January 11, 2007 - CORRESP: A correspondence can be sent as a document with another submission type or can be sent as a separate submission.

January 11, 2007

VIA EDGAR

Mr. James B. Rosenberg

Senior Assistant Chief Accountant

Division of Corporation Finance

Securities and Exchange Commission

Mail Stop 6010

100 F Street, NE

Washington, D.C. 20549

Re:

RenaissanceRe Holdings Ltd.

Form 10-K for Fiscal Year Ended December 31, 2005

File No. 1-14428

Dear Mr. Rosenberg:

On behalf of RenaissanceRe Holdings Ltd. (the Company), I am responding to the comments of the staff (the Staff) of the Securities and Exchange Commission (the Commission) pertaining to the Companys Form 10-K for the fiscal year ended December 31, 2005, as discussed with Mr. Mark Brunhofer on October 13, 2006, December 15, 2006 and January 9, 2007 (the Comment Letter).

I have set forth each of the Staffs comments below, as communicated verbally by the Staff to the Company on October 13, 2006 and discussed again on December 15, 2006 and January 9, 2007, with the Companys corresponding response.

Defined terms used herein without definition have the meanings ascribed to them in the Companys Form 10-K for the fiscal year ended December 31, 2005, as filed on March 3, 2006.

* * * * *

Securities and Exchange Commission

January 11, 2007

Page 2

Form 10-K for the year ended December 31, 2005

In the Companys December 31, 2006 Form 10-K and subsequent Exchange Act filings, please expand your property catastrophe reinsurance reserving disclosure to provide more information on the ground up process that management undertakes to estimate these reserves as well as clarify the use of industry and statistical information sources used to estimate your property catastrophe reinsurance reserves.

In the Companys December 31, 2006 Form 10-K and subsequent Exchange Act filings, please expand your disclosure to describe more specifically how management completes its quarterly reserve review in the Companys specialty reinsurance and Individual Risk businesses.

In the Companys December 31, 2006 Form 10-K and subsequent Exchange Act filings, please identify and quantify the key reserving assumptions that materially affect managements estimates of reserves for claims and claim expenses. For each of your key assumptions, quantify and explain changes in these assumptions and describe what gave rise to these reserving assumption changes over the periods presented.

In the Companys December 31, 2006 Form 10-K and subsequent Exchange Act filings, please describe whether and to what extent the actual reserves for claims and claim expenses have significantly deviated from your initial reserve estimates over the periods presented.

In the Companys December 31, 2006 Form 10-K and subsequent Exchange Act filings, if you are unable to provide sensitivity analysis for the Companys December 31, 2006 Form 10-K, please disclose that you will provide such sensitivity analysis in future filings.

In the Companys December 31, 2006 Form 10-K and subsequent Exchange Act filings, please provide a sensitivity analysis that includes quantified and narrative disclosure of the impact that reasonably likely changes in the key assumptions identified may have on the Companys financial position for the Companys property catastrophe reinsurance business.

In response to all comments above, please refer to Exhibit A for disclosure the Company will include in its Claims and Claim Expense Reserves Critical Accounting Estimates discussion in its Annual Report on Form 10-K for the fiscal year ended December 31, 2006 and subsequent Exchange Act filings that addresses these comments.

Securities and Exchange Commission

January 11, 2007

Page 3

Should you have any questions concerning this letter please call John S. DAlimonte of Willkie Farr & Gallagher LLP at (212) 728-8212 or the undersigned at (441) 295-4513.

Very truly yours,

/s/ Mark A. Wilcox

Mark A. Wilcox

Senior Vice President
Corporate Controller and Chief Accounting Officer

cc:

Mr. Donald Abbott

Mr. Mark Brunhofer

Mr. Neill A. Currie

Mr. Fred R. Donner

Stephen H. Weinstein, Esq.

John S. DAlimonte, Esq.

Exhibit A

SUMMARY OF CRITICAL ACCOUNTING ESTIMATES

Claims and Claim Expense Reserves

General Description

We believe the most significant accounting judgment made by management is our estimate of claims and claim expense reserves. Claims and claim expense reserves represent estimates, including actuarial and statistical projections at a given point in time, of the ultimate settlement and administration costs for unpaid claims and claim expenses arising from the insurance and reinsurance contracts we sell. We establish our claims and claim expense reserves by taking claims reported to us by insureds and ceding companies, but which have not yet been paid (case reserves), adding the costs for additional case reserves (additional case reserves) which represent our estimates for claims previously reported to us which we believe may not be adequately reserved as of that date, and adding estimates for the anticipated cost of claims incurred but not yet reported to us (IBNR).

The following table summarizes our claims and claim expense reserves by line of business and split between case reserves, additional case reserves and IBNR at December 31, 2006 and 2005:


	Case Reserves	Additional Case Reserves	IBNR	Total
At December 31, 2006 (in thousands of U.S. dollars)

Property catastrophe reinsurance	$ xxx	$ xxx	$ xxx	$ xxx
Specialty reinsurance	xxx	xxx	xxx	xxx

Total Reinsurance	xxx	xxx	xxx	xxx
Indivisual Risk	xxx	xxx	xxx	xxx

Total	$ xxx	$ xxx	$ xxx	$ xxx

At December 31, 2005

Property catastrophe reinsurance	$ 544,750	$ 576,992	$ 207,087	$1,328,829
Specialty reinsurance	180,868	95,312	414,445	690,625

Total Reinsurance	725,618	672,304	621,532	2,019,454
Indivisual Risk	194,016	-	401,081	595,097

Total	$ 919,634	$ 672,304	$1,022,613	$2,614,551

Our estimates of claims and claim expense reserves are not precise in that, among other matters, they are based on predictions of future developments and estimates of future trends and other variable factors. Some, but not all, of our reserves are further subject to the uncertainty inherent in actuarial methodologies and estimates. Because a reserve

-1-

DRAFT

estimate is simply an insurers estimate at a point in time of its ultimate liability, and because there are numerous factors which affect reserves and claims payments but cannot be determined with certainty in advance, our ultimate payments will vary, perhaps materially, from our estimates of reserves. If we determine in a subsequent period that adjustments to our previously established reserves are appropriate, such adjustments are recorded in the period in which they are identified. During the twelve months ended December 31, 2006, 2005 and 2004, changes to prior year estimated claims reserves [decreased/increased] our net [income/loss] by $xxx.x million, reduced our net loss by $xxx.x million and increased our net income by $xxx.x million, respectively.

A 10% change to our reserves at December 31, 2006 would equate to a $xx.x million adjustment to net claims and claim expenses incurred, which represents xx.x% of our net [income/loss] [available/attributable] to common shareholders for the year ended 2006, and xx.x% of shareholders equity at December 31, 2006. During 2007, we intend to develop an analytical approach to quantifying reasonably likely changes to the variability of our key reserving assumptions embedded in our reserving estimation techniques. We will include the results of this analysis in our 2007 Annual Report on Form 10-K.

Our reserving methodology for each line of business uses a loss reserving process that calculates a point estimate for the Companys ultimate settlement and administration costs for claims and claim expenses. We do not calculate a range of estimates. We use this point estimate, along with paid claims and case reserves, to record our best estimate of additional case reserves and IBNR in our financial statements. Under GAAP, we are not permitted to establish estimates for catastrophe claims and claim expense reserves until an event occurs that gives rise to a loss.

Our claims and claim expense reserves are reviewed annually by an independent actuarial firm. The actuarial firm performs this work for the purpose of issuing an actuarial opinion on the reasonableness of the claims and claim expense reserves for each of the Companys insurance subsidiaries. The actuarial opinions are required to meet various insurance regulatory requirements. The actuarial firm discusses its conclusions with management and presents its findings to the Audit Committee of the Board of Directors of the Company. Although we do not explicitly rely on the work performed by the actuarial firm for estimating our reserves for claims and claim expenses, we compare our recorded claims and claim expense reserves to those estimated by the actuarial firm to determine whether our estimates are within the actuarial firms reasonable range of estimates. To date, our estimates of claims and claim expense reserves have been within the actuarial firms reasonable range of estimates.

Reserving for our reinsurance claims involves other uncertainties, such as the dependence on information from ceding companies, which among other matters, includes the time lag inherent in reporting information from the primary insurer to us or to our ceding companies and differing reserving practices among ceding companies. The information received from ceding companies is typically in the form of bordereaux, broker notifications of loss and/or discussions with ceding companies or their brokers. This information can be received on a monthly, quarterly or transactional basis and normally includes estimates of paid claims and case reserves. We sometimes also receive an estimate or provision for IBNR. This information is often updated and adjusted from

- 2 -

DRAFT

time-to-time during the loss settlement period as new data or facts in respect of initial claims, client accounts, industry or event trends may be reported or emerge in addition to changes in applicable statutory and case laws.

During 2005, we incurred significant losses from hurricanes Katrina, Rita and Wilma. Our estimates of these losses are based on factors including currently available information derived from claims information from our clients and brokers, industry assessments of losses from the events, proprietary models and the terms and conditions of our contracts. In particular, due to the size and unusual complexity of the issues relating to hurricane Katrina, meaningful uncertainty remains regarding total covered losses for the insurance industry and, accordingly, our loss estimates. Our actual losses from these events will likely vary, perhaps materially, from our current estimates due to the inherent uncertainties in reserving for such losses, the potential inaccuracies and inadequacies in the data provided by clients and brokers, the inherent uncertainty of modeling techniques and the application of such techniques, and complex coverage and other legal issues.

Because of the inherent uncertainties discussed above, we have developed a reserving philosophy which attempts to incorporate prudent assumptions and estimates, and we have generally experienced favorable development on prior year reserves in the last several years. However, there is no assurance that this will occur in future periods.

Our reserving techniques, assumptions and processes differ between our Reinsurance and Individual Risk segments, as well as between our property catastrophe reinsurance and specialty reinsurance businesses within our Reinsurance segment. Following is a discussion of the risks we insure and reinsure, the reserving techniques, assumptions and processes we follow to estimate our claims and claim expense reserves, and our current estimates versus our initial estimates of our claims reserves, for each of these units.

Reinsurance Segment

Property Catastrophe Reinsurance

Within our property catastrophe reinsurance unit, we principally write property catastrophe excess of loss reinsurance contracts to insure insurance and reinsurance companies against natural and man-made catastrophes. Under these contracts, we indemnify an insurer or reinsurer when its aggregate paid claims and claim expenses from a single occurrence of a covered peril exceed the attachment point specified in the contract, up to an amount per loss specified in the contract. Our most significant exposure is to losses from earthquakes and hurricanes and other windstorms, although we are also exposed to claims arising from other catastrophes, such as tsunamis, freezes, floods, fires, tornadoes, explosions and acts of terrorism. Our predominant exposure under such coverage is to property damage. However, other risks, including business interruption and other non-property losses, may also be covered under our property catastrophe reinsurance contracts when arising from a covered peril. Our coverages are offered on either a worldwide basis or are limited to selected geographic areas.

- 3 -

DRAFT

Coverage can also vary from "all property" perils to limited coverage on selected perils, such as "earthquake only" coverage. We also enter into retrocessional contracts that provide property catastrophe coverage to other reinsurers or retrocedants. This coverage is generally in the form of excess of loss retrocessional contracts and may cover all perils and exposures on a worldwide basis or be limited in scope to selected geographic areas, perils and/or exposures. The exposures we assume from retrocessional business can change within a contract term as the underwriters of a retrocedant may alter their book of business after the retrocessional coverage has been bound. We also offer dual trigger reinsurance contracts which require us to pay claims based on claims incurred by insurers and reinsurers in addition to the estimate of insured industry losses as reported by referenced statistical reporting agencies.

Our property catastrophe reinsurance business is generally characterized by loss events of low frequency and high severity. Initial reporting of paid and incurred claims in general, tends to be relatively prompt. We consider this business short-tail as compared to the reporting of claims for long-tail products, which tends to be slower. However, the timing of claims payment and reporting also varies depending on various factors, including: whether the claims arise under reinsurance of primary insurance companies or reinsurance of other reinsurance companies; the nature of the events (e.g., hurricanes, earthquakes or terrorism); the geographic area involved; post-event inflation which may cause the cost to repair damaged property to increase significantly from current estimates, or for property claims to remain open for a longer period of time, due to limitations on the supply of building materials, labor and other resources; and the quality of each clients claims management and reserving practices. Managements judgments regarding these factors are reflected in our claims reserve estimates.

Reserving for substantially all of our property catastrophe reinsurance business does not involve the use of traditional actuarial techniques. Rather, claims and claim expense reserves are estimated by management after a catastrophe occurs by completing an in-depth analysis of the individual contracts which may potentially be impacted by the catastrophic event. The in-depth analysis generally involves: 1) estimating the size of insured industry losses from the catastrophic event; 2) reviewing our portfolio of reinsurance contracts to identify those contracts which are exposed to the catastrophic event; 3) reviewing information reported by clients and brokers; 4) discussing the event with our clients and brokers; and 5) estimating the ultimate expected cost to settle all claims and administrative costs arising from the catastrophic event on a contract-by-contract basis and in aggregate for the event. Once an event has occurred, during the then current reporting period we record our best estimate of the ultimate expected cost to settle all claims arising from the event. Our estimate of claims and claim expense reserves is then determined by deducting cumulative paid losses from our estimate of the ultimate expected loss for an event and our estimate of IBNR is determined by deducting cumulative paid losses, case reserves and additional case reserves from our estimate of the ultimate expected loss for an event. Once we receive a notice of loss or payment request under a catastrophe reinsurance contract, we are generally able to process and pay such claims promptly.

- 4 -

DRAFT

Because the events from which claims arise under policies written by our property catastrophe reinsurance business are typically prominent, public occurrences such as hurricanes and earthquakes, we are often able to use independent reports as part of our loss reserve estimation process. We also review catastrophe bulletins published by various statistical reporting agencies to assist us in determining the size of the industry loss, although these reports may not be available for some time after an event. In addition to the loss information and estimates communicated by cedants and brokers, we also use industry information which we gather and retain in our REMS© modeling system. The information stored in our REMS© modeling system enables us to analyze each of our policies against a loss and compare our estimate of the loss with those reported by our policyholders. The REMS© modeling system also allows us to compare and analyze individual losses reported by policyholders affected by the same loss event. Although the REMS© modeling system assists with the analysis of the underlying loss and provides us with the information and ability to perform increased analysis, the estimation of claims resulting from catastrophic events is inherently difficult because of the variability and uncertainty associated with property catastrophe claims and the unique characteristics of each loss.

For smaller events including localized severe weather events such as windstorms, hail ice, snow, flooding, freezing and tornadoes, which are not necessarily prominent, public occurrences, we initially place greater reliance on catastrophe bulletins published by statistical reporting agencies to assist us in determining what events occurred during the reporting period than we do for large events. This includes reviewing Catastrophe Bulletins published by Property Claim Services for U.S. catastrophes. We set our initial estimates of reserves for claims and claim expenses for these smaller events based on a combination of our historical market share for these types of losses and the estimate of the total insured industry property losses as reported by statistical reporting agencies, although we generally make significant adjustments based on our current exposure to the geographic region involved as well as the size of the loss and the peril involved. This approach supplements our approach for estimating losses for larger catastrophes, which as discussed above, includes discussions with brokers and ceding companies, reviewing individual contracts impacted by the event, and modeling the loss in our REMS© system.

In general, our property catastrophe reinsurance reserves for our more recent reinsured catastrophic events are subject to greater uncertainty and, therefore, greater potential variability, and are likely to experience material changes from one period to the next. This is due to the uncertainty as to the size of the industry losses from the event, uncertainty as to which contracts have been exposed to the catastrophic event, and uncertainty as to the magnitude of claims incurred by our clients. As our property catastrophe reinsurance reserves age, more information becomes available and we believe our estimates become more certain, although there is no assurance this trend will continue in the future. As seen in the Actual vs. Expected Property Catastrophe Reinsurance Claims and Claim Expenses Reserve Analysis table below, xx.x% of our inception to

DRAFT

date claims and claim expenses in our property catastrophe reinsurance unit were incurred in the 2004 and 2005 accident years due principally to the losses from hurricanes Charley, Frances, Ivan, Jeanne, Katrina, Rita and Wilma. Due to the size and complexity of the losses in these accident years, there still remains significant uncertainty as to the ultimate settlement costs associated with these years.

Within our property catastrophe reinsurance business, we seek to review substantially all of our claims and claim expense reserves quarterly. Our quarterly review procedures include identifying events that have occurred up to the latest balance sheet date, determining our best estimate of the ultimate expected cost to settle all claims and administrative costs associated with those new events which have arisen during the reporting period, and reviewing the ultimate expected cost to settle claims and administrative costs associated with those events which occurred during previous periods. This process is judgmental in that it involves reviewing changes in paid and reported losses each period and adjusting our estimates of the ultimate expected losses for each event if there are developments that are different from our previous expectations. If we determine that adjustments to an earlier estimate are appropriate, such adjustments are recorded in the period in which they are identified. During the twelve months ended December 31, 2006, 2005 and 2004, changes to our prior year estimated claims reserves in our property catastrophe reinsurance unit [decreased/increased] our net [income/loss] by $xxx.x million, reduced our net loss by $xxx.x million and increased our net income by $xxx.x million, respectively.

Actual Results vs. Initial Estimates

The table below summarizes our initial assumptions and changes in those assumptions for claims and claim expense reserves within our property catastrophe reinsurance unit. As discussed above, the key assumption in estimating reserves for our property catastrophe reinsurance unit is our estimate of ultimate claims and claim expenses. The table shows our initial estimates of ultimate claims and claim expenses for each accident year and how these initial estimates have developed over time. The initial estimate of accident year claims and claim expenses represents our estimate of the ultimate settlement and administration costs for claims incurred from catastrophic events occurring during a particular accident year, and as reported as of December 31 of that year. The reestimated ultimate claims and claim expenses as of December 31, 2004, 2005 and 2006 represent our revised estimates as reported as of those dates. The cumulative favorable (adverse) development shows how our most recent estimates as reported at December 31, 2006 differ from our initial accident year estimates. Favorable development implies that our current estimates are lower than our initial estimates while adverse development implies that our current estimates are higher than our original estimates. Total reserves as of December 31, 2006 reflect the unpaid portion of our estimates of ultimate claims and claim expenses. The table is presented on a gross loss basis and therefore does not include the benefit of reinsurance recoveries.

DRAFT

Actual vs. Expected Property Catastrophe Reinsurance Claims and Claim Expense Reserve Analysis

(in thousands of U.S. Dollars)

Accident Year

Initial Estimate

of Ultimate

Claims and

Claim Expenses

Reestimated Ultimate Claims and Claim Expenses as

of December 31,

Cumulative

Favorable

(Adverse)

Development

% Decrease

(Increase) from

Initial Ultimate

Total Reserves

as of December 31,

2006

% of Claims

Unpaid as of

December 31,

2006

2004

2005

2006

2001 and Prior

$xxx

xx.x%

$xxx

xx.x%

2002

xxx

xx.x%

xxx

xx.x%

2003

xxx

xx.x%

xxx

xx.x%

2004

xxx

xx.x%

xxx

xx.x%

2005

xxx

xx.x%

xxx

xx.x%

2006

xxx

xx.x%

xxx

xx.x%

Total

$xxx

xx.x%

$xxx

xx.x%

[Comments on above that respond to comment 4 in this Comment Letter and comment 4 in the original comment letter dated July 28, 2006.]

Specialty Reinsurance

Within our specialty reinsurance business unit we write a number of reinsurance lines such as catastrophe exposed workers compensation, surety, terrorism, medical malpractice, casualty clash, property per risk and other specialty lines of reinsurance, which we collectively refer to as specialty reinsurance. We offer our specialty reinsurance products principally on an excess of loss basis, as described above with respect to our property catastrophe reinsurance products, and we also provide some proportional coverage. In a proportional reinsurance arrangement (also referred to as quota share reinsurance or pro-rata reinsurance), the reinsurer shares a proportional part of the original premiums and losses of the reinsured. We offer our specialty reinsurance products to insurance companies and other reinsurance companies and provide coverage for specific geographic regions or on a worldwide basis. We expanded our specialty reinsurance business in 2002 and have significantly increased our presence in the specialty reinsurance market since that time.

Our specialty reinsurance business can generally be characterized as providing coverage for low frequency and high severity losses, similar to our property catastrophe reinsurance business. As with our property catastrophe reinsurance business, our specialty reinsurance contracts frequently provide coverage for relatively large limits or exposures. As a result of the foregoing, our specialty reinsurance business is subject to significant claims volatility. In periods of low claims severity our results will generally be favorably impacted while in periods of high claims severity our results will generally be negatively impacted.

Our processes and methodologies, in respect of loss estimation for the coverages we offer through our specialty reinsurance operation differ from those used for our property catastrophe-oriented coverages. For example, our specialty reinsurance coverages are

DRAFT

more likely to be impacted by factors such as long-term inflation and changes in the social and legal environment, which we believe gives rise to greater uncertainty in our claims reserves. Moreover, in reserving for our specialty reinsurance coverages we do not have the benefit of a significant amount of our own historical experience in these lines. We believe this makes our specialty reinsurance reserving subject to greater uncertainty than our property catastrophe reinsurance unit.

When initially developing our reserving techniques for our specialty reinsurance coverages, we considered estimating reserves utilizing several actuarial techniques such as paid and incurred development methods. We elected to use the Bornhuetter-Ferguson technique because this method is appropriate for lines of business, such as our specialty reinsurance business, where there is a lack of historical claims experience. This method allows for greater weight to be applied to expected results in periods where little or no actual experience is available, and, hence, is less susceptible to the potential pitfall of being excessively swayed by one year or one quarter of actual paid and/or reported loss data. This method uses initial expected loss ratio expectations to the extent that losses are not paid or reported, and it assumes that past experience is not fully representative of the future. As the Companys reserves for claims and claim expenses age, and actual claims experience becomes available, this method places less weight on expected experience and places more weight on actual experience. We reevaluate our actuarial reserving techniques on a periodic basis.

The utilization of the Bornhuetter-Ferguson technique requires us to estimate an expected ultimate claims and claim expense ratio and select an expected loss reporting pattern. We select our estimates of the expected ultimate claims and claim expense ratios and expected loss reporting patterns by reviewing industry standards and adjusting these standards based upon the terms of the coverages we offer. The estimated expected claims and claim expense ratio may be modified to the extent that reported losses at a given point in time differ from what would be expected based on the selected loss reporting pattern. Our estimate of IBNR is the product of the premium we have earned, the initial expected ultimate claims and claim expense ratio and the percentage of estimated unreported losses. In addition, certain of our specialty reinsurance coverages may be impacted by natural and man-made catastrophes. We estimate claim reserves for these losses after the event giving rise to these losses occurs, following a process that is similar to our property catastrophe reinsurance unit described above.

Within our specialty reinsurance business, we seek to review substantially all of our claims and claim expense reserves quarterly. Typically, our quarterly review procedures include reviewing paid and reported claims in the most recent reporting period, reviewing the development of paid and reported claims from prior periods, and reviewing our overall experience by underwriting year and in the aggregate. We monitor our expected ultimate claims and claim expenses ratios and expected loss reporting assumptions on a quarterly basis and compare them to our actual experience. These actuarial assumptions are generally reviewed annually, based on input from our actuaries, underwriters, claims personnel and finance professionals, although adjustments may be made more frequently if needed. Assumption changes are made to adjust for changes in the pricing and terms of coverage we provide, changes in industry standards, as well as our actual experience, to the extent we have enough data to rely on our own experience. If we determine that adjustments to an earlier estimate are appropriate, such adjustments are recorded in the period in which they are identified. During the twelve months ended December 31, 2006, 2005 and 2004, changes to our prior year estimated claims reserves in our specialty

DRAFT

reinsurance unit [decreased/increased] our net [income/loss] by $xxx.x million, reduced our net loss by $xxx.x million and increased our net income by $xxx.x million, respectively.

Actual Results vs. Initial Estimates

The Actual vs. Initial Expected Ultimate Claims and Claim Expense Ratio table below summarizes our key actuarial assumptions in reserving for our specialty reinsurance business. As noted above, the key actuarial assumptions include the estimated ultimate claims and claim expense ratios and the estimated loss reporting patterns. The table shows our initial estimates of the ultimate claims and claim expenses ratio by underwriting year. The table shows how our initial estimates of these ratios have developed over time. The actual experience reflects the combination of how actual experience in the initial expected ultimate losses and expected reported losses compares with our initial estimates. The initial estimate is based on the actuarial assumptions that were in place at the end of that year. A decrease in the ultimate claims and claim expense ratio implies that our current estimates are lower than our initial estimates while an increase in the ultimate claims and claim expense ratio implies that our current estimates are higher than our initial estimates. The result would be a corresponding favorable impact on shareholders equity and net [income/loss] or a corresponding unfavorable impact on shareholders equity and net [income/loss], respectively. The table also shows how our initial estimated ultimate claims and claim expense ratios have changed from one underwriting year to the next. The table below reflects a summary of the weighted average assumptions for all classes of business written within our specialty reinsurance unit. The table is presented on a gross loss basis and therefore does not include the benefit of reinsurance recoveries.


Actual vs. Initial Expected Claims and Claim Expense Reserve Analysis
Estimated Ultimate Claims and Claim Expense Ratio

Year		Estimates of Expected Ultimate Claims and Claim Expense Ratios
Year	Initial Estimate	Reestimate as of December 31, 2004	Reestimate as of December 31, 2005	Reestimate as of December 31, 2006
2002	xx.x%	xx.x%	xx.x%	xx.x%
2003	xx.x%	xx.x%	xx.x%	xx.x%
2004	xx.x%	xx.x%	xx.x%	xx.x%
2005	xx.x%		xx.x%	xx.x%
2006	xx.x%			xx.x%

[Comments on above that respond to comment 4 in this Comment Letter and comment 4 in the original comment letter dated July 28, 2006.]

Individual Risk Segment

We define our Individual Risk segment to include underwriting that involves understanding the characteristics of the underlying insurance policy. Our principal contracts include insurance contracts and quota share reinsurance contracts with respect to risks currently including: 1) commercial and homeowners property coverages, including catastrophe-exposed products; 2) commercial liability coverages, including

DRAFT

general, automobile, professional and various specialty products; 3) multi-peril crop insurance; and 4) reinsurance of other insurers on a quota share basis. We expanded our Individual Risk operations in 2003 and started writing a significant amount of this business in that year.

We use the Bornhuetter-Ferguson technique to estimate claims and claim expenses within our Individual Risk segment. The comments discussed above relating to our reserving techniques and processes for our specialty reinsurance unit also apply to our Individual Risk segment.

During the twelve months ended December 31, 2006, 2005 and 2004, changes to our prior year estimated claims reserves in our Individual Risk unit [decreased/increased] our net [income/loss] by $xxx.x million, reduced our net loss by $xxx.x million and increased our net income by $xxx.x million, respectively.

Actual Results vs. Initial Estimates

The Actual vs. Initial Expected Ultimate Claims and Claim Expense Ratio table below summarizes our key actuarial assumptions in reserving for our Individual Risk unit. As noted above, the key actuarial assumptions include the estimated ultimate claims and claim expense ratios and the estimated loss reporting patterns. The table shows our initial estimates of the ultimate claims and claim expenses ratio by underwriting year. The table shows how our initial estimates of these ratios have developed over time. The actual experience reflects the combination of how actual experience in the initial expected ultimate losses and expected reported losses compares with our initial estimates. The initial estimate is based on the actuarial assumptions that were in place at the end of each fiscal year. A decrease in the ultimate claims and claim expense ratio implies that our current estimates are lower than our initial estimates while an increase in the ultimate claims and claim expense ratio implies that our current estimates are higher than our initial estimates. The result would be a corresponding favorable impact on shareholders equity and net [income/loss] or a corresponding unfavorable impact on shareholders equity and net [income/loss], respectively. The table also shows how our initial estimated ultimate claims and claim expense ratios have changed from one fiscal year to the next. The table below reflects a summary of the weighted average assumptions for all classes of business written within our Individual Risk unit. The table is presented on a gross loss basis and therefore does not include the benefit of reinsurance recoveries.

DRAFT


Actual vs. Initial Expected Claims and Claim Expense Reserve Analysis
Estimated Ultimate Claims and Claim Expense Ratio

Year		Estimates of Expected Ultimate Claims and Claim Expense Ratios
Year	Initial Estimate	Reestimate as of December 31, 2004	Reestimate as of December 31, 2005	Reestimate as of December 31, 2006
2002	xx.x%	xx.x%	xx.x%	xx.x%
2003	xx.x%	xx.x%	xx.x%	xx.x%
2004	xx.x%	xx.x%	xx.x%	xx.x%
2005	xx.x%		xx.x%	xx.x%
2006	xx.x%			xx.x%

[Comments on above that respond to comment 4 in this Comment Letter and comment 4 in the original comment letter dated July 28, 2006.]

2005 Reserve Reviews

We announced in early 2005 that during 2005 we would review the processes and assumptions for establishing and evaluating our reserves during 2005. Starting in 2002 and 2003, we significantly expanded our specialty reinsurance and Individual Risk businesses, respectively, which resulted in significant growth in our reserves for claims and claim expenses for these lines of business over the subsequent three years. When establishing our initial reserves for these lines of business, we placed significant reliance on industry data as we did not have the benefit of our own historical experience. Prior to 2005, we started experiencing favorable development on the run-off of our reserves. This favorable development was principally due to our specialty reinsurance and Individual Risk businesses performing better than historical industry averages and paid and reported claims developing better than expected in our property catastrophe reinsurance business. We elected to perform our reserve reviews in 2005 principally due to: 1) the growth in our businesses and reserves for claims and claim expenses; 2) an increase in then-recent periods in favorable development in reserves for our specialty reinsurance and Individual Risk businesses, which had performed more favorably in what was effectively their first several years of operations than historical industry data would have anticipated; 3) recent favorable development with respect to our property catastrophe reinsurance reserves; and 4) a determination to consider enhancements to our reserving processes and assumptions.

We completed reviews of our property catastrophe reinsurance reserves, specialty reinsurance reserves and Individual Risk reserves in the second, third and fourth quarters of 2005, respectively. As a result of these reviews, we reduced prior year reserves within our Reinsurance and Individual Risk segments by $248.1 million and $1.1 million, respectively. Within our Reinsurance segment, the decrease included a $118.2 million decrease in reserves for our property catastrophe reinsurance business and a $129.9 million decrease in reserves for our specialty reinsurance business. After adjusting for the impact of minority interest in DaVinci, our 2005 net loss was reduced by $226.9 million as a result of these reviews.

DRAFT

Within our Reinsurance segment, we experienced a decrease in our property catastrophe reinsurance reserves for claims and claim expenses principally due to reestimating the ultimate expected cost for both small and large catastrophes. This was due to a reassessment of our reserves for claims and claim expenses in light of historical paid loss trends and reported loss activity for these catastrophes. In coming to this conclusion, we reviewed claims files, requested updated claims information from brokers, and developed some new tools to help us assess trends in paid and reported losses. As a result of this review, we determined that there were trends of paid and reported claims for these catastrophes that were coming in less than originally expected and therefore the ultimate expected losses for these catastrophes were decreased.

The review also included a reassessment of our actuarial techniques and assumptions for our specialty reinsurance business. We considered whether we should continue with the use of the Bornhuetter-Ferguson actuarial method in reserving for our specialty reinsurance business or whether a different actuarial method might be more appropriate. As a result of the review, we determined it was appropriate to continue with the use of the Bornhuetter-Ferguson actuarial method.

To enhance the reserving for our specialty reinsurance business, we further segmented this business during the review with the aim of grouping risks into more homogeneous categories which respond to the evolution of actual exposures. This became possible as the volume of this business increased over the three preceding years. This further segmentation required the selection of loss reporting patterns to be applied to these new groups. We also updated our assumptions for our original loss reporting patterns based on a combination of new industry information and actual experience accumulated over the three preceding years. The assumptions for the new loss reporting patterns were applied to all prior underwriting years. In addition, we made explicit allowances for commuted contracts whereas previously these were considered in the overall reserving assumptions. We also reviewed substantially all of our case reserves and additional case reserves. The result of the foregoing was a decrease in our specialty reinsurance reserves for claims and claim expenses.

The reserve review for our Individual Risk business followed a similar process to the specialty reinsurance review noted above. The changes within our Individual Risk segment as a result of the reserve review were insignificant.

DRAFT