Aspen Insurance Holdings Ltd (AHL) 2005 Q4 法說會逐字稿

Good morning, ladies and gentlemen. My name is Sandra and I will be your conference facilitator today.

At this time I would like to welcome everyone to your Aspen Insurance Holdings fourth quarter and year end 2005 financial results conference call. All lines have been placed on mute to prevent any background noise. After the speakers' remarks, there will be a question and answer period. [OPERATOR INSTRUCTIONS]

It is now my pleasure to turn the floor over to your host, Mr. Noah Fields. Sir, you may begin your conference.

Thank you and good morning. The presenters on this morning's call are Chris O'Kane, Chief Executive Officer of Aspen Insurance Holdings; and Julian Cusack, Chief Financial Officer of Aspen Insurance Holdings.

Before we get underway I would like to make the following remarks: Yesterday afternoon we issued our press release announcing Aspen's financial results for the quarter ended December 31. 2005. This press release, as well as corresponding supplementary financial information, can be found on our website at www.aspen.bm.

This presentation may contain, and Aspen may make from time to time, written or oral forward looking statements within the meaning under and pursuant to the Safe Harbor provisions of the U.S. Federal Securities laws. All forward-looking statements will have a number of assumptions concerning future events that are subject to a number of uncertainties and other factors. For more detailed descriptions of these uncertainties and other factors, please see the Risk Factors section in Aspen's annual report on form 10-K for the year ended December 31, 2004 and current report on form 8-K dated October 4, 2005, filed with the SEC and on our website.

Finally, this presentation will contain non-GAAP financial measures which we believe are meaningful in evaluating the Company's performance. For a detailed disclosure on non-GAAP financials, please refer to the supplementary financial data posted on the Aspen website.

Now I will turn the call over to Chris O'Kane.

Thanks, Noah. Good morning and welcome to Aspen's year-end 2005 earnings conference call.

In this morning's call I will cover our performance in the fourth quarter and in 2005 as a whole and comment on the marketplace as we currently see it following the January renewal season. I will also bring you up to date on some of our newer risk management initiatives concerning catastrophe risk, which we commented on at some length during last earning call.

2005 was an extremely challenging year for the U.S. insurance industry. In the fourth quarter of 2005, Aspen reported net income of $30.3 million or $0.32 per share, with a combined ratio of 104.8%, while gross written premiums increased 13% to $245 million.

For the twelve months ended December 31, 2005, we posted a net loss of $177.8 million with a combined ratio of 117.2%.

Despite the most costly hurricane season on record, Aspen posted positive earnings in three of our four business segments.

We have always considered our diversification by product line and by geography to be a core competitive advantage. In 2006 we will continue to diversify our business away from Property Reinsurance and emphasize our growing leadership position in our Specialty line segment and we will continue to leverage our strong position in Casualty Reinsurance.

However, Property Reinsurance remains an important part of our overall business and is an area where we believe we have considerable expertise and we will remain flexible to ensure that we're able to take advantage of the more favorable market trends that we expect to develop in this area as 2006 progresses.

Overall, the market at January 1 behaved as we thought it would, with three out of four of our business segments' experience in price movements in line with our expectations. Property Reinsurance is the exception, where we expect the best pricing to become prevalent ahead of the mid-year renewals or even in 2007.

As you know, we measure rate relativity in all our lines of business, and we have achieved an 11% rate increase across the part of our portfolio that renewed at 1/1 on a premium weighted basis.

Our Specialty line segment, that is to say, our marine and aviation insurance and reinsurance lines, got off to a particularly healthy start this year, with prices increasing 18% on average for business written in January. The Specialty line segment comprised about 14% of our gross end premium in 2005 and is expected to increase to an excess of 20% in 2006.

The most exciting story, as we expected, concerns our offshore Energy Physical Damage account. While we achieved rate increases outside the Gulf of Mexico of about 20% on average, the risk situated in the Gulf of Mexico itself, the actual rate increases were in the order of 4.5 to more than 6 times. This needs to be seen in the context of a dramatic reduction in aggregate limits for named windstorms, where limits in the range of 1 to $2 billion of aggregate for named windstorms were available in 2005, but in 2006 we've yet to see as much as $200 million of wind limit being achieved. Thus we're seeing much more rate for a lot less exposure.

Partly as a knock-on effect of the energy market, we're also seeing good increases in our Marine Liability lines of about 15% and Marine Hull is also performing well, with increases of about 4%.

Turning to Aviation insurance, this book does not see much renewal activity in the first two months the year. The airports and noncritical products business, we did rise at 1/1, increased by between 5 and 10% on the back of continued deterioration in this sector's past year results.

Our Casualty Reinsurance business represented about 25% of gross end premium in 2005 and also approximately 25% of our anticipated planned earned income in 2006. A big proportion of this business doesn't [inaudible] on the first of January and overall we achieved an average rate increase of 3%. Rates have held up well, with smaller reductions on the international account and better than expected increases in U.S. casualty.

Our international casualty reinsurance account experienced a modest 1% average rate increase in the Australian, Canadian, and London markets. Competition is continuing to behave rationally and a number of [cedents] are opting to retain more risk, which we view as positive, as in this excessive loss reinsurance account we are thereby further removed from loss frequency.

Our U.S. casualty reinsurance book has seen price increases of 5% on average, with workers' compensation reinsurance programs renewing close to expiring and medical malpractice reinsurance programs renewing at a nearly 6% increase.

U.S. casualty facultative reinsurance business has seen rates firm or slightly increase.

Our insurance segment represented 21% of gross end premiums in 2005, and we expect it to comprise approximately 18% of our anticipated planned earned income in 2006. Our insurance account does not typically experience a great deal of renewal activity during 1/1, with the first major insurance renewal date expected at the end of March.

U.K. commercial property insurance rates have come under pressure and have decreased by approximately 4% in our book. This is much better than the market as a whole has seen.

U.K. liability insurance saw average rate reductions on renewed business of approximately 7%.

Our U.S., excess and surplus property lines business has seen catastrophe exposed business increase by 24% in December and we expect further increases as the year progresses. Casualty U.S. rates, on the other hand, continue to register modest single-digit reductions.

Turning now to property reinsurance, the market is still coming to terms with the impact of the 2004 and 2005 hurricane seasons. This, regarding the effect of reinstatement premiums, our property reinsurance segment comprised 37% of our gross earned premiums in 2005 and this will trend down considerably this year.

Ahead of the renewal season our thesis was that we would see significant movements in pricing, terms, and conditions in favor of the sellers' reinsurance as it regards U.S. hurricane exposed nationwide Florida business. The reality was somewhat different, the price increases varying significantly from as low as 15% to highs of over 100%, depending upon loss experience, terms, data quality, and the nature of the original business. The upper limit of rate increases was achieved in programs covering commercial exposures which had sustained hurricane losses in 2005. We expected that this phenomenon would have a knock-on effect on pricing for U.S. regional programs, and this did occur, although with price increases of 10 to 15% it was much more muted than we would have hoped.

We had doubts as to what extent the international business would be affected by events in the U.S., and these doubts were well founded. Rate increases in Europe were driven mainly by losses from European floods, with loss effective programs seeing increases of 15 to 30% but programs without losses experiencing only single-digit percentage increases.

Now, if we felt this is as good as it gets in this business, we would be disappointed. My interpretation of these market dynamics is as follows: when Aspen rebased its catastrophe pricing models last October to recognize increased hurricane frequency and severity as well as potential for a given unit of exposure to sustain a greater quantity of damage in a given storm, we appear to have been one of the first reinsurers to do so. It is important to stress that all of our 1-1 risks were priced on the basis of our new model, as well as many of those incepting in the latter part of 2005. I believe this put us ahead of the market in what we were demanding in terms of price and data quality.

During the second quarter of this year, when the new releases of vendor models are expected to be made, incorporating assumptions which we believe are very similar to those which we have already deployed, we expect to see a sharp correction in market pricing.

There will be four drivers of this. Firstly, the new models will demand more price per unit of risk. Secondly, many sellers of reinsurance will recognize that a unit of risk carries more exposure than they thought and we expect that they will become more risk averse, i.e., they will accept less risk despite higher prices. Thirdly, the buyers of reinsurance will find that they need to buy more dollars of protection for the same amount of exposure to meet both the demands of the new pricing models as well as the constraints placed on them by the new rating agency models. Finally, retrocession cover, which at times in the past has effectively operated as a disguised subsidy to reinsurance pricing, will become scarcer and more expensive and hence will have the opposite effect and thereby drive prices up still further.

At Aspen we reduced critical learned catastrophe exposures that renewed at 1-1 by between 40 and 50% on a like-for-like basis. This is considerably more than the correction that we need to make for the year, but it is a good tactic to do this because we're confident that we will be able to replace a significant part of this lost exposure by the middle of this year at much better prices.

Looking forward further into 2007 and beyond, we expect that the effects of these dynamics will also be reflected in increasing primary insurance rates, especially in coastal areas, in keeping with the latest scientific thinking on the frequency and severity of hurricanes.

I am now going to turn the call over to Julian for a review of the fourth quarter and year end results.

Thank you, Chris.

I will begin with an update of our hurricane loss estimates. In our press release on 6 of December, we increased our combined estimates for losses from Katrina and Rita by $27 million. We have since increased this by an additional $9 million, with the result that our cumulative net losses for Katrina and Rita now stand at $435 million and 50 million, respectively.

The figure for Katrina is now net [inaudible] deduction of $18 million in respect to the estimated fair value of our potential recovery under our catastrophe swap derivative contracts. I will explain more about the CAT swap shortly.

In the same press release we advised that at that time we expected the net of tax losses from Wilma to be in the range of 20 to $35 million. Our current estimate is $22 million.

As referred to earlier, we have recognized a pretax credit of $25 million in the quarter in relation to our CAT swap contracts. This contract provides for up to $100 million of cover in relation to industry hurricane losses on a linear basis, triggering at an industry loss of $39 billion and paying the full if industry losses reach $47 billion.

For this purpose industry loss is based on the estimates announced or to be announced by PCS. At the current time PCS made two preliminary estimates, with its latest estimate announced on December 6th, being at $38.1 billion, just below the trigger point of the contract. This is a derivative contract and we are therefore required to place a fair value on it. In arriving at our valuation we have analyzed how previous PCS industry loss estimates developed over time, and based on this analysis and taking into account all other factors that could affect the fair value, we have arrived at the valuation of 25 million or 18 million net of tax.

Our diluted book value per share was $18.73 at December 31, 2005, compared to $17.53 at September 30, reflecting capital raising in the quarter. Closing shareholders equity rose to $2.04 billion, with total capital now at $2.3 billion, including long-term debt.

The capital raised in the quarter, which for the exercise in June 2006 -- sorry, January, 2006 -- of the over-allotments option in respect to the preferred shares, increased total net proceeds to $819 million.

Of the approximate 790 million raised in the quarter, 560 million was downstreamed as equity to Aspen Insurance Limited, a Bermuda operating company, prior to the year end. We also contributed $15 million to Aspen Specialty, our Boston-based surplus lines company.

In January we contributed $150 million to Aspen Insurance U.K. limited. The balance of the capital raised, amounting to approximately $65 million, is currently being held in the holding company.

Taking into account these capital injections, both our U.K. and Bermuda operating companies now have net assets of just over $1 billion. Aspen Specialty has a GAAP surplus of around $110 million.

I will now turn to our underwriting results on a segmental basis. The combined ratio for the property reinsurance segment in the quarter was 176, mainly driven by hurricane Wilma. Excluding the major hurricanes, the combined ratio would have been 54%. This underlying ratio is consistent with Q4 2004 and Q2 2005, reflecting the absence of any other major losses in the period. For the full year ended December 31, the combined ratio was 172%.

The casualty reinsurance segment has not been materially affected by the hurricanes. There has been a small reduction in the loss ratio which relates to continuing favorable developments on the business written in 2002 and 2003, now at international casualty accounts, and in our 2004 U.S. workers' compensation catastrophe book. This has given rise to releases in relation to prior periods of $12 million in this quarter. The losses on this account take a long time to be notified and develop, but we are now seeing favorable trends that enable us to make a small reduction in some of the held reserves as explained above. Casualty reinsurance loss reserves at the end of the quarter have risen to over $670 million.

The Specialty lines segment has a combined ratio for the quarter of 69%, benefiting from prior period releases in the quarter of just over $5 billion and $26 million in the full year. The segment has suffered one major loss affecting both the insurance and reinsurance segments of the account, where a reserve of $80 million has been established in relation to an explosion in the United Kingdom.

The Property and Casualty insurance segment has a combined ratio of 57% in the quarter. Within these accounts are our worldwide property accounts and our U.S. excess and surplus lines accounts written out of our Boston office. These accounts have both suffered losses from hurricanes Katrina, Rita, and Wilma.

There has also been favorable developments, however, on our U.K. commercial property and U.K. liability accounts, which have resulted in releases from prior years of $2 million in the period and $27 million in the full year.

Net investment income in the quarter of $39.3 million was 78% higher than in the fourth quarter 2004, due to both rising portfolio book deal from favorable movements and interest rates and positive cash flow, including the additional capital raised in the fourth quarter.

Cash and invested assets increased by 47% compared to the corresponding period last year, and the invested portfolio book yield at the end of the quarter was 4.08% with a market yield at 4.58%. The duration of our invested portfolio at December 31, 2005, was 2.9 years versus 2.7 years as at September 30, 2005.

The aggregate portfolio, including shorts and investments in cash, had a book yield of 4.08% and a market yield of 4.45% the duration of 2.2 years at December 31.

At December 31 our gross reserves for losses and loss adjustment expenses were approximately $3.04 billion, of which 52% represented estimates for losses incurred but not reported. This is down from the 75% IBNR ratio at September 30, mainly because a large proportion of the hurricane losses within IBNR at September 30th have now been notified.

Reinsurance recoverables have increased significantly to around $1.2 billion at December 31, 2005, up from $893 million at the end of September. This increase is primarily driven by the reinsurance recoveries in relation to the hurricane events. We have presented in our earnings supplements an analysis of this [inaudible] by both A.M. Best and S&P current ratings.

Following December capital raising, our financial leverage as measured by long-term debt to total capital is 11%. After the exercise of the over-allotments option in January, our preferred shares will stand at approximately 10.8% of total capital. This is treated as equity by S&P and A.M. Best for rating purposes and attracts a 50% equity credit from Moody's.

I would now like to discuss some of the factors that are likely to influence our results for 2006. Overall, we would like to be opportunistic if the market continues to develop favorably, and at this stage in the year our provisional overestimate of growth in GWP, as it is most likely to be in the range of 5% to 10% with property GWP being the hardest to call at this stage.

Overall, we expect the combined ratio to be between 85% and 95%, barring major losses, but this includes around 10 points of expected annual average CAT load.

Investment income, which rose to $121 million in 2005, is expected to rise further in 2006. Based on the year-end balances and book yields, we can reasonably expect a minimum of $180 million before tax. Reinvestment rates are currently higher, at around 4.75% to 5%, which together with positive cash flow could increase this to $200 million. The greatest uncertainty relates to the rates at which we will pay hurricane related claims.

Interest payable, including dividends on our preferred shares, would increase from $16.2 million in 2005 to $29 million in 2006. For 2006 we currently expect a tax rate in the low 20s on underwriting results and in the range of 18% to 20% for investment return.

I will now turn the call back over to Chris.

Thank you, Julian.

During the last earnings call I discussed our changing view of the world. After the 2005 hurricanes we took a long, hard look at our processes and have initiated a number of steps to enhance risk management in catastrophe risk pricing, accumulation control, and risk governance. We're now applying higher loadings for hurricane frequency and severity, including adjustments differentiated by exposure type and hurricane zone.

In addition, we have reduced our risk tolerance for 100-year and 250-year exposure in peak zones from 20% to 17.5% and from 30% to 25% respectively.

Another risk management tool which we've used in the past is the purchase of a comprehensive reinsurance or retrocessional program. For 2006 we have successfully placed well in excess of $400 million of cover for a reasonable increase over last year's cost, given current market conditions. We anticipate buying at least $100 million of additional cover this year and we will continue to review the market for opportunities and, depending on price and availability, we may buy further cover, but possibly in the cat bond market.

The retrocession market is a very difficult one indeed for new players, but Aspen has benefited from a long-term investment in building and maintaining strong relationships with certain key partners. Our view, therefore, is that we continue to have a considerable potential advantage from our reinsurance and retrocession arrangements.

I hope you have found this review of recent events useful, and with that we would be happy to take your questions.

Thank you. [OPERATOR INSTRUCTIONS] Your first question is coming from Tom Cholnoky of Goldman Sachs.

Good morning. I guess one topic which Chris didn't touch on, but not that you should have, but obviously last night with PXRE being downgraded to below an A minus, what impact do you think that potentially has on the market as we go forward and -- let me leave it at that.

First of all, Tom, I think another powerful reminder of the power of the rating agencies. We will see how it unfolds, but I think most people assume that with an A in the rating reinsurers can trade successfully. With a rating with a B in it, it is very, very, very tough to keep a portfolio of clients together confidently.

Would you expect some opportunities to arise out of this?

I think that traditionally that company's strong points have been in the retrocession area, and I think if they cease to be acceptable security to some of their buyers, then retro, which is already very scarce, is going to be more scarce still. I think it is going to be that further impetus to higher prices as the year goes on. But to be fair, they're not a huge company. They're sizable in that business, but not huge, so I don't think the impact will be huge, but significant.

Okay. And then just if I could just follow up on your appetite for property business, what really has to change a lot more, just so I understand, of what could actually increase your appetite for that?

I guess what we're really trying to do here, Tom, is build as far as we can a diversified, a balanced portfolio, if you like. So clearly, when you look at property, and some of it, I think, is very well priced indeed. We're not happy to take as much volatility as it appears to cover. So, for us write more we either need more of all those other non-commensurate risks so that the thing stays in balance, or the price has to increase to such an extent that the extra premium we receive means that we can allocate more capital to it.

Prices have moved a long way for the most dangerous stuff, really, that -- I guess the Florida southeast wind, and a few of those nationwide accounts actually paid at 1-1 price increases between 2 and 3 times. That's pretty good, year on year, but it is dangerous stuff and I think that's not where we want to take the business. We would like to be writing more International. We would like to be writing more U.S. regional, and there the prices can be okay. They can be acceptable. But they're not very exciting.

The other point these days, and I think a lot of people last time out sort of found that some of the companies, especially some of the big commercial writing companies, are not really on top of their exposures. The data they can report to their reinsurers tends to be a bit lacking in detail, and that means that our ability to see through and price it correctly is challenged. Our view would be, if we can't see through and get the right data, we don't write it in all.

So, enhancement in data quality in general across the industry I think is in everybody's interest. I think from the point of the view of the buyers of reinsurance they're going to have a better handle on their exposures and from the point of view of sellers, theyr'e going to have more confidence that what they're actually pricing is exactly the risk they're getting. That really hasn't been the case these last couple of years.

Okay. And then, sorry, just one other numbers question if I can. Julian, I missed on the favorable reserve development, the 15 million or so, can you just sort of break that down of where it came from again?

Yes, I can. The fourth quarter reserve reliefs is currency reinsurance, 12 million; specialty lines 5.6 million; insurance lines, 2.2 million; and then there is a 4.5 million efficiency on property reinsurance, rising mainly from risk excess account.

Okay. Great. Thank you.

Thanks, Tom.

Thank you. [OPERATOR INSTRUCTIONS] Your next question is coming from Ron Bobman of Capital Returns.

Hi, good morning. I had a reinsurance program question and as it relates to your net. What are your largest -- for the '06 business year, what are your largest net retentions going to be on an account or risk basis? And what line of business would that be?

I think the largest retention is going to be on the catastrophe type event that has the potential to affect more than one line of business. In other words, maybe it can hit our property reinsurance account, hit our property insurance account, maybe hit the marina, the energy account as well. An earthquake in California is going to affect marine, non-marine, casualty too, that sort of thing. And when I answer this, I answer it a simple underwriter's way, i.e., before the effect of reinstatement premium, before the effect of taxation, before all those other accounting issues come, and our basic retention is in the region of $100 million.

We're still working on that. It depends which combination of losses we have, and we also have a major program which is our marine program, which is renewing just now, and it ought to be done in the next week or two. If that successfully completed, it has a potential to bring that clash exposure issue down for us. If it weren't completed successfully, it could move up.

So it's not, Ron, a very precise answer. That's partly because some of our programs come up for renewal during the year and so we haven't quite gotten there yet.

Just so I understand it correctly, so there could be a treaty with one particular insurer where you're providing protection to them where an event could produce a -- again, prereinstatement, pretax, net loss to the Company of up to $100 million?

That wouldn't be from one treaty. That would come from our whole account.

I'm sorry. I was interested in on a -- whether it be a treaty basis or a particular account basis.

Maybe I am not getting the question, because what would concern me is not so much how much we lose in a single treaty but the fact that two or three or four treaties go in the same event. But on the individual lines, if you want to focus there, we are usually buying down to something between about a $1 million and $20 million, depending on the individual line of business.

Okay.

I really want to caution you. I don't want you to go away thinking they only lose a million bucks in a given event, because, you know, you can lose that million quite a few times.

No, no. I understood that. Thanks again.

Pleasure.

Thank you. Your next question is coming from Stephen Peterson of Citadel Investments.

Good morning. I was just wondering if you might be able to return to your market comment, and you gave us sort of a brief list of reasons that you thought the market was going to continue to harden throughout the year. Did you hold back capacity in order to take advantage of some of that as the year progresses and did you get a sense in your interactions with the marketplace that others may be doing so as well?

The answer the first part of the question, I suppose, is yes. I think this is especially true for hurricane risk, because every year one sees more and more paranoia in the market place the closer you get to the hurricane season. It follows, I think, that sensible people don't want to sell their hurricane capacity in January. They want to sell it in May or June when the market is more favorable for the seller. I think this year, because of the new model releases that are likely to come out in the second quarter, I think that's going to be accentuated still further. So when we were quoting prices at 1-1, we were finding the market didn't want to go with that price, so we dumped a lot of exposure. We don't feel bad about that. I think as I said on the call earlier, we will be able to write some more back at better prices later on, and if the prices don't come, we won't do it. That's okay, too.

The extent to which other people did it I think is questionable. I don't know exactly what other companies do, but all I can say is, I think a lot of programs where we cancelled our line did get placed, and in some cases, where programs didn't get placed, we began to see them coming back to the market on a short fall basis. That means it is basically the same deal that was maybe 70% placed to 1-1, but now back in the market today playing maybe 20 or 30% more money. One or two of those we've actually looked at and we've written already because it was the same deal that a lot of guys took two months ago, much better today.

The other phenomenon that's quite interesting, and we're wise to this, is we're seeing a lot of people whose renewals come up in June or July, coming into the market right now. What that's saying to us is they, too, think the market is going to be tougher in June or July, so they're trying to get in in February and get easier pricing. Well, that's not going to work with us.

Okay. Terrific. Thank you for taking the question.

Sure.

Thank you. [OPERATOR INSTRUCTIONS] Your next question is coming from Alain Karaoglan of Deutsche Bank.

Good morning. I have a few questions. The first one, Chris, is if you had in place your new view of the world today at the beginning of 2005, what would your results have looked like in 2005? You mentioned that your exposures have decreased by 40%. But do you have an idea of what the results would have looked like in terms of earnings?

That's a very tough question, Alain. I mean, it's not too much -- use a crystal ball to look backwards.

One way of answering it, and I don't know if it's going to go as far as you want, is to look at those 2005 losses which across all the hurricanes on a gross basis add up to approximately $1.45 billion. And if we apply our pricing criteria, the new attachment point that we want to attach at, our very much reduced risk appetite for the higher, more remote, events, and nothing else was different, than our new underwriting criteria would produce losses more in the order of $900 million. I think actually a little under 900, about $860 million on our analysis. So you can just see on a gross loss basis we would have done something very differently.

One of my regrets, of course, is that we didn't do this, but frankly, until we had the losses we couldn't look at how the models were under-representing the loss potential, so, sadly, we're learning from experience, we didn't learn ahead of experience.

So I think in a way if you think about that much less gross loss, you could do those calculations to see how that would feed into net loss and post tax loss, but it would be a very, very different situation.

But I would assume you would have also gotten more premium for that same exposure, given your view of risk.

Well, you would, but I mean, the problem with that, of course, is if that had been our view of risk but no one else in the market shared that view, we wouldn't have been able to write the business at all because it wouldn't -- our pricing wouldn't have been available in the marketplace. I think there is an awful lot of variables to try and flex in answering a question like that.

Okay. The gross written premium of 5 to 10%, Julian, how would that translate into net premiums written? I am not sure -- what should we think about ceded premium or reinsurance costs for 2006?

I think that you should look at 2005, but adjust that for the instance of reinstatement premiums in 2005 and just to give you some figures, the ceded, as a percentage of gross in 2004, was 14%. That rises to 21% in 2005, but about 7% of that relates to reinstatement costs. So like the underlying ratio in 2005 is more like 14%, and I think that it is difficult to be sure because, as Chris said earlier, we haven't placed or completed all our program, but we would expect probably that to rise a little, but not significantly above that.

Okay. And then the third question relates to the combined ratio expectation of 85 to 95%. A year ago in 2005 your expectation was for 80 to 90% combined ratio for the year. This seems to have gone up by 5%. What is driving that? Is it the additional mix of casualty? Because I would assume the additional -- your new view of the world, you're going to price for that, so that's going to be reflected.

Yes. It is more reflective of the change or projected change in business mix. As Chris has said, deemphasizing the property reinsurance line, which, certainly on a long-term average basis, has a lower combined ratio expectation than other lines of business that we do.

Thank you very much.

Thank you. Your next question is coming from Sacket Cook of Menemcha Capital.

Hi, good morning. A couple of questions. I appreciate the difficult questions you just answered about what would happen to the gross loss if you kind of had your new exposures in place, and I guess I would be interested in taking that one step further, if you could, and I guess it relates to the reinsurance program that you had in place this year versus the one you expect to have next year.

Should we expect the same type of -- and again, all else being equal that you had the same losses, which I know is a tough assumption, but I am just wondering if by reducing your reduction in exposures that you are all now feeling that you don't need to buy so much reinsurance or your dependency on reinsurance has become less, and I guess the reason why I ask the hypothetical scenario is just to kind of see how that would play out in the numbers. That's my first question.

Okay. It is a good question. It is a little easier, actually, to get your arms around that one. Not all of the program renews at 1-1, so I can't be totally definitive about the bit that hasn't happened. But if I look at what we were doing at 1-1-2006, we have placed or are just about to complete placing about $508 million of cover, which potentially all could respond to the same major event, so [fire away.] Now, if the event doesn't affect the marine, the insurance, and the reinsurance in the right way, then the recovery would be less. But if it affects all programs, you have a potential recovery and what we've done this year, or more or less competed, for 508.

Last year on the same basis the figure was about 636 million, so you can see we got about 130 million less reinsurance cover than we had last year. Alongside that, you need to say, well, three things, really. One, we got really quite a lot more surplus this time this year than we did last year, so, you know, we can run more risk against our balance sheet.

Second, our exposures have come down deliberately and intentionally, so we don't need to buy as much reinsurance or retro as we had before, and obviously what we're trying to do is manage our exposures down roughly in the same way as our reinsurance or retrocession buying is coming down. The two things need to be done together and kept in balance, and we're working very hard to do that. So far I think we have.

Then I think the third thing is the price of the risk. And while the prices have gone up, they haven't necessarily gone up as much as we think they should. That in itself is making us a little bit more risk averse.

So, to date -- how to sum up: To date we're buying a little less cover. We're paying a bit more for it. We're taking in more money, more rate to pay for it, and overall I think the equation adds up reasonably. I think the only other thing I'd draw your attention to, and I think one of the earlier questions, maybe it was Ron Bobman asked, how much could we lose to a single event, and that number is up a bit on last year, that number of something in the region of about $100 million, that [pop] is a bit bigger than last year.

Just to follow up -- I appreciate that and that makes a lot of sense. You keep suggesting that you're going to -- you have some powder dry, as the year goes on, and I come away from the call thinking these guys have reduced their exposure on a gross basis by 50 to 60%. However, they say they still have powder left for some opportunities later in the year. What do you -- talking to the rating agencies and figuring out what you kind of have done with your retro, how much room do you have?

So what I am trying to ask is the 50 or 60% exposure cut that you have right now, if there are opportunities in Florida or wherever, can you reduce that to 30% reduction? Is that how much room you have to kind of be opportunistic?

I don't think I can put a number on it, Sacket, in that way exactly. What I'd take it back to maybe is in the call we said in the past we didn't want to lose more than 20% of surplus at a 100-year event. The new number is 17.5% of surplus at 100-year event. So as long as we're tracking at or below that, that's how much powder we're keeping dry.

But how much powder we actually want to take out and put in the gun and start firing is really going to depend on how good the opportunities are. The ability to write it, if it's there, is one thing. Seeing enough business you want is a big question. The market hasn't moved enough yet. I think it will, but it hasn't moved up yet. I don't think I can be more precise for you than that.

Okay. And one follow-up question, because I know you've mentioned it and I think it is important, is just maybe talking a little bit more about the offshore energy market, and I guess you talked about price. I would be interested in, number one, any changes to terms, because I know that there is kind of a built-in CAT exposure that people have kind of woken up to there. And I guess my second question is, what does the offshore energy market look like? Do you like it? Is it better looking as a primary company or as a reinsure?

For us, that business is almost entirely on a primary basis. This is an insurance operation for us.

Okay.

And I think it looks much better as an insurance proposition. Very, very tough to take those exposures and reinsure them. Very tough indeed.

So, in terms of -- what I sort of summarized on the call was outside of the Gulf of Mexico increases in the order of about 20%. Inside the Gulf of Mexico, price increases ranging from about 4.5 to over 6 times. This is big stuff, okay?

Now, in the past those -- the price had to pay for like a full amount of hurricane exposure on the rig, so if the rig value was $2 billion, you'd have $2 billion of hurricane exposure on that rig, and what we've seen so far this year is no one, no single rig in the regular market has been able to buy as much as $200 million.

Wow.

So it is very dramatically different. Now, 2 billion, that's a big rig. There wouldn't be many of those, to be fair. Typically you're talking in the several hundred million dollars. But the guy who might have been buying a $500 million policy last year, maybe for hurricane he is looking to get 100 million. 75 or 100 million this year.

The beauty of this story to me is price determined by experience, hurricane experience, on an ongoing basis you don't have the same potential to have that exposure, that experience again, because the hurricane risk is down. Around the side of that, I think some of the rig owners are going to, like, a secondary market to try and buy extra hurricane cover, top op cover. We're not at this point selling that. Again, because I think the prices on this stuff are going to go higher still. Not many rig owners are actually buying in the sort of January/February period. We're talking here on the basis of a fairly small sample. So our guess is as the year goes on and the demand gets greater, then the prices are going to be more attractive, and that's the point where we might want to sell a little bit of extra cat risk in the Gulf, but not so much there.

All right. Thanks very much.

Thank you. Your next question is coming from Michael Whitney of Taylor Investments.

Hi. Good morning. I am a little bit confused about something regarding your combined ratio guidance. If your business in 2006 is presumably going to be better business, lower risk business, more profitable business, because you're shifting your mix, why, then, is your guidance for combined ratio actually lower than what you posted last year, excluding the hurricanes?

I think you've got to look at the -- actually, the six of our different lines of business. So, for example, in casualty reinsurance, the combined ratio there has been in the sort of 90 to 95 area, and the extent to which that and also our casualty insurance lines are more emphasized in the business mix, that will have a greater impact on the overall combined ratio. So I think what you're seeing is moving to lower risk business or business that we hope will be lower risk, is actually means also moving to business, specifically on the casualty side, which has higher combined ratios.

But on the other hand, the casualty business, as long as it is directly reserved, does have the potential to build significant balance sheet asset positions, which will continue to boost our investment income into 2006 and beyond.

So does that mean you're deemphasizing the more specialty products you used to do in favor of a more generic casualty type thing?

No, I wouldn't say that. I think Julian's point is that -- back up a little. We really only started writing the Marine Liability business about twelve, eighteen months ago, and that proportion of our book has been building, and that's a business which we're very happy if it runs at a 90 combined because we think the downside is modest and it generates terrific cash flow, terrific investment income opportunity. So that's okay, and that slice of the cake is getting bigger.

But I think there is another point I wanted to sort of add to what Julian said to you. And that actually applies to the property side itself. When we talk about this new model, saying that the business is riskier, we're assuming that every year you're going to have more cat events so the expected loss ratio of the cat book, we're saying it's higher than we used to say. Even though there is more rate, we're getting more money for the product, we're just assuming that wind is going to blow a bit more often and a bit more fiercely and the damage is going to be greater.

You still have the potential to have a year like 2002 or 2003 when the wind doesn't blow at all, in which case the upside surprise in our property could be greater than it used to be, because we've built in more expected attrition to our numbers in advance, which I think is a prudent thing to do, but it does mean that the potential for the upside surprise is bigger than it used to be. That's also a contributor to our overall combined ratio view.

Okay. That makes more sense. Thank you very much.

Thank you.

Thank you. Your next question comes from Bin Laurence of SuttonBrook Capital.

Hi. I may have missed this. What is the 28 million other income?

That is the pre-tax uplift in the valuation of our catastrophe short derivative contracts that I mentioned in connection with the hurricane estimates.

Okay. Thanks.

Thank you. [OPERATOR INSTRUCTIONS] Your next question comes from Robert Rowell of RiverSource Investments.

Hi. Good morning. Just to follow up again on the combined ratio, taking it another step further, what -- last year and last year's range of 80 to 90%, what was the expected attritional CAT loss load in those numbers versus this year? I think you said 10 points this year, but what was it last year?

Rob, I think that the way I'll answer that is to say that based on our modeling and our view of risk a year ago, that average CAT component was actually about the same number of percentage points on a combined ratio at that time. But if we were to rebase that now, based on our assumptions about greater frequency and greater severity, then in fact it would have been significantly higher than that. So on a like-for-like basis we've seen it as a reduction in the expected CAT load.

Okay. So if you were using last year's models, your load would be less than ten points, based on your current exposures?

Yes. Exactly right. Yes.

Okay. And then, there is no prior period reserve adjustments baked into that number, is there? That's just an accident year --

That's based on accident year, average expectations, yes.

Okay. And then just on the last question about the CAT swap: could you remind us exactly -- so, it's 28 million after tax has been revalued up? I think it was 100 million previously, if the PCS number got to 39 billion, could you just remind us where those moving parts are? If I missed it?

The 100 million would be recoverable only at the top of the range, which is $47 billion. The trigger level of loss is $39 billion. And what we've done in putting a fair value on the contract is to make what we believe is a cautious estimate of the potential, but on actual Katrina recovery of $25 million, and then other changes in the fair value of derivative, take that to $28 million, which is the figure pretax reported in the income statement.

Okay. And where is the -- technically, where is the PCS number right now?

$38.1 billion was their announcement on December the 4th.

Okay. Thanks very much.

Thank you.

Thank you. Your next question is a follow-up coming from Ron Bobman of Capital Returns.

Hi. Thanks a lot. I was wondering -- I had two questions. One, do you have any existing [PX] 3 recoverables, and did they participate in the $400 million cover that you've purchased so far?

Yes. Our position on the existing recoverables of [PX] 3s that we have, approximately $53 million, of [inaudible] outstandings with that company based on 2004, 2005 hurricane events. And in relation to the 2006 program, that we have I think currently a participation in our program of around $17 million.

Okay. And then, when you have an IBNR recoverable balance, I assume the contract has some sort of ratings trigger where you can see collateral. If it does, do you expect to initiate that?

Certainly if any of the run-up contracts do, we would certainly consider that.

Okay. Thanks. My other unrelated question was, earlier you compared, I think talking to Sacket, the 508 soon to be, hoped for, completed reinsurance program, I guess the 400 plus a pending 108, and you compared it to 636 bought last year. Did the 636 bought last year include -- I think the number was 100 million derivative contract that we've talked about a little bit today?

No, it didn't. That's not treated for this purpose as a reinsurance because it is a derivative contract.

Okay.

Just to be clear on what I was giving out there was reinsurance or retrocession we purchased in January/February this year versus January/February last year. The overall totals in both cases are likely -- I know last year it was a bit higher and this year it could be higher again, but we can't predict the future in terms of what we're going to buy because it's a function of price, et cetera. So it was trying to be a like-for-like comparison on where we are in January/February.

Okay. Thanks for the candor. Best of luck.

Thank you.

Thank you. Your final question coming from Bob Sayles of LMK Capital Markets.

Hi. You blanked out a little bit when we were talking about comparing with some assumptions, the impact of another repeat '05 to what it would look like with the risk changes in '06. And I wrote down that the total losses in '05 were 1.45 billion without reinstatements and prior to tax, and it would compare to 860 this year. Is that correct?

Those are the correct numbers, yes.

Here is my question: What does the 1450 net out to on a pre-tax basis with reinstatements for '05?

On a pre-tax basis and before deductions for the upward revision in the fair value of the derivative contract, that would be 595 million.

595.

Yes.

And then, so I guess the follow-up question, could we assume the same netting out on a percentage basis on the 860?

No. I wouldn't make that assumption, and I think what I said to the original question is this is looking in the crystal ball backwards, there is too many moving parts I think. I can see what you'd like to do, and I would as well. There are just too many moving parts to be able to do what you're looking to do. Don't apply the same percentage discount.

But I'm driving to a final question. My final question is when you look at your modeling, given the series of hurricanes, Katrina, Rita, and Wilma that occurred in '06, what type of probability would you assign to the '05 hurricane season? That's really what I am driving at and the reason I asked the questions about the numbers.

Very tough one to answer. I will tell you, in the old model a hurricane effecting the U.S., mainland, anywhere, and costing sort of, say, 55 billion, came in the old model at about once every 103 years. On the new model I would expect a hurricane costing about that to occur once every 30 to 40 years, 30, 35 years, I think. Okay? That's just for one of them.

Clearly that multi-strike thing pushes the probability out a lot further. If you think hurricanes are discrete events that don't have a common cause, you get one answer then. If, on the other hand, you believe in global warming, climate change, then you would say there is a common cause. You would start saying, once you get one of these big storms there is a good chance you're going to get loads more in the same season, in which case you're going to get a different number for the probability of sort of three strikes in a year. And, you know, I read a lot of stuff on this and we have a lot of guys in the Company smarter than me who look at this stuff. No one has given a definitive answer to your question.

Yes, I realize it's very conceptual. I just wanted to understand your thinking more than a definitive answer. My next question is, have you in the call guided to or do you guide to a target return on equity?

No. Our guidance has been restricted of late to GWP combined ratio, investment income, and other expenses that are material. There are too many moving parts.

Yes. And the last question for you is, your Wilma loss, in the face of some of the other competitors or like companies, was fairly modest, and I am curious about Wilma, not so much with your loss, but let's take the Mexican hit out of the picture. Why was Wilma as costly -- more costly than it seemed like it should have been for some of the players relative to the hurricanes that came through in 2004? And I know that doesn't impact you, but I am looking for just more of a macro view from your perspective.

Wilma seems to have affected a very, very wide area of Florida. Initially I think people thought it was a very similar track to one of the '04 storms, I think it was Charlie. But the width of the storm seems to have been very, very big, so the individual number of claims seems to be a lot larger than any of the '04 storms. I haven't got that data to hand, but I think you can maybe get it off the PCS website or something like that. A lot of individual instances of loss, and it is picking up a lot of insurance portfolios. That seems to me to be the key driver.

I don't think that the wind speed was necessarily that great and I don't think the damage ratio per unit, or damage risk ratio per risk is actually necessarily all that high, but an awful lot of risk was actually affected. That seems to be the difference.

Okay. Thank you very much.

Thank you.

Thank you. At this time there appears to be no further questions. I would like the turn the floor back to management for any closing remarks.

Well, thank you for your time and attention. Good bye.

This does conclude today's teleconference. You may now disconnect and have a wonderful day.