旅行家集團 (TRV) 2006 Q3 法說會逐字稿

Good morning, ladies and gentlemen, and welcome to the third-quarter earnings review for St. Paul Travelers.

We ask that you hold all questions until the completion of the formal remarks, at which time you will be given instructions for the question-and-answer session.

At this time I would like to turn the call over to Mr. Michael Connelly, Vice President of Investor Relations.

Mr. Connelly, you may begin.

Good morning.

Welcome to the St. Paul Travelers discussion of third-quarter 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.st.paultravelers.com under the Investor section.

Today with Me I have Jay Fishman, CEO;

Jay Benet, CFO;

Brian MacLean, Chief Operating Officer;

Joe Lacher, Head of our Personal business; and other members of senior management.

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 their 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 following.

Our presentation today includes certain forward-looking information as defined by the Private Securities Litigation Reform Act of 1995.

All statements, other than statements of historical facts, may be forward-looking statements.

Specifically, our earnings guidance is forward-looking, and we may make other forward-looking statements about the Company's results of operations, financial condition and liquidity, the sufficiency of the Company's reserves, and other topics.

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 our current expectations 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 Securities and Exchange Commission.

We did not undertake any obligation to update forward-looking statements.

Also in our remarks or responses to questions, we may mention St. Paul Travelers' operating income, which we use as a measure a profit, and other measures that may be non-GAAP financial measures.

Reconciliations are included in our most recent earnings press release, financial supplements, and other materials that are available on the investor section of our website, stpaultravelers.com.

With regard to the regulatory matters affecting our industry, the investigations described in previous quarters and our most recent 10-Q and 10-K remain ongoing, as have the Company's efforts to cooperate with the regulators.

Therefore, we will not report further on these matters at this time.

With that, I'm going to turn it over to Jay.

Thank you, Mike.

Good morning, everyone, and thank you for joining us today.

Aided by very benign weather, this quarter marked a continuation of our solid performance, both on the top and bottom lines.

The Company posted $1.037 billion of operating income, or $1.46 of operating income per diluted share, and an operating return on equity of 17.4%.

Our top line continued to grow nicely in each segment.

Net written premiums increased 6% over last year's third quarter, excluding the Company's runoff operations and adjusting for the estimated impact of transitioning to six-month policies for a large portion of our personal automobile new business.

We also recorded another quarter of net favorable prior-year reserve development, even including the reserve strengthening for asbestos and environmental exposures.

Due to a number of factors which Jay Benet will discuss in a moment, we were able to complete our annual asbestos review this quarter.

The $155 million pretax charge for asbestos represents 3.7% of our September 30 carried reserves.

It is worth noting that over the last several years, we have experienced a stabilizing asbestos environment, and in addition, over that time we have successfully negotiated a number of significant settlements.

Consequently, we believe that the potential volatility of our asbestos exposure has been reduced from where it was just a few years ago.

As it relates to current market conditions, consistent with last quarter, in our commercial businesses, retentions were once again at historically high levels and margins remain attractive.

Renewal pricing continues to increase for catastrophe-prone exposures, particularly in the Southeastern U.S.

Elsewhere, renewal pricing is generally flat.

Since retentions are high across the industry, new business flow for our commercial businesses is somewhat limited and pricing on new business remains more competitive than on renewal business.

Nevertheless, we are pleased with the growth that we are experiencing.

In the Personal Insurance segment, we are generating significant profitability and above-average industry growth, driven by the continued success of the Quantum Auto product and our leadership position in homeowners.

We advanced a number of strategic initiatives to position us to continue to drive profitable growth.

Since we spoke with you last, we announced a realignment of our business segments to a structure which we believe will maximize our broad franchise potential, while maintaining our important focus on industry and product specialization.

This is in addition to expanding the roles of our regional executives to provide them with more authority at the point of sale.

We believe that these changes will better position us to provide producers and customers easier access to our substantial array of products and services, which should ultimately produce a meaningful competitive advantage.

Turning to page 3 of the webcast, we are increasing our 2006 full-year guidance to a range of $5.50 to $5.65 per share.

The increase primarily reflects our strong operating performance in the quarter, combined with the very low level of catastrophes.

The assumptions underlying our current guidance include catastrophe losses of $35 million after-tax, or $55 million pretax, for the fourth quarter of 2006, and no additional prior-year reserve development, either favorable or unfavorable.

Average outstanding diluted shares are assumed to be 717 million for the full year before the impact of any share repurchase activity in the remainder of the year.

It is important to note, as Jay will further elaborate, that our year-to-date performance is such that we intend to meaningfully accelerate the pace of our share buybacks from that of previous quarters.

But to reiterate, our guidance excludes the impact of any share repurchase activity in the fourth quarter.

I will now turn it over to Jay Benet and then come back at the end for a few summary comments.

Thanks, Jay.

Page 4 of the webcast contains a summary of our third-quarter 2006 financial performance.

Let me first note that on the top line, our earned premiums increased by 6% over the third quarter of last year, reaching a record level of $5.26 billion.

And for September year-to-date in 2006, operating income per quarter has averaged slightly more than $1 billion, as we have benefited in this nine-month period from cumulative catastrophe losses of only $54 million after-tax, net favorable prior-year development of $155 million after-tax, a combined ratio of 88.6%, and record net investment income.

I would also note that our average diluted share count of 715 million was up 4.5% of the third quarter of last year due to the August 2005 issuance of 15.2 million shares in connection with our equity-linked notes, partially offset by share repurchases made in the most recent two quarters in connection with our $2 billion buyback program.

Sequentially, from the second quarter of 2006, our average diluted share count was down almost 1% due to share repurchase activity.

During the third quarter, we repurchased 2.7 million shares at an average cost of $44.33, bringing our year-to-date totals to 8.4 million shares at an average cost of $44.36 per share.

As the data on page 5 indicates, by far and away, the most significant reason for the difference between consolidated operating income and GAAP combined ratios in the third quarter of 2006 as compared to the third quarter of 2005 was the amount of after-tax catastrophe losses -- $10 million in 2006 due to tropical storm Ernesto versus $1 billion in 2005 due to Hurricanes Katrina and Rita.

In addition to catastrophe losses, the dollar amounts of other items that we point out as being included in operating income -- prior-year net favorable reserve development and the beneficial impact of reestimations of current year loss ratios, all of which is provided to aid in your understanding of current and prior period results -- were at fairly comparable levels for all periods shown.

Page 6 shows a buy segment and consolidated 2006 quarterly and year-to-date prior-year reserve development, inclusive of asbestos and environmental charges recorded in the third quarter.

We are extremely pleased to have experienced net favorable prior-year reserve development on a consolidated basis for each quarter thus far in 2006 and net favorable prior-year reserve development for each of our business segments year-to-date.

In the third quarter, net prior-year reserve development was unfavorable in our Business Insurance segment by $46 million pretax.

This amount included pretax charges of $155 million and $120 million for asbestos and environmental reserve development, respectively, which were partially offset by $229 million pretax for other net favorable prior-year development, primarily due to better-than-expected frequency and severity in our general liability and property businesses.

Also in the third quarter, net prior-year reserve developments was favorable in our personal insurance segment by $132 million pretax, mostly due to a reduction in 2005 catastrophe loss estimates.

It should be noted that this favorable adjustment represents approximately 5% of previously recorded 2005 cat losses, net of reinsurance and pretax, for Katrina, Rita, and Wilma, and came about due to normal levels of uncertainty in the reserve estimation process.

Page 7 provides a roll forward our asbestos and environmental reserves.

Taking into account incurred loss and paid activity, asbestos reserves now stand at $4.2 billion and environmental reserves at $466 million.

Page 8 shows this year's asbestos reserve addition broken down into the same categories reported to you in prior years.

Let me first say that we were able to complete our annual ground-up review of asbestos exposures in the third quarter this year as compared to the fourth quarter in prior years due to several factors, all of which made the work less time-consuming to perform -- the emergence of more stable payment trends for a greater portion of policyholders; a decrease in the number of new claims received; a decrease in the number of large asbestos exposures reflecting ongoing settlement activities; a decrease in the number and volatility of asbestos-related bankruptcies; and the absence of new theories of liability or new classes of defendants.

While the review was completed earlier, the level of analysis performed was entirely consistent with that of prior years.

For example, the review considered active policyholders in litigation cases with potential product and nonproduct liability, including the ongoing litigation related to ACandS.

Developing payment trends for policyholders in the home office and field office, as well as for assumed and international exposures, were also analyzed.

I would like to offer certain observations from the review.

First, as indicated on page 9, the asbestos environment is more stable than it was three to four years ago, mostly due to the many judicial and legislative initiatives that have taken place in that time period and the extent of settlement activities that have also taken place.

As a result, we are experiencing a significant reduction in new claim filings, as well as a reduction in indemnity and defense costs in the non-settlement categories.

While the reduction in indemnity payments is in line with our previous expectations, defense costs are not decreasing at the rate we had expected, leading to almost all of our $155 million, or 4% pretax, asbestos reserve increase this quarter.

Approximately half of the $155 million adjustment was due to an increase in projected defense costs for only ten policyholders.

Additionally, $15 million of the pretax reserve increase was attributable to a delay in the approval and expected payment of the previously announced PPG settlement.

The remainder of the reserve increase was primarily due to continued litigation against smaller, peripheral defendants.

Page 10 summarizes our asbestos reserve position at the end of third quarter '06 in the same format used at the end of fourth quarter '05.

The number of policyholders for which we are settling asbestos-related claims now stands at 1788, up less than 1% since December '05.

And for 65% of these policyholders, we currently estimate our exposure for each to be less than $100,000, consistent with December '05.

IBNR comprises $2.6 billion, or 62%, of the reserve, also consistent with December '05.

And the gross reserve now stands at $4.9 billion and the percentage of reinsurance recoverable is 15% of the gross reserve.

Page 11 describes the current status of the environmental landscape, which is generally consistent with where it had been in the prior year.

As anticipated in the prior year, we continue to see reductions in policyholders submitting environmental claims for the first time.

The severity of new claims has gone down and the exposure base has been greatly reduced from what it once was.

Claim activity principally relates to small to midsize businesses, not chemical manufacturers and petrochemical companies, as once was the case.

However, we did increase our environmental reserves this quarter, based principally upon two factors.

First, current estimates of defense costs have risen above previous expectations, as they have for asbestos-related claims.

Second, based on our recent settlement experiences, where we have seen settlement costs rise above previous expectations, we have reevaluated the settlement value of our existing exposures.

Page 12 updates the two graphs we provided last year, the first showing the number of policyholders by year of first claim notice, and the second, gross paid losses by year, with single large settlements split out, each for the period 1998 through September year-to-date 2006.

Notwithstanding the current quarter's reserve addition, the overall takeaway is of a greatly reduced exposure base for environmental-related claims.

With that, I will let Brian comment on our updated cat loss estimates and provide further insight into our operating results.

Thanks, Jay.

Before I get into the business results, I'll turn to page 13, where we have updated the probable maximum catastrophe loss estimates that we originally presented in June.

As you recall, these are our estimates at different exceedence probabilities of the probable maximum loss for a single hurricane or earthquake, net of reinsurance and tax.

These are obviously estimates which involve a wide range of assumptions, but fundamentally two changes were made to the June disclosures.

First, we updated for the 2006 RMS version 6.0 cat model; and secondly, we updated our exposures as of June 30.

As you can see, our estimates for 1-in-100 and 1-in-250 likelihoods have decreased for a single hurricane.

This is primarily due to better-than-anticipated frequency and severity assumptions in the RMS model, especially in the Northeast.

This improvement more than offset our increase in Personal Lines exposure count, except in the 1-in-50 category, where our estimate is up slightly.

Overall, our commercial business exposures are down.

You can see a more detailed explanation of our assumptions on page 14, but bottom line, the numbers have not changed dramatically from June.

Now let me turn to page 15 for the financial results of the Business Insurance segment.

Business Insurance earned $613 million this quarter with a combined ratio of 92%, a terrific result and, obviously, the weather -- or I should say an absence of bad weather -- had a lot to do with it.

But even without the weather good fortune, the underlying margins in this business continue to be strong, loss trend continues to be moderate, with favorable loss frequency and stable severity.

We're seeing these favorable trends across virtually all of our products and businesses.

Our expense ratio is up slightly due to investments in our commercial business platform and corporate initiatives, most notably our national advertising program.

On page 16, you can see premiums for Business Insurance.

In the new segment format, we have aligned our businesses based on the customer type, how the business is marketed, and the manner in which risks are underwritten.

Let me take a minute to walk through the six primary buckets.

Select Accounts, which consists of our small-business customers.

Commercial Accounts is our middle-market customers, which we previously referred to as Commercial Field.

National Accounts is our large-account, fee-based insurance services and loss-sensitive insurance priced businesses.

Industry-Focused Underwriting includes construction, technology, oil & gas, and other industry specializations.

Target Risk Underwriting is mostly national property and our marine businesses.

Finally, Specialized Distribution includes those businesses primarily distributed through wholesalers.

Excluding Business Insurance Other, our runoff operation, net written premiums in the segment grew by 4%.

The growth was primarily in Commercial Accounts, Industry-Focused, and Targeted Risk Underwriting.

In Commercial Accounts, the growth was driven by strong retentions and low single digit renewal price increases.

Growth in the National Property and Inland Marine businesses within Target Risk Underwriting is also due to strong retentions and significant price increases, driven primarily by coastal exposures.

The Industry-Focused growth was strongest for our construction and oil & gas businesses, where general economic conditions have been favorable.

Turning to small commercial, after adjusting for the transfer of some small-business insurance programs from Select to our Specialized Distribution Group, business results were essentially flat with last year.

Our National Accounts premiums are lower this quarter.

A significant amount of this business is priced on a loss-sensitive basis, as certain states implement workers' compensation reform, loss experience improves and less premium is needed.

In addition, less business is being written in the involuntary residual market, also contributing to the National Accounts decline.

Page 17 provides the production details we normally disclose.

The data is broken out in three groups, Select, Commercial Accounts, and the remainder of the specialized business units.

You can see the same trends that we have been showing for over a year, very strong retentions and renewal pricing slightly positive.

But as we have talked about previously, there is a coastal/noncoastal story underneath this data.

Along the cat-exposed coast, retentions and new business retentions are lower than our national levels and new business is down slightly, while renewal pricing is up significantly.

Away from the coast, we have the opposite situation, with retentions and new business improving and renewal pricing basically flat.

Across all the Business Insurance markets, new business levels are modest, which is appropriate, given that retentions have remained high across the entire industry and, accordingly, competition for new business is very strong.

We do believe, however, that as we improve our coordination across our broad business capabilities, we can do better without just competing on price.

Moving to the Financial, Professional, and International segment on page 18, operating income is up 67% over last year.

Financial, Professional, and International includes Surety, prime and financial liability, which primarily is credit-based underwriting processes, along with the P&C insurance business predominantly marketed on an international basis.

The improvement was driven by the absence of cat losses in 2006, partially offset by a smaller prior-year benefit compared to the prior year.

Even excluding cat losses and prior-year reserve development, however, the overall results still improved due to volume increases in Bond, in Canada, along with increased investment income.

The overall loss ratio was down over 3 points to 54.1.

When we adjust the loss ratio for the impact of cats and prior-year reserve development, it is actually up slightly, but still providing very solid underwriting margins.

The expense ratio improved significantly from prior year.

This is in part due to the reinstatement premiums we paid in 2005.

However, even excluding this impact, the expense ratio improved about 1 point.

Net written premiums were up 8% for the quarter.

Half the increase is due to the cat-related reinstatement premiums last year.

After this adjustment, growth is still very solid, particularly in Bond and International.

The increase in Bond is due to the growth in both Executive Liability and Construction Surety.

International's increase was driven by healthy organic growth in Canada and the UK, as well as the previously mentioned cat reinstatement premium dynamic.

Financial Professional Services, Lloyd's and Ireland were flat year-over-year when we adjust for some timing differences.

Looking at the production stats on page 19, Financial and Professional retentions of 84% were the highest they have been over the last seven quarters.

Pricing at 6% was strong, although down from 10% in 2005.

Both '06 and '05 reflect exposure increases due to some account rounding.

New business is down slightly.

For International businesses excluding Lloyd's, retentions of 86% were again at the highest levels they have been in the last seven quarters.

Pricing at negative 1% was consistent with prior quarters and reflects the competitive international environment.

New business growth came from both the UK and Canada.

So in summary, all of our commercial businesses had very strong results for the quarter.

The business is performing well.

Margins are excellent.

Retention is holding at very high levels and renewal pricing is acceptable.

New business is challenging, but we believe reasonable, given the current market conditions.

So overall a great result.

Now let me turn it over to Joe Lacher for the Personal Lines perspective.

Thanks, Brian.

Our Personal Insurance segment delivered very strong earnings this quarter, with operating income of $341 million and a 78.4% combined ratio.

I will point you to page 20 to discuss some of the drivers.

As you would expect, the quarter benefited from light catastrophe activity, with only a $10 million charge.

We also had favorable prior-year reserve development of $132 million pretax, or $85 million after-tax, driven by a couple of sources.

The largest was a lowering of our loss estimates from the 2005 storms.

We continue to monitor and manage the claims from these events with great care.

After reviewing our experience and incorporating all the currently known legal, regulatory and environmental factors, we concluded that our loss estimates, particularly in terms of average severity, were high, and adjusted those total estimates this quarter.

In addition, we had other favorable prior-year developments related to prior-year reserves and favorable current year reestimation, driven by the ongoing results of our claim initiatives, which continue to bear fruit.

Our results also benefited from increases in business volumes, driven by our continued profitable growth.

Net written premiums for the quarter grew 6% over prior year.

That is down from the second quarter's growth of 10%, and I wanted to take you through why.

Jay mentioned a little earlier and in last quarters call I mentioned that we were shifting a large percentage of our Auto new business from 12-month policy terms to 6-month policy terms.

This has allowed us to better position Quantum Auto for more the frequent rate tuning needed in a multivariate rating program.

As we discussed, this should have no impact on policy counts or earned premium, but will slightly depress written premiums for two quarters.

The impact this quarter is roughly $40 million.

Adjusting for that impact, net written premium growth for all of Personal insurance and for Auto would have been about 9% and in line with prior quarters.

As you look page 21, you can see that policy in force growth has remained generally consistent with prior quarters.

You can also see that new business premiums for Auto were down about $40 million from the second quarter.

Again, this is driven by the impact of the 6-month policy changes.

As we look at our Auto results more fully, we had a very strong quarter.

The Auto combined ratio was 89.7% for the quarter, and policy in force growth was 11%.

Retentions remained strong and stable, and renewal price change was steady at 1%.

We continue to see and experience low single digit loss cost trends that are somewhat outpacing renewal price changes.

The significant increase in policies in force continues to be driven by the impact of our Quantum Auto product.

At the end of the quarter, Quantum Auto was active in 37 states and the District of Columbia.

As I mentioned before, this represents the bulk of the impact of the program's rollout.

We have a number of states where we may introduce the program in the future, but due to operational and regulatory constraints, we have delayed their implementation.

We continue to monitor and tune the program actively.

It has been embraced by our agents, it has allowed us to access a broader cross-section of the marketplace and a broader cross-section of risks, and it has enabled us to increase our growth.

Overall, the program continues to meet or exceed our expectations.

As we look broadly across the auto marketplace, our view has not changed a lot since last quarter.

Profitability is strong for the industry, most carriers appear to have an appetite for growth, and we are seeing signs of increasing competition for business.

However, we are not seeing aggressive pricing being used as the principal tool.

There seems to be a bias towards advertising and marketing activities first.

We're pleased with our Auto results and we anticipate continued profitable growth going forward.

Let me shift for a minute to the Homeowners line, where our GAAP combined ratio was 64.1% for the quarter.

This was heavily driven by a light catastrophe quarter and the favorable development on the 2005 storms.

Excluding the impact of those items, the combined ratio was strong and generally in line with prior periods.

Homeowners' production continued its consistent performance.

Retentions remained stable at 87% and renewal price change was generally consistent with what we have seen the last four quarters at 6%.

Our policies in force growth increased to 9%.

New business premiums were very strong, $150 million, and continued to be generated across diverse geographies.

The marketplace in Homeowners continues to be a complex one.

In Atlantic wind exposed areas, capacity is at a premium.

Most carriers are executing a combination of exposure reduction, changes to terms and conditions, or rate activity.

We continue to execute very locally to ensure that our risk/reward equations are appropriately balanced in these areas.

Outside of these wind-exposed areas, the marketplace is considerably more stable.

We are observing a competitive environment that I would describe as disciplined.

Again, we're pleased with our Homeowners results overall and we anticipate capitalizing on our strength to continue to profitably grow this business.

With that, I will pass it back to Jay.

Thanks, Joe.

As you can see from page 22, average invested assets grew by $1.6 billion to $72.1 billion in the quarter.

Operating cash flows of $1.7 billion were exceptionally high this quarter, up from $1.4 billion in the third quarter of last year, due to growth in both premium volume and net investment income.

Average invested assets were up $4.5 billion from a year ago.

The quality of our investment portfolio remains very high, the low investment-grade securities were down to 2.7% of the fixed maturity portfolio, and duration was reduced to 4.0.

After-tax NII, as shown on page 23, was $668 million in the quarter, up 7% from last year's third quarter due to very strong fixed-income results.

The fixed-income portfolio benefited from increased invested asset base, as well as higher rates.

On average, the rates on long-term asset purchases in the third quarter averaged approximately 70 to 80 basis points higher after-tax than the rates on asset sales and maturities, adding further momentum to NII going forward.

The non-fixed-income portion of the portfolio, while performing within our range of expectations, did not perform as well as in recent quarters, resulting in a 10 basis point reduction in the after-tax yield of the entire investment portfolio, from 3.8% in the second quarter of '06 to 3.7% in the current quarter.

Lastly, as page 24 indicates, we ended the quarter with over $24 billion of common equity, up 11%, and book value per share ex FAS 115 of $35.10, up 12% since the beginning of the year, each after the impact of $371 million of share buybacks and dividends of $520 million.

Our debt-to-total capital ratio of 21.2%, or 19.3% excluding fourth quarter 2006 prefunded debt maturities, was well within our target range.

That surplus increased by 12% since the beginning of the year to just under $20 billion, and holding company liquid assets were almost $2.7 billion, or $1.9 billion ex the prefunded debt.

Given current market conditions and barring unforeseen events, we expect to continue to generate more than enough capital to support our business growth.

And consistent with what we had said in prior quarters, we remain committed to both growing book value and returning excess capital to our shareholders.

Our continued strong operating performance throughout this year and our very strong financial position is allowing us to meaningfully accelerate our share buyback activities in the fourth quarter over second or third quarter levels.

With that, let me turn it back to Jay Fishman, for his concluding remarks.

Thanks, Jay.

Quickly in closing, this was a terrific quarter and a continuation of a very strong year.

This is our third consecutive quarter with earnings at around $1 billion in operating income.

Even after recording the asbestos environmental reserve charges this quarter, it is worth noting that we experienced another quarter of net favorable reserve development.

Our book value ex FAS 115 has increased 12% through the first nine months of this year and our top-line growth in the last two quarters we believe is indicative of the successful positioning of our franchise.

We have a lot going on and momentum continues to be very strong.

With that, we will open it up for questions.

(OPERATOR INSTRUCTIONS) Matthew Heimermann, JPMorgan.

A couple questions.

First, the reserve study.

Defense cost has been -- I think the last three years has really been the issue that has caused you the most pain in terms of increasing LAE reserves.

I was just curious, can you just give us a sense of what your current assumptions are, to better understand what changes in the environment would result in this occurring again next year?

I just want to put that into context.

This is Jay Benet.

I would not say it is easily described in terms of an assumption.

The way we analyze this particular element is to look at, on a policy-by-policyholder basis, what we think, based on current activities and expectations of future trends, what these costs are going to be.

And they are projected out over periods of years in many cases.

So when you start looking at those things and make an estimate on a policyholder basis and then adjust it upward because you thought it was going to come down to a certain level and it has not come quite down to that level, you end up with a multiplier effect.

So there is no one assumption.

All it is really doing is saying each time we're looking at it, it is just at a different level than we had previously estimated.

And that is just the nature of the (indiscernible) and estimation process.

Maybe as a practical example, if you think about on the asbestos side, those 10 policyholders that accounted for 50% of the increase.

Is that related to just an extension in the litigation period or other factors that are driving it?

It is a funny situation, because there's quite a few of those where the way the analysis was done last year, we evaluated the loss cost estimates along with the litigation costs.

And based on where the policy limits were, we expected to reach the policy limits over a certain period of time; and once those policy limits were reached, we would be cutting off the litigation expenses.

What has happened in some of these cases -- this is where good news begets some additional cost -- the asbestos environment having improved has caused the loss costs to actually go down and be delayed in terms of reaching the policy limits.

So that is the good news.

But we had always expected to reach the policy limit; we are still expecting to reach the policy limit.

It is just on a delayed basis.

In the meantime, as a result of that, we think we're going to be incurring additional litigation expense.

Okay, that's fair.

Changing gears, second question is you announced at the analysts' day earlier this year the launch of the advertising program.

And I was just curious if you could talk about what the early results are and how that might be comparing to both your expectations and also in terms of your goals, in terms of the message you were trying to express and how clearly that is being received.

Sure.

First, our effort is really -- I would call it at least intermediate to long-term in nature.

We think Travelers is a terrific brand-name that has just been allowed to kind of wither on the vine.

And this investment is specifically designed to increase recognition at point-of-sale and to increase the level of enthusiasm from our agents, the people who sell our products.

Certainly with respect to the second one, it has been really a very strong response.

We have gotten lots of responses from our agents, appreciative of the advertising and supporting the effort.

So that has been really very good.

The brand tracker data that we are watching is still very early, but the results are encouraging.

It looks as though the recognition of the brand-name is going up.

Now again, it is very early and we are into this all of about five months, so no one here is declaring victory.

But so far, so good.

And as we think about, again, our long-term effort to reinvigorate the brand and brand recognition at point-of-sale, I think we are off to a good start.

All right, thank you.

David Small, Bear, Stearns.

A few quick questions.

The first is was there any favorable development in the Auto line this quarter?

Yes, there was some.

The bulk of the prior-year reserve development came from the third-quarter storms or last year's storms, the 2005 storms.

But there was some in the Auto line.

We typically don't disclose the breakdown by line.

And Joe was obviously answering for Personal insurance.

Yes, I assume that is where the question was.

That was the question.

My other question is, you did indicate that it is a tale of two markets.

And perhaps you could just give us some better sense on the renewal pricing.

So if you took out the cat-exposed property, what would that look like?

Away from the coast, essentially renewal pricing is flat.

And obviously renewal price is made up of both a rate component and exposure change.

The rate component is actually very slightly negative, as in 1%, 2% type of change.

The exposure change is a little bit positive.

And so it is netting out to something very close to zero price change --actually, probably a slight positive when you look across our total commercial book in all the markets.

I think it is useful just to add in that regard that obviously there's lots of anecdotal conversation in the market about changing pricing.

If you go out to our field folks and ask them, you'll hear about the individual account that was taken by someone else off a meaningful percentage.

And just our observations would be is that what is happening is the anecdotal experiences of the kind of one-off transactions seem to be sitting in people's heads more than the actual real data suggests.

What we're looking at here is in fact real data.

It is not anecdotal.

It is not based on stories.

It is the actual renewal price change on business.

So we're looking at real facts, and what we are generally seeing across the market, away from the coast, is renewal price change about zero.

So we do not dispute the anecdotal dynamics that you hear and see in the marketplace.

Our data is just what it is, and tells a somewhat different story.

Okay, thank you.

Then the last question would be, maybe just on a new business, are you seeing much of a difference within new business pricing and renewal pricing?

I think two levels to the question, because there's always a different between new business pricing and renewal pricing.

It is probably a little more dramatic than we normally see in the marketplace.

So we talk about it as being a competitive new business environment.

So yes, we are seeing a widening spread there.

Okay, so in that type of environment, does that impact how you look at growth?

Yes, and I think if you look at our data, it is kind of evidencing itself.

Our growth numbers are really being driven by very high retentions and slightly positive price and new business numbers that are fairly modest.

I think, though, it is important to remember that we are speaking relatively here.

So the fact is while new business might be returning a lower margin than an equivalent new business risk, the fact is it is still very much at attractive levels.

There is a tendency in these discussions to discuss business as though somehow it is good or it is bad, and there is an absolute dynamic.

This is very much a continuum.

And the renewal book for all of the reasons you all well understand -- the fact is it is subject renewal evaluations periodically, accounts are repriced -- over time, a renewal book produces superior returns.

The new business starts off without the benefit of that scrubbing and will produce a lower return.

But frankly, even at these levels, the business that we're taking on -- I said very specifically my comments -- we are very pleased with.

It is not as though we're writing business at return levels that are inappropriate or silly to do.

Quite the contrary.

Okay, thank you very much for the answers.

Bill Wilt, Morgan Stanley.

I will follow up on the last question, acknowledging it is tough to generalize.

Can you quantify even broadly the difference between new and renewal pricing in the commercial classes?

It is difficult because each risk is unique, and so it becomes very difficult to compare what one risk would price that if it were new verses renewal.

Historically, we have done lots of broad-based analysis where we look at blocks of business and we age them over time.

And then we compare the results from a block of new to a block of renewal, almost on a vintage basis.

Historically, at times we basically saw a 5% to 10% -- now let me make sure I get this right -- is it loss ratio points in a differential?

In other words, we were -- to the extent that we would see -- and I'm using examples here -- a 60% loss ratio on a renewal book, it would not have been unusual to see 65 to 70 in the new book.

Sometimes that gap --.

Price change, not loss ratio, but 5 to 10 points difference.

We look and we say there is a renewal to new business pricing gap as best as we can measure it, as Jay said, because all risks are different, is somewhere between 5 and 10 points of price, not of loss ratio.

And the further experience has been that has narrowed sometimes.

There are times where the analysis suggests that -- and we refer to that difference as simply a new business penalty, is the term we used internally.

There are times where that gap gets much more narrow and there are times where it widens out.

And Brian was responding that instinctively -- because again, we don't know until we can look at it in retrospect -- we seem to be at the higher end of that range now.

That is helpful.

I guess following up on that same thread, it would be helpful to get your assistance in interpreting -- I'm looking at across the business segments the year-to-date accruals combined ratio, X'ing out cat and reserve development, basically unchanged '06 versus '05.

Now, clearly the business earned premium is much more heavily weighted to renewal than it is to new, but I guess I would have thought that with manageable loss cost inflation, moderate renewal pricing, and then this new penalty which you described on new business, I would have thought that the year-to-date '06 accrual ratio would be inching up higher than the year-to-date '05.

Our plan works where we do indeed assign a higher loss ratio to that portion of new business, and again I am being -- I am making it sound like it is very specific.

It is less so, but our plan and the way that we record our loss estimates does indeed reflect an estimated new business penalty against that new business.

What we have found -- and I think this is part of what you're seeing -- is that the claim initiatives that have been at work here for several years are really beginning to bear fruit, and Joe has described some of these before.

He has talked about reducing the inspection time on cars and claim dramatically, and the work that Doreen has done in claim to get the organization very focused on high-volume claims.

We're seeing some real benefits from it, so we've got a bunch of factors going on.

Basically, pricing is flat, again up in renewal in coastal zones, southeastern coastal particularly.

Loss costs have generally been fairly benign, although it is somewhat positive.

We've got a benefit in the loss ratio that is coming out of some of these investments.

And you're right, I think on balance, we have basically got margins this year ex all the prior period stuff that are flat to last year, and that is really how it works out.

And there have been some environmental issues.

I mean, comp reform in certain states has been a positive.

Court reform in certain environments has driven things.

So there are environmental issues that we think are healthy.

That have generally been positive.

Right -- severities and frequencies both very much in line.

That is very helpful.

Thanks.

And I will, if I may, a second.

View on acquisitions as a strategic matter, setting aside past press releases or media, just looking forward --perspective broadly 2007 in a (multiple speakers).

We've got a strong focus on organic growth.

We've got all the products that we need.

We have got all the skill that we need.

And we are extremely focused right now on converting an extraordinary set of conditions, which is the product breadth that we have -- it is incomparable, truly incomparable -- converting that condition into a competitive advantage.

And that is absorbing all of our time and all of our attention, and it really does relate to the way in which we conduct business at the point of sale.

So we are, I would say, virtually exclusively focused on organic growth and that's got our attention.

Thanks very much.

Jay Gelb, Lehman Brothers.

I had a question on the pricing to follow up.

I believe it was just mentioned on the call that the overall commercial book is down around 1% to 2% on pricing.

And the latest CIAB data that came out showed the overall market in commercial lines was down around 5%.

So I am just trying to understand how St. Paul Travelers is able to get that better pricing power versus the market.

This gets back to the comment I made earlier.

The CIAB data -- and we read it, obviously -- is largely anecdotal, meaning they go out to agents and producers and they ask them questions.

What is happening to rate?

So the people who respond are responding without the benefit, I think, of real data.

And as a consequence, what we think tends to happen is that what you're getting is responses that are largely to new business flow, exclusively to new business flow, that they are really not thinking about their renewal book.

And they are telling you what their impressionism is of their new business marketplace.

We all believe that it is good faith, it is done with good intentions, but it is fundamentally anecdotal, not fact-based.

And I do not think that we are necessarily getting particularly better -- I am not making the statement that we're getting better rate performance, because we do not have the data to compare it to in the market.

We know what our data is and it is what it is.

I think that if there was a venue, an opportunity to compile industrywide data the way that we do, my guess is that the data would look more as the data we're showing.

We are big enough that we reflect the marketplace.

We're doing $20 billion plus in premiums now.

It is not as though we're getting isolated instances.

We really are getting the marketplace.

And I suspect what you're seeing is just the difference between fact-based information and anecdotal responses.

Okay, that's an interesting point.

Then two other quick ones.

The recent asbestos and environmental review, did that include all reserves for other issues like D&O, backdating of stock options, and lead paint, and other legacy liabilities?

We look at all of our reserves on a quarterly basis.

So there is nothing that about the subjects that you discussed that is waiting analysis for a future quarter.

These are all part of the constants update that we do.

But I think to answer the questioning more specifically, we certainly did, in the course of looking at our third-quarter reserves, evaluate the impact and updated our analysis of the backdating issue, and at the moment have nothing to report in that regard.

So no change from the investor day in terms of the outlook there?

The facts have changed a little bit.

The number of claims are up somewhat.

And we have all that data.

The facts are different, but the conclusion is the same.

Obviously, when we went through last quarter, we looked at the number of claims we had in the door etc., but we also made an estimate of what we thought we were going to see, both in number of claims, the type of defendants, the estimated values of the claims.

And the changes we have seen to date have been well within those estimates.

Okay, then a quick one for Jay.

The other investment income was $68 million in the third quarter, which was down from $105 million in the second quarter.

I know that is lumpy, but what do you think is a reasonable expectation of investment income and that line going forward?

And think I will repeat what I've said in past quarters.

You're absolutely right.

It is lumpy.

There is a good component of that that has been very favorable, private equity returns that we've had in prior quarters.

And of course there's hedge funds and real estate in there.

We, for our internal purposes, just kind of average things out and roll it forward.

But it will be what it will be, unfortunately.

Were there any onetime issues in the most recent quarter or it just was what it was?

There is nothing that stands out.

If you go back a couple of quarters, we have just had very, very strong private equity returns.

And this quarter, I would say we had good private equity returns, but they were not at the exceptional level that we had before.

And the hedge funds will vary from quarter-to-quarter, as well, depending upon market conditions.

Thanks very much.

Alain Karaoglan, Deutsche Bank.

I have a few questions.

First, thank you for getting the asbestos study done early and getting it done in the third quarter.

In terms of reserve releases, Jay, I think you mentioned that there was some releases in the commercial lines and the liability lines.

Could you tell us how much and what year did they pertain to, what accident years?

They tend to be from some of the more recent years.

And what is taking place is more of the rolling forward of some of the very positive results that we have been seeing in prior years.

So you feel comfortable releasing reserve from the recent years on the liability side, even though there isn't that seasoning?

Well, yes.

All of the reserves represent our best estimates, and we do not feel that there is any stretching on our part in looking at this.

This is -- the pieces that we look at are seasoned, Alain.

Some of this is actually arithmetic, meaning that we establish a loss pick for a given year, in part based on our experience for the preceding year.

So to the extent the book seasons and that older year generates some positive reserve development, it tends to roll through arithmetically to the current years simply by the way that we establish the current year pick.

So while it does impact the more current years, it is based fundamentally on changes that relate to more seasoned books of business.

And within the liability book, not everything is very, very long-tail either.

So these things tend to be, even if they are in the more current years, the pieces of the reserves that you can look at and say with some higher degree of confidence that this in fact is a change that needs to be reflected.

And pieces of -- and we're not going to do a whole reconciliation -- but there were some significant pieces of the favorable reserve news in the Business Insurance segment that were not liability (indiscernible).

Okay, the second question relates to some of the segments in the analysts' day that you have shown.

You had shown some segments on a graph, and some lines of business or segments -- and I wouldn't call them segments -- some sublines of business were not producing returns on capital that were adequate.

Most of the business was by far producing great returns on capital.

Could you give us an update on what happened to these sub-performing business?

Are they now performing?

What are your expectations going forward?

My recollection from that investor day is that we actually had one business unit that was below the threshold.

And without the benefit of having the chart in front of me, my recollection is that there still remains one, but it is a different one.

The fact is that the business that was there, as a result of pricing and underwriting actions taken, is now pricing at levels that are within it.

And a business that was within it, for competitive dynamics reasons, has now fallen below it.

So I think the thing to take away from that is the way in which we use that data as a tool.

We use it to maneuver and manage our pricing and underwriting philosophy virtually continuously.

So that is really I think the state-of-the-art right now.

The last question is on capital management.

Clearly, you're mentioning that you are going to accelerate a little bit the share repurchase program.

But any thoughts on doing something a little bit more dramatic?

Your premium to surplus ratio is below 1.1-to-1.

If you take it to 1.3-to-1, which some of your overcapitalized competitors are at, that is a few billion dollars of excess capital.

You have $2 billion at the holding company.

You're generating terrific earnings.

And based on the scenarios of catastrophes, very few events are likely to wipe out more than a year of earnings, if any.

Any thoughts on doing something more dramatic than $2 billion over 2006 and 2007?

In terms of the capital management structure, I think what we have said before is the least important statistic is premiums to surplus ratio.

That is a result.

It is not an indication of how it is managed.

What you do not see are all the results associated with the rating agency capital models that then gear you towards particular rating characterization.

So I would not characterize us as being in a position where our premiums to surplus ratio is where it is, therefore we must have a lot of excess capital.

I would say that is an indication of our strength.

But really, it comes to the capital models.

Having said that, what we are committed to do is manage the place to have the amount of capital we need to run the business, and then beyond that, to return the excess to shareholders.

And that is what we're saying we're going to be doing.

With regard to looking at the second quarter, we came up with an estimate of what we thought it would be a reasonable amount to do that.

We were very cautious in the third quarter, given that it was cat season, so we reduced our share repurchase activity accordingly.

Having gotten through not just the cat season, but having the very robust earnings we have and the very strong financial condition we have, we are, as Jay said and I said, we're going to be accelerating the buybacks in the fourth quarter.

So we're not committed necessarily to doing the program over the initial date that we said.

We said we would return capital as it is generated, and we are accelerating accordingly.

If and when we exhaust that program and we have excess capital that we're generating beyond that, we're going to be going and asking the Board for additional capacity to do more.

I would just add, Alain, I think it is likely that we will go through the program sooner than we had generally spoken about a couple of quarters ago.

And it is really not based upon the catastrophe season particularly, because that is just not how we manage this long-term risk.

Cat risk is fundamentally a long-term issue.

The fact that we had a benign wind season is interesting, but, gosh, I do not hear any weather experts changing their views about the long-term implications of weather changes, and we're going to continue to manage our business with that in mind.

It is really based upon the fundamental operating earnings performance of our business over time.

We are now, again, three quarters -- round numbers -- of $1 billion of operating income.

And at that level, we're simply generating considerably more capital than we need to fuel our growth.

And as that occurs, we will accelerate the program.

Thank you, and congratulations on a great quarter.

Brian Meredith, UBS.

Two quick questions here for you.

First, Joe, can you talk about on the Personal Auto side what do loss costs and pricing look like?

Are they in line, are loss costs above pricing?

Where are we in that kind of continuum?

I'll sort of answer the question in two ways.

One, from a pricing perspective, you can see our renewal price changes have run about 1% and have been consistent there for awhile.

That is generally in line with what we'd expect over the near-term.

Loss cost trends continue to run generally favorable to where they have historically.

We're seeing a favorable frequency environment; severity ticking up.

I think we would probably describe the macroenvironment as one that would have low single digit loss trend.

The twist on that is that we have been very successfully executing a variety of claim initiatives, which I think are giving us a little bit better performance than we see generally in that macroeconomic view.

As those initiatives are coming to fruition, we're seeing that favorably work its way through our view of loss cost trend and help us deliver a little bit better view than we might otherwise.

But I think what you probably should see when you aggregate it all together is loss trends maybe a little bit outpacing pricing.

Great.

And were there any reserve releases this quarter in the Auto insurance area?

There were.

We spoke about those generally a couple minutes ago, but we do not typically describe the split of prior-year reserve between auto and property.

The bulk of prior year did come from the 2005 storms.

Got you.

Lastly, on the reinsurance, just curious, given that there may be some more capacity in the property reinsurance market going into 1/1 renewals, would you think about trying to maybe fill in some holes or top up on your catastrophe program, even though it was a 7/1 renewal?

First, we certainly have not heard or seen any indication, particularly, the capacity for 1/1 is going up.

In fact, just in conversations that we have had with some of the reinsurers at industry meetings, our view is that they have as aggressive a view towards 1/1 is they had towards July 1 last year.

So I'm not sure that opportunity will exist.

Having said that, even if it did for us, I think the answer is no.

I do not think we have any holes.

We feel very good about generally our gross exposure.

I had described in June that there was really one territory, and that was the Northeast, where we would over time attempt to manage our gross exposure down somewhat.

It returns appropriate levels, so it is one of those interesting conditions where you actually like to ride it at those returns, you just want to make sure that you're not overexposing your business.

And we're not a business that is reliance on the property catastrophe reinsurance business for its working model.

So I just kind of like our position here.

We're not reliant on really anybody.

Last question.

Given that your expectations of your PMLs here going forward are a little bit lower than originally anticipated, do you anticipate maybe getting a little more aggressive or at least looking for some more opportunities in some of the catastrophe-exposed areas, maybe on the commercial line side, try to be more domestic?

Honestly, looking at that PML data, that is so broad -- it is much more directional than it is directly influencing any underwriting strategy.

So no, that has not changed our view of how we want to underwrite.

I don't particularly care for the term even opportunistic.

Again, what we're dealing with here, at least -- and again, we are being held to these standards; we are being held to these standard by rating agencies; we're being held to these standards, in effect, by the modeling firms that produce the data -- is a perception of a substantially increased average annual loss.

And from our perspective, ultimately, we're going to try and look at the potential returns that are available.

And I don't see us trying to take advantage of anything or being opportunistic in a sense in any way.

We love to write insurance.

We're happy to do it anywhere when the returns and the risk have appropriate balance to it, and that is kind of where we are.

Great, thank you.

Jay Cohen, Merrill Lynch.

First, just to go back to an earlier question, I thought I had another explanation.

Bill Wilt asked about the underlying combined ratios, first three quarters this year versus last year.

But I recall in the fourth quarter of last year, you did re-evaluate the first three quarters of last year.

Would that have meant those numbers would've been a little bit lower?

In other words, if you're asking, Jay, if we had rolled back the adjustment to the loss ratios that were at that point in time, the first two quarters would've been lower.

But the same can be said this year -- we made a cumulative adjustment in the third quarter as well.

That's true.

Fair enough.

I would have to go back and actually restate the loss ratio in last year's first and second quarters and this year's first and second quarters and actually compare them, Which we can do; we just haven't done that.

I don't think it is a huge number anyway, so --.

The other question I had is on Quantum, when you first introduced that product, you were, as you recall, accused by others of being a bit aggressive in certain states.

Have you gone back and revisited the initial pricing assumptions you made to see in fact how that has played out?

We are constantly looking at all of our pricing assumptions and very actively monitoring all of those components, looking at what our actual loss experience is, what our frequencies are, where things are deviating from our expectations, where they are deviating from historical data, and where they fit competitively.

So it has really got nothing to do with anybody accusing us of anything or pointing things out.

It is us very actively and appropriately managing the product and the program.

And we would do so regardless of anybody trying to help us out in the marketplace.

And we feel terrific about where it is positioned.

We have made a wide variety of rate filings and rate changes, and we've talked about those before.

But we would have expected to make a wide variety of rate filings and rate changes, and we feel very good about where the product is positioned.

One of the, we believe, misperceptions that exists in the marketplace generally is that somehow Quantum was a product that was designed uniquely to lower prices and therefore become competitive.

That is not what Quantum did for us.

It actually enabled us to write a product competitively in segments of the Automobile business that we were not writing in before Quantum.

We were simply not a factor or a force -- for expertise reasons, for lots of reasons, but we simply were not there.

And what Quantum did was enable us to begin to market to a pretty decent size of the automobile insurance business that we were not quoting before.

So much of the growth -- a substantial amount of the growth that you're seeing is not because we have lowered prices in areas where we were previously doing business.

It is that we have established competitive prices in areas where before we simply were not a market factor at all.

So we've become -- and I said this before -- a market alternative that did not exist before.

And I think that is a very important dynamic underlying Quantum's success that really is not understood by much of the kind of commentary community.

And that really is important.

There certainly were states where our pricing was initially our best estimate.

It turned out to be in effect hotter than it probably should have been.

We have made changes.

That is the whole notion in these multivariate models, which is you set up pricing and you watch what comes at you.

And you adjust the pricing on almost a continuous basis based upon the flow.

There have been areas where we were too aggressively priced; they have changed.

There were areas where we were not aggressively priced enough, and we have changed it.

So I think if people were commenting on specific circumstances where an individual area or state might have been aggressively priced, yes, that could very well have been the case.

And not necessarily intentionally, although sometimes intentionally so, particularly in the early days.

And now we make adjustments and we continue to be a legitimate market alternative.

And I think that is what you're seeing.

A complete answer.

Thanks.

Next topic, Equitas.

What kind of exposure from a reinsurance receivables standpoint do you have to Equitas?

And does the potential Berkshire deal vary your view of the creditworthiness of that receivable?

I'm going to turn it over to Jay on the receivable.

I Actually would make just another observation about the Equitas transaction.

I have not spoken to anybody at Berkshire Hathaway or at Equitas or at Indemnity.

But it would seem to me that a transaction like that would in fact speak to the same dynamic that we are seeing in the asbestos arena, which is a stabilizing environment.

That is a very smart, very intuitive person, really thoughtful guy, who is looking at a market opportunity based upon a changing environment.

And I just found it interesting that that would come on the heels of essentially what we have been experiencing.

So we tend to see the transaction generally almost as a validation of the trends we have been seeing.

Jay is going to get to the specific numbers.

Basically, Travelers had long ago settled out with Equitas, and there was no receivable.

St. Paul had a relatively small receivable from Equitas.

Yes, the remaining amount is roughly $25 million, so it's not a big number.

So it is nice that there is a stronger credit behind it, but we never worried about it when it was $25 million from Equitas.

So I would say not much of a deal for us.

Yes, the bigger numbers are on a Travelers' side.

And as Jay said, we did settle that a few quarters ago.

Years ago.

Great.

I just want to sneak a quick one in.

When does the Board reconsider the dividend again?

I don't think anything is ever cast in stone.

We will look at it periodically.

Again, I view the dividend, I think we all do, as another way of capital management.

And we have spoken before about excess capital and the commitment to return it, so we will look at it periodically.

Great, thank you.

Joshua Shanker, Citigroup.

My first question -- or really all my questions regard Quantum, just following on Jay's questions.

The first part of it, what percentage of the business that you are writing in the Auto lines on the Personal side is currently under the Quantum pricing model?

The program is only operational for new business right now in the majority of states.

There's a couple of smaller states where we've moved some of the renewal books.

And I am looking for the actual statistic as we're talking.

Why don't you go ahead and ask the second question and I'll make sure I get you the right number while we are working.

Great.

The second question involves trying to figure out, in terms of nonrenewals, churn, pricing changes, PIF is up more dramatically than premium is up on Auto lines.

I'm trying to figure how to reconcile that.

I realize you cannot speak in minute detail to every number, but if you can talk about what is happening in terms of people shopping, what is happening in terms of renewal business that you have changed the pricing on.

How do we reconcile the change in PIF and the change in premiums?

You get a couple things going on there.

I think if you adjust, as we talked about earlier, for the impact of 6-month policies and our switch towards a higher percentage of our policies being 6-month policies, that impact is about $40 million from a new business perspective.

And if you adjust for that, our premium growth in Auto would be about 9%, which is much closer to being in line with the PIF growth.

The balance around that difference comes from geography, typically, and mix of business.

If you write a policy in, say, downstate New York or you write one in Kansas, they have very different average premiums.

And we have spoken before about a broader geographic diversification is running through our book and an increase in new business coming from states off of the eastern coast.

So that tends to drive a slightly lower average premium per policy, but consistent margins.

So that will cause us to be a little different in those stats.

I would also add that just in terms of your comment about churn, we just not only are not seeing but have not seen it.

If you go back to page 21 of the webcast, we've got retentions at 84% flat in the Auto business going back to the second quarter of '85, and the first quarter of '05 was 83.

So there has been a remarkably consistent book of business here that just is not going through the type of churning that you're describing.

Very good.

The final question involves adoption of alternative claims solutions for Auto.

What are you seeing in terms of your centers for getting claims?

Are you seeing people adopting it?

Where is that coming along?

Help me -- I'm trying to understand the question.

Alternative claims?

You're trying to consolidate where people go to get their cars fixed and what not.

Our concierge claim centers?

That's exactly it.

They are working terrifically and we are seeing customers -- anybody who goes there -- very high customer satisfaction rates.

They are very much a function of geography.

You're not going to drive 40 miles to deal with one.

It is a function of whether or not when you have your accident and where you live, are you geographically close to them?

So to the extent that folks are, we are seeing reasonable adoption rates.

It varies geographically, but it is in line with our expectations.

It is not an enormous percentage of our book at this point.

I do not have the exact number, but it is less than 15% of our claims that are running through there at this point.

Any thoughts if I live within ten miles of a concierge center what my likelihood that I'll go there with my cars?

I just don't have that sat.

I am not sure we have calculated it that way.

It varies state to state, in terms of how hard we can pitch it.

Some states struggle with you pitching anything in that capacity.

So we will describe the program and offer it and point out its benefits, and we are seeing reasonably high adoption rates.

I just do not have the number for you.

As Joe reminds me all the time, in the longer run, we've got to do an analysis to determine whether it is worth the investment.

Does it matter?

Does the higher customer satisfaction, does it produce either higher retentions, greater consistency, lower cost of claim?

Ultimately this is an investment, and it will be evaluated like any other investment.

Thank you.

Just also say congratulations on that snowball commercial.

I think it is terrific.

Thank you.

We know we are running over and we will take one more question.

I know other companies have their calls.

We just want to be respectful.

Larry Greenberg, Langen McAlenney.

You might have already touched on this, but was there any additional rigor this quarter relative to normal on looking at non-A&E reserves?

Or was it just a typical quarterly review?

Well, I would describe it as a typical quarterly review, which is a very rigorous process that we go through.

So information flows constantly to our claims people and to our actuaries and our finance people, and we are constantly updating and evaluating our reserves accordingly, Larry.

Great.

So there was not anything unusual in this quarter.

No, not in terms of the rigor in which we always go after looking at the reserves.

Great.

Thank you.

Please.

Thank you all for dialing in and your interest.

We appreciate your interest in our Company and we look forward to speaking with you next quarter.

And obviously Mike and the rest of the Investor Relations team are available to take any questions you might have.

With that, thank you.

Thank you for your participation in today's conference.

This does conclude the presentation and you may now disconnect your lines.

Everybody, have a wonderful day.