Allstate Corp (ALL) 2005 Q3 法說會逐字稿

Good afternoon ladies and gentlemen and welcome to the third-quarter 2005 earnings conference call.

At this time all participants are in a listen-only mode.

Later we will conduct a question-and-answer session and instructions will follow at that time. (OPERATOR INSTRUCTIONS) As a reminder, this conference call is being recorded.

I would now like to turn the conference over to your host, Mr. Robert Block, Vice President of Investor Relations.

Mr. Block, you may begin.

Thank you and good morning everyone.

I am sorry for the delay.

We were trying to make sure that we got as many connected on this call as possible.

Joining me today are Ed Liddy, Tom Wilson and Dan Hale who will comment on our third quarter results that we released last night.

During the Q&A session that will follow the comments we would appreciate it if you could keep to one question and a follow-up so that we can hear from as many people as possible in our time together.

After the completion of the call, Phil Dorn, Larry Moews and I will be available to handle any further inquiries you may have.

And now for the legal disclaimer.

The following discussions may contain forward-looking statements regarding Allstate and its operations.

Actual results may differ materially from those projected in the forward-looking statements.

For information on important factors that could cause such differences, see the forward-looking statements and risk factors affecting Allstate's section in our most recent 10-K as well as yesterday's press release.

In this call we may discusses some non-GAAP measures for which you will find reconciliations in our press releases and in our quarterly investor supplements that are available on our website under Investor Relations.

This call is being recorded and your participation in the call will constitute consent of the recording, publication, webcast, broadcast and use of your names, voice and comments by Allstate.

If you do not agree with these terms, please disconnect now.

A replay will be available following the conclusion of this call.

All of our remarks are current only as of the date and time of this call.

Now I will turn the call over to Ed Liddy.

Good morning all and thanks for joining us again; we apologize for the slightly delayed start.

As you know, and you can tell by our earnings release, it has really been an extraordinarily eventful quarter with the unprecedented impact of the Gulf Coast hurricanes on our customers, our employees, our agencies and of course our shareholders.

First let me cover some of the numbers to make sure that we are all in the same place.

Losses from Katrina are estimated at about $2.4 billion after-tax, and losses from all catastrophes this quarter are estimated at just over $3 billion after-tax.

The operating loss per diluted share of $2.52 which we posted this quarter, eliminated about 1.5 quarters of earnings for the year.

This size loss and the level of exposure is simply unacceptable to us, and we will talk in just a minute about our mitigation plans for the future.

But beyond the financial impact to Allstate, the human impact of the storms really is overwhelming, and I just want to spend a quick moment on that.

I think you all realize that helping restore peoples' lives after something bad happens is what we do, and we do it very well.

Earlier this month Tom Wilson and I visited the Gulf Coast, and we saw some of the damage from the hurricanes.

We spoke to our customers and agencies' employees as well as the Louisiana Insurance Commissioner.

I will tell you that you simply cannot imagine the devastation unless you witness it firsthand.

Seeing it on television or in a newspaper simply doesn't do it justice.

In the face of that, Allstate people are meeting the challenge of these events.

Our agents in those areas are more concerned about the policyholders than they are about themselves, and many have lost much or everything.

In Baton Rouge, Tom and I met with about 190 people who now work in our temporary claims office there.

Many have lost almost everything they own, yet there they were, working hour after hour and shoulder to shoulder with their colleagues to help Allstate policyholders.

It was a very encouraging meeting.

Our overall response, both the degree of effort and the number of customers we are helping is unprecedented.

No other storm in U.S. history comes close in comparison, and as you all know and almost inconceivably hurricane Rita struck Texas while the industry was assisting hundreds of thousands of Katrina victims.

We have dedicated more resources to these events than any other in our 75-year history.

And to put that in perspective we have deployed about 4100 catastrophe specialists and approximately 30 mobile response units.

We are routinely the very first company to arrive to begin the claim process.

For hurricanes Katrina and Rita, we expect to handle more than 300,000 claims.

We've already visited tens of thousands of homes to move forward with the claim process, and we are closing literally thousands of claims every day.

We remain absolutely committed to helping our customers return to some semblance of a normal way of life as soon as possible.

Now the hurricanes have taken a lot of our attention over the last several weeks, and they will continue to require our attention for many months to come.

But I don't want that to overshadow all that has been accomplished during the course of this year and continuing in the third quarter.

Our core businesses are performing well and that is what I want to take just a minute to cover.

So let me highlight something about that performance to the numbers.

Our core property casualty lines of standard auto and homeowners, excluding catastrophes, performed spectacularly, continuing a multiyear trend.

For the quarter our Allstate protection combined ratio again excluding Cat was 78.2, an improvement of 1.4 points versus last year.

Our Allstate brand standard auto combined ratio excluding Cat was 81.5 or 3.8 points better than last year's third quarter.

Frequencies continued to improve throughout the quarter, and we saw no indications of a reverse into the mean that some analysts have forecast and other companies have hinted at in their quarterly calls.

Severity increases continued at moderate levels and well within our pricing assumptions and our retention ratio remains strong in the quarter.

Allstate brand homeowners had a combined ratio excluding Cats of 63.7, a 1.9 point improvement over last year's third quarter and the retention ratio remained at a consistent level of 88.5.

PIF growth in our core lines was moderate to good, 3.6% in standard auto and 4.4% in homeowners.

While moderate, PIF growth continued to exceed industry norms and remains concentrated in the higher value IFS tiers, which has a very positive implication for future profitability.

As we see it the competitive marketplace remains disciplined and rational.

Insurance customers clearly are shopping less given only modest rate activity.

Allstate Financial continued to make progress with operating earnings of $156 million, up 3.3% versus last year.

I think you all know that we have intentionally slowed top-line growth particularly in the fixed annuity and the institutional productlines to focus on improving our returns.

Sales of financial products to the Allstate agency channel continued to grow, increasing 5% in the quarter and 7.8% on a year-to-date basis.

Our capital position and balance sheet remains very strong.

We repurchased $717 million of our stock in the third quarter, bringing the total of 2.2 billion in repurchases out of our current authorization of $4 billion.

We anticipate completing this program in 2006 as originally authorized.

Despite this quarter's loss, our book value per share excluding the impact of unrealized capital gains on our fixed income securities, remains slightly ahead of prior year.

The success and rollout of Your Choice Auto or YCA continues to equal or exceed our expectation.

This product has now been introduced in 15 states, and we expect it to be in a total of 27 states by year end 2005.

We continue to add to our agency force, having increased our exclusive agency count by 250 so far this year.

That also continues a trend that we saw over the last couple of years.

The success of Your Choice Auto and our agency expansion initiative and continued enhancements to our tiered pricing structure should help drive profitable premium growth in the future years.

Now with respect to hurricane losses and future actions, it certainly does not show in this quarter's results but we have made much progress in reducing our hurricane and earthquake PML exposure around the country over the last couple of years.

As those of you that have followed us know, we have worked hard on this initiative for about the ten years that we've been a public company.

The California Earthquake Authority, wind storm pools in Florida and other states and additional reassurance in many areas are all evidence of that.

In fact, our property policies in force in Louisiana's hurricane risk management zone was reduced over 15% over the last several years.

However, the simple fact is our progress wasn't fast enough or deep enough to protect us from an extraordinary event such as Katrina.

We are going to be doing much more operationally and legislatively in those areas where we are exposed to these historically unprecedented events.

And Tom Wilson will take a couple minutes and talk about that.

Tom.

Thanks, Ed and good morning.

Let me start by reiterating what Ed said, which is that reducing our property exposure to major catastrophes continues to be a top priority for our Allstate protection business.

And we've been at it very aggressively for the last three years and really the last ten years.

We are clearly disappointed by the size of our losses this quarter particularly in Louisiana, although that doesn't really negate all the progress we have made in many other areas over the last decade.

And our game plan is really to continue to reduce it but at an accelerated rate using as many of the same tools that we've used before, which include first just reducing the size of our business in coastal counties.

In many of the coastal counties around the country we are reducing the number of policies in force we have.

As our existing business rolls off with people moving or switching to competitors, we are not writing as much new business, and we have restrictions on what type of new business we will write.

And most notably in some of the big states like Florida, like California and New York.

So for example in Florida right now we're writing virtually no new business.

It is really just customers moving from one place to another.

In California and New York our new business in the third quarter this year was down in the 25 to 30% range versus prior versus last year, and we are doing it in many other areas around the country, as well.

Secondly, we are purchasing reinsurance.

We've increased the amount of reinsurance that we've purchased dramatically since 2002, and expenses are up about sevenfold.

We put out an 8-K last month which does describe the reinsurance coverage in detail by some states and the same information was in the press release that went out yesterday.

And we're going to keep looking at reinsurance.

It is a good vehicle for us to shed our risk immediately before we can take operational action.

And of course we did not have reinsurance in Louisiana for Katrina.

We had allocated our reinsurance money to areas with higher probabilities of losses and in states with larger capital requirement.

And Dan will take you through some of that logic in a couple of minutes.

We were projected, thankfully, by reinsurance in Texas and you will see that when you look at the difference between our gross and net losses on hurricane Rita, which were a couple hundred million dollars.

Florida, of course, continues to be a challenge in and of itself, and we continue to get smaller there.

We've been in the process of transferring 95,000 policies to Universal.

We have a reinsurance agreement that where that risk is already transferred but we are physically moving the policies to Universal over the next year, and as I said, we are really writing virtually no new homeowners there.

We are heavily reinsured in Florida.

We've purchased external reinsurance which goes above the reinsurance we get from the Florida hurricane Cat fund.

We have disclosed that before; people have questions about it I would be happy to talk about it.

But there is more to do in Florida.

It is not a place where we want to continue with our current exposure levels and at our current rates.

And so we're going to continue to get smaller in Florida unless there are some changes in the economic or regulatory structure, which enable us to get a proper return down there.

Fourthly, we have a large number of risk management tools that we've developed over the last couple of years, which give us the ability to look with great precision at which individual coastal risks we want to take.

And so we can reduce our PML very surgically.

And our fifth strategy of course to continue to push for an appropriate price for the risks we take.

We have received a price increase in Florida of about 9%.

We had split that into two pieces.

There is another 18% which was recently denied.

We believe we deserve that price increase as evidenced by the losses we've had this year and last year there, in that state.

There have been some questions about whether this is a cost we will spread amongst other states; that is not the way we do pricing.

We look at the risk in each individual state and determine what we should charge in that state.

That said, as we learn more and as we develop an increased understanding of the amounts of risks we have, we will be factoring that into our pricing.

Outside of all those actions, which are totally within our control, we believe there is an opportunity for some positive legislative action to address this issue of infrequent high severity catastrophes that you just basically can't predict by any actuarial method.

So we would like to see a series of state funds to cover major catastrophes, which are then backed up by a federal government program that would enable consumers to have adequate protection at the right price.

We are starting to get the attention of some legislators, particularly after the events over the last couple of years.

I would say we are into the legislative discussion phase as opposed to legislative action.

So we are continuing to pursue our operational initiatives and the legislative solutions at the same time.

Until then we will just keep pushing forward to keep reducing our property exposure in the high Cat prone areas.

Now I would like to turn it over to Dan who is going to talk a little more about our third quarter catastrophe losses.

Before providing more detail on the catastrophe losses we want you to know that we purposely delayed making any announcements about the losses until we were convinced that we developed our best estimate.

As you know, hurricane Katrina has caused the largest losses ever by a single hurricane with industry estimates indicating that Katrina completely destroyed at least ten times as many homes as Andrew in 1992, or the four hurricanes last year.

And of course Katrina was quickly followed by hurricane Rita on September 24th.

So our estimating process required substantial resources and a more complex set of computations than has been needed traditionally.

Our estimates are based on claim adjustor reviews, investigations and estimates, reported claims, interpretations of external facts and circumstances, as well as internal factors which include our experience with similar cases and historical trends of claim payment patterns.

However, in areas where we have not yet been able to complete inspections, or we have reason to believe that our historical loss development factors may not be predicted, we relied on analysis of actual claim notices received compared to the total policies in force, as well as on visual, governmental and third-party party information which includes aerial photos, area observations and data on wind speeds and flood depth to the extent available.

Our estimates were based on the coverage provided by our policies and our homeowners policies specifically exclude coverage for losses caused by flood.

As I said, we believe the estimated impacts of all four hurricanes this quarter had been appropriately developed and recorded.

However, with events of this magnitude you obviously know that actual results may differ materially from the amounts recorded for a variety of reasons.

As Ed mentioned after-tax catastrophe losses for the quarter amounted to $3.06 billion, $4.67 per share or 1.95 billion more than in last year's third quarter.

This year's losses include 37 million pretax of expected assessments from our residual market pools, principally in Louisiana, some of which will be recouped from policyholders.

The breakdown per storm is as follows.

For Katrina total gross losses were 3,684,000,000.

We had a very small $1 million reinsurance recoverable related to the Florida portion of Katrina where we brokered 95,000 policies and reinsured them, so the net for Katrina, 3,683,000,000; on an after-tax basis that would be 2,394,000,000.

The Katrina losses are based on an expected 216,000 total claims, and todate we've received 175,000.

For Rita the gross losses 1,055,000,000, reinsurance recoverable 205 million for a net of 850 million or an after-tax total of 553 million.

Rita losses are based on expected claims totaling 90,000, 62,000 have been received todate.

As Tom mentioned one question we would like to address upfront on the call is why we did not purchase reinsurance for our exposure in Louisiana.

As you may recall, we purchased reinsurance for our exposure in Florida, Texas, New York, Connecticut, New Jersey and to a lesser extent in North and South Carolina.

As we mentioned on previous calls, we now evaluate our catastrophe risk as part of our overall enterprise risk management process.

Catastrophic risk is a key risk but only one of the risks we evaluate as part of the required economic capital needed to support the business.

The amount of capital required based on our statastic (ph) economic capital model is driven to a significant extent by the size of potential catastrophe losses in key catastrophe exposure areas.

By the so-called extreme tale events arising when simulating 125,000 scenarios in the future.

We purchased reinsurance where we attained the most benefit in terms of economic capital requirements, as well as where we could recoup part or all of the cost through rates.

And we look at our exposure on an aggregate annual basis countrywide.

That approach is obviously appropriate when you consider the fact that four hurricanes made landfall in the U.S. in each of the last two years.

We did not buy reinsurance in Louisiana because our exposure there was not as great as in the coastal areas of Florida, Texas and New York.

We also did not buy reinsurance in Mississippi, Alabama, Georgia, Virginia, Delaware or Maryland.

We invested our reinsurance premiums where the probability of payback was greatest.

The single event probability of the industry experiencing losses at or above Katrina's level as a result of a hurricane hitting in the specific area where Katrina made landfall, was 0.2% based on the AIR worldwide corporations catastrophe model.

That is a one in a 500-year event.

Going forward, we will be working with the Cat modeling companies to go beyond the traditional approach of using a 110 years of history that averages periods of both high and low-frequency storm activity.

We will try to determine the capital requirements and pricing impact of entering into a period of what appears to be higher frequency and/or severity of storms in the Atlantic, Caribbean and the Gulf.

And then to evaluate what that means from a PML management point of view.

In the meantime as Tom and Ed commented, we're going to accelerate our PML reduction actions while at the same time we will actively participate with other interested parties in efforts to develop a more efficient and effective private/public national solution to the major catastrophe problem this country faces.

Now changing topics briefly, a few comments on the reserve re-estimates reported this quarter.

Net favorable prior reserve re-estimates for the quarter totaled $150 million.

Favorable prior year reserve re-estimates of 284 million were reported in Allstate protection to reflect favorable prior year reserve development.

Principally due to lower claim severity trends and late reported loss development than was anticipated in previous reserve estimates.

Also during the quarter we strengthened reserves with discontinued lines and coverages following our annual comprehensive ground-up review.

This included 139 million related to re-estimates of asbestos reserves primarily for product related coverage.

The provision was a result of claim activity being reported by excess insurance policyholders with existing active claims because the new claims from excess policyholders and because of re-estimates that liabilities were increased assumed reinsurance sessions, since seeding companies also experienced additional claim activity.

And of course increased reported claim activity reflected our estimate for future reported claims.

As we pointed out in the past, as a provider of excess policies and assumed reinsurance, we don't have access to claims that are filed at the primary carrier level until such time as the carrier decides to seek recovery under excess policy or assumed reinsurance contract.

However, we are somewhat encouraged that the pace of industry asbestos claim activity seems to be slowing.

Perhaps reflecting various recent state legislative actions and increased legal scrutiny of the legitimacy of claims.

At the end of the quarter we had a balance of 1.1 billion on our net reserves for asbestos and our three-year survival ratio after adjusting for commutations, policy buybacks and settlements is now 29.2 years.

On an unadjusted basis it is 11.7 years.

And finally, a few comments on guidance.

Following the events of the third quarter we are now expecting operating income per diluted share for the year, assuming the level of average expected catastrophe losses used in pricing for the remainder of the year, in the range of $2.35 to $2.50 compared with the previously announced range of $6.00 to $6.40, the reduction obviously due to the unprecedented level of losses this quarter.

Now, Bob, I think we are ready for questions.

Operator, if you like to start a Q&A session.

(OPERATOR INSTRUCTIONS) Tom Cholnoky of Goldman Sachs.

I guess I am going to turn to auto for a second.

On the frequency and severity trends, there was obviously a very noticeable drop in frequency and then a bit of a spike in severity, and I wonder if you can just go a little bit more into what you are seeing there and what may have been responsible for those two trends.

I will talk for a minute about frequency, and then maybe Bob Block can talk a little about severity.

We said for a while that there are speculative factors that are driving improvements in frequency for the entire industry.

We are getting more than our fair share.

We see that continuing in the third quarter.

So it wasn't a matter of all the frequency improvements just happening in September.

Some people speculate that higher gas prices could accelerate what we've already seen.

We have over the last couple of years shifted our book to a very high-quality book of business.

We think that that will continue to provide some very good frequency benefits.

My comments on severities, I'll have to go back, Bob, I don't remember any one particular severity line being substantially out of line.

So I'm not sure what you're looking at.

But all of our severity numbers are well within our pricing guidelines.

Yes, it is actually negative and PD was positive 2.9 for standard auto, Allstate brand.

Yes, I was looking at total auto paid severity which was up I guess 4.6%?

Right, the BI paid severity tends to bounce around a little bit.

Tom, if you look over the last eight quarters it has been anywhere between 2.5% or 4.5%, up to 3% down over that.

And I think in this particular quarter the increase was influenced by a couple of claims in a few markets.

So it wasn't anything of a wide spread trend.

If you look at our bodily injury severity over the last three or four years on a paid basis, you will see results which so far surpassed anybody else in the industry.

Our physical damage coverages have been kind of in the moderate inflation area so that in between 3 and 4% generally, bouncing around there.

So we are priced for what we have in severity increases and very comfortable with our operating results particularly as it results in comparison when you compare those to everybody else in the business.

And then just as a quick follow-up I realize you are doing a lot to remediate or reduce your PML in a lot of these Cat areas.

But frankly your actions, these are going to take time and I would assume that if we had another Katrina occur next summer in a similar area like where we had this year, your losses are still going to be very significant, I assume.

Unless are you doing -- do you think you are going to do something else in those states that could mitigate the potential for another large loss?

My answer would be we have 12 months to keep working on it.

And each month counts; we are after it every month.

We would hope there wouldn't be another event like Katrina down there, but we're not going to wait around to see.

We will as Tom mentioned earlier, Tom Cholnoky, we will invest probably even more on the reinsurance side, and we will have some very intense debates with insurance commissioners that those costs ought to be passed on to our policyholders or to policyholders in general.

Great.

Thank you.

David Vargo, Fidelity Investments.

Question is withdrawn.

Greg Peters, Raymond James.

Good morning, everyone.

Well, it has been an eventful quarter.

One disclosure question for you.

In your press release, you pointed out that there was 13.2 billion of securities that you would consider selling to meet the requirements of the claims payment process.

I'm wondering what went into that allocation and identification of debt portfolio.

And I was wondering if you could comment on operating cash flow for the quarter and through the first nine months.

As usual we have strong operating cash flow as you can see by the underlying performance of the business numbers we just talked about.

In terms of designating intent to pool versus intent to sell, that is something you may have noticed in the last few Qs that we had done.

As you recall we reduced the overall duration in our AIC portfolio Allstate protection from the 5 plus range down to about 4.5, I think we ended the quarter about 4.7.

And so doing that to make sure from an accounting perspective we designate the portions of the portfolio that may be sold and to the extent there are losses there we have to record those.

As part of looking at making sure we have adequate liquidity and that it is not an issue whatsoever for us is the result of the hurricanes we did the same thing this quarter.

And we made sure that we had the capability to sell some securities as necessary going forward.

I'm not sure the operating cash flow number for the quarter, did you answer that question?

We don't have that number yet for you but it continues to be strong on an underlying basis.

But that usually is something we have available in a few days.

If you follow-up with Bob we will provide it to you.

All right.

Let me just -- my follow-up question is repair cost inflation.

In the past that has come back to nip you guys in the estimate of your Cat loads or your Cat assumptions for events.

I wondering what have you changed this time to try and prevent that from coming back and hitting you in subsequent quarters?

I think that is really Tom's point.

Is we have nine months to go before this ugly thing called the hurricane season starts all over again.

And we will probably increase the Cat loads in some of those areas.

We will attempt to increase our pricing to reflect reinsurance.

We are going to do everything we can, Greg.

What is frustrating to us is that this kind of lost makes it look like we haven't made as much progress as we have.

If Katrina had been a Florida event we would be talking about a much, much, much smaller number.

So it was an unusual event.

Very low probability.

We're going to keep working at the extreme cases and doing something to reduce that exposure.

As part of our reserve estimated process this time we did learn from the four hurricanes last year and part of that information and the process we used, the methodology, takes into consideration that increase in demand costs, if you will, from last year.

So we have learned from that and it is part of our estimate in this third-quarter.

Greg, last point, remember that when you poke at operating cash flow for the year-to-date or for the quarter that actual payments on either Katrina or Rita so far would be very, very small.

Katrina was an August 29th event.

Is that correct?

And Rita, I'm sorry I do not remember the exact date but payments would not have begun to flow out after three to five weeks on either of those events to any material extent in the third quarter that we are reporting.

Thank you for the answers.

Bob Glasspiegel, Langen McAlenney.

I am with you 1000% with your position that your contracts that exclude flooding should not have to pay for flooding.

And I'm rooting for you guys on that point of view.

Could you put that in an Amicus brief?

But, there is a but coming.

If you're wrong on that, how much sensitivity is there to that assumption?

And more importantly, is your 10-Q going to have any sort of material risk factor, financial risk statement in it on that subject?

Bob, we would not quantify what that could possibly be.

I have to tell you and it is probably not a good idea to comment too much on open litigation.

But the contracts and the interpretation of the contracts are very, very tight.

This issue has been plowed before, flood insurance has been in existence for 40 years.

We think that we will prevail on this.

We think the lawsuit that the Attorney General in Mississippi, which is interesting, it has not been followed in other states.

You never saw anything like that in Florida last year; we think it is without merit, and we will over the long run prevail on it.

So I would never want to minimize risk.

I would never want to minimize the chances of things that can happen when you're in the court system, but we think that this case is meritless.

Okay, and your discussion on Louisiana and Monday morning quarterbacking, seems to suggest you did good analysis, you just were unlucky and sort of just got bit by a really unlikely event.

And you've been cutting back in Louisiana.

With all that said, there is no sort of rethinking, was 22% marketshare Louisiana overweight the state assumed that just on a pre-reinsurance basis?

You know, Bob, we use several different modelers.

One in particular that we have an expanded relationship with that we pump all of our data through, and as Dan or Tom said before, this under any modeling scenarios this falls out at the extreme low end.

I have to tell you having said that when you look at this quarter's events and last year's third-quarter events, and you tack onto that hurricane Andrew and you tack onto that Northridge earthquake, we just simply cannot -- it's not good stewardship of our shareholders' capital to have these kinds of losses in extreme events.

So we just need to get better, smarter and get smaller in some of these coastally exposed areas.

We haven't been growing in Louisiana, taking more risk.

We have been shrinking in Louisiana.

As I mentioned before I don't recognize the 22% number but I will accept it at face value.

And we are going to do everything we can to get smaller in those coastal areas so we have less exposure.

Thank you very much.

Dan Johnson, from Citadel.

Thank you very much.

As you think about cutting back on your coastal exposure over time, can give us a sense of the amount of capital that that unlocks?

Meaning that in your discussions with rating agencies, regulators and your own capital model, how much extra road does a dollar of premium in a Cat exposed arena carry versus somewhere in the central United States?

Dan, I'm not going to be as forthcoming as your question would like me to be.

Let's just -- a couple of basics, homeowners insurance in general is written at a premium of surplus ratio of one-to-one.

Auto insurance is written at a premium to surplus ratio of somewhere around 2 to 2 point 2 to 1.

But to the extent we had less exposure we would probably free up capital.

The issue becomes at what rate can we free it up.

I don't view it as an attempt to free up capital to do something else.

I view it as good stewardship of the capital that we have.

But if you cut back in areas where you're probably writing at less than one-to-one, that should generate excess capital, or am I missing something?

No, your basic logic or thought pattern is correct.

We just haven't quantified it and aren't prepared to do that right now.

It's a good question, Dan.

As I indicated earlier, we will be working with the modeling companies to see what kind of requirements should be going forward.

Maybe that changes, as well.

So there are issues that are plus and minus.

But net net we are going to be taking down our exposure, and you would expect over time that would free up some additional capital.

I think you all know that we have been working with the rating agencies on a broader-based enterprise risk management process, a broader way of thinking about how much capital the industry requires, something more sophisticated than merely a premium to surplus ratio.

We've actually made good progress with them.

I think that they are endorsing our way of thinking about this and we are aligned with others in the industry.

So we will keep pushing how much capital do we need to run the business theme (ph).

Thank you very much.

David Small, Bear Stearns.

You mentioned a few times the AIR model talks about this Katrina event of being a 1 in 500 year event.

I guess has your thinking on this changed since the storm, and does that change your view on reinsurance?

And I guess the second thing would be does this also change your view on the technical rate that you need to price this business?

And have you had discussions with regulators about that?

As we said before, this is a single event in that specific location for the industry kind of a number.

And as I said, what we look at now is annual aggregate exposure.

We will continue to do that, and look at the various areas where we have the most exposure.

And again, look at all of the different levers for reducing that exposure going forward.

But this was a looking back AIR catastrophe model exposure number 1 in 500 years; as we work with them on whether or not the frequency severity kind of patterns should be changing in the models going forward.

That clearly may change, and so that will impact the amount of capital required, the amount of reinsurance, the amount of exposure reduction actions that we want to take.

I think we are positioned as legislatively as a much broader problem than just the insurance industry.

The trends in America are that risks are going up.

And so that is more people living in coastal areas in places where there is earthquake.

The home values have been rising quite rapidly, so part of the reason for an increase in our probable maximum loss has been increased home values, which as you know along the East Coast and in California have gone up dramatically over the last four years.

And then there is more of these events, and the water is warmer so there are a lot more hurricane these days, and so risk is going up.

At the same time, coverage has been going down.

We and other people in the industry have been taking actions over the last 10 years really to reduce our exposure, whether that's changing the contract to raise deductibles or lower the amount of additional living expenses or that is to move out of certain areas or not sell earthquake insurance at all in California like we do not, coverage has been going down.

And what that has done is put consumers' net worth at risk because most of it is in their home.

So we're trying to position this as a broader countrywide issue that involves not just the insurance industry, but also other financial players as well.

For example, the mortgage industry benefits substantially from the insurance that we provide down in Louisiana this time around.

And while some of the cost of that risk should be borne by consumers, as we learn more and as our country becomes smarter about the size of that risk, there is no reason why some of that risk shouldn't be spread amongst other financial players as well.

So we are learning as we go.

So your question about when is the model accurate; the models are as accurate as the data that goes into them and the assumptions that have been made.

We don't think they are as accurate as we would -- they're certainly not as accurate as we would like them to be.

Otherwise, we wouldn't have had the losses we had last quarter.

So we are continuing to refine those.

But given that they are not that accurate, we will continue to take our coverage and exposure down because we have no moral or legal obligation to provide this kind of coverage to people.

Great, thanks.

If I could just ask you one other question.

On the nonstandard auto, PIP growth was down I think about 12% in the quarter year-over-year.

Just maybe is that part of the market a little more competitive than other parts?

David, it is just not a focus of ours.

It is more competitive, but our PIP isn't down because of that.

We have better opportunities in what historically has been referred to as standard auto or your choice auto, with more of a standard auto product.

So we find the hunting to be much more to our liking and the profitability to be much more to our liking in the standard or preferred categories.

Those designations are over time becoming meaningless.

There'll be a point in time when we have a more effective nonstandard product offering.

But we think we have had great opportunity in the standard preferred areas.

And David, we made a call about a year and a half ago when we looked at the profitability of that line, the price sensitivity of the customers in that risk category, that given where profitability was we didn't think it was a great time to rush into the market when we thought profitability in that area would get squeezed sooner than other places.

That may be turning in the next year or so.

We will see, and when it does, it will be business that we are clearly capable of pricing well and handling the claims and customer service well.

Terrific.

Thank you very much.

Jay Gelb, Lehman Brothers.

Thanks and good morning.

First on the management of catastrophe exposure I noticed in the press release and the Homeowners Catastrophe Management Strategy you talk about establishing an exposure limit based on hurricane and earthquake losses that have a 1% probability of occurring on an annual aggregate countrywide basis.

Can you put some dollar numbers around that and what it might be versus where you are currently?

I'm trying to get a better sense on the PML.

You got the approach right, formula right.

We are not disclosing numbers on that yet.

We are still refining and will be again with the modelers to what the right numbers will be going forward.

And how to focus on which areas to take down exposure.

But we are not disclosing yet what that number is.

Let me take a second to help you understand the analytical issues around the probability.

One in 100 years, if you use 100 years worth of data you will get a certain amount of frequency of hurricanes.

If you use the last two years you would obviously get a different frequency.

So if you load two years worth of data into your model and say that is the frequency trend, you would get one thing.

If you landed in 100 years you would get a totally different and obviously much lower number because 7 of the countries 10 largest catastrophes have occurred in the last for years.

So that is the issue that our actuaries constantly, continuously work through to try to solve it on an economic basis.

But that is why you don't -- it isn't really -- doesn't create a great conversation to be sharing all that information as we learn it because then as numbers change, as periods of data that you load in change, you end up doing all kinds of reconciliation.

I think we would like you to just understand that we are serious about continuing to reduce our risk.

Dan has goals for us on certain measures that we're going after quite aggressively.

As he gets new measures and those numbers get lower, we just go at them back again and put our helmet on and go get it.

So it is just if you take it under -- understand that we're working very aggressively to reduce our exposure I think that gets the point across.

I see.

And then on a separate issue, can you update us on your comments with the rating agencies now that you have disclosed the catastrophe losses?

Sure, Jay.

As you would expect, have been keeping rating agencies informed of our -- first of all the methodology and progress and development and our Cat cost estimates, that we've covered our capital adequacy and our excellent liquidity position within the fact that we fully expect to be able to complete our current share repurchase program during 2006 without having to raise capital.

We are keeping our financial leverage at a very conservative level.

We expect that we will be hearing from them.

They will be disclosing in a few days to a week where they are with us and probably many other insurance companies.

But those discussions have gone well.

Right, and no sense at this point that you need to raise any capital, let alone diluted capital?

Correct.

Great.

Thanks very much.

Bill Wilt, Morgan Stanley.

Another question on Cat management.

Do you anticipate that the insurance commissioners in Louisiana and Mississippi will allow meaningful reductions in exposures?

And do you have the sense for or any ability to quantify the amount of nonrenewals you might undertake?

It is too soon, it really is.

You know the way Tom and I have been thinking about it is I guess if push came to shove and you had to actually stop writing homeowners in some states in order to reduce the PML, we would do that.

In many states it will not come to that.

Some of the insurance commissioners, particularly Robert Wooley in Louisiana is a pretty reasonable guy; he understands that you have to have availability and affordability in order for the marketplace to function.

But insurance companies just can't continue to bear these kinds of risks.

So I think we will have some intense debates.

I think the more enlightened commissioners will be just fine with where we are going.

Okay, thanks.

And on personal auto, and admittedly a 30,000 a foot question, but as you look forward to 2006 in personal auto, do you anticipate charging on balance more, the same or less per policy?

You know, not focused on it right now; we are focused on getting through the third quarter.

I don't see a major change in trends maybe is a better way to talk about it.

I think as you and I have chatted personally and I have chatted with so many of you on the phone, if you look at the things that drive frequencies, cars are getting safer, drivers are getting safer, highways are getting safer.

Nothing goes straight down.

And you all get very nervous if you get any kind of perturbation in the frequency improvements.

But I think the secular trends that drive frequency improvements will continue.

Severities, as we said, severities are performing well within our pricing guidelines.

We feel very good about the fourth quarter and going into 2006.

Very good.

Thanks.

Kelly Nash, KeyBanc Capital Markets.

My first question was just regarding New Jersey, your comments in the press release, and also if you could touch on perhaps new auto business growth in terms of if you are going to be decreasing your homeowners exposure, how does that potentially impact your auto growth in states such as Florida or other states where you might be decreasing exposure?

We will do them in reverse order and Tom will have some comments also.

Obviously if we shrink on a more rapid basis on our homeowners policies in force, it presents a challenge in terms of us growing PIF.

That is a good trade and we will take that trade any day of the week.

I will also tell you that it means our agencies have more time, more energy and more focus on homeowners.

So on auto.

So that is a good thing.

You know, we have begun to see Florida bounce back.

We had not only some issues when we offloaded some of our policies but we also had a rating plan that was suboptimal.

As we fixed that rating plan and got it better, things have begun to stabilize in Florida.

So it's fair to draw the connection.

We think it is a good trade.

If it means less PIF in homeowners in order to improve our profitability and add consistency to our profitability, we think it makes good sense.

I didn't catch the New Jersey reference, can somebody help me.

The New Jersey reference, Kelly, is to GEICO.

GEICO entered the market in New Jersey.

They picked up some share.

We have marketing programs around that.

But we have intentionally backed off on certain risk characteristics that we think are not as well priced as other ones.

So we have in a ceded some of that growth to them and we're working hard to make sure we keep the business we want.

In New Jersey we're really focused not only market share, but on share profitability.

And then if you could just comment on your comment on consumer shopping trends, consumers aren't shopping as much and we have seen that kind of throughout the year.

What do you see that might change this trend going forward?

Change in the trend -- what you've seen this year, Kelly, is there has been a dramatic increase in marketing dollars.

And when you look at particularly at us, Progressive, GEICO and State Farm, we drove our marketing way up in 2003.

They raised theirs in '04.

That will tend to stimulate some shopping in the marketplace.

Clearly the stable pricing environment keeps people from shopping some.

From our standpoint we have been working hard, not just on competing in the marketing basis, but also in improving our customer satisfaction.

We've made dramatic improvements in customer satisfaction over the last year, moving from below average to above-average both by our internal surveys and JD Powers.

Obviously happier customers tend to stay with you more and don't change for $20 or $50.

So we are working hard not just on making sure we compete aggressively for those who do shop, but on keeping more of those who are our existing customers.

So it sounds like on a longer-term basis consumer shopping might continue to be challenging in terms of getting new business.

If pricing is what stimulates the shopping, yes, we'd like to think that our marketing stimulates some people particularly the ones who are interested to come to us.

We think things like Your Choice Auto which are new and nobody else has in the marketplace will cause people to shop, and then we think we protect our home base by making sure people are happier.

But we don't see the decline in shopping behavior as a reason that should stop our growth from occurring.

We are continuing to grow.

Our quotes are up and it is just a question of which business we want to take.

Thanks.

Gary Ransom, Fox-Pitt Kelton.

I had a question on homeowner's frequency.

Looking at the frequency there excluding catastrophes and kind of looking back in time, it has declined fairly significantly over a long period of time.

And I just wondered if you could comment on what has been going on there more broadly and whether those trends are likely to continue.

The first thing I think that has occurred has been as homeowner prices have gone up, consumers have shifted to self-insure more, which means higher deductibles.

So when you look at the percentage of deductibles that have moved up from 250 to 500, 500 to 1000, it is a meaningful shift.

We do track that, and that means the small claims go away.

And what you will see is an offset in the cost of the severity.

So when you look at the average severity of homeowners you will see that that has been up at higher than inflationary rates.

And that is because when you're looking at the average if you can take out small guys you're still left with the big guys and so the average goes up.

So you really have to look at the two of those combined.

When you look at our paid peer premium in homeowners excluding catastrophes it has been a good line for us.

One of the -- some of the pushback we got last year and the year before as we were aggressively reducing our Cat exposure was (indiscernible) run into 75 to 80, to 85 combined ratio and a line that has hardly any tail to it and is highly predictable.

Why would you not want that business.

Well, this quarter showed us why we were smart in doing what we did, and we wish we would have done more of it in Louisiana.

But as Ed said, declining frequency really has been due to higher step up.

Some people would argue that there is something called claiming behavior.

It is not a number you can prove, which is when there are fewer people selling homeowners insurance, people are less likely to file a claim.

Or when you have surcharges for claims out there people are less likely to file a claim.

That is not a number that anybody can prove.

We hope that all of our customers who have a claim call us.

I guess that means your deductibles have been going up at fairly steady pace for several years because this decline has continued actually for several years.

The mix has been shifting, yes.

Just one follow-up question.

In the holding company can you update us on how much -- how many funds are in the holding company available for buyback right now?

What you are asking, Gary --.

As opposed to cash that is in the Allstate Insurance Company you have upstream money up into the holding company using that for buybacks.

I am just --

Somebody will have to give me an exact number, but there is somewhere between 2.5 and $3 billion of money that has been dividended up over time.

From the regulated entity of Allstate Insurance Company into the unregulated entity of Kennett Capital.

That's an aggregate cumulative number over time?

Is that what you are saying?

Is that what is up there right now?

That's what there is now.

Thank very much.

Alain Karaoglan of Deutsche Bank.

The question is on catastrophe management.

It seems that you do a good job of protecting against events that have already happened in the sense that you mentioned that if Andrew were to happen again your losses would be significantly better than they were.

And I would assume the same is true if Katrina were to happen two or three years down the road.

Are you working as hard on reducing your earthquake exposure because you haven't had a big one in a while, and it seems to be a risk that people aren't focusing as much on as, of course, when given that you have recent losses.

A couple of comments -- and then Tom is closer to it than I am.

The creation of the California Earthquake Authority is specifically in California, took a lot of the risk of earthquakes off of us and put it in the California Earthquake Authority; to the extent of course that it is solvent and can handle those losses.

We are very much aware of things like the New Madrid Fault and some of the exposures up in the great state of Washington.

And we are in fact attempting to manage those down.

I don't know, Tom, if you want to add more to that.

I would say we are managing them down.

We are doing that in a variety of ways.

Some are dramatic increases in the cost of earthquake insurance, over 100% in some areas.

So you have a lot of people who don't buy it when you raise price.

Some are just not renewing or not selling it in other places.

We are also looking at reinsurance as we speak for California.

We still do have some exposure to earthquake in California through our assessments in the California Earthquake Authority, and potential losses in fire following events which to your point are very far out on the probability.

To have fire following losses after an earthquake which are covered by a homeowners policy -- not only do you have to have an earthquake -- not only do homes have to be damaged that you insure, but they have to be damaged by fire.

And but that is a risk.

And so we're looking aggressively as we speak at whether we want to purchase reinsurance for our CEA assessment and fire following an earthquake.

A question on the life business.

Does the merger of Lincoln National and Jefferson Pilot lead you to think the position of Allstate Financial and its competitive position in the life market?

No, it doesn't cause us to rethink it at all.

That's an interesting acquisition.

It is -- or merger.

It is two companies that really have made good progress in increasing the returns, and that is what we're working on.

But have been struggling with increases in operating income, whether this helps them or not I guess is anybody's guess.

Remember Allstate Financial a good chunk of it is sold through the Allstate agency channel so the distribution focus is entirely different.

It is different than what Safeway had.

It is different than what Safeco had -- I'm sorry, it's different than what Jefferson Pilot or Lincoln had.

We are encouraged by where we are with Allstate Financial.

We think over time we will improve those returns.

We think the debate on Social Security is probably a healthy one and maybe Americans will refocus their energies on saving.

And the Allstate brand plays very, very well in the savings, retirement, life insurance category.

It is a big business.

We've been at it for 50 years.

We know it well; we think it's an opportunity for us.

We're going to continue to pursue it.

And just one numbers question.

Do you have the property casualty statutory surplus at the end of the third quarter?

No.

It is too soon to have that; we would have that probably for another two weeks or so.

Thank you very much.

Nick Pirsos, Sandler O'Neill.

Just a question first on expenses.

It was a great quarter and I was curious in the context of your massive claims effort, would we expect to see any upward pressure on that, or how should we think about that?

When you look at the underwriting expense line that excludes of course our claim expenses.

It excludes any claims expenses.

The reason for the great performance this quarter was a reduction in management compensation as a result of the losses associated with the hurricanes.

So we are continuing to invest in marketing.

We are continuing to invest in technology.

We haven't changed those strategies or that funding at all.

It has really been we're paying people a lot less because we're not making any money this quarter.

And so we need to adjust for that.

So that is the change there.

The claim expense piece, that does get factored into the Katrina and Rita estimates.

We have lower -- if you look at court claim expenses on an ongoing basis in comparison to progressive, GEICO or State Farm, we are better than they in all those companies.

Our claim expenses will be rolled into the Cat numbers.

Those will be slightly higher because we're using some outside adjusters because we have 4, as Ed says we have over 4000 people working down there.

We don't have 4000 extra people sitting around here so we have a relationship with some outside vendors to bring contract adjusters in.

That will cost us more money.

But all of those incremental costs which are higher than our base run rates that I talked about before are priced in to our (indiscernible).

And on the auto side with regard to future claims factor possible adjustments, just given some of the disruption in the auto parts sector would you anticipate any changes there?

You talking about reserves?

Just claims factors exactly for reserves thinking more specifically what is going on with Delphi and the bankruptcy there.

The cost of parts is of course significant in our physical damage costs.

The other thing that drives some of it is used car prices.

We have benefited over the last three or four years from a low or even declining used car prices, which gets into what you pay for a car, what you pay for parts and that kind of stuff.

Most of the parts that we're talking about are crash parts which are metal.

And I don't expect that you would see the bankruptcy of Delphi or anybody else impact those prices in metal parts.

They are way too high as it is.

Great.

Thank you.

J.F.

Tremblay, HSBC.

I hear you loud and clear that you want to lay off more of your risk in the reinsurance market and potentially public entities.

At the same time I think the risk is real that you might have to increase your capital allocation to Cat exposed areas if you decide to keep doing business in.

Now here's my question.

To what extent you think you might have to revise your return expectations in light of this new Cat management strategy?

Don't expect to do that.

We prefer to solve the numerator issue I guess rather than the denominator issue.

You know, our business is a well-run business.

And you all have heard me say this many times a well-run property casualty companies generate good cash flow, which we have done.

And they generate reasonably good ROE's.

When you have the scale and capability and capacity that we have.

So we would prefer to solve for the PML issue rather than the lower return issue.

Paul Newsome, A. G. Edwards.

Just a correction, maybe if you could comment little bit on competition pricing in the business in general for your major lines.

Yes, you know, as I said in my opening comments, it's a competitive marketplace; it has been for the better part of 18 months.

Strategic risk management, which is our tiered pricing system developed by Fred Cripe and our pricing people serve us very, very well in this kind of a marketplace.

We think the competition is rational and disciplined.

We are encouraged by that.

Don't see any people losing their way and doing silly things.

So I see more of the same.

Do you think that the effort that goes into paying the claims to Katrina and Rita could change the progress that you and your competitors -- probably more the competitors -- are making towards implementing some of these new underwriting systems and generally improving the statistical methods that are used in pricing?

I would say no.

In general I would say no.

Let me just expand briefly.

When you do four major hurricanes in the state of Florida you learn a lot about it.

And we are smarter today than we were last year, and we were smarter last year than we were five years ago.

So there are all those learnings that go on that get filtered back into underwriting processes.

But it is a secondary or tertiary effect, not as direct as I heard in your question.

Great.

Thank you very much.

Ron Frank, Citigroup.

Two things, if I could.

First, there is this ongoing dilemma that some of the largest markets in the country, Florida, California, New York, Texas are all Cat prone.

You can't really rely on reinsurance pricing and capacity to meet all the needs consistently, and you can't offset shrinking in those states by growing in Montana or what have you.

So realistically are we at the point we're getting to Dan Johnson's earlier point -- it is just going to be less PIF growth and more capital generation unless and until you get the government involvement you want, Ed?

Maybe yes, maybe no, Ron.

I'm sorry for the less than firm answer.

As I said earlier, clearly as we attempt on an accelerated and more aggressive basis to manage our PML down, it will have negative implications on our PIF in some of those large states.

But I want to go back to what Tom said.

This is people want to view this as an insurance issue.

It is much more than that.

It is an economic vitality issue.

And if you don't think I have lost my mind, let me explain.

The state of Mississippi, when they lose the casinos that were a very large source of the state's revenue, or you look at what has happened to New Orleans and people aren't down there and property taxes aren't down there, it is a big issue for those states, and the economic vitality in those areas.

I think Jeb Bush has woken up to this.

In the state of Florida.

And therefore, we think the whole system needs to be re-thought about I think the number is 55% of Americans live in harm's way.

In harm's way of these hurricanes.

And that is the states where everybody wants to move, about 1000 families a day move into the state of Florida.

So maybe one silver lining in this kind of a very catastrophic event is there is a greater willingness on the part of state governments and state regulators and the federal government to recognize that this is not insurance companies whining.

It is the basic economic vitality.

So we think there is some reason to believe that some of the states will in fact wake up and get it.

Florida is a lot better off today than it was before, a lot better off over the last 10 years.

So it sounds like you are not so much challenging my premise as you are expressing more optimism that you will get back government help.

I think we will.

Nobody gets government help easily, and that is why I want to make you concentrate on Tom's point.

This is not just an insurance issue.

It is a real estate brokers, it is mortgage bankers, it is everybody who has got a stake in the real estate business.

The American dream is to own a home, and so much of peoples' net worth is tied up in home values, there has got to be a broader solution.

I am optimistic that that will take place; it isn't going to happen by Monday.

And as a follow-up, could I just ask on frequency, you said the secular trend remains favorable.

There was nothing unusual in the quarter.

So as we look at auto frequency down 5% in the third quarter and down 0 to 2% roughly the last three quarters, which number is in line with that secular trend that you spoke of?

I don't know, I tend not to think about it.

I tend not to think about it that way; maybe it is someplace between the two.

Fair enough.

Thank you.

Kevin Lampo, Edward Jones.

Just taking a look at, you talk about Katrina being a 1 in 500 year event, but then when I look at the Cat losses excluding Katrina, if I am half right, about 15 points for both losses in the quarter.

And that is historically pretty high so I am curious how you if you can help me understand maybe how the rest of your business you had been working on reducing some of the (technical difficulty).

Kevin, we are going to take a crack at that question;

I apologize to you, you were breaking up as you were going through it, so maybe Dan you can try to respond and if we heard it correctly, great.

We will come back.

I'm not sure I heard the question.

So I am not sure how to respond.

Could you repeat it?

Yes, excluding Katrina you still have Cat losses of 15 (ph) points in the quarter which historically is pretty high.

So I am trying to understand how your efforts to reduce your risks of catastrophes, excluding Katrina, it still looks like you haven't made a whole lot of progress (technical difficulty)

Let me see if I can address it.

When you look at the results for the quarter clearly we've lost more money than we would have hoped or more money than we expected to lose based on the programs we had in place.

But what you have to do is look at the risk reduction on a smaller geographical basis than just in total.

So if you look at Florida, and in Florida with all of the losses we had, if it was before we created the Florida hurricane Cat fund they got the changes we made last year and all that, if we had $1 billion loss we would have had a $1 billion loss.

Today a $1 billion in property turns into $300 million in property losses.

Which is a pretax number.

Yes, pre-tax.

So we have done that in Florida.

In Texas we purchased reinsurance, in New York we purchase reinsurance.

The fact that it is by some peoples model a highly unusual event has led us to do less work on those highly unusual events than we have on the more likely higher probability event.

So if you just look at Florida and Texas, they both stick into the Gulf.

So they are kind of sticking out so you are just -- physics would tell you that you've got a bigger probability of them being -- the models also tell you that.

So we've spent more of our time focusing on those things that stick out like Florida and Texas or things that have very high capital requirements but are still relatively low probability like New York where we have a large share.

So while it was a low predicted probability, what we have said is we are now focusing not only on the higher probability events, but we are also going to be more, to the lower probability events in working on those as aggressively as we have on the higher probability areas around the country.

Is that helpful to you?

It helps some, maybe if I could follow-up looking at your Rita, the losses from Rita, can you break out what those losses were by state as far as Louisiana versus Texas?

Rita was about 50-50 Texas versus Louisiana.

And again, the recovery, insurance recovery is about 205 million.

Okay, so on a growth basis 50-50 (inaudible)

Of course reinsurance recovery piece is only on the Texas loss, not on the Louisiana loss.

We were kind of hoping it would go west, young man.

Thank you.

Brian Meredith, BankAmerica.

Good morning everybody, I have one brief question here.

Looking at the flood situation in Louisiana, one of the lawsuits that are around right now is the Scruggs lawsuit.

It is focused more on I think an agent E&O issue.

The question I have is now that your agents are independent contractors, do you have any exposure to any E&O risk with your agents?

We don't believe so.

We are going to fight this thing tooth and nail as I said earlier.

It is very clear what the federal flood program covers, and it's very clear what our policies cover.

Those issues have been tested in many, many places.

And we feel that we will prevail.

Now there could be some comings and goings within the Mississippi or within the federal court system but we feel pretty confident that we will prevail over the long run.

Okay but the point is you don't know legally whether you have exposure to the E&O with your agents now where you may have --.

No, there has been plenty of notice, conversation, the national flood insurance program does advertising around this.

Many insurance commissioners do public service ads around this.

They all put out brochures around this.

The concept of flood insurance has been out there since 1968, and I would point out that Dickie Scruggs owns flood insurance himself.

So he obviously was well-educated on it.

Great and one other quick question can you just talk about the how the premium products are going in the Your Choice?

You know, a little too soon to declare victory, but we sure like where we are right now.

As I mentioned, the mix of products is more on the premium and the goal side, and that's good for us because it represents a revenue increase.

A premium increase.

What is really good about the product, though, is that it gives the customer more control over his selection to match up to his or her own risk appetite.

So we are liking what Your Choice Auto is doing.

As I said, it is in 15 states now, and it will be in a total of 27 states by the end of the year.

It is producing the kinds of premium switches that we thought it would.

And we think it is it will have great impacts on our profitability in the future.

Great.

Thank you.

(indiscernible) Second Curve Capital.

Congratulations on the continued strong court performance.

I just had one quick question for you guys.

It looks like a lot of your competitors have even adjusted for shifts in mix or geography are experiencing already sort of flat to some cases declining average premiums per policy.

I remember Allstate, actually you guys still are able to generate increases on that front, on auto, that is.

I just want to know from a competitive standpoint what do you believe is driving that?

What is it that has helped Allstate to still get the increases on average premiums?

Well we get -- you get some increase in average premiums just as people buy new cars and the cars are a little more expensive.

In insurance vernacular it is called symbol drift.

Just that people buy more expensive stuff.

Secondly, in comparison and I am not sure which competitor you're looking at, but in comparison to some competitors we haven't been as aggressive in lowering prices to get new business.

Because we have said we are going to stick with our discipline; our goals last year the year before and next year are one, to reduce our capacity exposure, two to protect our margin and three to grow.

I believe our growth is off some this year because we have not decided to chase price down.

And that has cost us on our close rates, if you look at our close (indiscernible) but our close rate is down a little bit.

But we intend to continue to be competitive in the marketplace.

So we are not -- we don't have our head in the sand and if our price is not competitive, we will get competitive as long as we have the kind of margins we want to have.

If it is a question of growth or margin, we're going for margin.

Thank you so much and one quick follow-up on your choice.

I guess how important is potential increases in policy life expectancy retention to that equation?

When you look at the benefits of Your Choice, we think clearly one of them as Ed mentioned is higher average premium.

We also think that it will have a good impact on retention.

We are a little too early in the game to figure it out.

We think it will have a pretty significant move on first term retention, which has always been a challenge for everybody in the industry getting them to stay just once.

The longer they are with you the easier it is to keep them as a customer.

And so we are optimistic that it will do that, particularly in the platinum product where you get a credit on your next insurance policy.

So not only does somebody have to beat our out-the-door price, they have to beat that plus the credit somebody got from having a policy with us for the last six months.

And I should mention that we are working to patent protect that process and those features so that we have a competitive advantage.

Great.

Thank you very much.

Todd Buechs, Sanford Bernstein.

Good morning, guys.

Just listening to the range of these questions, I am a little concerned that people are still trying to think about cat management too precisely.

I mean, if you're playing the game of craps, even though 7 is the number that's going to come up the most often, 2 occasionally comes up.

And that is kind of how I think we need to think about what to expect from catastrophes.

I mean, you can tighten the range of outcomes by buying reinsurance or managing your exposure, but you're still going to produce a range of outcomes.

Even your own quarterly disclosures on catastrophe show that even normal catastrophes have quite a range of outcomes that would matter a lot to any quarter's earnings.

So I am wondering if it isn't time for a new voluntary disclosure, not just from you but from most companies, whereas you work with the rating agencies and work with the modelers, you come up with some key, high-profile scenarios and you lay out your current distribution of outcomes for those, based on your current reinsurance program and current exposure program.

Something like a range of probable losses for a new Madrid event, a California earthquake, a windstorm in the Gulf, similar to what Lloyds does.

I know that would take some time to develop, but I think it would be really helpful and I think it would be better than PML's or even putting out your reinsurance program, which is just too difficult I think for investors to ripple losses through.

What do you think?

If there is a human cry for that kind of disclosure we would consider it at some point in time doing it.

We have moved over the last six to seven years, I think to being one of the most transparent companies in financial services to make certain that people can understand what the risks are and what the opportunities are in investing in our stock.

So we will look at that and make no commitments to you.

My instincts are that you are correct.

We tend to talk about these PML exposures with great precision and we know, we use a 1 in 100 year, or 1 in 500 year and you know what guys, the models just are not that accurate.

They just are not.

So anything we can do to be a little bit -- to make you feel better about the risk that we are taking, may make some sense.

I would also tell you, Todd, you have absolutely put your -- hit the nail on the head with -- each quarter I struggle with how do you give guidance when you can have such variability around normal catastrophes?

Because normal can have a very substantial range attached to it.

We will keep thinking about that issue.

In terms of your hue and crying, I am huing and crying.

I will tell my clients to hue and cry with me.

We have let this go on a little longer than usual because we wanted to make sure that everyone got their questions raised and answered.

As I said when I started, this obviously has been a difficult quarter for us.

But I am extremely proud of the way our agencies and our employees everywhere rallied to face these challenges, working to help restore our policyholders lives as quickly as possible.

You know, this hurricane season reminds us that while we have made real progress on PML management over the years there is still much to do and we will do it.

It also has clearly demonstrated the need for a much better solution to these mega catastrophe events for our nation, and we are actively involved in crafting and leading that solution.

That said, this quarter is a real statement about the strength of our franchise.

I think in two ways.

First, we have the scale, we have the know-how, the processes and most important we have the people to effectively handle events of this magnitude.

We do a good job in these events.

It is our business, and we do it very well.

Second, the fact that we can experience over $5 billion in pre-tax catastrophe losses through the first nine months of the year and still make three-quarters of $1 billion in net income, is a very powerful statement.

In that time the first three-quarters of this year we generated significant cash flow.

We used that to grow our business.

We used it to fund shareholder dividends and to repurchase stock.

We have a very strong high-quality balance sheet with really only a 20% debt to capital ratio.

And more than adequate liquidity to cover our catastrophe losses without raising additional capital, and we fully expect to complete our current $4 billion share repurchase program during 2006.

Do not overlook the fact that our core lines of business are performing very, very well.

But also don't misunderstand me.

We will be taking more actions to reduce our exposure to catastrophes.

But our franchise is very, very strong.

Have a safe year end and we will talk to you in February.

Goodbye.

This concludes today's conference.

Thank you for your participation, and have a wonderful day.