Amerisafe Inc (AMSF) 2010 Q3 法說會逐字稿

Good morning, ladies and gentlemen. Thank you for standing by. Welcome to AMERISAFE's third quarter earnings conference call. During today's presentation, all participants will be in a listen-only mode. Following the presentation, the conference will be open for questions. (Operator Instructions) This conference is being recorded today, Friday, November 5, 2010. At this time, I'd like to turn the conference over to Mr. Ken Dennard with DRG&L. Please go ahead, sir.

Thank you. And good morning, everyone. We appreciate you joining us for the AMERISAFE conference call to discuss third quarter results. We'd also like to welcome our Internet participants as this call is being simulcast over the web. Before we turn the call over to management, I have the normal housekeeping details to run through.

You could've received an email of the earnings release yesterday afternoon, but occasionally there are technical difficulties. So if you didn't receive your release via email, or you'd like to be placed on the email distribution list, please call our offices at DRG&L. That's 713-529-6600. Also, there will be a replay of today's call. It will be available via webcast, by going to the Company's website. And, of course, that address is AMERISAFE.com. Also, there'll be a telephonic replay for seven days, 24 hours a day. And that information is in the press release. Please note that information on this call speaks as of today, November 5, 2010. And therefore, you are advised that time-sensitive information may no longer be accurate as of the time of any replay listening.

Also, statements made in the press release or in this conference call that are not historical facts, including statements accompanied by words such as will, believe, anticipate, expect, estimate, or similar words, are forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995 regarding AMERISAFE's plans and performance. These statements are based upon management's estimates, assumptions, and projections at the date of this call, and are not guarantees of future performance.

Actual results may differ from these results expressed or implied in the statements as a result of risk, uncertainties, and other factors, including, but not limited to, the factors set forth in the Company's filings with the SEC, including AMERISAFE's 10-K for the year ended December 31, 2009, and future and other filings. AMERISAFE cautions that you do not place undue reliance upon forward-looking statements contained in the release or in this call. AMERISAFE does not undertake any obligation to update or publicly revise any forward-looking information or statements to reflect future events, information or circumstances that may arise after the date of the release and call. For further information, please see the Company's filings with the Securities and Exchange Commission. Now, with that behind us, I'd like to turn the call over to Allen Bradley, the Company's Chairman and Chief Executive Officer. Allen.

Thanks, Ken. And good morning, ladies and gentlemen. Thank you for joining us for our third-quarter 2010 conference call. As usual, I'll make a few comments about the quarter before turning it over to Geoff Banta, our President and Chief Operating Officer, who will in turn introduce Janelle Frost, our Chief Financial Officer, for additional details.

The third quarter was disappointing for us. Just as we're starting to see some signs of pricing stability in the market place, we've experienced a deterioration in our current accident year claims experience. As a result of this experience, we raised our current accident year loss ratio to 81.2%. This increase is due to a rising frequency of lost time claims for $1 million of earned premium, as well as a number of severe claims experienced during the quarter. Consequently, we reported a net combined ratio for the quarter of 101.7%. This is the first quarter since becoming a public company five years ago that AMERISAFE has failed to generate an underwriting profit. And clearly, this is not an acceptable result.

In my opinion, the significant external factor that is driving the increase in frequency of claims per million of earned premium is a cumulative impact of multi-year reductions in lost cost of premiums. According to the National Council on Compensation Insurance, the seven-year cumulative lost costs, or premium reductions, totaled 26.7% countrywide. In some states, that decline has been much more. Exacerbating this lost driver has been a reduction in pricing, reflecting the excess capacity in our industry, as well as reduced payrolls as a result of the national economic circumstances, particularly impacting our targeted industries.

While AMERISAFE has strived to avoid much of the pricing competition, we have reduced pricing levels over levels of 2007 and prior. That decline in pricing is now at an end. We have launched a number of initiatives designed to improve our risk selection, and boost our pricing. For competitive reasons, I will not provide specifics on these initiatives. But I can say, that we're taking actions impacting where we write business, what risk -- what business we will write, and how much we charge for those risks.

Earlier, I mentioned there were a number of severe claims during the quarter. Let me give you some color around that. As of September 30, AMERISAFE had received 26 claims during the 2010 accident year with incurred losses of greater than $500,000. I picked that number because in our 10-K, we identify claims with a $500,000 or greater limit as severe claims. 17 of those 26 claims occurred in the third quarter. And 14 of the 17 claims that occurred in the third quarter happened on or after August 2. In fact, we had one very bad day on September 15 where three claims in that category happened on the same day in three different parts of the country. While not unprecedented, this rash of claims was disappointing.

On a positive note, there are signs that the industrywide pressure on pricing is easing, not going away, but becoming less intense. Additionally, prior-year loss development for AMERISAFE is continuing to be very favorable, as reflected in the $5.3 million favorable prior-year development. Finally, AMERISAFE remains vigilant in managing our underwriting expenses. With those initial comments, I will pass it on to Geoff to discuss our operational metrics in greater detail. Geoff.

Thank you, Allen. And good morning everyone. I'll make a few comments about our overall operating performance before turning things over to Janelle. From an operating standpoint, we had our first quarterly underwriting loss since becoming a public company. In reinforcing the comments Allen just made, we attribute the loss to the continued decrease in our premium writings, and to an increase in lost time claims that were reported to us in the third quarter. In turn, both of these factors led to a substantial increase in our premium-based claims frequency, a trend that we saw emerging and gaining momentum in the first and second quarters. We are naturally disappointed at having to report an underwriting loss, but we are also committed to taking steps that we believe will return us to underwriting profitability.

I'll now take you through the major components of our operating results for Q3 2010, beginning with premium. In the third quarter, we experienced a 13.8% year over year decrease in our deck sheet premium for voluntary workers' comp, a decrease that amounted to $9.1 million. Counterbalancing this lower deck sheet premium was a significant slowing in our negative premium adjustments, including EBUB. The net of these two factors was a decrease in growth premiums written of 5.3% for quarter three 2010. This is the lowest year over year decrease in our top line since the first quarter of 2009. Small consolation I realize, but improvement nonetheless.

In terms of new business after a 4.4% increase in new deck premium in the second quarter, our third quarter new deck premium decreased year-over-year by 12.8%. This decrease was the result of a lower new business quote rate, and a lower average premium for new business.

In the third quarter we increased our pricing slightly to an effective LCM of 1.42, or 142% of the approved loss cost in the states that use that mechanism for pricing. This compared to an ELCM of 1.41 in the second quarter, and 1.44 for the year ago quarter. We study our pricing and underwriting every week, looking for opportunities to write new business in states and industries where our performance has been good, but we also increase pricing if we find states and industries that are not performing well. As a result of our underwriting results in the third quarter, we have taken more pricing industry -- increases than decreases, and we expect our pricing to continue to increase throughout the fourth quarter.

In terms of renewal business, our retention results continue to be excellent. In the second quarter, our policy retention was 93.2% versus 92.6% for the third quarter of 2009. By the way, this 93.2% retention level is the highest in our history. As a result of the strength of our policy retention, our in-force policy count has increased 4.7% over the last 12 months. We are proud of the fact that we are retaining the great majority of our customers in the current, very tough market environment.

Regarding losses. As a result of the third quarter increases in claim frequency and percentage of lost time claims, we increased our 2010 accident year loss ratio estimate from 74% through six months to 81.2% through nine months. On the positive side, we experienced favorable loss development for accident years 2008 and prior during this third quarter. In her comments, Janelle will provide further color around our loss ratio and its components.

Reported claims were up significantly over the year ago quarter and for the nine-month period. Year-to-date reported claims in 2010 are up almost 10% over the same period in 2009. In terms of severe loss experience, we had 17 claims in the third quarter with estimated incurred of more than $500,000 versus nine severe claims for the first and second quarters combined. Seven of those claims estimate -- have estimated incurred totals over $1 million. The largest incurred total for any single claim in Q3 2010 stands at $2.1 million. Similar to the second quarter, our claims closure rate was down slightly in the third quarter year over year.

On the other hand, our paid to incurred ratios for the three most recent accident years remain at very low levels, which we view as an indicator of case reserving adequacy. That concludes my comments. I'll now turn this discussion over to Janelle for details regarding our financial performance.

Thank you, Geoff. And good morning, everyone. In the third quarter of 2010, AMERISAFE reported net income of $4.4 million, $0.23 per share compared to $15.1 million, or $0.74 per share in the third quarter of 2009. Gross premiums written in the third quarter of 2010 were $52.2 million compared to $55.1 million in the prior year period, a 5.3% decrease. That represents the smallest year over year decline in quarterly growth premiums written that we've seen in quite a while. Net premiums earned decreased 6.4% from the year ago quarter as a result of declines in gross premiums written over the last 12 months.

Our net investment income totaled $6.6 million, a 4.5% decrease in the prior year period. Cash and invested assets decreased slightly from the year ago quarter end to $806 million, primarily due to cash used in our share repurchase program. The tax equivalent yield for our investment portfolio was 4.6% compared to 4.7% in the prior year quarter of 2009.

In total, revenue in the third quarter of 2010 was $60.5 million, down 9.9% from $67.2 million last year. Our current accident year loss ratio increased to 81.2% for 2010, which resulted in a 95.5% loss ratio for the quarter. This compares to 69% in the third quarter of last year, and 73.8% for the full year of 2009. Our incurred loss and loss adjustment expenses totaled $46.7 million, which included $5.3 million of favorable prior-year development, primarily attributable to accident years 2006 and 2007. This was partially offset by unfavorable development for accident year 2009. In comparison, incurred losses in last year's third quarter were $33.4 million, which included $6.7 million of favorable prior-year development.

Total underwriting and other expenses decreased 33.3% to $8.4 million from $12.5 million in the third quarter of last year. Expenses included $5.1 million of salaries and benefits, $4.1 million of commission, and favorable underwriting and other costs of $834,000. Underwriting and other costs were favorable, primarily due to rate reductions in certain loss-based assessments. The expense ratio declined to 15.4% from 21.6% in the same period a year ago. In total, our combined ratio was 101.7% in the third quarter of 2010, and 93.4% for the nine months ended September 30, 2010 compared to 79.3% in the third quarter of 2009, and 84% for the first nine months of 2009.

Return on average equity for the third quarter of 2010 was 5.5% compared to 19.3% for the third quarter of 2009. Book value per share at September 30 was $17.26, an increase of 8.3% from $15.94 at September 30, 2009. Keep in mind, not reflected in our book value or earnings were $35.2 million of net unrealized gains on our investment portfolio, classified as held to maturity.

In the third quarter, we repurchased 213,000 shares at an average price of $17.51, including commissions. As of September 30, 2010, we had spent $10.4 million of the $25 million authorized in our share repurchase program. However, on November 2, our Board extended our share repurchase program to December 31, 2011, and authorized a new limit of $25 million, effective October 1, 2010. Finally, statutory surplus at September 30 was approximately $312 million. That concludes my prepared remarks on the financials. I will now turn the discussion back to Allen.

Thanks, Janelle. In my initial remarks this morning, I shared my concerns regarding state mandated lost cost, and the impacts of the economic recession, as well as competitive forces that are impacting our financial performance. To be sure, these factors do impact our performance. But I do not offer these to you today as excuses for these quarterly results. All of our competitors participate with us in the same marketplace, with the same loss cost, and on many of the same accounts. Ultimately, it's our responsibility as the management team at AMERISAFE, to ensure that we do our best to select and price risk appropriately, regardless of mandated loss cost, competitive pressures, or economic conditions.

In that regard, this quarter was probably a bit of a wake-up call. And I can assure you we are wide awake. We're confident that the underwriting and other implemented initiatives that we have launched will return AMERISAFE to underwriting profitability. With that, let's open the call for questions.

Thank you, Sir. (Operator Instructions) Our first question comes from the line of Beth Malone with Wunderlich Securities. Please go ahead.

Okay. Thank you. Good morning. A question on this trend of losses that you experienced in the third quarter. How does that change your assumptions, or does it change your assumptions -- actuarial assumptions on the book going forward? Would this require -- does this mean there's less excessed reserve available to release going forward given that you might want to hold more back?

All right if you -- I assume that you're talking about with respect to our remaining open claims, Beth? How we look at those in terms of development?

Yes. Thank you.

It's been very interesting as we look back at those. The 2008 and prior years are developing very nicely where a lot of the reserve development you see is coming out of reductions in case incurred developments. We do not see the extension or as much elongation or duration expansion in those older claims as we do in the 2009 and 2010 year. So the increase in reserves on 2009 and 2010 to a [75.7] I think for 2009 accident year, and then [81.2] for 2010 reflect what we feel are the dynamics in those claims. We don't think we're going to have -- we believe that our reserves for all prior years are very adequate.

But I'd like to add, Beth, that Allen mentioned 2008 and prior. I'd like to add that even in 2009, still a very green year, that from a policy year standpoint, is not completed. We have seen versus past years extremely slow case development given the greenness of that year. So it could be that -- my belief is that that trend that we see in 2008 is continuing even in 2009 and even in the fact that we increased the reserve in 2009 just a bit. We are seeing development that for an accident year, that green is fairly unprecedented.

Okay. And then one other question. On the impact of the economy and unemployment, can you explain if someone goes into -- on a workers' comp claim or whatever and there's not a job for them to go back to because of the economy, is that going to lengthen the workers' comp claim?

The workers' -- the determination of a continuation of the duration of a claim is largely a medical consideration, not a question of whether or not they have a job to go back to. Having said that, we need to be a bit more realistic about what happens in the marketplace. With some claims, you have soft tissue claims, you have claims that are difficult to quantify from a medical perspective. And so there does tend to be an elongation. You have to have more doctor visits and there's more prescriptions, and then more indemnity payments. That's the duration I was speaking of, Beth. But strictly speaking, the length of a claim is not dependent upon whether or not the injured worker has a job to return to.

So to just follow-up. So these changes in this pattern of severity wasn't related to the economy?

The changes -- we have had an increase in the number of lost time claims per $1 million of earned premium. We've noticed that most notably, of course, we had some severe claims. We have severe claims from time to time. It's a lumpy business that we are in because of the severity. But we have noticed an increase in the lower level indemnity claims. And those are the $10,000 to $100,000 range. We saw an uptick of those in the third quarter. Some of that may be reserving increases or reserving pessimism on the part of our own employees. Some of it may be the economic conditions pushing some of those claims into the $10,000 claims. At the same time, we've seen a decrease in the medical only claims that have been initiated. So that, too, kind of pushes that severity out there.

Okay. All right. Thank you.

Yes ma'am.

Thank you. And our next question comes from the line of Mark Hughes with SunTrust. Please go ahead.

Yes. Thank you. Geoff, I think you had suggested that 2009 was still developing in a very favorable way, but you had to strengthen the reserves. Did I hear that correctly? What accounts for the difference there?

Well, we did -- we have seen very benign case development. On the other hand, this being a green year, and what we have seen also in 2010, we were just not comfortable enough to put up IBNR that is more in line with some of our past years. So you've got a situation where you're seeing real reserve takedowns on a case basis that are fairly unprecedented for a year that green, with -- is there still a shoe to drop since we're not even through the policy year. So I realize that sounds a bit contradictory, but on the case side, very, very low development in what's happened in the claims organization.

On the aggregate side where we set bulk reserves, gee, should we be reacting to that when the year is really not done from a policy year standpoint. So I guess we took a -- even though what we saw in case reserving was positive, we said this is too early for us to take this seriously. And we actually bumped up IBNR as a result. I hope that makes sense.

Mark, let me try to add a little bit to what Geoff said. If you take the case development, forget the total reserves, talk about just the case development, the development of 2009 is at levels significantly below 2008 and prior at the same points in time. The paid to incur ratios are much more -- are much lower in 2009 than those prior years at the same points in time. So while we believe the case reserves will probably prove out ultimately to be more adequate, we have to be mindful of the need to make sure that we have adequate IBNR. So that's the change you're seeing.

Right. And how long in the process do you get a good look at that where you have some sense that IBNR is adequate?

I would say 30 months from the inception of the policy year, which will be the June 30 of 2011.

Okay.

Try not to take any drastic favorable moves on any year until it's at least reached that age. And part of the reason for that is the way the statistical reports work on accident years and accidents.

I think you said that reported claims up 10% year-over-year, was that in Q3?

The 10% was nine months to nine months.

What is -- do you have the Q3 comparison?

Yes. 25%, approximately.

Thank you. And our next question's from the line of Matt Carletti with JMP Securities. Please go ahead.

Thanks. Good morning.

Good morning, Matt.

Just had a few numbers questions, actually. The first one is on the tax rate. Just wondering given that the accident year has come up, and so I guess more of the pie seems like it will be coming from investments and munis, should I look at like the nine-month tax rate as the new go-forward level?

That's absolutely right, Matt. The quarter is basically -- when you look at the tax rate, the quarter is basically a catch-up to get you through your year-to-date effective rate, so you need to be looking at the year-to-date.

And similar question on the expense ratio. It sounds like from the second injury funds, the rate assessments, would that be a sustainable savings in that those are set to new levels? And I would imagine that there's some sort of accrual catch-up in the quarter. But again, should I look at kind of the expense ratio for nine months, and that be a new level?

It is. I would say on the nine-month basis, it's probably maybe two percentage points, the rate adjustment on the assessment. Keep in mind those assessments come in every year. It's our two largest loss base assessments, Georgia and South Carolina. And both of those funds are in run-off. So, I mean, obviously I don't know actuarily how they're going to develop. But one would think, hopefully, they're on the decline.

Okay. And then lastly, you mentioned -- I think Geoff mentioned actually the paid to incurred ratios for some of the more recent accident years are at very low levels. Can you -- do you have those numbers handy? I think it was like 2008 and prior?

No, I don't.

Okay. No problem. Thanks alot.

Thanks, Matt.

(Operator Instructions) Our next question is from the line of Bijan Moazami with FBR Capital Markets. Please go ahead.

Good morning, everyone. Did you guys notice any kind of difference between the lost time claims in logging versus construction and trucking businesses?

No, we have not. The logging now is about 4% of our business. And that's for all lumber related businesses. So, I wouldn't say we've seen a change there. The one area of some uptick in terms of claims, and I suspect this may be work activity, though I don't have those numbers, is in the trucking subsegment of our business. In fact, unusual statement. But we had two losses in the trucking subsegment this quarter where individuals were hired in one day and were deceased by the end of the second, the next day, by trucking accidents.

Okay. In terms of the -- why do you guys think that this time around is a little bit different than the experience that you had when you reported your fourth quarter results back in February?

I would say the difference here -- there's a couple of differences that we see as we look back to last year. Number one, last year, when you looked at claims in the fourth quarter, we made significant increases in the third based on claims from $100,000 up. This year, we've had a little bit intense year in the $500,000 and up, but claims between $100,000 and $500,000 have actually decreased. We've had an increase in the number of claims in the $10,000 to $100,000 range. That was the increase of loss time claims. So, as strange as this sounds, at this particular point in time at 9/30, you just look at average case reserves, average case incurred amounts on claims, the average case incurred amount on claims in 2010 are actually less than they were at this point last year by more than 10%. So, you had an increase in frequency, but the average claims are down in terms of size.

The average lost time claims distribution has moved from more the severe claims to, like Allen said, the $10,000 to $50,000 to $100,000 indemnity claims.

Thank you. Our next question's from the line of [John Grimstead] with Piper Jaffray. Please go ahead.

Good morning.

Good morning, John.

Just one question around the 17 severe claims that you all experienced this quarter. In terms of -- any trends around that group of severe claims, states, types of business? How do you all think about this going into fourth quarter? Or are these just kind of isolated geographically diverse different industry claims?

Well, John, that's a good question. One of the areas where we have seen an uptick in those severe claims are trucking. That is true in this quarter. It was very noticeable in the third quarter. None of the other industry groups seem to have any kind of huge fluctuation. The other trend that we saw -- or say trend, this is is something that we see frequently, but it was very noticeable in the third quarter. And that is the average tenure of the workers that were injured in severe ways. Historically in our industry, for our experience, people that are on the job with the current employer for less than 12 months are much more likely to get injured than others.

But as I mentioned with those two trucking accidents, and we see -- we saw repeatedly during this quarter, there were new employees. Employees that worked certainly less than a year. Many of them worked less than 90 days. So, does that raise the question of whether or not you've got some retiring in the workplace? And what are the standards by which people are rehiring? Are they spending time to train people? Are they hiring people that are less expensive, and therefore, less experienced? That's something we're looking at very closely from a safety perspective. And I think, Geoff, you probably have some comments on that.

Sure. And John, realize that we're not necessarily talking about credible data. We don't have enough of this kind of data to draw I'll say a statistically valid conclusion, but it is definitely an anecdotal trend we're seeing. It was strong enough in the third quarter that we have alerted all our safety and underwriting people, and actually changed some procedures to focus on new employee hiring practices and training practices.

We expect that will make a difference if -- but what's interesting to me, and I only speak for myself on this, and the NCCI unfortunately will be probably two or three years late on reflecting this. But if -- and today's employment numbers were very good. If employment starts to go down, and employees start coming back into our industries in mass, there is going to be an uptick in frequency for that, and that criteria alone because, as Allen said, new employees, employees with tenure less than a year, are going to get hurt in the work comp environment more frequently than the more experienced employees. So that's going to be really interesting to watch if the unemployment rate falls in our industries and we see a statistically valid influx in new employees. And we're going to be watching that like a hawk.

Geoff, Allen, thanks a lot.

Thank you, John.

Thank you. Our next question comes from the line of Adam Klauber with Macquarie. Please go ahead.

Thanks, good morning.

Hi, Adam.

How many -- you mentioned you had 26 claims so far this year above $500,000. How many were there last year?

At this point in time, I'm going to say the number is not significantly different. But this year the difference is the largest claim is on a case basis, $3.45 million. And last year at this point in time it was much larger than that.

Okay. And then how much of the increase in accident year lost ratio from 74% to 81% is due to these larger claims?

I don't know that, Adam, that we went through and broke that out as much as we looked at the overall trend of frequency and the more indemnity claims and the burn cost as we call it, you know, the estimated earned, or the actual earned, divided by the case incurred.

Okay. Okay. And then you mentioned you're putting in a range of pricing and underwriting actions. Is your hope within 12 months or so because obviously it takes a while to put those actions in, that the accident year ratio gets back down to the mid-70%s? Or do you think we're above 80% going forward?

Well, I -- we don't want to give forward-looking guidance. I can tell you that there were some initiatives put in earlier in the second quarter that were designed to encourage more profitable business, business that has been good for us long-term. More of the initiatives this time are slanted toward elimination of certain businesses, and certain risk and certain categories for the pricing of those risks. Remember that when we report the effective LCM, that's an index. So what happens to the underlying lost cost is a question. If the underlying lost cost continues to trend down, then even an increase in the effective LCM will still generate the same premium per $100 payroll.

Okay. Thank you very much.

Thank you. Our next question is from the line of Raymond Iardella with Oppenheimer. Please go ahead.

Good morning guys. I guess you guys had mentioned a lot of commentary about the reported claims during the quarter. And I was just wondering if you could maybe talk about the claims that you closed during the quarter? I mean has that number been pretty steady? Or has it been increasing?

The closure rates are down just a bit, Ray. In just about -- looking year-to-year for the last -- we focus on the last three or four years mainly, the closure rates are down slightly. And I think I made a comment about that in our -- in my opening remarks. But not materially, number one. And number two, we would expect the closure rates to be somewhat slowed down because of the difficulty in the return to work scenario, getting people back to work when there are no jobs for them to go back to. So think about the fact that, in the pie of all our claims, we have more lost time claims. And then think about the difficulty in getting those employees with those lost time claims back to work. If nothing else changed but that, we would expect duration to increase. And it has -- at duration, well, we would expect duration to increase. We would expect closure rates to decrease. Closure rates have decreased very, very slightly.

Okay. Thanks. That's very helpful. And then I guess changing gears a little bit. On the buyback program, can you maybe talk about how you think about that I mean sort of as a capital management tool and whether -- what -- how you guys view price, and when you're -- and your decision to buy back stock?

Ray, this is Janelle. I will tell you our primary goal with the repurchase program is not to dilute the common shareholder. And we usually look one to two quarters out in terms of book value.

And I guess is that book value -- is that GAAP book value? Or does that include sort of your unrealized gains on your portfolio?

GAAP book value.

Okay. Thank you guys very much.

Thank you. Our next question is a follow-up from the line of Mark Hughes with SunTrust. Please go ahead.

Yes. The claims up 25% in the third quarter compared to up 1%, I think, you talked about in the second quarter. Can you give us some sense through the month how that trended? It seems like such a dramatic turnaround when you look at the claims frequency and see that kind of movement. Aside from the large loss claims, what do you think is going on? And then have you reached out to other high hazard peers to see what they're seeing?

I'll speak for myself, Mark. I have not reached out to other high hazard peers. We have been analyzing this backwards and forwards, as Allen stated. There may be some phenomenon regarding new employees. There may be -- a lot of it obviously is the decrease in premium. We are still analyzing this. We're still trying to figure out why the frequency is up so much in the second -- in the third quarter. It did accelerate, I want to say this is from memory, maybe Allen has the exact figures. It did accelerate as we moved from July through September.

That's right. I do have the numbers. We'll give you percentages basically. The increase -- most notable increase, Mark, was in August. In August, the losses moved up -- the reported losses moved up about 40%.

Right. Any update on -- you touched on competitors maybe without naming them in times past. But any thoughts on how they're behaving now? Are you seeing them act any differently Q3 compared to Q2?

We've seen, with respect to national carriers, a lot more restraint in terms of pricing aggressiveness. We've seen in various parts of the country even regional carriers backing away from certain class codes, certain underwriting dynamics of certain claims, of certain risk. You will see people make changes with respect to experience rating modifier types of claims. Use that as a basis. Or an industry group. You see a lot of changes in the trucking industry as far as underwriting appetite for that. You see obviously some for construction. There are people that restrict underwriting for heights that they did not previously do. We see some of those across the country. You see a changing appetite by many carriers with respect to new business and what they will accept in terms of parameters for new business.

Having said that, when you get into states and you deal with nonpublic carriers, and smaller carriers, and maybe carriers that are a single state or right in one or two or three states, they still tend to be fairly aggressive. And we've watched that phenomenon. It does seem that there's more stable -- stabilized pricing environment out there right now. But on the claims counts, I want you to understand that we always have an uptick in every year in the number of claims in the third quarter and in the second quarter than the first and the fourth because there are full employment quarters within that year. The differences I'm telling you about were third quarter this year compared to third quarter 2009.

Right. Thank you.

Mark, your point is really well taken about are we going to see any different dynamics in terms of duration, in terms of closure rates, in terms of average reserves on these -- this very high reporting of claims in the third quarter? We're going to watch that like a hawk with this third quarter, in particular, to see if there was anything else in there that says this might have been a different type of claim versus past quarters, even versus last year in the summer. So, we'll be -- that's definitely one thing we want to look at and analyze, and make -- draw conclusions from to see exactly what this was.

Thank you. Our next question is from the line of Mike Grasher with Piper Jaffray. Please go ahead.

Good morning, everyone. Just want to transition the topic here a little bit just around your balance sheet. A very strong balance sheet. Allen, you're talking about some pricing maybe improving here. Are you seeing any signs where maybe some of the competition wilting to the point where there's opportunities to pick up a balance sheet, or some other opportunities out there that maybe earn distribution?

We are seeing a lot of smaller carriers whose leverage ratios have increased dramatically who have actually shrunk their statutory surplus. And we are very mindful of that. We've watched and we've seen stress on self-insurance funds that do not function as well over a protracted soft market. We are seeing -- some underwriters, quite frankly, are not as optimistic about the workers' comp business that write multi-lines, Mike. They have workers' comp business they are not as comfortable with. So we are very interested in doing renewal rights transactions. Our parameters for those sort of transactions would be industries which are consistent with our targeted industries. Their pricing structure which would give us the greatest opportunity that they are either at our level, or close to our level. So we'll have a good opportunity to renew that business.

Geographic distribution, while it's not essential to us doing a renewal rights transaction, if we can do one in an area where we have an existing infrastructure, that is all the better. And policy size. We want those that we can apply our business model to. Additionally, we would be interested in lost portfolio transfers in those situations where we felt that there would be an opportunity for our claims department to bring some of the medical costs containment, and other cost containment mechanisms to the forefront.

Okay. That's very helpful. It sounds like just an outright balance sheet acquisition in this environment just would not make sense. Fair to say?

We might do an outright corpus acquisition. It's going to have to be favorably priced. In this environment, there is probably an enhanced risk of assumption of anybody's reserves. And, you know us, Mike. I mean we're not really interested in betting the farm on any transaction. So, we don't want to take somebody else's troubles on our balance sheet.

Understood. Thanks very much.

Thank you. Our next question is from the line of Robert Paun with Sidoti & Company. Please go ahead.

Good morning.

Good morning.

Your policy retention rate remains pretty high, over 90%. Do you expect that rate to come down as you sort of review and re-underwrite some of the book?

I will just tell you I do. Hopefully not materially because we feel like if the market turns -- if we've got -- if we're in good shape both from the standpoint of in-force policies, and a relatively high effective lost cost multiplier, we'll be much more quick to react to the hard market, and to take advantage of the hard market. Having said that, given what we saw in the third quarter, given the increased loss ratios in 2009 and 2010 and the fact that we're going to take some pricing actions to forestall and hopefully never have to speak of underwriting loss again, I would be very surprised if that policy retention count increases any more. Good question.

Robert, maybe this might help identify our attitude toward this. We are more concerned at this point about frequency than severity. So the renewal retention should reflect those accounts that we renew that have demonstrated solid frequency patterns. And that is the first primary screen as we go through the renewal process.

Having said that, you don't want to punish people that have been those helping along, and providing a tailwind to the company from an underwriting perspective. Those on the other hand that have exhibited frequency issues will probably be the ones that will absorb the greatest -- not probably, will be the ones that absorb the greatest increases. But I would also tell you to remind you this. Our renewal book of business is not really at fault in the 81.2% underwriting loss ratio at this point. It is -- our new book of business drives the higher. And the spread between renewal and new is significant. As an investor, you want us to keep more of our renewal business and be more finicky about our new business.

Thank you. Ladies and gentlemen, there are no further questions at this time. I'd like to turn the conference back to management for any closing remarks.

Thank you. Thank you, ladies and gentlemen, for joining us today. We are disappointed in the results of this quarter. But we do realize that the industry -- the workers' comp industry in 2009 reported a calendar year loss ratio of 80%, and a combined ratio of 110%. That still does not provide a parameter for AMERISAFE to operate, but it does indicate that there are pressures as a whole on the industry.

On a completely different note, unrelated to AMERISAFE, I think anyway, today is November 5, 2010, and, as usual, my father is sitting in the room to make sure that I do the right thing during earnings call. And today, he is 91 years old. So, happy birthday, Dad. Thank you, ladies and gentlemen.

Thank you, ladies and gentlemen. If you'd like to listen to a replay of today's conference, please dial 303-590-3030 using the access code of 4373781, followed by the pound key. This does conclude the AMERISAFE third quarter earnings conference call. Thank you for your participation. You may now disconnect.