Arch Capital Group Ltd (ACGLO) 2015 Q3 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the quarter-three 2015 Arch Capital Group earnings conference call. My name is Emma and I will be your operator for today. At this time, all participants are in listen-only mode. We will conduct a question-and-answer session towards the end of the conference. (Operator Instructions) As a reminder, this call is being recorded for replay purposes.

Before the Company gets started with its update, management wants to first remind everyone that certain statements in today's press release and discussed on this call may constitute forward-looking statements under the federal securities law. These statements are based upon management's current assessments and assumptions and are subject to a number of risks and uncertainty.

Consequently, actual results may differ materially from those expressed or implied. For more information on the risks and other factors that may affect future performance, investors should review periodic reports that are filed by the Company with the SEC from time to time.

Additionally, certain statements contained in the call that are not based on historical facts are forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. The Company intends the forward-looking statements in the call to be subject to the safe harbor created thereby.

Management also will make reference to some non-GAAP measures of financial performance. The reconciliation to GAAP and definition of operating income can be found in the Company's current report on Form 8-K, furnished to the SEC yesterday, which contains the Company's earnings, press release, and is available on the Company's website.

Now I'd like to turn the call over to Mr. Dinos Iordanou and Mr. Mark Lyons. Please proceed.

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

Our third-quarter earnings were driven by solid reporter underwriting results, while investment returns were impacted by decline in the equity markets.

Group-wise and on a constant-dollar basis, our gross written premiums increased by nearly 4% in the third quarter over the same period in 2014, while net written premium was up approximately 1% as underwriting actions in our insurance and reinsurance business were offset by growth in our mortgage business.

Changes in foreign exchange rates reduced our net written premium on a US dollar basis by approximately $21 million or 2.4% over volume in the quarter.

On an operating basis, we earned $126 million or $1.01 per share for the third quarter, which produced an annualized return on equity of 8.6% for the 2015 third quarter versus a 9.7% return on equity in the third quarter of 2014.

Looking at it from a trailing 12 months ending in September 30, 2015, after-tax operating income available to Arch common shareholders produced a 9.9% return on average common equity, while net income available to common shareholders produced an 11.6% return on average common equity.

On a net income basis, Arch earned $0.60 per share this quarter, which was lower than operating income primarily due to realized investment losses. Net income movements on a quarterly basis can be more volatile as these earnings are influenced by changes in foreign exchange rates and realized gains and losses in our investment portfolio.

Our reporter underwriting results remain satisfactory as reflected in our combined ratio of 89.7, and were aided by a low level of catastrophe losses and continued favorable loss reserve development.

Net investment income per share for the quarter was $0.54 per share, up a penny sequentially from the second quarter of 2015. The strengthening of the US dollar impacted total return on the Company's investment portfolio, which declined 31 basis points for the 2015 third quarter.

Our operating cash flow, excluding Watford Re, was $359 million in the third quarter as compared to $319 million in the same period a year earlier, primarily reflecting a higher level of premium collections. A book value per common share at September 30, 2015, was at $47.68 per share, a slight increase from the second quarter of 2015 and an increase of 8.3% from September 30, 2014.

With respect to capital management, we continue to have capital in excess of our target levels. However, we did not find many opportunities to repurchase shares in the third quarter that would meet our previously stated criteria for share repurchases. As you may recall, our philosophy with respect to share repurchases is based on the relationship of expected return to the premium-to-book value with the exception that we could earn the premium back over a reasonable period of time. Considering the underwriting environment we're operating in and the returns we're achieving, we believe that it would take more than three years to recover the premium paid.

With respect to overall market conditions, reinsurance industry pricing remains under pressure. Today, we have not yet seen significant erosion in the insurance business with the exception of certain lines which I will discuss in a moment. We believe, however, that the ability to buy reinsurance by us and our competitors on favorable terms will eventually lead to more competitive conditions that [caused] the insurance industry in the future.

As we have discussed in prior calls, there are several areas in the insurance sector that are experiencing increasingly more price competition. They are E&S, property, global property, large accounts, professional liability line, including D&O, especially in the foreign markets, as well as marine, aviation and energy -- business lines in which we continue to reduce our exposure and our participation.

Turning back to our quarterly results, the insurance segments, gross written premium on a constant-dollar basis grew 5.6% and 2.8% on a net written basis in the quarter over the same time period in 2014, with most of the growth coming from our construction and national accounts, travel and accident and health business units. Mark will comment further on premium volumes in a few minutes.

In our reinsurance segment, we responded to soft underwriting conditions by reducing gross written premiums on a constant-dollar basis by approximately 2% over the same period of a year ago. Increased sessions, primarily to Watford Re, in the quarter led to a further reduction in our net written premium, also on a constant-dollar basis, of 6% for the quarter-over-quarter comparison.

Our mortgage segment includes primarily mortgage insurance written through Arch MI in the US, reinsurance treaties covering mortgage risk written globally, as well as GSE credit risk-sharing transactions. Beginning in 2015 third quarter, the current quarter, new credit risk transactions now follow insurance accounting, which Mark will discuss in a few minutes.

Gross written premiums in the mortgage segment were $74.7 million in the third quarter 2015 or a 12.5% increase than in the same period, same quarter, in 2014, driven primarily by growth in Australian mortgage reinsurance premium along with Arch participation on GSE credit risk-sharing transactions. Net written premium grew 14.3% over the same period to $66.4 million as we retained a higher percentage of Australian business written in 2015.

Our US mortgage insurance operations produced approximately half of the segment's net written premium in the third quarter with $24 million coming from the credit union channel and approximately $8 million of premium written from the bank channel. Of note this quarter, underwriting income for the US operations moved into the positive territory with about $2 million of underwriting income in the quarter reflecting the slow but steady progress we're making in this area.

We continue to make progress in the expansion of the bank channel as Arch MI has approved 835 master policy applications from banks, and more than 350 of these banks have submitted loans to Arch MI for underwriting.

In the third quarter of 2015, we reached a modest milestone when new insurance written within the bank channel of $1.8 billion surpassed our credit union production of $1.4 billion of new insurance written. While we continue to see good opportunities in mortgage reinsurance, the GSE risk-sharing transactions issued by Fannie Mae and Freddie Mac are becoming an important component of our mortgage segment's revenues. We pioneered one of the first of these structures over two years ago by building upon the expertise of our mortgage team and working with the GSEs to provide insurance coverage.

Since we're talking about innovation in the sector, it might be worth a minute now to discuss the introduction of RateStar to mortgage lenders last week. We began this project approximately five quarters ago with a goal to develop a more robust risk-based pricing tool that in many ways will parallel what we do across our entire enterprise and is a hallmark of the Arch underwriting group approach.

To us, the current rate card approach, which uses just two factors -- FICO scores and loan-to-value ratios, while very important variables oversimplifies the issue. Our team reviewed the very rich proprietary data that we acquired from PMI and other available industry data that would allow us to establish an appropriate price for the risk exposures assumed. RateStar will enable us to more effectively allocate capital and calculate the appropriate risk-based returns on mortgage insurance at the individual loan level. While these may be a relatively recent innovation for the mortgage insurance industry with United Guaranty, the first to introduce a risk-based pricing tool, this approach has been utilized in our other lines of business within Arch for many years.

As many of you know, the return on risk-based capital serves as the basis for our incentive compensation plan. Our underwriters are paid on the basis of what profit they produce on allocated capital. We are pleased that we're now at the point where we are making the appropriate filing and where appropriate seeking regulatory approval for mortgage insurance and expect to introduce this into the marketplace during December of 2015.

Let me now turn back to the overall market conditions. Across all of our markets -- insurance, reinsurance and mortgage, conditions are competitive to varying degrees. However, Arch diversified mix of business and a willingness to exercise underwriting discipline should allow us to continue to generate acceptable returns. Group-wise, we believe that on an expected basis that the present value ROE on the business we have written this year will continue to produce an underwriting year ROE in the range of 10% to 12% on allocated capital.

Before I turn it over to Mark, I will also like to discuss our PMLs. As usual, I would like to point out that the cat PMLs aggregate reflect business bound through October 1, while the premium numbers included in our financial statements are through September 30, and that the PMLs are the reflected net of reinsurance and retro sessions. As of October 1, 2015, our largest 250-year PML for a single event remains in (inaudible) fees at $509 million or approximately 9% of common shareholders' equity. Our Gulf of Mexico PML decreased to $473 million as of October 1, and our Florida tri-county PML also decreased to $414 million.

I will now turn it over to Mark to comment further on our financial results, and after his comments we will come back and take your questions.

As was true on last quarter's call and the last few quarters, my comments that follow today are on a pure Arch basis, which excludes the other segment, that being Watford Re, unless otherwise noted. So same as in previous calls, I will be using the term core to denote results without Watford Re and the term consolidated when discussing results that includes Watford Re.

The core combined ratio for this quarter was 89.7%, with 2.3 points of current accident year cat-related events, net of reinsurance and reinstatement premiums, compared to the 2014 third quarter combined ratio of 88.5%, which reflected a lower level of cat at 1.6 points. Losses recorded in the third quarter from 2015 catastrophic events, net of reinsured recoverables or reinstatement premiums totaled $18.8 million versus $14.2 million in the corresponding quarter last year, primarily emanating from the Chilean earthquakes and the California fires, along with various smaller events.

The 2015 third quarter core combined ratio reflects 7.1 points of prior-year net favorable development, net of reinsurance and related acquisition expenses, compared to 8 points even of prior period favorable development on the same basis in 2014's third quarter. This results in a core accident quarter combined ratio, excluding cats, for the second quarter of 94.5% compared to the 94.9% in the third quarter of last year.

In the insurance segment, the 2015 accident quarter combined ratio, excluding cats, was 95.8% compared to an accident quarter combined ratio of 98% even a year ago. This 220 basis point improvement was driven by 150 BPs reduction in the loss ratio and a 70 basis point reduction in the expense ratio, with the loss ratio decrease reflecting lower large loss attritional activity than was the case in the third quarter of last year.

Taking this into account, the insurance segment accident quarter loss ratio was slightly higher this quarter versus the third quarter of 2014. The reinsurance segment 2015 accident quarter combined ratio, again excluding cats, was 94.6% compared to 90.6% in the 2014 third quarter.

As noted in prior quarters, the reinsurance segment's results reflect changes in the mix of premiums earned, including a continued lower contribution from property catastrophe and other property businesses. This quarter, most of the combined ratio increase relative to the third quarter of 2014 stemmed from the expense ratio and from a marginally higher level of larger attritional losses.

The mortgage segment 2015 accident quarter combined ratio, excluding cats, was 82.5% compared to 88% for the third quarter of 2014. This decrease is predominantly driven by the continued low levels of reported delinquencies benefiting the loss ratio associated with the CMG business we acquired in 2014, along with excellent credit experience to date on business written since the acquisition. Some of the benefit on the CMG business is offset by the contingent consideration earn-out mechanism negotiated within the purchase agreement.

As we commented on last quarter, an accident quarter approach to the mortgage business has the same meaning it does on the PC side because of the way the business works and the way the accounting works.

The insurance segment accounted for roughly 16% of the total net favorable development this quarter, and was primarily driven by medium- and longer-tailed lines, predominately from the 2007 to 2012 accident years.

The reinsurance segment accounted for approximately 78% of the total net favorable development in the quarter, excluding associated impact on acquisition expenses, with approximately 39% of that due to net favorable development on short-tailed lines concentrated in the more recent underwriting years, and the balance due to net favorable development on longer-tailed lines, predominantly from underwriting year 2009 and prior.

The remaining 6% of the net favorable development emanated from the mortgage segment, which reflects the continued improvement in the US book delinquency rate.

Approximately two-thirds of our core $7.3 billion of total net reserves for losses and loss adjustment expense are IBNR and additional case reserve, which still continues to remain fairly consistent across both reinsurance and insurance segments.

The core expense ratio for the third quarter of 2015 was 34.2% versus the prior year's comparative quarter expense ratio of 33.5%, partially driven by a 3.6% decrease in net earned premiums, and I will discuss each segment's expenses shortly.

The insurance segment's expense ratio decreased 70 basis points to an even 31.0% for the quarter compared to 31.7% a year ago. The net acquisition ratio decreased 90 basis points, whereas the operating expense ratio increased by only 20 basis points. The insurance segment net acquisition ratio reduction continues to reflect materially improved treaty ceding commissions on an earned basis associated with quota shared contracts ceded. It's important to note, however, that on a written basis, the front-end gross commission ratio worldwide actually decreased 50 basis points, whereas the average quoted share cede commission ratio improved a substantial 260 basis points which, as you will recall, is identical to the benefit achieved last quarter. These overall net acquisition improvements, however, will continue to be felt as these ceded written premiums are earned over the next few quarters.

The reinsurance segment's expense ratio increased from 32.6% in the third quarter of 2014 to 35.6% this quarter, primarily due to a 12.2% lower level of net earned premiums, a higher level of treaty cede commissions and a slight increase in operating expenses, although serially the expenses actually dropped compared to last quarter. The net acquisition ratio increased 100 basis points due to market forces, whereas the 200 basis point increase in the operating expense ratio is almost exclusively driven by net earned premium, a reduction mentioned earlier.

As I commented on last quarter, separating components of expense ratio can be a little fallacious because of the accounting does not go back and reflect the reimbursement of operating expenses contemplated in the cede commission itself.

The ratio of net premium to gross premium on our core operations in the quarter was 73.1% versus 75.5% a year ago. The insurance segment had a 72.2% ratio, compared to 74.2% a year earlier, whereas the reinsurance segment had a net-to-gross ratio of 72% in the quarter compared to 75.8% a year ago, primarily reflecting increased sessions to Watford Re as a reinsurer.

Our US insurance operations saw a 60 basis point effective rate decrease this quarter, net of ceded reinsurance. As commented on the last couple of quarters, the pricing environment is quite different for short-tailed lines versus longer-tailed lines, as Dinos also referred to.

Our short-tailed first-party lines of business had an effective 4.4% rate decrease for the quarter compared to a 30 basis point effective rate increase for the longer-tailed third-party lines, both on a net and ceded reinsurance basis.

Looking more deeply, some lines incurred rate reductions, such as an 8.9% decrease in property and an 8.1% decrease in high-capacity D&O business. While others enjoyed healthy increases, such as plus 6% in our low capacity D&O lines and a 4.1% increase in our program businesses. Also, our lower capacity D&O lines have now achieved 17 consecutive quarters of rate increases.

Turning to our continuing market cycle management, the insurance group worldwide reduced net gross written premiums in the highly competitive and volatile line of E&S property and global property by 12% and energy and marine by 15% quarter-over-quarter. By contrast, lower volatility lines of contract binding and travel expanded north of 20% on a gross basis, partially offset by a decline in program business due to purposeful underwriting actions.

I stated in last quarter's call, some volume impacts were a result of underwriting actions taken on two programs, whereas another program administrator has been purchased by a competitor and the premium loss impacts will be felt beginning next quarter.

Lastly, as Dinos has already stated, the insurance segment's construction business saw growth this quarter; however, much of this book has project policies and odd time policy terms which can result in lumpy premium volume quarter-to-quarter.

The reinsurance group only had 9% of its net earned premium represented by property cat this quarter, and property cat net written premiums were reduced by another 11% quarter-over-quarter, reflecting our view of that marketplace.

Additionally, the property other than property cat had a net written premium decrease of roughly 6% this quarter. And the reinsurance group also reduced net volume again in motor quota share and [crop] [hale] by approximately 20% in response to market conditions.

The mortgage segment posted a 75.2% combined ratio for the calendar quarter. The expense ratio, as expected, continues to be high as the operating ratio related to our US primary operation will continue to be elevated until proper scale is achieved.

The net written premiums of $66.8 million in the quarter is driven by the $31.2 million from our US primary operation and $35.6 million of net written premium for our reinsurance mortgage operations primarily. This segment also had $3.6 million of other underwriting income for the quarter versus approximately $1.0 million in the comparative quarter last year due to our GSE credit risk-sharing transactions. This quarter marks the first time that we have reflected some mortgage risk-sharing transactions with insurance accounting rather than derivative accounting treatment.

The net written premium this quarter under insurance accounting totaled $2.2 million, whereas legacy risk-sharing transactions shall continue to be accounted for as derivatives -- that is, reported in other underwriting income. One should also note that mortgage reinsurance premium growth is driven by the fact that the Australian business is a single premium market as opposed to the United States, which is predominantly a monthly premium market.

At September 30, 2015, our total mortgage segment risk in force is $10.3 billion, which includes $6.5 billion from our US mortgage insurance operation, $3 billion even through worldwide reinsurance operations, and approximately $800 million primarily composed of the GSE risk-sharing transactions.

Our primary US mortgage operation is [now] $3.2 billion of new insurance written during the quarter, which was approximately 57% through the bank channel and 43% via credit union clients.

The weighted average FICO score for the US primary portfolio remains strong at 737 and the weighted average loan-to-value ratio held steady at 93.2%. In those states, risk-in-force represents more than 9% of the portfolio. And our US primary mortgage insurance company is operating at an estimated 10.2 to 1 risk-to-capital ratio as of the end of September.

The other segment, that being Watford Re, reported a 99.4% combined ratio for the quarter on $125 million of net written premiums and $99.2 million of net earned premiums. As a reminder, these premiums reflect 100% of the business assumed rather than simply Arch's approximate 11% common share interest.

As for business sourcing, approximately 29% of the $131 million in gross written premium this quarter was written directly on Watford paper, with the remainder ceded by Arch affiliates. It should be known, however, that this sourcing mix can vary materially quarter to quarter.

The total return on our investment portfolio was a reported negative 31 basis points on a US dollar basis this quarter, primarily reflecting declines in most areas other than investment grade fixed income. Total return was negatively impacted from the strengthening US dollar on most of our foreign-denominated investments.

Excluding foreign exchange, total return was a positive 4 BPs in the quarter. On a year-to-date nine-month perspective, total return was a positive 76BPs on a US dollar basis and a positive 173 BPs, excluding the effects of foreign exchange. Our embedded pretax book yield before expenses was 2.1% as of September 30, compared to 2.18% at December 31 of 2014, while the duration of the portfolio lengthened slightly to 3.42 years.

The current duration continues to reflect our conservative position on interest rates and the current yield environment and tactical moves in the fixed income portfolio.

Reported net investment income in the quarter was $0.54 a share or $67.3 million, versus $0.53 a share in the 2014 third quarter or $72.2 million. As always, we evaluate investment performance on a total-return basis and as such, invest in asset sectors which may not generate above the line net investment income.

Interest expense for the quarter on a core basis was $12 million, which is more consistent with our normal quarterly run-rate versus last quarter and the third quarter of 2014 that were affected by periodic adjustments for certain loss portfolio transfer.

Our effective tax rate, our pretax operating income availability to our shareholders for the third quarter, was an expense of 5.7% compared to an expense of 2.5% in the third quarter of 2014. Approximately $1.8 million or 22% of this quarter's tax represents a true-up to bring the first half of the year to this now higher effective tax rate. Reflecting this, the nine month or annualized effective tax rate is 4.5% of pretax operating income.

As always and as demonstrated this quarter, fluctuations in the effective tax rate can result from variability in the relative mix of income or loss that occurs or is projected by jurisdiction. Our total capital was $7.05 billion at the end of this quarter, which is virtually flat with total capital as of June 30, 2015, and December 31, 2014. Approximately $522 million remains under our existing buy-back authorization as of the end of this quarter.

Our debt-to-capital ratio remains low at 12.6% and debt plus hybrids represents only 17.2% of our total capital, which still continues to give us significant financial flexibility. And as Dinos has mentioned, we continue to estimate having capital in excess of our targeted position.

Book value per share was $47.68 at the end of the quarter, up 4.6% relative to the end of the year of 2014. The change in book value per share this quarter primarily reflects the Company's continued strong underwriting result.

With that said, we're now happy to take your questions.

(Operator Instructions) Amit Kumar, Macquarie.

Congrats on the quarter. Just maybe one or two questions. Number one is going back to the discussion on rate card there was some confusion in the marketplace when the press release came out as to what it means for pricing and your competition. Can you talk a little more about it and, without obviously giving away the secret sauce, talk about the expected ROEs? And maybe talk about how should we think about the adoption rate off RateStar going forward?

As Coca Cola will never reveal their formula, we won't reveal our formula either. But at the end of the day, we're in the underwriting business and I think the more robust analytics you have in the way you allocate capital and price, the risk of exposures, the better off you are as an organization. These (inaudible) is two words that go. We're trying to go from a more simplistic approach to pricing mortgage risk to something a bit more sophisticated that we introduce other variables in the decision making and, in essence, affecting the pricing.

It doesn't mean we're going to abandon the rate card. The rate is out there and there is some bank channels, some customers, they prefer that, and basically they will only do business on that basis. We will continue to do that, but also there are other channels that they prefer to go to a more sophisticated pricing methodology that more appropriately allocates the right premium to the exposure and we're going to go forward with that where appropriate.

So you're going to continue to see us having both -- the rate card and RateStar. And only the marketplace will tell us as to how much of which is going to be used over time.

That's helpful. The only other question I have is going back to the discussion on capital management and again, it's a high quality problem. I'm not sure the capital is burning a hole in your pocket. Would you consider other avenues to return capital or are we not there yet? How should we think about that?

That's the million dollar question. At the end of the day, yes, we always consider other avenues. Having said that, there is also the unknown that you might want to have a little bit of ammunition in case opportunities come as the market turns. So it's more of a complicated issue for us. What was not complicated in this quarter was that we usually stick to our knitting, and when we made the calculations we felt that it might take four to five years to earn back the premium we're going to pay when we repurchase shares and we said: let's not do that, let's see what other opportunities we have or other avenues.

Having said that, we're going to have those discussions both internally as a management team and also with our Board when we meet. And we'll make determinations at that time.

Okay, fair enough. That's all I have. Thanks for the answers.

Michael Nannizzi, Goldman Sachs.

A couple hopefully quick ones on the MI business. Can you talk about how much of your NIW in the quarter was singles versus monthly premium?

It was approximately 24% that would have been singles.

Okay, and the rate start is relevant to the monthly business I take it?

Yes, predominantly yes. You can apply it on both sides because even when you do singles you get granular mortgage-by-mortgage attributes, so you can apply that. But at the end of the day -- when you go to single you try to look at your return, what kind of a price you're going to get and that's why you saw a significant reduction in us for the quarter as to how much we wrote in singles.

Got it. And then if we were to think about the RateStar versus the rate card, I'm guessing for some types of business it's going to be cheaper and for others it's not going to be. Is there any way to kind of think about --?

This type of business is exposure, Michael; it's exposure. Let me turn it over to you and you tell me the difference, right? If you have two loans that are both 750 FICO and 90 LTV but one borrower has a coverage ratio of 35 and the other one 25. Which loan would you prefer? And at the end, how do you reflect that in your pricing? I'm not going to get into all the algorithms that we have because then -- it's not only you listening; our competitors are listening.

I understand.

Introducing additional variables. You've seen it in a lot of other P&C lines. You've seen it a lot on the selection of risks in the automobile business. Progressive is very good and famous for it, Geico, et cetera. And at the end it makes for a better return for shareholders and probably a fairer charge to the consumer based on their own risk characteristics.

Okay. And just a last one on that, not about the algorithm, but for the players or for your customers that do accept or prefer RateStar, have you seen a meaningful change in submission volume?

We haven't yet introduced it to them. We finished the project. We made the press release. That's why I talked about it. And our sales force is in discussions with the marketplace and starting to introduce it. At the end of the day, we're going to continue having both rating engines available and it will be up to our customers to choose which one they prefer.

Got it. Thank you so much that it, Dinos. Appreciate it. And then really quick, Mark, I was just looking at Watford written premiums versus reinsurance ceded premiums and there's a sort of growing gap there. Is Watford or is the insurance sub or segment ceding business to Watford or is Watford picking up business from outside of Arch as the difference?

As I commented, on a gross basis here's the way to look at it: 29% of it is coming natively on their paper. But we're continuing to get Arch Re affiliates and Arch insurance to be sending over either a retro session or reinsurance. I think this quarter there was slightly more proportionately from the insurance segment.

Got it. And then last one. On the reinsurance expense, just trying to think about that a little bit. It sounds like the expense ratio to procure business for reinsurance is going up. So that probably a tail wind to the insurance expense ratio so that's going to raise the acquisition costs I guess for reinsurance. But then you have an offset from Watford because I'm guessing the same dynamic exists between Watford and the reinsurance company. How should we think about -- do these things neutralize each other or is there more of a head wind or more of a tail wind from those sort of intercompany transactions?

Your observations are right. The net impact is really market-force driven, and there is some element of Watford that's reflected on the fees, reflected in acquisition expense. So as Watford continues to grow, and sessions, that will continue to be more meaningful as an offset, which I think is the question you were asking. It's probably close to neutralizing but not quite. So you still could perhaps see a net increase but nowhere hear the increase it would be without the existence of those.

Let me add something to your question from a different perspective. At the end of the day, our intellectual factory that produces great results is the underwriting talent that we have within the reinsurance group. This management team, me down to Grandisson and Lyons and Papadopoulo, et cetera, we strongly believe that we have very good underwriters talent and underwriters and independent if the market might not allow us to utilize them at 120%, which we usually do. We're not willing to send those underwriters back into the marketplace for our competitors to hire, et cetera. I never saw a company have problems because their expense ratio went up maybe a point or two. I've seen all companies having a lot of difficulty when their loss ratio balloons by 5, 10, 15 or 20 points.

You got to understand. That's our philosophy. Yes, we expect our managers to manage expenses, and there is attrition within the organization. But we're not willing to let go good talent just because we can't utilize the factory at full capacity. That to us, the underwriting staff we have is our intellectual factory that produces a profit and I'm going to hold on to that.

And one other technical point, Michael. What Dinos just talked about is a core principle really for us. But on the technical side, a little bit of a difference in shift. The treaty business is falling off a bit more, whereas the facultative is not, and that has a direct sales force. So you get a little bit of that waiting pushing it up as well.

Got it. Thank you both so much for the answers. Really appreciate it.

Ryan Tunis, Credit Suisse.

Dinos, your point on the tiered pricing is that it allows for better risk selection. That makes sense, but there still seems to be a concern in the market if you look at how some of the stocks have acted in the past week or so that if you're successful at implementing tiered pricing, competitors may follow. And that would then lead to broad pricing pressure. I'm just curious if that's also a concern of yours and do you think more tiered pricing in general for the industry could lead to pricing pressure?

I don't believe it will because the other factor you haven't factored in is what expected returns different us and our competitors are looking for. Better selection doesn't mean that you have lowered your return expectations. All you're doing is pricing more appropriately the type of exposures you're getting.

This is not about reducing pricing in the marketplace. This is about assigning the right price to the right exposure. And at the end of the day, our return characteristics are no different if we use the rate card or RateStar.

So having that in mind, it would tell you that basically the whole effort was to improve how we think from an underwriting point of view, not to gain market share as some people -- I've heard comments to that effect. If we wanted to do market share or reduced prices, the easiest way to do it is take the rate card and you shave a few BPs in each one of the categories and I don't have to be spending a lot of brain power with a lot of our people over thousands of man hours in developing something that is more sophisticated.

So I think there is misunderstanding in the marketplace but eventually for those who know Arch and know our underwriting approach, they will understand that at the end of the day we're trying to be better in the way we're going to select and price risk appropriately, which is the foundation of this Company.

Got it. My follow-up then is just talking about assigning the right rate to the right exposure. In doing that, is there a segment of the marketplace that you envision Arch MI becoming quite a bit less competitive in that comes to mind?

Yes. There's going to be segments we're going to become more competitive and segments we're going to become less competitive. If you expect a certain return from the pie and now the pie is cut a little differently, you're going to have the pluses and the minuses. The question is are you getting a lot more on the pluses and a lot less on the minuses, and what kind of return you're going to have with that. Only time will tell, but we're more comfortable with our ability to price the exposures better by using more variables than just FICO score and LTV.

Thanks.

Sarah Dewitt, JPMorgan.

The GSE growth opportunity sounds pretty interesting for you. How could we think about sizing that? If we look out over the next five years, what percent of overall earnings do you think that could be?

Well, over the next five years -- my crystal ball doesn't go five months. (laughter) But still, we do view it as a positive opportunity. And I think one way that you should think about it is that the advent now of Fannie joining Freddie on this, and it seems that because they're expanding and looking for others to participate in this, they're going through the effort to establish a broader market, which gives credence to the fact that they're here to stay and it's not just a transitional thing.

So with Freddie continuing to do this and Fannie continuing to do this, we do think it's an exciting opportunity. So it's definitely going to be a growing piece. Now as long as pricing stays sane, we will continue to be participants in that growing marketplace. So far, on the Fannie deals, they've all had the same structure. They've all been 2 1/2 points excess of 1/2 point on the subject loans that are out there.

This is why it's difficult. We don't know how those structures are going to change over time. Are they going to be higher attachments, lower attachments? So it's very difficult to put your thumb on volume let alone how much of these are going to be pushed out into the marketplace. But we view it as an exciting opportunity for us.

Yes, and they might change to go to a first loss and right now it's excess of loss using insurance terms. But this is an evolving area but the demand is robust.

Okay, great. Thanks. And then just on MI broadly, are you able to be more competitive on price because you have a diversification advantage versus your model line competitors? Or is that not a consideration?

That is not a consideration.

Great, thank you.

Vinay Misquith, Sterneagee.

The first question is RateStar once again. What percentage of lenders do you think will use RateStar and is it the smaller lenders versus the larger lenders?

On your first question, I don't have a clue. I can't even project that. On the second question, I would say most likely the small lenders will be more adapting to the RateStar than the larger lenders because the larger lenders they like to have more simplicity of the rate card and they have a lot of power in the marketplace, et cetera. Smaller lenders are trying to find niches so they can penetrate the market. But that's purely forward guesses on my part. Only time will tell once we introduce this this year.

Okay. So that means that this thing could take some traction to get through the book just because large lenders I guess make up a bigger portion of the total business. Correct?

That's correct.

Okay. My view of this was that the high FICO scores were subsidizing the lower FICO scores. So the new RateStar will sort of reduce pricing for the higher FICO scores and raise prices for the low FICO scores. Do you worry that since 60% of your business is in the higher FICO score business that this could lead to higher competition among some peers and sort of reduce the profitability for the larger pieces of the business?

No. You're going in the wrong direction, Vinay. If it was just FICO scores, you didn't need to go and make all these efforts to create a RateStar with multiple algorithms. It's other characteristics. There is restyling the loans be provided to us by the lenders. Who is the borrower, the co-borrower? What location is the house? Blah, blah, blah, blah. I'm not going to get into all of the stuff that we think. If it was purely FICO score, you don't need to make -- you have LTV and you have FICO score and then if you want to make higher FICO scores cheaper, you take a few points off your rate card and you accomplish that.

So that's not what it's all about. I'm surprised as to how much confusion is in the minds of people as to what this is all about. This is a product that will allow us to take other characteristics of the loan and find what we believe is a more appropriate price for the exposure that we're assuming.

Yes, Vinay, this is just a more sophisticated class-rated plan, just like you have on the PC side as an analogy. But let's not lose the fact that there's already been discrimination between risks between each MI of -- I'll use your example -- of high FICO people. The analogy is schedule rating. You have a file plan, you have schedule rating where you can deviate for individual risk characteristics. And I think that's been pretty meaningful up to 20% or 25% to reflect characteristics of each risk.

So it's already been occurring with the old rate card that there is discrimination between risks. This is simply we believe a better way to do it and a more consistent way to do it.

Okay, that's helpful. And just as a follow-up to this mortgage insurance. I just noticed that the premium growth has slowed a little bit recently, especially on the owned premium side and also on the written premium side. Curious as to what's happening there since you now recording the GSE premiums also as written. Correct?

The part you didn't mention is our reduction in the singles. The change in the trajectory I would say became 100% out of our reduction in the singles.

And it's a good time, Vinay, for me to correct something that I said before. I had said that singles were 24% in the quarter. I misspoke; it's 21%. So I believe the trajectory and our view of that continues to drop. I agree with Dinos' comment.

Do you have a sense for what percentage of the business it was last year? Was it a much higher percent last year?

Yes. I don't have an exact figure in front of me, but it was substantial.

It was the first time we did it last year. And actually pricing on singles a year ago, it was more acceptable to us as the last three, four quarters emerged it became a more competitive marketplace because we have some competitors that are trying to gain market share through singles. We don't view that a good place to be and we're disciplined when it comes to underwriting.

Okay, thank you.

Jay Gelb, Barclays.

I may have missed it, but did you mention your [Tengen] loss?

Insignificant.

First of all, it's not a cat, so we didn't reflect that within the cat load. It's just not that large for us, Jay.

If it was anything notable we would have put something out, but it's not notable within our numbers.

But there is some exposure from the reinsurance side and the insurance side, and as you know the uncertainty surrounding these things is quite large. The ability to get in and check things out has really just begun recently, so there's a lot of volatility around it, so you never know.

Okay. Did you add some IBNR just in case?

We always do, believe me.

But it's contained within our standard attritional IBNR.

Okay, perfect. Thank you for that. The other question I had was on the tax rates -- 13% in the third quarter. You said that was a true-up. What do you feel a normalized tax rate is going forward since historically it's been in the low single digits?

First off, longer term implies I know what jurisdictions are going to give me any profits on a go-forward basis, and that really does fluctuate from quarter to quarter. But you're looking at the tax rate on that income as opposed to the tax rate on operating. Just the simple arithmetic of it -- you're got the tax rate on pretax operating income and to convert over to net income it's really the net realized losses. So you have the same pack of dollars with a smaller denominator. That's the arithmetic of it, the push to get up to 13.

Our current view on operating of like a training 12-month type view on operating income is likely to be fourish -- fourish percent, fiveish percent.

I would say be doing between 4% and 5%, and that's a better way to look at it. Don't look at net income in one quarter --

Because realized gains --

Look at it from a trailing 12 months. Then you've got more of the net income, and then you can add all the tax and then it would give you a better feel as to what the percentage is.

Okay, that's fine. It just bounced around a little so I just wanted to quantify that. And then on the buy-back. With the stock now trading around 1.6 times book, it sounds like Arch is really not going to be in the market for buy-backs --

I didn't say that. I said something different.

Sorry. I must have misinterpreted.

I never said I'm not going to be in the market, right?

Okay. So that valuation seems to be a pretty important parameter. How should we think about it?

Yes, valuation is very important. We look at it based on the prospects of what returns we get on the business we write. If my recovery period elongates and we started getting uncomfortable over three years, then we shy away from it. It has nothing to do with how we feel about the stock. It's just purely our approach to it, and that approach might change. I don't know what my discussions with wiser guys -- that's why I got pretty wise guys on my Board to give me advice -- and what perspective they're going to have. Based on what I said, we might sit on excess capital because there also might be opportunities for us to deploy in a different fashion in the marketplace.

And also, Jay, as you know we moved our mix of business and our capital around depending on what the opportunities are. And we talked earlier about this. There are real opportunities in the GSE, credit risk-sharing space. So hypothetically, if that increased at a higher rate than we anticipated or we have other opportunities around the world, that mixture might increase our view of forward ROE by 200 basis points or something, which is going to come into the equation of time to pay back.

Of course. If the stock were valued instead at say at 1.5 times book, would that have been within your range of viewing it within the three-year payback period?

It could be. It's hard for me to project into the future, right? It's like you're describing a guy who can read the obituary pages five years from today and he finds his name there. That's not a good place to be.

We make those decisions and we're flexible on a quarter-to-quarter basis. And we have unknowns, and when we have unknowns sometimes we hold back a little bit. For example, the mortgage GSE opportunities -- and I think Sarah is the one who asked the question and I couldn't answer it because I know the opportunity is big, the demand is there, but I don't know how big it's going to be. And I don't know how much of our capital we want to allocate to that.

So when I have unknowns, I'd rather say: I've got unknowns and here is how we're thinking, but they might be opportunities. Because I know you want to build your models and project year out and quarter out and all that, and I don't operate on that basis. I'm trying to make sure that our underwriting unit is they make the right decisions based on the latest information that they have. If I have excess capital, it's not burning holes in my pocket, unless somebody is trying to put his hand in my pocket and take it, and then I will cut it, that's not a problem. It's a good problem to have.

That's right. Thank you.

Kai Pan, Morgan Stanley.

Do you have any potential exposure to Volkswagen and Hertz and [Patricia]?

Insignificant. Insignificant.

Okay. That's great. And then probably cat gen one pricing. What's your outlook and you have reducing the business quite a bit over the past few years. If the market stabilized, would you become more interested in writing more business there?

Listen, a hallmark is putting the right price based on the recent exposures we underwrite. If that market improves, we're going to write more. Our appetite has not disappeared. There were times that we were committing 20%, 21%, 22% of equity capital to that line on a PML basis and we're down to now 9% and actually for Florida and Gulf of Mexico, which is less than that. So our appetite will depend on the market pricing.

Now predicting what's going to happen on January 1, who knows? If I had to guess, it would probably be as stable where it is today because I think, even for those that participate, that new capital that comes in, even with no real cat their returns are not super juicy. And that's a sad statement to say. When there is no cat, you don't have super juicy because what do you do when you have the cat, right?

Kai, I would also add you use the term if it stabilizes. It depends what you mean by stabilizes. Improving doesn't mean stabilizing to me. And if the rate cuts stay -- there's no more rate cuts and it stays at 0% change, we've been shedding volume given that relative level. So I wouldn't expect our business to increase if it stays at the levels it is today. It would have to improve, not merely stabilize.

Okay. So you're not expecting any sort of meaningful price increases from current levels?

Nothing I see on the horizon now that is telling me to anticipate pricing increases. But we don't make decisions on anticipation. We make decisions on what we see in the marketplace.

Okay, that's great. Then on sort of management succession. Dinos, you've been running -- had a great track since you founded the Company, and I'm sure you are excited running the business as of today. But I don't know if the Company has mandatory retirement age, but is the Board considering a succession planning, and how do you think about it?

We have succession planning in every senior position we have. It's part of our process within the comp committee. Their responsibility and my responsibility as CEO is not only preparing my successor but also each one of key positions has one or two successors ready from within. And that process is not new; it's been in place now for over 10 years.

Having said that, if you read my contract it goes all the way to the end of 2017, actually March 1 of 2018. So I can sign the 10-K if I decide to just become the Chairman, but no decisions have been made. There is an existing succession plan within the Company that is part of the responsibility of our Board, and they take it seriously and we talk about it at least once a year in the comp committee.

That's great. Thank you so much for all the answers.

Meyer Shields, KBW.

Dinos, I think you did a great job of laying out the point of the RateStar program, but if you're allowing lenders to choose between RateStar and the rate card, doesn't that just invite averse selection?

If you allow them to have both, yes. But basically what we're telling lenders you can either have one or the other.

Okay, so they don't --

Yes. You can't have the rate card and then price it that way and then price it on the other way and go back and forth. It's either you choose to participate with us on the rate card or you choose to participate with us under RateStar.

Okay, that helps. What is the lose trend of the insurance segment that corresponds to the 60 basis points sort of premiums or rate decline?

I'm sorry, could you --

The 60 BPs rate decline --

Right, trying to get the -- (multiple speakers)

It's probably short-lined at weighing the increases we get on long-tailed lines.

That's exactly what it is. So as I stated, it's 4.4% down on the first-party lines. And I think I said 30BPs or 40BPs up on the third-party lines.

And our volume on the short-tailed lines is small, so you've got to do weighted average, right?

Right. I'm just trying to get the sort of weighted average loss cost trend that corresponds to that.

If you remember your Algebra I, Meyer, you got all the information you need.

I'm sorry. I'm probably more dense than --

The 4.4% is not in excess of loss trend; it's the pure effective rate change. You need to layer on top of that, to do a loss ratio conversion from period A to period B what your estimate of loss trends is. But as we said in the past, it varies widely by line of business.

Okay. Thanks a lot.

And we go through those calculations. When say 60BPs, there's a lot of work behind it to come up to that. I'm being surrounded by actuaries. We're pretty technical when it comes to that stuff.

That sounds like a nightmare, but good luck.

Jay Cohen, Bank of America.

Maybe a bigger picture question on the mortgage business. I believe, Dinos, in the past you've said that when this business gets to scale that I think the segment earnings could be as much as a third of the overall Company's earnings. Is that still a view that you believe is accurate?

Yes, that's an accurate view. But I also said that it will take three to five years to get to that point. We believe that, yes, we have potential to be earning $150 million, $200 million annually from the mortgage business but it's got to get to maturity and we're not there yet.

Right. And then maybe a bit more technical. When I look at the -- as you get scale in this business the expense ratio comes down. I'm assuming the bulk of that shows up in other operating expense ratio. Should the acquisition expense ratio also improve over time or should that be relatively stable?

That should improve because on the US mortgage side it's really a sales force that's there. So you're going to have those fixed costs and you write more volume, it should drop.

A sales force is constant, right? We're not adding -- once you get to a steady state on your sales force, maybe you add one person here and there. And then there is a little bit of increase, a cost of living adjustment, et cetera, incentive compensation. So that's more of a steady number. And then as you're building volume, your expense structure on a percentage rate basis is going to improve.

And Jay, just to clarify on your first question, yes, with longer-term view mortgage could be a materially significant piece of our net income or our underwriting gain or loss. But that's the mortgage segment in totality. That is not necessarily US MI. You have the reinsurance segment and, as we said, the increasing contribution from the GSEs, it's in totality.

That's what he's asked. He didn't ask about anything --

Yes. No, it was the segment. That's what I figured.

Right. Right.

Okay.

Very helpful. Thanks, guys.

Brian Meredith, UBS.

Just quickly, Mark, I don't know if I caught it but last quarter when you talked about the difference between on the insurance side your ceded benefit that you're getting on the ceded on the acquisitions or commission ratio versus the increase that you're paying -- I think it was 60 basis points on the written. Was it similar this quarter, the spread?

Yes. On the quarter-share treaties that dominate the sessions, it was a 260 basis point spread -- or actually, let me restate that. Improvements in the ceding commission by 260 basis points from 3Q to 3Q and last quarter 2Q to 2Q had exactly the same improved spread difference.

And at some point in time that would disappear because once we cycle over four quarters it's over until -- in comparison, quarter to quarter.

You keep getting the gain, but the difference will go away.

Right. So therefore your acquisition expense ratio should probably continue to come down in the insurance space year-over-year for at least the next couple of quarters.

On the earnings, yes.

Barring no change on the front-end direct commission.

Got you. And then my second question I guess bigger picture also. If I look at your overall business -- reinsurance, return on equity probably continue to kind of come down here with the rate pressure insurance may be flat as to down. Is the increase in the mortgage insurance that you're seeing growth when you look out here, is that enough to continue to offset the decline you're seeing in the reinsurance ROEs to keep it stable?

It's a difficult question to answer because it depends on the volume. Two things I got to tell you. Don't underestimate how good our insurance guys are. They're finding other avenues. Not all of their business is these what I would say large client under a lot of pressure business. They're finding niches here and there to still be relevant and have good returns.

And at the end of the day, yes, if our mortgage business continues to grow it might offset it. But I don't know that because I can't project volume. We don't spend time thinking about volumes and that's why -- we're not trying to be avoiding the questions but to us future projections are not -- we don't spend a lot of time on those. What we spend a lot of time is to analyze what we have and how we're going to behave quarter to quarter based on the market conditions that we see every quarter.

Let me just add a little bit to that, Brian. The insurance group when it comes to math is 60%-65% of the net written, which will find its way into earnings. And their margins have continued to improve and was this quarter as well. And some of that is on the loss ratio side that we've seen and some of that is on like the question about the cede commission overrides. So we expect continuing contributions from the insurance group which, as I said, is 60% to 65% of the weight.

Got you. Great. Thank you.

Ryan Byrnes, Janney.

Just trying to figure out why your [Tengen] loss was immaterial. It seemed to kind of affect multiyear competitors. Just want to see if you guys avoided certain risks or coverages that kept you away from these losses or if it was just simply luck. I imagine it's more the former.

Well, you can call it luck; you can call it good underwriting or a combination of both. When you give me the choice, I'd rather be lucky than good, but I think we're both -- lucky and good.

Okay. Thanks, guys.

[Rob Path], Wells Fargo Securities.

When I look at your balance sheet, it looks like your revolving credit borrowings went up by about $239 million in the quarter, but when I look at your calculation of leverage, you don't seem to be including those in your leverage numbers. Are these Watford borrowings or is something else going on here?

You found it.

I love it when you answer your own questions. (laughter)

Okay.

That's exactly right. It was $239 million increase in borrowing from revolver [unlocker] but since we consolidate, of course, we have to reflect that on our balance sheet. You're also correct that our capital composition exhibit is for non-Watford. So you hit it exactly.

Okay. And I don't know if you can discuss it -- what they need the money for and if these borrowings are nonrecourse to Arch.

Nonrecourse to Arch is their borrowings and they're using them for investments. Their business plan always included I think 1 1/2 times leverage, up to 1 1/2. It's a company with over $1 billion of capital, so they will probably borrow up to $400 million-$500 million and use it in their investment strategy. We're not responsible for the investments. We're only responsible for the underwriting side.

Okay, right. Thanks very much.

Ian Gutterman, Balyasny

First to follow up that last question, is the reason Watford chooses to use debt to get to their asset leverage because they're behind plan on float and then they thought they would have [been] on float and they were replacing it with straight debt or --

These are questions for Watford, but what I'm telling you is that they believe there might have been opportunities now based on what they've seen in the market and they say: Hey, we can put some leverage on it buy some stuff.

Right. Got it. Okay.

And Ian, just to add to that, use of leverage was there from day one on the initial business plan.

No need to go out and borrow when you haven't even deployed your own capital yet. And don't forget we're starting to generate a float for them. Our premium plans we've been hitting based on the original plan so there is float coming in from our underwriting activities.

Exactly. That's what I wanted to make sure about. Okay, good. The first question I was going to ask before that was on the capital discussion, can you remind me -- you guys have obviously never paid a dividend, whether it be ordinary or special, remind me sort of why you guys are averse to dividends. Is it a tax thing? Is it just you don't want the commitment or it or -- (multiple speakers)

You're forcing a tax bill to your shareholders, right?

Right.

And once you give the money if the next day you get an opportunity then you got to go and borrow to take advantage of it. We always like to have a little bit of excess capital. Maybe we have a lot of excess capital, but right now it's not at the level that is really giving me a lot of angst, even though excess capital is only earning 2.5%, 3% or thereabouts. It is what it is, but like I said -- we talked to our investors and some of our investors are opposed to a special dividend. They think that over time it might not be in the next few quarters, it might be in the next year or two, we'll find the right opportunity and deploy capital.

Don't forget, we were talking about excess capital, et cetera, our mortgage business is a new business for us. We only started it about four, five years ago and it really is getting scale now. If I didn't find that opportunity with our guys -- it was predominantly Marc Grandisson who discovered it based on our discussions with our investment department and me, et cetera. We wouldn't have that opportunity to deploy it today in excess of $0.5 billion of capital into that business.

Everybody tries to say: Oh, if I had this magic balance sheet that is always in balance, that would be utopia. But I'm a realist. There is no such thing as utopia. We try to do the best we can.

No, very fair. Just wanted to make sure I was remembering correctly. Then just a couple quick things. One, I don't think this has come up yet. I believe there is a lot of talk about just pricing changing in the bank channel. I think it's more in the community bank channel if I'm correct, that some of the banks are sort of I guess jealous of the credit unions success and saying since the crisis things have changed and the credit union is not necessarily a better channel than a regional bank anymore given changes in lending standards and why are we charging so much more for MI in the bank channel. And therefore we should cut rates to bring it more in line with the credit union experience. Is that happening or is that being discussed and if so, just what are your thoughts on that?

I don't know if it's being discussed because I haven't really specifically talked to our sales force about that. You're right. Some of the community bank experience has been better than what I would call the large regionals or the nationals. And the data shows it. And that's why I said before that maybe some of the rates store might be more adaptable to these community banks, which have similar characteristics to the credit unions. They're closer to their customers. They know them well. They're in the community; they know who is who. And to a great extent they spend more time and effort in approving mortgages. So how do you reflect that? That's our [secret] to us.

Got it. Very fair. And then just lastly on the RateStar thing. What's interesting to me -- I'm not asking to give us the secret sauce here -- but I always thought about FICO -- again, I know it maybe had some missteps in the crisis but that arguably was because of lending standards maybe more than FICO itself, right? And when you look at auto insurance that FICO is the best predictor of whether you're going to get into an auto accident. It seems like it's a pretty powerful variable. What do you see as --? (multiple speakers)

Absolutely. But we're not eliminating FICO --

Right. Right.

FICO is a very powerful -- (multiple speakers)

(inaudible) Okay --

But there is other attributes that have predictive ability and value. So by ignoring them, is it two borrowers or one co-borrowing? Is it coverage ratio? Is it 30, 40 or 50 -- and I can go on and on and on into the other things. What territory are you in? What do you think about the housing market in that territory, et cetera, et cetera?

And I'm not going to go and tell everybody as to what we've done with this, but we've done a lot of work. We believe that it's -- I wouldn't say smartest because that arrogant -- I think it's a different way of looking but I think it's a better way in our view to assign the right price to mortgage risk.

Maybe if I ask at a slightly different way. Is your sense that FICO is maybe explaining -- I'm just going to make up numbers here? Is FICO so good that it was explaining 90% of the difference in borrowers and this gets you the last 10% or was it maybe two-thirds and this give you a whole other third. You know what I mean? I'm just trying to get a sense of significant --

I don't know from the work I've seen. I've seen some of their work and I participated in some of their discussions. So I can't put a percentage or predictability on any one attribute, but I can tell you FICO is a very important piece. LTV for some risk is very important. For other risks it might not be.

Like a young couple -- two MBA students, college sweethearts. They both have pretty good jobs and they can only scrape together a 5% payment because they want to live in a bigger house because they have a lot of income. So the LTV may not be as critical and they might have super FICO scores and these other attributes, and you might price that loan differently than a simple rate card.

Got it. Makes sense. Thanks for the explanation.

Charles Sebaski, GMO Capital Markets.

The first is on the GSE business and, obviously you can't predict how the flow on that risk sharing is going to come in the future or how much. At the current pricing and structuring level, is there any other constraints other than the flow from the GSE for how you guys would participate at current pricing? Is there aggregation or other issues that might halt that as it comes online?

It's two issues you got to think about this. The number one issue is the GSEs, they're going to put this in the market. That's known. There's a lot of pressure by Congress to [derisk] and not be the credit providers for loans beyond the mandatory 20% down payment -- maybe all the way down to 40%.

That's why we've been hesitant on volumes. A lot of this goes to the capital markets, and it depends what pricing they're getting from the capital markets. What both GSEs, Fannie and Freddie, are doing, they're developing two parallel markets. They're developing the insurance/reinsurance market. It fluctuates sometimes. They allocate 20% to 30%, and then the rest of it goes to the capital market. But we have no control as to what those allocations. If the capital markets become expensive, maybe they will start allocating 30%-40% to the insurance markets. And believe me, what happens in the capital markets also affects the pricing that comes onto the insurance and reinsurance markets.

The reason they're doing that they believe that by creating two avenues and two different sources of capital responding to these credit risks is good in the long run. It might create more stability for them because they have two different paths to share credit risks into the private domain instead of the government taking it.

So I don't know which way it's going to go but right now we believe that with insurance accounting being introduced and the innovations that we have worked very closely with the GSEs there and their willingness and they're talking to a lot of others within the insurance and reinsurance business -- I don't know how many have the expertise to do it, but some do. I think this is going to be a new market for the insurance/reinsurance business and it can be substantial over time.

And Chuck, the other thing that makes it difficult to predict, as Dinos said, Dinos was describing more how Freddie Mac has done it. Where it's the same notional base of loans and capital markets and the insurance/reinsurance industry share on that same set. Fannie has done it a little bit differently, but they're using capital markets and the reinsurance market but it's a different pool. So what's gone out to the Fannie deals has been exclusively a pool that went to the reinsurance industry. And a separate pool may have gone to the capital markets. They may not continue doing it that way. They may wind up doing it similar to Freddie. There's a lot of different parameters that make the projections difficult.

It's a young emerging market and a lot of it is because Congress in general they want Fannie and Freddie do the risk and for that reason we feel optimistic that the demand is always going to be there. Now, is it going to go 100% to the capital markets? I doubt it. What percentage comes to the insurance/reinsurance versus the capital markets is in their hands, and you got two big customers here and they hold all the cards.

I'm not asking you to predict what they are going to put out. What I was trying to understand is what is your constraint? Conceptionally, Fannie and Freddie could put our more risk than you guys could possible take or the insurance market just to the size of the portfolio. Why is your guys' constraints if we read that Fannie is accelerating their --?

As we do with every line of business that we have, we have -- think of it as a PML and how much of our equity capital we want to risk. So there is a constraint and we have developed -- actually maybe we're the only ones. I don't know if our competitors do that or not. I have no idea. I'm sure from a risk management point of view they do something of that sort. But we do calculate on a quarterly basis what the PML values we have for the mortgage business. That will be a constraint at some point in time. When we reach the upper limit of the available PML, that will be a constraint for us. But we've got other vehicles. We might create a (inaudible) at that time. We might use our knowledge and ability and underwriting ability in the systems we have. Don't forget, you've got to have good systems to price loan-by-loan, et cetera, to introduce other capital providers into the sector with us. As we've done with Watford, we can do a mortgage Watford, so to speak.

So we have a lot of flexibility. We're nowhere near yet of having that constraint so for the time being we've got freedom to operate and we've got plenty of capital to deploy and it's not violating any of our PML criteria that the Board sets as to how much risk you're going to take in any particular -- I don't care if it's cat risk or mortgage risk or D&O risk, we have, in our risk management principles, we have limits that we want to take.

And finally on RateStar, are you guys first in trying to use a more automated, multi-variant pricing model here? And if you are, what's the lag time or the lead time. If you guys are pitching this out into the market to be originators and your competitors go: Oh uh, Arch is a leg ahead of us here now on this. What's the lead time you guys have on this kind of product?

I don't know. We're not the first. United Guaranty, part of AIG, introduced risk-based pricing first. Probably they've been out for about a year now. They were ahead of us maybe longer than a year. Basically we agree with the approach. It's fundamental to underwriting. Now how acceptable this is going to be to the marketplace and all that, I don't know. But it seems that United Guaranty, they have some penetration and they have acceptability of it for a quite a few of the states from an approval point of view. So we're optimistic.

I appreciate all the answers. Thank you very much.

Now I would like to turn the call over to Dinos Iordanou for closing remarks.

Thank you for listening to us. It was a little longer. It was mostly mortgage. I almost forgot that I'm in the insurance/reinsurance business. (laughter) But we're looking forward to be speaking to you next quarter. Have a wonderful afternoon.

Thank you for your participation in today's conference. This concludes the presentation. You may now disconnect. Have a good day.