need to push as much? That's question number one. And then secondly, if you could describe the dialogue you're having with agents and brokers. Obviously, your effort is to improve profitability, improve returns, theirs is to do the best for their clients. How much pushback or are you getting from the agents and have you sensed a change in that pushback? Jay S. Fishman: Jay, I'll take on the first one. No, our strategy now remains as is. We're going to continue to do this selectively, thoughtfully targeted, but we are going to continue. And our goal -- we get asked all the time, "What's the target?" We don't have one. And that's partially because some of the conditions that we're responding to were just not as clear to us as they might be. I'm speaking specifically about weather. I'm speaking about investment returns. And so we're going to try and drive our products and our returns back to the historical kind of broad-based targets. We still have a mid-teens ROE as a -- long term. We talked about it over time, but we still very much have that as a long-term target. We are most certainly, from an individual pricing basis, this is important now, I'm not speaking about the return on equity in the whole place, I'm getting very granular with you in talking about the returns on allocated capital of individual products based upon the rate that we have in place today and our immediate view of loss trend. We are getting closer. We are making real progress in getting back to the levels that will ultimately convert it to mid-teens ROE on a GAAP basis. So we got some to go yet but we are encouraged and we're going to keep driving until we find whether we can or not get back to our historical kind of return thresholds. So that's, I think, the answer with what does 2012 feel like from a strategic and tactical perspective. [indiscernible] Bill Cunningham to talk about the agent arena. William E. Cunningham: Good morning. I would say on overall basis, we look at our agent relationships as very strong partnerships and as being a strong partner, not surprising them is critical. So being upfront early in the process with our agent in terms of what we're doing on an overall basis, but more importantly, as Brian mentioned, this is not a one-size-fits-all. So every account communicated to the agent as early in the process as possible because our ultimate goal is together, to be out with the client, explaining why we need a given rate increase on a given transaction. At the same time, we do -- from time to time, agents feel a need to test where we're coming from, so they give them time to do that. But I would tell you on an overall basis, it has been very positive in terms of the feedback we received and how we've handled it. And so again, partnership early in the process, no surprises, together sitting down with the customer whenever possible and not putting them in a position where their back is against the wall is key to our partnership.

Our next question comes from the line of Brian Meredith of UBS. Brian Meredith - UBS Investment Bank, Research Division: A couple of questions here for you. First, going back to the auto insurance results that you're seeing in the quarter. When you talked about the 98.5 and need to get it down to a 96, I guess, the question I have there is, one, kind of what are you doing to get it down there? Is it just pure rate? And is the market kind of receptive of the type of rate that you need to get it back down to those 96 levels? And then I have one follow-up.

Brian, these is Greg Toczydlowski. In terms of the auto strategy, predominantly, a rate strategy that we're focused on automobile. When Jay and Brian talked about all the other levers that we're looking at from an underwriting in terms of conditions, we're focused a little more on the property side of the house with that. So we're monitoring our competitive position all the time and given the regulatory environment that we live in, it takes a little bit of time before we can get that rate grow, but we're feeling that we're continuing to see good flow and good retention numbers and we'll continue looking at that as we go forward with the business. Brian Meredith - UBS Investment Bank, Research Division: But obviously, if you're writing at a 98.5, 99, that's not your kind of an acceptable return level. They can't get a double-digit ROE off of that, right?

That is correct. We are shooting closer to that combined ratio that you threw off [ph] in 96 and we're going to be doing that through pricing. Jay S. Fishman: And some of the dynamics, Brian, that we've been experiencing we suspect are more broad-based. This is just one. But it's -- some unusual -- the pattern has developed with the slowdown of the new automobile business generally. Price of used cars has risen. If the price of used cars has risen, what's been interesting is the number of accidents that result in total losses, where the car produces salvage, have actually gone down. The car relative value up. Auto repair cost much less so more cars are being put back on the road. And a consequence, we've seen of that is an increase in sheet metal cost. I'm not talking about new sheet metal. I'm speaking now about replacement parts. And that's not unique to us. That's going to be a condition that's going to affect lots of people. Now it may impact us a tad more because we tend to write more comprehensive coverage than some other carriers do, so we may be on a more extreme side of that. But the kinds of things that we're seeing are not things that -- it's very important, are not things that indicate problems with underwriting or risk selection. They are more episodic issues with respect to specific costs that are not, I would say, unique to us . So our hope is, and of course, we are going to find out, that we can go to the market and drive the kinds of rate increases that will bring this back, just like we did in Business Insurance. We started extremely slowly. We started 18 months ago asking fundamentally if our folks can get 1 point. That's what we said. Let's see if we can get 1 point, and we've built on that every month. We're going to take that same systematic, thoughtful and patient approach to auto insurance and homeowners and see what progress we can make. Brian Meredith - UBS Investment Bank, Research Division: Great. And then question for Brian. When you mentioned pure rate also up 8%, can you define that? Is that just rate less cost inflation? Brian W. MacLean: No. When we say pure rate, we're looking at the components of the premium change. Jay S. Benet: No exposure. Brian W. MacLean: So it's the rate charged per -- exposure on that, on the average renewal. Brian Meredith - UBS Investment Bank, Research Division: Got you. And then how would that compare to what loss cost inflation kind of looks like in the quarter? Brian W. MacLean: We don't peg the exact number, but we've said we're meaningfully exceeding it. Brian Meredith - UBS Investment Bank, Research Division: Great. Excellent. And then just lastly, Interest rate assumption when you're talking about getting rate, getting close to your kind of historical kind of double-digit ROE targets, what are you thinking about the interest rate environment? Jay S. Fishman: Mechanics on that are that all new money is discounted at the current yield curve, basically, where we invest money today, and the surplus embedded in the business in that calculation is a more historical element premised on the fact that we have surplus that rolls off from old products and then comes on to support new. But importantly, we're not kidding ourselves the investment return. That's all new money at all new -- at current yield curve today.

Our next question comes from the line of Vinay Misquith with Evercore Partners. Vinay Misquith - Evercore Partners Inc., Research Division: The first question is on pricing. We've seen pricing and workers compensation and commercial auto up sharply because of high loss cost trends. Curious about what are trends you've seen in pricing and other lines that you're pushing this year versus last year? Brian W. MacLean: So a couple of just clarifications on that. The comp and the auto, we have the strongest rate improvement. It's not just because we're seeing -- or it's not because we're seeing sharp increases in loss cost there. Obviously, pricing always has something to do with the loss cost trend, but it's not a sharp increase. We saw a fairly significant increase of a couple of points, really, on almost every single line of business we write. So this isn't an example where comp is spiking up dramatically. Auto's up a lot and everything else is kind of flat. I would say every line in the commercial space is moving comp and auto more dramatically, but it's not completely out of pattern with everything else. Vinay Misquith - Evercore Partners Inc., Research Division: Are you talking about rate? Brian W. MacLean: Rate. I'm talking about rate change. Jay S. Fishman: And as Brian reminds all of us, and you as well. In many of our lines of business. It's very much a state-by-state story. There tends to be this view that it's one national rate. It's not. There are states where either frequency or severity underlying trends are either more concerning or less concerning than others. So our reaction with respect to rate and what we're trying to accomplish is very much driven by the state environment, specific to that particular risk. Vinay Misquith - Evercore Partners Inc., Research Division: Fair enough. That's helpful. As a follow-up, what is the impact of non-cat to weather for all the segments? I believe for the personal lines, it was 2 years for the full year. Brian W. MacLean: Non-cat weather? I'm looking at Greg. Non-cat weather in PI for the full year. Can we disclose that level of granularity, I guess, is the question.

It's not something that we've disclosed previously. We obviously have some view of it, but the substance of that answer is functionally dependent upon whether our assumption of normal weather is right or wrong or high or low. So rather than trying to get into a debate of how much was non-cat weather, we just -- that's just -- it's internal, and I'm not sure the data is all that meaningful. Vinay Misquith - Evercore Partners Inc., Research Division: Okay. Fair enough. And just one last thing, if I may, what -- to what amount of rate increase do you think on the personal auto line do you think will be necessary for you to get to your level of profitability that you're targeting? Jay S. Benet: Well, the arithmetic is actually pretty simple. In Personal Insurance, if we want to drive the combined ratio down by 2 points, we got to get about 3 points of rate above and beyond what we otherwise would, because we are obviously in effect, I'd say, always getting rate. That's not universally true. But fundamentally, it's a business where the renewal price change has always been positive. I can't recall that it ever went to a [indiscernible]...

Up 2 to 3 points. Jay S. Benet: So that's really the dynamic of kind of 3 points of rate above and beyond we would otherwise get, and under the presumption that loss trend stays about where it is. So there's an all other things being equal dynamic. But the answer to that is it's not as far away as it might otherwise see. Now the marketplace will have to say whatever it wants to say and we're obviously aware of that. And our influence, our market consequence, our breadth in Business Insurance is I think one of the attributes that's enabled us to accomplish what we did in PI whether -- we'll see whether that same dynamic holds true in personal in the same way. But it's just not that far off. Brian W. MacLean: Just jumping back to his first question because we have answered part of this in the past, so on the non-cat weather. I mean, we said last quarter in Business Insurance, and it's still probably about the right number, is there's a point in the combined ratio of variance above a normal non-cat weather number for the year. I said -- in the quarter, it was pretty flat in BI, but for the year, it' still about 1 point. In auto personal lines, I said there was a point of clearly identifiable non-cat weather it might actually be more in auto. And the Whole number, I don't remember off the top of my head.

They are going to go on a same basis. Full year property would be closer to 2 to 3 points. Brian W. MacLean: Variance from a non-cat weather. Jay S. Fishman: From whatever we [indiscernible] coming back to the notion of... Brian W. MacLean: Whatever was normally our view of normal and weather always been.

Our next question comes from the line of Jay Gelb with Barclays Capital. Jay Gelb - Barclays Capital, Research Division: Two questions for you. The first is on the share buyback pace for 2012. Should we be thinking about buybacks being equal to net income, operating earnings, retained earnings, can you just clarify? Brian W. MacLean: Well, what we were trying to say was, it should be looked at based on earnings, so not retained earnings for sure. And then net income versus operating income, they are both income and they have tended to be relatively similar. So we think mostly in terms of operating income but if they were large realized gains or something, we would take that into account as well, but that would be more episodic. Jay Gelb - Barclays Capital, Research Division: Okay. I just want to confirm. And then what's the sort of maximum comfort level on debt to capital x the unrealized gains? Jay S. Benet: We've talked in the past about -- for AA company like us it's being in the 15% to 25% range. We are at a little over 23%. Part of the 23% is actually driven by prefunding $250 million of debt that's going to mature in 2012. So that's worth about 7 basis points in that calculation. But being within that range is a very comfortable place for us. Jay Gelb - Barclays Capital, Research Division: Got you. And then the final one is on the non-fixed income, after-tax earnings, it was down in 4Q versus 3Q. I'm just try to figure out what you feel would be a reasonable run rate heading into this year. William H. Heyman: Jay, Bill Heyman. For what it's worth, the quarter came in just about exactly where we thought the quarter would come in on October 1 as we look to the matrix of factors which affect returns in the non-fixed income portfolio. I suspect the best way to make predictions about the future is to look at the bar graph. Back a few years, try to draw a mean, I figured that's about what we'll make because if we couldn't make that, we probably wouldn't maintain the portfolio. For what it's worth, there were no real surprises in the fourth quarter. This was about what we expected. Jay Gelb - Barclays Capital, Research Division: Okay. So if I look at that over the past couple of years, probably averaging around $50 million to $65 million a quarter?

I'd have to look at the bar graph. I think on the low side of that. Jay S. Benet: This is Jay. It's one of those things, as you know, it's really difficult to predict. So whatever you put in your model, there is a level of uncertainty. We have great visibility into the fixed-income part of the portfolio but this is going to be so dependent upon what the economy does.

Our next question comes from the line of Greg Locraft with Morgan Stanley. Gregory Locraft - Morgan Stanley, Research Division: I wanted to dig into the workers comp line for a bit, just get some color on the market, and I guess, more specifically, one of your competitors in Hartford has had -- has had some trouble -- some emerging troubles grow through 2011 in that line. Some other cracks are appearing from other competition. I know you don't believe in the cycle but what are you seeing specifically in that line? How do they manage their books of business different? And why are you so confident that this won't begin to have some issues into '12 and beyond? Jay S. Fishman: Let me talk for a second about -- in a little more specific, what we're seeing. I'm not going to comment on how other companies might be managing their business but we can talk certainly about how we manage ours. So just to be clear on workers comp, what we saw this year, in the aggregate, continues to be a pretty strong picture for us. So if you look at the prior years, '02 through '09, every year continued to perform better than we had expected. So positive in every one of those years. In 2010, we did see... Jay S. Fishman: Bill said to me yesterday, and make sure you still believe this, that if you look back to '02, the current expected loss outcome in each of the year '02 through '09 is currently lower than the original estimate at the time we made it. Jay S. Benet: Better. Lower loss ratio, better result, right. Jay S. Fishman: '02 through '09. So in 2010, we did see some continuation of the increase in late reported claims that we discussed in the second quarter. And we believe this is really a direct result of the impact of the great recession on workers comp claim reporting patterns. And so the 2010 year was out of pattern. But overall, all years, including 2010, we continue to have positive development on our overall workers comp book. The impact of all of that is fully reflected in not just 2010, but rolled into how we look at 2011 and what we booked there. And most importantly, is completely rolled into our pricing the products today. And given the pricing improvements that we've gotten, we feel good about the returns and the profitability levels of the comp business we're writing today. Now as Bill said before, it's a state-by-state game and very, very granularly managed. Jay S. Fishman: And I've got a couple of other facts that I think would be helpful. First, the magnitude of the 2010, I'll call it, excess claim activity, to put a number on it, it was under $100 million after-tax. That's the -- that's the magnitude of that change. And again, what we see -- everything that we see tells us that those were occurring to dates, incidents in '10 that didn't manifest themselves until 2011. That's an unusual dynamic and something we haven't seen before. Maybe it's a function of the economy and how people feel about job security and whether they raise their hands or not. And again to Brian's point, we've rolled that forward into the 2011 loss estimate, book that fully and are currently incorporating that into our 2012 estimate for pricing and reserving. And this is -- I think it's important because it will, again, help quantify it. The combination of the rate gains and the loss trends and the positives and negatives and everything else, our estimate of that allocated return on capital measure that we use to measure the profitability of the products in workers’ comp all-in now is low double digits. It's low double digits. So it's, I think, it's helpful to put a perspective on what we saw. Now our view on that 2010 development is that it's more, I'll call it timing rather than a permanent change. And of course, we'll see whether that's the case or not. To get back to your other question about -- and of course, you all know this, you're sophisticated analyst, you get it. The issue of how one company reports versus another. The critical element in all this is what did the company generally assumed? And so whatever experience we have in the loss arena is sort of one thing. But what original assumptions were made is a critical element in determining the embedded health of the business and I'd come back to the ‘02 to '09 period as at least indicative of the way we view the line. Importantly, severity remains as expected, that's an important element so the actual inflation dynamic there is not an issue. And lastly, and again, this just speaks to the mechanics of the business, and I think the integrity of the process here, the delta, the change that we had assumed from the '10 into the '11 year, that turned out to be a good -- at least so far, remains a good assumption. The base of your [indiscernible] change some, but the delta into '11 that we had anticipated turned out to be, at least, so far consistent. So a tiny bit of a complicated story but you've got all the news.

Our next question comes from the line of Michael Nannizzi with Goldman Sachs. Michael Nannizzi - Goldman Sachs Group Inc., Research Division: Just a couple questions here. One thing you haven't talked about for a little while is the direct initiative. You mentioned and it looked like in the release about 80 basis points on a combined ratio, which something about $0.30 a share, if my math is right. Clearly, more significant than it was as a percentage of earnings when you launched it. Just kind of what do you look at in the development of that business to let you know that you're creating value and that you're hitting your objective so that it eventually starts to bring in earnings as opposed create that pressure? And just one follow-up. Jay S. Fishman: I'll let Doreen and Greg answer as well, but I picked -- the one element that we continue to work on from an operational dynamic is we're getting pretty good at getting people through the quote process and getting quotes out. And our next step is to do better. Next step, we're working on more than anything else right now is to do better in converting those quotes into actual sales. And it's not immediately apparent to us that it's a function of price, although, it won't surprise any of you. If you listen to all the ads, the comment is those who switched saved whatever the number is, $400. No great shock. $400, no great shock if you quote someone $400 more than they're currently paid, the likelihood of switching is actually pretty low. So to comment, price is not an insignificant element there. We have a -- I don't know exactly what to call it, but we have a conversion curve. We know -- relative to at least what the customer discloses to us, their current prices is, we know how far, either plus or minus to that mark, we -- in a large number of accounts, how close we need to be to convert. But we've got more work to do in the conversion process so that the dollars that we're spending in the media -- and we're not spending nearly what anyone else and the big players, but to maximize that investment, just to get more of the quotes converted into sales. So that's our dynamic. Some of it is, we suspect, I'll say, reputational. I think there are other companies that have, as a result of doing this for years, found a reputation for being less expensive. Whether that's true or not is a whole other question, but they've seen to have earned a reputation for it. And we have to continue to build a reputation dynamic around the value proposition that we give. We're never going to be great at selling half a coverage sandwich, which $6. We're not comfortable selling half the sandwich for $3.10. That's just certainly not that who we are culturally. We're much more comfortable providing customers with the coverage they need rather than the dynamic of how much they can afford. That's just not us. So we're going to keep working at this and we got our fingers crossed, but we're well on our way. Michael Nannizzi - Goldman Sachs Group Inc., Research Division: Got it. And then just a little bit more on workers comp. Just the premium line was pretty bumpy during the year. I'm just curious. Is that a result of pure seasonality? Is that a result of maybe some change in the competitive dynamic or just the market itself or any strategic change on your end? Brian W. MacLean: It's actually more just pricing and audit premium running through the numbers. So a decent piece of that is movement of the underlying exposure of the accounts and what payrolls have been and how those have moved up. And especially when compared to a year ago when many comp accounts had some fairly significant return premium dynamic. So you're going from a return premium environment to an audit premium plus environment, and that's probably the biggest single driver of the movement in that line. Jay S. Fishman: And from our perspective, it's a good thing.

Yes. Brian W. MacLean: Oh, yes, that's definitely true. Michael Nannizzi - Goldman Sachs Group Inc., Research Division: So it's not initially a policy count, it's more of a exposure per unit? Brian W. MacLean: Yes. Jay S. Fishman: Payrolls. The initial premium is based on an estimate. We're coming down the road. When the policy expires, we do an audit. If the actual payroll was less than had originally been disclosed to us, then we have a return on premium if the payroll turns out to be more than what it was originally disclosed to us, we have an additional billing during the sort of depths of that great recession we were in the return of premium arena as customers shed payrolls far more aggressively than they have anticipated they would. And now they reached the stage where they were flattening out. That brought the audit premium back to 0, and have been growing modestly. And so what we're seeing again is now a change into the build arena of audit premium. Michael Nannizzi - Goldman Sachs Group Inc., Research Division: I mean, any change in the competitive dynamic just over a year, just in the larger case market that you've noticed? William E. Cunningham: This is Bill Cunningham. Nothing discernible. I mean, in a large national... Jay S. Fishman: When you say large, let's make sure... William E. Cunningham: You're talking large national accounts less responsive business? Michael Nannizzi - Goldman Sachs Group Inc., Research Division: Like your commercial accounts business, not the small, not the Select or Select Express, not the small and the commercial but kind of the down-the-middle commercial accounts business. William E. Cunningham: I would just say over the course of the last 12 months, some of the crack that you've described, and we have seen competitors pushing price and moving price in another direction in a number of states that we compete with them.

The next question comes from the line of Josh Shanker, Deutsche Bank. Jay S. Fishman: I'm sorry, let me -- before I could take that question, let me just -- because Bill used words that we understand here internally. I think what he was saying -- you should correct me if I'm wrong, but I think what you're saying is when you say cracks, what you're seeing is other competitors struggling with workers comp line and not an insignificant increase in rates by other competitors as well. We never deal with other competitors. We get anecdotal observations from agents. But that's what Bill is speaking about, that we're, particularly, in the states that have been more challenging from a loss trend environment where we see other -- we understand, again, trying to -- we never talk you never talk with other competitors about rates in any way but we see them reaching for rate as well in that environment. That's what our comment was.

Okay. Next question -- Josh Shanker, question, please. Joshua D. Shanker - Deutsche Bank AG, Research Division: During the Goldman Sachs Financial Service Conference Call, you dissuaded investors from taking a look at Schedule P. Obviously, now we're going to look at that. But maybe you know more than we do, and if you can walk the math on why our look Schedule P would lead us to the wrong answers, and maybe if you can talk about what your reserve are going to look like when we do that following all of the 2011 numbers? Jay S. Fishman: First, I don't think we ever dissuaded anybody from looking at the Schedule P. I just -- go back to that Goldman Sachs conference and we were talking about the complexity of making, I'll call it somewhat simplistic assumptions based upon kind of overall views. But our Schedule P is going to reveal the things that Brian shared with you earlier. You're going to see the 2010 accident year for workers comp had more claim frequency than we originally anticipated. You're going to see, in the auto liability line, particularly, I guess not particularly, but commercial auto for sure is going to see 10, 9 and -- 9 and 10? Brian W. MacLean: A tiny bit in 8 but mostly 9 and 10. Jay S. Fishman: 9 and 10. And I think those are going to be the things that you'll see that -- and that's why we're providing the explanation here today, so that people understand what it is they'll be looking at when they see it. Joshua D. Shanker - Deutsche Bank AG, Research Division: And then so far from what you're seeing, looking at the beginning of the year of 2011 to where you are today, those things you're seeing in 2010 have sort of reverted to the mean are not showing up this year? Jay S. Fishman: Well, we booked the change into 2011, first of all. In other words, the base year, what we do, we develop loss estimates for a given year. We started out with the immediate preceding year. We'll make adjustments for frequency and severity that we anticipate. So when 2010 had that $94 million after-tax increase in claim activity that I spoke about, we obviously rolled that forward into our 2011 loss estimate, and that's been booked and recorded. As we look to 2012, we obviously are starting out with the base, again. But now you're asking the complicated question, of that $94 million in that year, how much is a timing difference that we see just from one year to the next, and how much is permanent difference? But we certainly are going into 2012 eyes wide open with respect to the 2010 accident year and what its effect was. That gets me back to my comment earlier that the estimated return on allocated capital for comp broadly speaking. I mean, before [ph] across the whole book is at the moment in the low double-digit range, so it's -- and that's incorporating that 2010 development into that. Jay S. Benet: And if I could add a couple of things. I mean, there's a story line that you have to follow in our Q's where we're describing things that are taking place each quarter, some of which, as Jay said, impact us as we look at the business going forward. And that's when we talk about base year movement and our adjustments for current year loss picks and ultimate results for a particular year. There are other things that take place that get reflected in Schedule P that don't impact the numbers going forward. So if you look at some of the property line, as an example, what you'll see is disclosure in our 10-Qs talking about some development that's taking place, one that comes to mind were the Arizona hail storms that impacted us from 2010 where we made loss estimates associated with what we have been seeing. And then in 2011, saw a further development on those losses that get reflected in our property lines. Those are losses that are episodic, and therefore, don't impact us when it comes to looking at loss fix. So we say it's complicated, those are the kind of things that we're talking about. The granularity of information that we go through to make a determination, number one, are our reserves fairly stated at the end of the quarter which are at always best estimate. And then two, what do this information tells us about the past as well as the future and then trying to very granularly and thoughtfully build those into our ongoing analysis. Jay S. Fishman: The important point I think we made the important point on the Goldman Sachs Day was that there seems to be a view somehow that current year reserve adequacy is a function of late. And that concept is just foreign to us. We don't understand it actually. We make estimates of losses and we make those estimates of losses, frequency, severity, and so the notion that we would not be making our best estimate because a couple of years ago, rate was more challenged relative to 5 years ago when it was less challenged. It was just completely foreign to us. We don't get that. So to us, at least, and I'm speaking just for Travelers, the way we establish loss reserves are independent of whatever the rate is, that's being very granularly determined and they are constantly being reviewed by actuaries internally. And you've heard us say, when we make re-adjustments to current year loss picks, current year loss picks, not prior year, it's because we're seeing activity either up or down that was different from what we had originally expected that loss activity to be, so that was the comment. By all means, dig into Scheduled P. Joshua D. Shanker - Deutsche Bank AG, Research Division: I appreciate, and thank you, very granular answers. Just a quick numbers question. Brian went through the math on personal auto. What is the run rate for this U.S. thing for annual cats per year in the loss ratio, normalized? Jay S. Fishman: In auto? Joshua D. Shanker - Deutsche Bank AG, Research Division: You said you were targeting 96. You got to 98.5, but that doesn't include any normalized cats. So I just want to also get that part. I think, at least, maybe a bit. Jay S. Fishman: No. That -- go ahead answer the question. Brian W. MacLean: Yes, Josh. It leads us [indiscernible] really insignificant for the auto side and put this in normal expectation.

Our next question comes from the line of Matthew Heimermann with JPMorgan. Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: A couple of questions. One, just to start clarifying. Your 99.5 to 98.5 run rate, was that Personal Auto Or Personal Insurance? Brian W. MacLean: That was Personal Auto and the delta is the 1 point of pretty clearly identifiable weather-related activity above that pretty low normal activity. Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: Okay, that's fair. I just wanted to make sure I heard that right. And then within FTI, obviously, you've been exiting the personal lines of business in Ireland. You talked about that from a premium standpoint. Just curious whether or not that exit is having any positive impact on the loss ratio, and if so, how much. Brian W. MacLean: I don't know the exact number. It should have a positive impact, but it will be pretty small. Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: Okay. And then just kind of timing wise, how far are we -- I mean, my sense is there might be one more quarter of kind of cycling through. Is that right in terms of timing? Brian W. MacLean: Maybe 2, but I think the overall personal lines of premium in 2011 was about $24 million, $25 million. It's pretty small and should tell us in the next couple of quarters. Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: And then, I guess, a bigger question, how customers are adapting to the rate increases? Obviously, we know exposures are up on an underlying basis with the economy. Which puts some purchase -- price pressure, at least dollar spend pressure, on customers who are raising rates on top of that. Are you seeing customers either change the limits they're buying? And I recognize not all your customers may have the flexibility to do so, but are you seeing any change in buying behavior? Either limits, deductibles or other things, clients are dealing to try to mitigate kind of the dollar impact on them? William E. Cunningham: This is Bill Cunningham. I would say, we're not seeing any meaningful change in the commercial space on terms of conditions either limits or deductibles. Where we are seeing changes in deductibles would be on the larger property end in wind exposed areas, and that would on the margin, but I would say across the board, clients are opting more for the price increase than a deductible increase. But again, that's largely anecdotal. But the numbers would support that it's more about the price and less about the terms and condition. Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: Okay. Then can you just help, on a written basis, connect the dots between what you reported for rate increases, exposure increases in Business Insurance relative to what we're actually seeing show up on the net line in terms of written premium increases? I mean, I get retentions down, but it doesn't seem to be down relative to the positive push from those 2 things to explain kind of the delta. William E. Cunningham: It's a retention and I would say new business dynamic, the new business being off of it. Brian W. MacLean: I'm not sure I got the question there, Matt. I'm sorry. Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: Well, based on your chart, it looks like you're saying Business Insurance pricing are up 6%. It looks like based on your chart, you're showing exposure growth of somewhere between 2% and 4%, which would make the range 8% to 10%. Retention's down 4%, so that takes that to 4% to 6%. And I get new business is down a little bit but that feels more like 100 bps, so I would net-net expected better growth. Brian W. MacLean: Okay. So you got audit premium spread. Why don't you... Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: That was a positive delta year-on-year too, though. Brian W. MacLean: Yes. Right. So if you're trying to reconcile written premium year-over-year, you've got the rate as a big improvement on a full year basis about 5 points. You got 2.5 points of exposure. You got retention down 5 points plus you've also got new retention business, a change in new business. Jay S. Fishman: The biggest driver offsetting the price gains is retention in new business the quarter. Brian W. MacLean: And then you've got audit premium swinging from negative to positives. Matthew G. Heimermann - JP Morgan Chase & Co, Research Division: Maybe I'll follow up offline because I'm not exactly following.

Our next question comes from the line of Josh Stirling with Sanford Bernstein. Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division: I appreciate all the colors and disclosure on the reserve issues both in change. And you really substantial disclosures in your third quarter stat filings, which, I think, are better than a lot of your peers. The one question I'd have basically the follow-up on Josh's conversation and all the sort of talking we've done about workers' comp and commercial and personal auto and the challenges there. Recent past couple of quarters have seen a lot of releases from surety, as well as I think disclosures on your third quarter regulatory filings suggests that the asset funds and the liability, so claims made in occurrence, are funding much of the recent reserve favorable development, as well as -- it potentially sound like you're maybe a UA [ph] reclassification. And I think we use the word episodic before, I'm curious, should investors be looking at these as the sorts of things that we should be plugging into our models for next year as favorable offsets to potentially sort of modest favorable development and going forward in the other lines where we're talking about potential challenges? Jay S. Fishman: This is Jay Benet. We've answered this question a number of times in the past. We, as a management team, have this very, very robust processes each quarter to evaluate, what our reserve changes look like, and to come up with our best estimate of reserves. We have no guidance. We have no views that we would have either internally or to share externally as to how reserves are going to develop. So I think, you, as an analyst, can feel free to think of whatever patterns you've seen, whatever you're hearing from other companies. As Jay said, it's all dependent upon what a company's assumptions are, what the starting point is, what are the loss ratios that they have at this point in time and how things are going to develop. So one company's development can look a lot different than in other companies. But I think you see our history, you see what we've done in terms of putting up reserves in the past, how those reserves have developed. And as I said, our view is we're always at best estimate and you can come up with whatever you view you think is appropriate. Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division: Yes, that's certainly fair and expected, something like that. Just in the context of sureties, obviously, sort of an episodic, credit-related reserve and then the commentary you guys had suggested around the excess coverage, which is related to favorable traditional environment over the past number of years. And we're just wondering if we should look at this as kind of a onetime thing and maybe you've already answer that so... Jay S. Benet: We're always building in the current view as to how an environment has changed. So if we have certain views of the toward environment, let's say, several years ago that we're driving a certain level of reserves and that sort of environment has improved to date, we'd look at, okay, where is it today and we have some view as to where we think it might go in the future and that would be the basis upon which we reserve. But what we don't do is we don't say, gee there some continuum of improvement. Let's book 1/4 of it now, we'll book a quarter next quarter, and we'll keep -- and keep going. It's always based on facts and circumstances as they exist and it's the best estimate at that point in time. Jay S. Benet: And if you go back in time in our disclosures, what you'll see is the description of what's generating positive development, it's changed. It's been this line or the other line or that year in property or liability, or just -- it's a very big business. We're making estimates for multi-billions of dollars of losses every given year and things just workout somewhat differently than we had originally expected. But it's not as if there is a pattern there that somehow is reliable. It's just changed over the years and different line show up at different times. If we had -- as we said at the Goldman Sachs Day, our goal is to get it right. It's not to -- I know it feels good when you have that favorable reserve development in the period that it's reported, but we're pricing product every day. And if we're going to make the most of what we got, either growing our business or recognizing lost trend or adjusting, the only way we're going to make really thoughtful decisions, really good ones, is if we get it right. So that's our goal. Our goal is to get it right. It's obviously better than -- everybody says favorable, better than unfavorable. But -- and maybe that's true at the moment, but when we have favorable development, that means that we over-costed our product in the year in which -- from the year it relates to, and maybe we could have sold more if we had a different view of costs. So we're very driven to try and get it right. Josh Stirling - Sanford C. Bernstein & Co., LLC., Research Division: Yes, I think that's right. I personally grew up in business always thinking that the right answer is the right answer, and if they could lead you to better decisions, I think you guys, obviously, sort of do a good job of explaining, but we shouldn't count on these things going forward. And so I don't think there's any issues there. If I could just ask one final numbers question.

We'd like to move on to the next question. Thanks. We kind of gave you your tip.

Our last question comes from the line of Cliff Gallant with KBW. Cliff Gallant - Keefe, Bruyette, & Woods, Inc., Research Division: Just on the investment portfolio. Do you anticipate any changes in strategy? And can you comment on the outlook for the non-fixed income portion of the portfolio, please? Jay S. Benet: Well, no changes in strategy. We have a strategy, which we think accommodates a very wide range of conditions. In terms of the outlook for the portfolio, am I allowed to give an outlook? Jay S. Fishman: Well, we sort of did. Someone else had asked, about the non-fixed income... Jay S. Benet: Well, that was what we expected on an ongoing basis. I think I'm being asked for an outlook for the quarter. So the question is -- I guess, typically, we don't an outlook for the quarter.

That was it? Jay S. Fishman: Yes, thank you.

Okay. Well, thank you for joining the call. We appreciate the time given that we went a bit over. But if you have any further questions, please contact myself or Andrew Hersom in the IR department. Thanks, and have a good day.

Ladies and gentlemen, that does conclude the conference call for today. We thank you for your participation and ask that you please disconnect your lines.