Selective Insurance Group Inc (SIGIP) 2007 Q1 法說會逐字稿

Good day, everyone.

Welcome to the Selective Insurance Group's first-quarter earnings release conference call.

At this time for opening remarks and introductions, I would like to turn the call over to Vice President Investor Relations, Ms.

Jennifer DiBerardino.

Good morning and welcome to Selective Insurance Group's first-quarter 2007 conference call.

This call is being simulcast on our website and the replay will be available through May 25, 2007.

A supplemental investor packet, which includes GAAP reconciliations of non-GAAP financial measures referred to on this call, is available on the investor's page of our website at www.selective.com.

Selective uses operating income, a non-GAAP measure, to analyze trends and operations.

Operating income is net income excluding the after-tax impact of net realized investment gains or losses.

We believe that providing this non-GAAP measure makes it easier for investors to evaluate our insurance business.

As a reminder, some of the statements and projections that will be made during this call are forward-looking statements as defined by the Private Securities Litigation Reform Act of 1995.

Forward-looking statements are not guarantees of future performance and are subject to risks and uncertainties.

We refer you to Selective's annual report on Form 10-K filed with the U.S.

Securities and Exchange Commission for a detailed discussion of these risks and uncertainties.

Please note that Selective undertakes no obligation to update or revise any forward-looking statement.

With us today are several members of Selective's senior management team; Greg Murphy, our CEO; Jim Ochiltree, head of Insurance Operations and members of his senior leadership team including Ed Pulkstenis, Chuck Musilli, John Marchioni, Mary Porter and Jeff Kamrowski.

Also with us are Ron Zaleski, Chief Actuary and Kerry Guthrie, Chief Investment Officer.

At this time, I would like to introduce our CFO, Dale Thatcher, to review first-quarter results.

Thanks, Jennifer.

In the first quarter, we achieved the 13th consecutive quarter of a statutory combined ratio under 100% and a 95.6% compared to an overall combined ratio of 93% in the first quarter of 2006.

Catastrophe losses added 1.3 points to the overall statutory combined versus 0.9 points in 2006.

In addition, we experienced property losses about $11 million higher than expected, largely driven by non-catastrophe weather events.

Operating income per diluted share was $0.50 for the quarter compared with $0.56 last year.

Net income per share was $0.62 on $37 million, down 3%.

New commercial business was up 6% to $76 million for the quarter, including a record 21% increase in small business seamlessly underwritten through our automated One & Done system.

Overall, commercial lines net premiums written were unchanged reflecting flat renewal pricing of 0.4%, including exposure, a slowing economy and lower retention of some larger accounts due to an increasingly competitive marketplace.

During the quarter, premium in our large account operation, which represents 11% of commercial lines premium, was down 7% to $42 million compared to first quarter 2006.

This was about $6 million below expectations as we lost business to carriers writing below technical premium levels.

On the other end of the spectrum, new business was up 15% and renewals were up 5% for accounts with premium $25,000 and under.

Retention in the first quarter held steady at 78%.

Commercial lines profitability remains strong with a 94% statutory combined ratio driven by solid results in all lines of business with the exception of business owners, which represents only 4% of commercial lines premium.

The ongoing implementation of our multi-faceted underwriting improvement plan for workers' compensation led to a 12 point improvement in the workers' compensation statutory combined ratio to a 98.2%, including 2.4 points of favorable loss development.

New workers' compensation business was up 19% and predictive model scoring and other key initiatives enabled us to retain more of the best accounts while targeting price increases to our worst performing business.

In the quarter, our reserve ratio was 17%, which, when coupled with minimal exposure growth, reflects a strong overall reserve position.

For the fiscal year ended March 31st, casualty loss costs were down about 3.5% versus average earned premium increases of 2.6%.

Our reduced loss costs were driven by lower frequency of 3.6% while severity was flat.

In addition, we experienced favorable reserve development of $4 million across our casualty lines of business during the quarter.

Personal lines net premiums written were down 24% to $47 million for the quarter.

This was partially the result of a one-time benefit of $11 million in the first quarter of 2006 due to the elimination of the New Jersey Homeowners Quota Share Treaty.

The decrease also reflected the loss of some New Jersey personal auto business, which our new MATRIX pricing program targeted at a higher rate.

These accounts are now with other carriers who may not have the benefit of more advanced pricing systems.

The total number of New Jersey cars we insure decreased to 75,000 from 77,000 at year-end 2006.

Although our personal lines performance remains below acceptable levels, MATRIX and other automation enhancements have helped us to establish a smaller, but stronger base from which to grow profitably for the long term.

The personal lines statutory combined ratio, excluding flood, was 110.3% compared to 103.5% in first quarter 2006.

Results were driven primarily by increased severity in homeowners' weather-related property claims.

The New Jersey auto statutory combined ratio for the quarter was 94.3%, a 6.4 point improvement from a year ago.

Although our personal lines combined ratio does not yet reflect the profitability improvements we anticipate, our efforts to improve results include the MATRIX pricing model, the personal lines service center, a three state personal lines expansion in our existing footprint and a reorganization designed to right-size the operations by leveraging the technology we have deployed.

Diversified Insurance Services revenue grew 7% in the quarter and generated a 10% return on revenue.

Flood operations grew 27% in the quarter bringing total flood premium service to $125 million.

Selective HR Solutions ended the first quarter with about 27,000 work site lives, up 7% from March 2006.

Growth was not as strong as anticipated due to a slower economy, but Selective agents are having more success marketing the rebranded employer protection program as demonstrated by a 56% increase in the sales pipeline compared to a year ago.

After-tax investment income rose a strong 11% to $31.2 million for the quarter, reflecting a 10% increase in fixed maturity and a 47% increase in short-term income.

Our overall annualized after-tax portfolio yield was 3.5%, unchanged from March 2006.

Duration, including short-term assets, was 3.9 years.

Invested assets remain essentially unchanged from year-end 2006 at $3.6 billion, while investment assets per dollar of stockholders' equity was up 3.6% to $3.46 from the end of the year.

The average credit quality of our $3 billion bond portfolio is "AA." The mortgage-backed securities and homeowners equity portfolio totals $661 million with an average credit quality of "AAA." Selective has no sub-prime mortgage exposure.

We remained active in our share repurchase program during the quarter, including the initiation of a 10(b)5-1 trading plan.

4.3 million shares were repurchased at an average price of $25.32, including three million shares in the first quarter and the balance through April 20.

We have roughly one million shares available for repurchase under the original 10 million share authorization, which has been extended to December 31, 2007.

Growth in book value per share was a strong 9% to $18.94 from $17.40 a year ago.

Now, I would like to turn the call over to Ed Pulkstenis, our Chief Commercial Lines Underwriting Officer, to bring you up-to-date on predictive modeling and our underwriting approach as we manage our way through the current pricing cycle.

Thank you and good morning.

We are in a cyclical business and we know that general price levels will rise and fall over time.

We do believe, however, that companies with effective tools in place to proactively manage and monitor their business exposure can continue to be successful even in tougher market conditions.

Over the last five years, Selective has been developing the cycle management tools and capabilities necessary to grow profitably in any market.

First, we reorganized the strategic business units to better align them with our best opportunities and improve our sales focus.

Second, we divided the commercial lines book into 80 business segments.

We monitor the profitability and growth of the segments and align our underwriting and pricing strategy with those results.

Third, we have significantly expanded our Safety Management capabilities by installing an automated system allowing us to direct specific safety management efforts to accounts that need them most.

Fourth, Knowledge Management was initiated to harness the tremendous collective knowledge of the organization and translate it into actionable underwriting information.

Last, but most importantly, we are launching predictive modeling for nearly all lines of business.

Selective's use of Knowledge Management allows us to analyze exposures and differentiate among risks in a much more disciplined and scientific way.

Our objectives as we embarked on this new capability were to better identify the best business to write, both new and renewal, and the specific price points we need to write it profitably.

Writing profitable business today starts with very sophisticated predictive modeling that identifies the characteristics of good and bad business and prices it for profitability.

Our first predictive model was for workers' compensation and was fully deployed in November 2006.

The model consists of over 30 variables and it allows us to rank business into five categories with five diamond business being the best and one diamond the worst.

In that context, through March 31, 2007, 94% of new workers' compensation business was the better three, four and five diamond business.

Historically, the distribution was that 80% of this business was in those categories.

We also non-renewed 39% of the low-performing, one diamond business, in line with expectations.

The one diamond business has historically represented 5% of workers' compensation premium.

The business owners or BOP model rolled out in fourth quarter 2006 and we are already seeing the benefits of scoring in the growth for this line of business.

In fact, since implementation, BOP new business increased 19% through the first quarter of 2007.

Our commercial package model has just been ruled out.

The commercial auto model is under construction and targeted for early 2008 delivery.

We expect new business growth and retention of our best business to be strong as a result of the full implementation of the remaining models, including personal lines.

We deliver the information from Knowledge Management directly to the desktops of our underwriters and agency management specialists or AMSs along with other key decision-making data such as Safety Management reports and agency information for each account.

Our field staff then uses this information within the context of the relationship they have with their agents to make the final underwriting decision.

In addition to having the models, we have built business analytic capabilities that deliver underwriting information on a granular level to our business units.

We use this information to identify trends and validate and enhance our models.

For example, over the last 12 months, our analysis shows that for our workers' compensation renewal model, the reported claim frequency for low-performing one diamond business has been 3.5 times the frequency for high-performing five diamond business.

This is an encouraging early indication of the capabilities of the model.

We believe that with our modeling capabilities, we are well-ahead of the majority of the competition and strongly positioned to generate profitable growth.

Now I will turn the call over to Greg.

Thanks, Ed and good morning.

In a quarter which commercial lines competition continued to intensify, we delivered solid, new business production, continued strong commercial lines profitability and double-digit growth in after-tax investment income.

Our approach to this marketplace centers on financial strength.

Notably Selective's A+ rating from A.M.

Best and strategic cycle management that leverages the benefits of strong agency relationships and underwriting discipline with new predictive modeling tools.

We remain clearly focused on delivering long-term shareholder value as evidenced through our growth in book value per share of 9% from the first quarter of 2006.

Due to our more aggressive share repurchase program, we are increasing our 2007 guidance by $0.05 to a range of $2.25 to $2.40.

Given the more competitive market, we now expect no commercial lines renewal price increases.

However, underwriting improvements enable us to maintain our combined ratio of guidance of a GAAP combined ratio of less than 98.5% and the statutory combined ratio under 97%.

Other assumptions supporting guidance remain, a normalized level of catastrophe losses of $14.7 million after-tax or $0.25 per share; after-tax investment income growth of 10%; and Diversified Insurance Services revenue growth of 12% and return on revenue of 10%.

Strength in new business was driven by a 21% increase in our automated small One & Done production system.

We achieved a record $206,000 in daily premium written in the quarter or $13 million overall as virtually every region and line achieved significant growth.

Our large account unit, SRM, is experiencing the most intense competition.

Several large renewal accounts were lost due to aggressive pricing behavior.

For example, a $1 million account up for renewal went to another carrier priced well below our pure premium levels.

We continue to have the discipline to walk away from such pricing.

However, we are still writing new business.

SRM's new business grew 19% to $8 million for the quarter.

However, it was $3 million below our expectations.

For the quarter, our AMSs generated $52 million of middle-market business, an increase of 2% from a year ago.

On average, each AMS generates $2.4 million in annual new business.

To drive profitable growth, we have established the following initiatives.

One, AMS deployment and productivity improvements.

Year-to-date, we have deployed seven AMS apprentices to service new and existing territories.

We currently have 13 AMS apprentices in the program, soon to be 17 as we have recently hired four AMS apprentices.

To increase AMS productivity, last year, we launched our agency integration technology xSELerate.

Year-to-date, approximately 19% of new commercial lines premium was entered through xSELerate.

Two, One & Done expansion.

The substantial new business growth in the quarter was due to enabling more of our existing products for One & Done issuance, expanding the pipeline and the successful implementation of Decision Support modeling.

With the expansion of this system, we anticipate an additional $25 million in annual premium opportunity.

Three, agency appointments.

We have appointed 37 new agents, bringing our total agency count to 803 or 1,620 storefronts.

We expect a new agent to generate about $1 million of new business in a three to five-year timeframe.

As part of our Greenfield commercial lines expansion into Massachusetts, we expect to appoint at least 25 agents this year.

Massachusetts represents a $5 billion commercial lines market opportunity.

Four, producer program.

In support of our excellent agency relations, to date, we have assisted agents in hiring 45 new producers.

Our 2007 goal is to hire 60.

Training these producers to succeed is also being provided.

Five, the personal lines growth plan is centered on the knowledge-based MATRIX pricing model for automobile.

We are seeing traction in new personal automobile business quoted and written through MATRIX, particularly outside New Jersey where direct premium written grew by 8%.

Later this year, personal lines will be expanded into three states within our footprint: Rhode Island, Minnesota and Iowa.

We believe our strong agency relationships and targeted growth efforts I just described will allow us to meet our goal of mid single-digit growth in 2007 despite the slower than expected first quarter.

This would exceed A.M.

Best's recent expectations of no growth in industrywide net premiums written for the year.

Now I will turn the call back to the operator for your questions.

(OPERATOR INSTRUCTIONS).

Greg Peters, Raymond James.

Good morning, everyone.

Morning, Greg.

Let's see, I will just focus on two areas.

First, you know, you talked about the non-catastrophe property losses and I am not sure if part of that flowed through the BOP combined ratio line, but I was looking at the BOP combined ratio and the trend is somewhat interesting.

I mean it bottomed out as a, on a combined ratio basis, around 83% in the first quarter '06 and it has been pretty much going up ever since.

I am curious if you think there is anything going on there or a problem within that specific product line.

I would tell you, Greg, right off the -- 14 of the large losses in that line -- in that total that we mentioned before were in the BOP line and that was about $3 million in the first quarter of 2007 and that was up against large losses a year ago of almost -$600,000.

So the big swing in the BOP performance is -- part of that reflects that severity of the $11 million, $11.5 million that Dale mentioned earlier.

That's the big movement in that line.

But the fourth quarter was -- you ran a pretty high combined ratio there as well and you know, I am just looking at the consecutive trend where the second quarter '06 to third quarter '06, it is running around 100 and then we jump up to 125 in the fourth quarter and now we are 119, 120 in the first quarter.

Are you seeing anything in your data that suggests there is something other than just some anomaly losses that are flowing through there?

Yes, I would say, first of all, you have got somewhat of the still turnover of all the underwriting initiatives we have deployed over the past several years that were in the line.

Now that we have almost stabilized the line, it still needs to grow to be able to I think better absorb some of the larger losses that we see from time to time.

But I would say that a lot of the severity in the line has been mostly driven by just large property losses for the most part.

I don't know whether anybody else has anything to add to that.

Yes, this is Ed Pulkstenis.

You know, the line of business is really one of our smallest commercial lines.

It is only about $13 million in premium for the quarter, so just the presence or absence of a single large claim, say $1 million claim, you know, which will occur from time to time, will swing that combined ratio by about eight points.

So you know, you will see volatility over time.

I will say that when we have quarters like that, we do go back through a very disciplined process.

We look at all of those claims and to the degree that we have a weather event or you have, like I said, a single large claim, you do see that volatility, but in those reviews, we have not seen any indication of underwriting deterioration or any other trends that would cause us concern.

Okay.

I mean your comment regarding size on a relative basis I think is very appropriate.

The other thing -- you know, I was listening to your prepared remarks and you talked at I think a couple of different points about productivity improvement, etc.

and I was looking at your statistical supplement and one of the stats you throw in there is net premium written per employee.

One of the things I am trying to square or foot here is you talk about productivity improvements, you talk about technology spends and yet, at least if I am reading it right, it looks like the net premium written per employee has been slipping.

It looked like it was 845 last year in the first quarter and slipped down to 827 by the year-end and then according to the supplement, was down to 800 at the end of the first quarter of '07.

And we're not seeing much movement in the top line, so I am wondering what is going on there and how does that foot with your commentary regarding technology spend, productivity improvement, etc.?

Sure.

What we are talking about in that regard is the productivity improvement with respect to our AMSs, the 86 people that we have in the field and what we are trying to garner in that is our average AMS writes, $2.4 million in new business a year and by increasing their productivity through xSELerate, and let me just explain the process for a second because this is where the big productivity gain comes and where you will see it ultimately in the top line is that a good portion of an AMS's day is spent entering business in some cases.

Now that we have the technology to seamlessly move that business from both agents who have Applied and agents who have AMS systems, they now no longer have to enter that business.

We just went through a demonstration the other day where a large 30 fleet of vehicles and drivers was entered into our system in a matter of minutes.

You know, that just saves an enormous amount of time inside the agency, as well as in some cases when that business is flowing over to us where we are entering it.

That is some of the productivity that we are referring to in that regard.

Our goal would be to try to push our average AMS volume from $2.4 million much higher.

So if we could get that volume to $3 million, for instance, think of the additional opportunity we would have to grow the top line.

With respect to the overall productivity drop, you are absolutely right and we measure this very closely.

Our fiscal year premium per employee is now at $800,000.

I think to some extent that reflects a much slower than expected top line.

But let me also comment that all of the new -- for the most part, our new business is up.

You know, all of the things that we basically have been working on for years now in terms of generating new business, new business production are there.

What caught us in the quarter I think was a much more competitive marketplace where we lost some large accounts and generally when Jim and I and other people, Chuck, talk to agents, universally you hear about the $50,000 to $150,000 style accounts that they lost this year and one or two or three accounts here.

Almost every agent that we talked to, we hear that kind of story.

I think that is where -- you know, put some of the lack of momentum in our growth for the quarter was some of that on the renewal side and then obviously, as I mentioned in my comments, we are now no longer expecting any renewal price increases in commercial lines.

So you know, some of the compression in the top line is a reflection of all of those things.

We feel very confident that we have the right tools out there for our inside underwriters to make decisions on renewal accounts.

We continue to fortify that.

We have got some very active growth in new business holistically in the Company.

We have got a lot of new agents coming on.

We have the greenfield strategy in Massachusetts that we will be working through and we still feel comfortable with our overall growth.

But I have got to tell you, the market change -- and it got a little bit, I say, a little bit irrational, particularly on the larger accounts.

I mean we had somebody that took an account that we believe to be $200,000 less than our technical premium levels or which you guys would call it our pure premium levels.

When you start losing big accounts where somebody is writing it below what our actuaries are telling us with the loss costs are pure premium are, I mean we can't compete on that anymore.

We feel we have got the right tools for our folks out there to make those decisions and I think that is what we are trying to do.

Just as one follow-up and I appreciate all the color you have provided there, Greg.

You know, if I look at just the decline, the rate percentage decline, it looks like the net premium written per employee is just growing a little bit faster than the top line and I can appreciate how the change in mix might affect that.

But I am just curious, as we continue on in this soft, softer market type of environment, where do you think that number might go down to if you have any sort of idea?

Better said, does it reach some point where it starts to raise alarm bells internally and you start to think about other structural changes you might make to the Company to keep that number from going any lower?

We are always looking at all those factors consistently.

So I would tell you that when that premium written per employee dips at that $800,000 level, we are definitely reacting to that number inside.

Excellent.

All right.

Thanks for your answers, guys.

Mike Grasher, Piper Jaffray.

Good morning.

I wanted to first ask about workers' comp.

I mean there's been dramatic improvement here.

Where do we go from here I guess?

Have we seen all the improvement and expect sort of the plateauing in here in terms of the combined ratio or is there more on the horizon?

I will just say that generally speaking, you are seeing obviously for the first quarter a little bit lower combined ratio than you will see as the year progresses partly because of -- we do -- there is a lot of seasonality in that line in terms of it's a high premium volume in the first quarter.

So on an accident year basis, we expect that to run somewhere around 102 for the year and when you look at that improvement year-on-year, you are seeing a substantial amount of core improvement based on the many initiatives we have, our predictive modeling that Ed articulated, a lot of the things that we have going on in our claim operation and then a holistic approach in terms of what we are doing on audit premium and other aspects of that line.

And last year, that number had a certain amount of adverse development in it.

This year, it has got a couple of points of favorable developments, so we expect that line to run this year in the 102 range and we believe that as we move out into 2008 and 2009, you will see ongoing improvements as a result of the structural change we are making in terms of the type of business that we are writing and all of the other controls that we continue to instill in that business to lower the loss ratio and lower the frequency and improve the profitability.

Okay.

That's helpful.

And then just to follow up on I think it was Ed's comments just with regard to the commercial package and commercial auto and some of the things you may be doing there.

As you sit there today or sit here today, do you anticipate seeing sort of the same impact on those lines as what you have seen on the workers' comp?

I'll let Ed comment on that that because he has been monitoring that pretty closely.

Yes, I think that we certainly have seen very good traction in the workers' comp.

I think in workers' comp, we had, you know, profitability that we wanted to address and predictive modeling gave us a tremendous tool, much better tools than we've ever had before to address that, along with other initiatives that we put in place.

As we go forward, we do expect continued improvement similar to what we have seen in the past on the package model and the auto model as those are deployed.

At the same time, we expect that not only will it help us to continue to monitor and continue to just improve that profitability, but it will also help us to continue to grow.

Predictive modeling gives us the tools that we need to really know exactly what the right price is at a very fine level of account detail.

So we are able to put out our best pricing upfront for the accounts that deserve it the most and so we expect to see that benefit as well as we roll forward with those lines.

Okay.

Thanks for that.

And then Greg, just a couple of follow-up or other questions here.

I am wondering if you could elaborate on the comment in the press release regarding the slowing economy impacting growth of existing accounts, maybe what you were seeing and where you were seeing that?

Yes, that was mainly in our Selective HR Solutions book where we normally get a fair amount of economic growth within our existing customer base and we are no longer seeing that from Selective HR Solutions and that is really -- a part of their business really, a good part of their business is also in Florida, so it is kind of disaggregated in terms of where we are.

In addition, what we were referring to as well -- normally we see a good pace of endorsement activity and our endorsement activity was down about $3 million and you know, that is another thing that affected our top-line growth and we attribute that to maybe slower, softer sales, you know, softer hiring and other aspects that's affecting maybe the business that we write.

Those are just two of the areas that's evidencing itself in some of the top-line revenue.

That's interesting.

And Florida is the primary here then that you speak of?

Well, in Selective HR Solutions, a good part of their business is Florida-based, yes.

Okay.

And then one other question, I caught the comment about moving into personal lines up in the Northeast.

Have you reflected upon or was there much discussion around what impact that might have in terms of volatility to earnings as we move forward as I suppose that book continues to build or becomes a larger part of the pie?

Yes, and I would say that really we expect - you know, and again, we are so tilted heavily in the commercial lines at almost that 86/14 level that we expect the ebb and flow of that, a better tuning of that could be more in the range of 80/20 and you know, now that we have the same kind of dedication into personal lines that we have had for commercial lines in terms of service center capability, having the absolute right pricing tools and getting that entire operation better fit for more seamless processing that we expect that that actually may better balance the ebb and flow of our earnings and actually reduce the overall volatility of the organization over time.

And again, we are looking more at something in the 80/20 style balance.

So it is not going to be tilted so far over that we will be so cat exposed that we will actually increase the volatility.

We are actually better managing our catastrophe exposure now, better managing our inventory in total to make sure that we are tuning that level so that business is actually going to get smaller and be a more profitable operation and then we are going to continue to grow it.

We are still working through the gyrations of New Jersey and that is really affecting our ability to grow our business in New Jersey because we are repricing our entire inventory and that has been problematic with our agents, but that is the only way that we really can deal in the personal lines marketplace in New Jersey.

I have got to tell you, I mean we have been with a lot of agents.

We have seen our pricing tools relative to all of -- most of our heavy competitors including Geico, Travelers, Allstate and others and our pricing stands up on new business very well.

It is just that we have to work through the noise of repricing business inside an agency's office and we are going to be done with that for the most part by August.

Okay.

Very good.

Thanks for your commentary.

Best of luck.

Thanks.

(OPERATOR INSTRUCTIONS).

Doug Mewhirter, Ferris, Baker Watts.

Hi, good morning.

Good morning.

Do you have any comment or idea of the magnitude of the damage of that northeaster that blew through the Northeast U.S.

during the quarter?

That was a second-quarter event and we are still accumulating data on that and we will have losses, but since it was also a flooding event, we will also have some level of income from our claims handling in our flood business, but we don't have data for release at this time.

Okay.

Also with regards to the -- I am assuming there was a net favorable development overall.

Do you have an overall number for your entire book of what the magnitude and the combined ratio effect was for this quarter?

It was $4 million favorable development for the quarter, which is 1.1 point on the overall combined.

Okay.

And last question, more of an underwriting-related question.

You talked about how your renewal rate for I guess one diamond customers was very low in the 30%.

I am assuming that you renewed any of your one diamonds at all is because you offer them a very high price and they took it.

There wasn't any other reason that -- maybe because it was part of a package or something like that.

It was just because you offered them -- they took your very high price?

This is Jim Ochiltree.

No, actually, we don't write monoline workers' compensation.

Very, very little bit of that.

So if that was -- if it was offered, probably it was because it was a part of a package account where the other part of the account was very good and we wanted to retain it.

Overall, the account we felt could be profitable.

And that was -- I think the number was 31% or 37% --

39% of low-performing --

39% non-renewed.

Non-renewed.

Okay.

So the way that the -- the way your workers' comp underwriting system works, it is not -- it doesn't have the final vote on the account, but it influences directly the overall profitability of the package and if that -- and if that sort of pushes it over the line then you may not renew it.

Whereas before, workers' comp hadn't -- underwriting decisions had less of an influence on the overall package.

I would say --

Is that the layman's way of explaining in?

I don't know.

I would say we are better equipped to make better decisions today.

I think that work comp has always been a part of it.

It's just that we have a much finer level of information to decide whether or not we can make money on the whole thing.

And there are different reasons why an account can score different ways.

We look at the reasons for the score, as well as just the pure score, so the model itself does not make decisions.

People take this information and they mix it with everything else we do and they make the decisions, but they are much better equipped to do that today than they were a year ago.

And I would kind of like to tie that into a thematic that came out in Ed's comments too in terms of the differences in the lines and the approach of the modeling because I think it will affect our ability to even grow more as we move through the year.

In comp, we had to take a harder stance somewhat on our worst-performing business and we did that because we needed to bring our combined ratio down and so that is why you are seeing non-renewal rates in the 40% range.

So in a sense, we are almost non-renewing almost half of that business and we are doing it because we need to improve the profitability of that line and we need to bring it down quickly.

That is why you are seeing -- that is one of the byproducts of why you are seeing a combined ratio today of the 98s in that line.

That is not the sole reason, but there are a number of things.

The attack that we are going to take on -- and you have got to remember, we write an agent's entire book of business.

We look at an agent's entire profitability.

70% of our agents are profitable.

In some cases, you know, we need to do some things for our agents as part of the holistic relationship.

I think as we sit and look at our information now, now we know better where we are doing that, how we are doing that and make sure that we are doing it for our best agents that are committed to us and giving us the best business.

In terms of these other models, they are really built around us growing our business now and growing a lot faster than we ever have before.

And then more tactically though, we will be able to take some approaches on the 30% of our agents that aren't profitable and then look at that and say, well, why aren't they profitable.

Let's do the segmentation of their business and now we will be much more surgical in our ability for a branch manager and an AMS to walk into an agent's office and say, look, you are not profitable with us.

These are the areas where we see that you are not profitable and now you can be very pinpointed in that discussion.

So as an entrepreneur that is being paid, the AMS that is being paid almost 40% of their compensation is based on the underwriting profit volume that they generate, they now have a much better tactical way to go in and improve the profitability for their underperforming agents.

And so that is how we are really using the other line's information and it is not like some wholesale reunderwriting of our book.

That sends chills down the spine of every agent when you have that conversation and we are just trying to be -- we were much harder on the comp because we needed to bring that down, but the other lines will be much more a new business story, as well as a very tactical renewal story for agents that aren't generating the long-term profitability we expect.

So that is how we are kind of using this information, but the business that we are growing now is going to be diamonds three, four and five.

It is going to be profitable business.

It is going to be priced correctly and where we are not doing, we will know why and for whom.

And then our monitoring tools up here to know who is not pricing business are incredibly focused, so our folks can dial in on what AMSs might need some supplemental training or other issues that they see on accounts that they -- that we might not have ever been even able to find before or would have found them only by happenstance.

Now, we are actually directed to those accounts and that is what we're focusing our attention.

So the educational opportunity out of this is tremendous as well.

Thanks very much for your answer.

[Carolyn Spiers], Fox-Pitt Kelton.

Hi, this is actually Dan Baransky calling.

I have a couple of questions.

On the other expense line item, I know the first quarter is sort of heavy for you on the compensation line.

Is there anything else we should think of that was impacting that this quarter?

The other item there was last year in the first quarter, we had some bad debt recoveries of about $500,000.

Whereas this year, we have a more normal bad debt expense.

It is not like it's anything big, but it just makes for a swing in the quarter of about $1 million.

So you have got the other expenses are up by $2 million.

Half of that increase is on the bad debt side.

And again, it is a very small kind of item, but it is just significant between the two quarters.

Then the other $1 million is the fact that both retirement-eligible individuals and our Board of Directors members as they get an annual stock compensation award, that has to be fully expensed.

So you see a heavier expense load in the first quarter.

Okay, great.

Was there any sort of -- was there any reserve impact in the first quarter of '06 so we can sort of look at an apples-to-apples accident year ratio?

You had a $600,000 favorable development in the first quarter of '06 compared to $4 million favorable development in first quarter of '07.

Okay.

Great.

The other thing is more ethereal.

Since you don't write any monoline workers' comp, is there any correlation between accounts that have unprofitable workers' comp and sort of the rest of the account or is there really no correlation there?

Sure.

This is Ed Pulkstenis.

That is information that we track and there is some correlation.

What we have found is that for workers' comp accounts that are unprofitable, a fairly good portion, probably about half of those accounts tend to be undesirable on the package side as well.

And that is probably due to the fact that some of the same management issues that caused the problems on the comp side are caused on the package side as well and then there is about half of the premium that would go in the other direction.

The comp is tougher, but the package is still pretty good.

So is there a sense as you have whittled away more of the unprofitable workers' comp and are sticking with the more profitable that sort of the overall book of business will benefit?

There is probably a very small effect -- that may be happening to a very small degree.

What I would say is that where we are really going to see the leverage is with the deployment of our package models because now rather than you know just lying on the workers' comp scoring, you are actually going to have the scoring for the package and so I think with that, you will see the very positive effects that we have seen on the comp, that we have seen in the BOP carrying over to the package in a much more direct way going forward.

Great.

And the last thing I had is you know you have sort of mentioned that you are underpenetrated in states, sort of your non-core states and I was wondering if you have a sense of what has the growth been like on a net run basis in sort of your non-top five states you know in the quarter or how have you been experiencing sort of in your non I guess core states we would say or your top states?

Yes, I could go through that a little bit.

I mean it is somewhat mixed.

I mean I think we are starting -- this was a tough quarter holistically so it is not a good benchmark, but my sense is from when -- what we talked to our folks, you know we are starting to get a lot more in our southern territory.

I think now that we have split the Mid-America into three different configurations, three sets of three states each, we are starting to see a lot more traction there.

But nothing obviously is really that evident in the number -- in terms of the numbers of the growth in every one of the regions, but I will let Jim and Chuck add any kind of color to that at this point.

This is Chuck Musilli.

I think the other future indicator of growth outside our core territories would just be the increase in AMS marketing territories that we have outside of New Jersey even since the end of the year and a lot of that increase in putting field people out there is based on the market planning information that we are utilizing to understand where our true growth opportunities are out in the market.

And if you take New Jersey out of the picture, we have added 14 new AMS territories.

Actually -- I'm sorry -- 12 new AMS territories in our existing footprint since the end of 2006 and while that doesn't have an immediate impact, what it does is it increases your future potential for growth in those territories.

So I think -- you know, as Greg mentioned, our regional reorganization.

Part of the impetus for that was to allow our leadership to get closer to the agents and the markets out in those states and I think that is beginning to manifest itself with the addition of these new territories.

Let me just add to that too.

I mean I would say and we spend a lot of time with agents all over our core and our core now soon to be 21 states and I would say that our brand continues to -- it continues to be stronger outside our core states and even in a state like Massachusetts where we have been out talking to some agents and harvesting some good positive leads, we are starting to get a number of calls inside because the word is out that we are coming into Massachusetts and that they want -- good agents want to have a Selective appointment in that territory.

So we are encouraged about the recognition of Selective and obviously as you crack some of these new markets, you know having the A+ rating by AM Best is really critical to these folks.

Just to give you an idea in some of our Mid-America regions, the Great Lakes regions is up 14% and the heartland region is up 10%.

You got a little bit of decline in the Big River region.

The Mid-Atlantic region is up 1.9%.

Some of those heartland regions that we just recently broke apart, obviously those numbers are off of small bases or smaller bases, but the overall heartland region -- if you look at the Mid-America region the way it used to be is up 5.7% for the quarter.

So we are showing some nice growth out there.

Okay, great.

That's all I have.

Thanks.

(OPERATOR INSTRUCTIONS).

Robert Farnam, KBW.

Hi there.

Just another question on the reserve development in the quarter.

Just can you give more of a breakdown by line and by accident year kind of the key leverage points as to what was causing the development, the favorable development?

Yes, I don't have it by accident year breakout, but we have got -- in the workers' comp, you have got about $2 million of favorable development.

General liability, you have got $3 million of adverse development.

Commercial, auto and personal auto both had $3 million of favorable development and the personal umbrella had $1 million of unfavorable or adverse development.

So that all totals out to the $4 million positive.

You don't have a breakdown by accident year?

No.

Okay.

That's all I had.

Thanks.

There are no further questions at this time.

All right.

Thank you, operator.

If you have any follow-up items, please get to Jennifer and Dale.

Thank you very much.

Thank you for participating in today's conference.

You may disconnect at this time.