The following is a discussion of the results of operations and financial position of IPC Holdings, Ltd. References to "we", "our", "IPC", or "the Company" mean IPC Holdings together with its wholly-owned subsidiaries, IPCRe and IPCUSL. This discussion should be read in conjunction with our Consolidated Financial Statements and related notes for the quarter ended September 30, 2008.

CRITICAL ACCOUNTING POLICIES

Our significant accounting policies are described in Note 2 to our audited financial statements for the fiscal year ended December 31, 2007, included in the Company's Annual Report on Form 10-K filed with the Securities and Exchange Commission ("SEC") on February 22, 2008. The following is a summary of the accounting policies for the three main components of our consolidated balance sheet and consolidated statements of (loss) income: premiums, losses (claims), including reserves and investments/investment income.

Premiums

Premiums are recorded as written at the beginning of each policy, based upon information received from ceding companies and their brokers, and are earned on a pro rata basis over the policy period. For excess of loss contracts, the amount of deposit premium is contractually documented at inception, and management uses this as its best estimate for accounting for these premiums. Premiums may be adjusted upwards or downwards as a result of changes in the cedant's actual exposure base and the original estimates thereof, although most contracts do provide for a minimum premium in the contract terms. We refer to such changes in premiums as adjustment premiums. For proportional treaties, the amount of premium is normally estimated at inception using data provided by the ceding company. We account for such premium using our initial estimates, which are reviewed regularly for reasonableness by management with respect to the actual premium reported by the ceding company. At September 30, 2008 the amount of premium accrued resulting from management's estimates for proportional treaties was 3.5% of total gross premiums written for the nine months then ended. Reinstatement premiums are recognized and accrued at the time we incur a loss and where coverage of the original contract is reinstated under pre-defined contract terms, and are earned pro rata over the reinstated coverage period. Such accruals are based upon actual contractual terms (including the associated rates on line-i.e. price) applied to the amount of loss reserves expected to be paid, and the only element of management judgement involved is with respect to the amount of loss reserves as described below. The amount accrued at September 30, 2008 for estimated reinstatement premiums on Reported But Not Enough loss reserves ("RBNE") and Incurred But Not Reported ("IBNR") loss reserves, as described below, was $30.0 million.

Loss Reserves

Under accounting principles generally accepted in the United States of America, we are not permitted to establish loss reserves until the occurrence of an event that may give rise to a claim. As a result, only loss reserves applicable to losses incurred up to the reporting date are established, with no allowance for the provision of a contingency reserve to account for expected future losses. Claims arising from future catastrophic events can be expected to require the establishment of substantial reserves from time to time.

Estimating appropriate loss reserves for catastrophes is an inherently uncertain process. Loss reserves represent our estimates, at a given point in time, of ultimate settlement and administration costs of losses incurred (including IBNR and RBNE reserves). We regularly review and update these estimates, using the most current information available to us. Consequently, the ultimate liability for a catastrophic loss is likely to differ from the original estimate. Whenever we determine that any existing loss reserves are inadequate, we are required to increase the loss reserves with a corresponding reduction, which could be material, in our operating results in the period in which the deficiency is identified. The establishment of new reserves, or the adjustment of reserves for reported claims, could result in significant upward or downward changes to our financial condition or results of operations in any particular period.

For proportional treaties, we generally use an initial estimated loss and loss expense ratio (the ratio of losses and loss adjustment expenses incurred to premiums earned) based upon information provided by the ceding company and/or their broker and our historical experience of that treaty, if any, and the estimate is adjusted as actual experience becomes known.

When a catastrophic event occurs, we first determine which treaties may be affected using our geographic database of exposures. We then contact the respective brokers and ceding companies involved with those treaties, to determine their estimate of involvement and the extent to which the reinsurance program is affected. We may also use computer modeling to measure and estimate loss exposure under the actual event scenario, if available. For excess of loss business, which is generally over 90% of the premiums we write, we are aided by the fact that each treaty has a defined limit of liability arising from one event. Once that limit has been reached, we have no further exposure to additional losses from that treaty for the same event.

As a broker market reinsurer, we are reliant on loss information reported to brokers by primary insurers who must estimate their own losses at the policy level. These estimates are sometimes derived from the output of computer-based modelling systems, and often based upon incomplete and changing information, especially during the period immediately following a catastrophic event. The information we receive varies by cedant and broker and may include paid losses and estimated case reserves. We may also receive an estimated provision for IBNR reserves, especially when the cedant is providing data in support of a request for collateral for reserves ceded. Information can be received on a monthly, quarterly or transactional basis. As a reinsurer, our reserve estimates may be inherently less reliable than the reserve estimates of our primary insurer cedants.

There is a time lag inherent in reporting from the original claimant to the primary insurer to the broker and then to the reinsurer. Reporting of property claims arising from catastrophes in general tends to be prompt (as compared to reporting of claims for casualty or other "long-term" lines of business). However, the timing of claims reporting can vary depending on various factors, including: the nature of the event (e.g. hurricane, earthquake, hail, man-made events such as terrorism or rioting); the geographic area involved; the quality of the cedant's claims management and reserving practices; and whether the claims arise under reinsurance contracts for primary companies, or reinsurance of other reinsurance companies (i.e. retrocession). Because the events from which claims arise are typically prominent, public occurrences, we are often able to use independent reports of such events to augment our loss reserve estimation process. Because of the degree of reliance that we place on ceding companies for claims reporting, the associated time lag, the low frequency/high severity nature of the business we underwrite and the varying reserving practices among ceding companies, our reserve estimates are highly dependent on management judgement and are therefore subject to significant variability. During the loss settlement period additional facts regarding individual claims and trends may become known and current laws and case law may change, which further increases the variability of our reserves.

IPC's controls in place require that claim payments and reserves must be authorized by an underwriter upon processing. Large claims must also be approved by senior management prior to a claims payment being made. While we have the right to audit client data, most of our claims result from events that are well known such as hurricanes or earthquakes, and in assessing the reasonableness of reported claims, our claims processors and underwriters ask follow-up questions as necessary. Since 2007, we also undertake a verification process of the completeness of our loss reserves, including the circularization of a number of our brokers. We also cross reference and verify amounts requested as collateral by ceding companies, in comparison to amounts previously reported to us.

For certain catastrophic events, there is great uncertainty underlying the assumptions and associated estimated reserves for losses and loss adjustment expenses reported by our cedants. Complexity resulting from problems such as policy coverage issues, multiple events affecting one geographic area and the resulting impact on claims adjusting (including allocation of claims to event and the effect of demand surge on the cost of building materials and labour) by, and communications from, ceding companies, can cause greater uncertainty in the reserve estimates reported to us by our cedants, as well as delays to the timing with which we are notified of cedants' changes to their loss estimates, resulting in greater uncertainty in our reserve estimates. For example, the initial estimate for hurricane Katrina had been based on estimates by cedants of their exposure, industry insured loss estimates, output from both industry and proprietary models, a review of contracts potentially affected by the events, information received from both clients and brokers, and management judgement, which includes consideration of the physical factors noted above, in aggregate. It has been assumed that underlying policy terms and conditions are upheld during our clients' loss adjustment process. However, the unique circumstances and severity of this devastating catastrophe, including the extent of flooding and resultant initial limited access by claims adjusters, introduced additional uncertainty to the normally difficult process of estimating catastrophe losses, which is compounded by the potential for legal and regulatory issues arising regarding the scope of coverage.

                                               Percentage increase of development from initial report
                                                                                                                       Total Development -
                                       After 6          After 1          After 2        After 3       Latest /       initial report to latest
                                       months            year             years          years        Final %                 $(000)
Cyclones Lothar/Martin                       61 %             66 %             71 %          73 %           69 %                       24,100
Cat # 48 (WTC)                                6 %              7 %              9 %           9 %            4 %                        4,500
2004 Florida hurricanes                     113 %            137 %            145 %         144 %          145 %                      118,000
Hurricane Katrina                           1.5 %            1.6 %           -0.4 %        -2.0 %         -2.0 %                      (16,000 )
UK floods (June 2007)                       -38 %            -43 %                                         -45 %                      (28,000 )

Generally, the most significant development arises within six to nine months of an event, due to the limited amount of information usually available immediately after the event.

Cyclones Lothar and Martin struck France and other parts of Europe in the last week of 1999. As such, many parts of the affected areas were still devastated, inaccessible and without power at the time we were attempting to establish reserves for 1999 year-end reporting. In many cases, our French cedants were unable to provide us with much information regarding their potential claims, and we relied more heavily on industry loss estimates, which themselves were based on very limited information. Consequently, there was significant development of our own loss, as well as for the reported industry loss. As an example, the reported industry loss for cyclone Martin increased 69% from the original estimate.

Similarly, for the four hurricanes that struck Florida over a six-week period concluding in late September 2004, not only was the initial estimation process made difficult by the proximity to the end of the third quarter 2004 reporting period, there were the added complexities of multiple events affecting one geographic area and the resulting impact on claims adjusting (as noted above) by, and communications from, ceding companies.

The initial estimation process for the flooding which impacted parts of northern England in June 2007 was also made difficult by the proximity to the end of the second quarter 2007 reporting period. The flood waters had not fully receded at the time that the initial estimate was prepared.

Particularly for extreme events such as the attack on the World Trade Center and hurricane Katrina, many excess of loss contracts that are impacted by the event incur full limit losses, on which there can be no adverse development. However, because of the uncertainties associated with hurricane Katrina noted above, there can be no assurance that significant development will not occur on contracts where the limits have not been exhausted, or that losses will not be reported for contracts for which we have not previously established a reserve. Generally, the size of a catastrophe is not necessarily an indicator of the amount of potential development that might occur. However, for larger catastrophes, a small percentage of development can result in a larger dollar impact on a company's results of operations, than a larger percentage development on a smaller event.

As noted above, our methodology provides us with an overall estimate of loss reserves for IPC as a whole. For information on historical development of IPC's overall loss reserves, please refer to the tables provided in Item 1, Business, Reserve for Losses and Loss Adjustment Expenses, in our Report on Form 10-K for the year ended December 31, 2007.

At September 30, 2008 management's estimates for IBNR/RBNE represented 59% of total loss reserves. The majority of the estimate relates to reserves for claims from hurricane Ike that affected various parts of Gulf coast states in September 2008, reserves for claims from hurricane Katrina that affected various parts of Gulf coast states in August 2005, the storm and flooding that affected parts of New South Wales, Australia in early June 2007, and the flooding events that affected various parts of the United Kingdom in June and July 2007. Given the magnitude and recent occurrence of hurricane Ike, delays in receiving claims data, the contingent nature of business interruption and other exposures, and other uncertainties inherent in loss estimation, meaningful imprecision remains regarding the estimated losses from this event. Our initial assessment of losses is based on a combination of our analysis and review of our share of estimated total industry loss, in-force contracts, the output of catastrophe modeling and a limited number of loss advices from clients. As discussed above, our reserve estimates are not mathematically or formulaically derived from factors such as numbers of claims or demand surge impact. If our estimate of IBNR/RBNE loss reserves at September 30, 2008 was inaccurate by a factor of 10%, our results of operations would be impacted by a positive or negative movement of $21.5 million. If our total reserve for losses at

Investments

In accordance with our investment guidelines, our investments consist of high-grade marketable fixed maturity investments, certain equity investments in mutual funds, and mortgage backed securities guaranteed by U.S. Government sponsored entities and, as part of the mortgage backed securities portfolio, to-be-announced securities ("TBAs"). Mortgage backed securities are typically traded on a to-be-announced basis. By acquiring a TBA, the Company makes a commitment to purchase a future issuance of mortgage backed securities. In addition, the portfolio may offset these purchases by taking positions in short TBAs (i.e. the Company makes a commitment to sell a future issuance of mortgage backed securities) but no net short TBA positions are allowed. As part of the mortgage backed securities portfolio, the investment guidelines allow the Company to enter into long and short TBA positions, which are deliverable within a month. No net short TBAs are allowed. By acquiring a TBA, the Company makes a commitment to purchase a future issuance of mortgage-backed securities. For the period between purchase of the TBA and issuance of the underlying security, the Company's position is accounted for as a regular-way security trade in the consolidated financial statements where delivery is taken. Where delivery is not taken, the Company's position is accounted for as a derivative in the consolidated financial statements. As at September 30, 2008, the Company had both long and short TBAs. Collateral is received or posted daily between the Company and the counterparty for the daily net gain (loss). TBAs are recorded in the consolidated financial statements at fair value.

In accordance with our adoption of Statement of Financial Accounting Standards ("SFAS") No. 159 "The Fair Value Option for Financial Assets and Financial Liabilities" on January 1, 2007 the investments are now reported as "Trading" under SFAS 115. Simultaneous to the adoption of SFAS No 159 we also adopted SFAS No. 157, "Fair Value Measurements" regarding fair value measurements, which required all unrealized gains and losses in our investment portfolio to be reclassified from accumulated other comprehensive income within shareholders' equity on our consolidated balance sheets to retained earnings as of January 1, 2007. This cumulative-effect adjustment reclassifying unrealized gains and losses was $128.0 million, which represented the difference between the cost or amortized cost of our investments and the fair value of those investments at December 31, 2006.

With the investments now being classified as "Trading" all subsequent changes to the fair value of our investment portfolio are recorded as net (losses) gains on investments in our consolidated statements of income (loss).

SFAS 157 established a hierarchy for inputs in measuring fair value that maximizes the use of observable inputs and minimizes the use of unobservable inputs by requiring that the observable inputs are used when available. Observable inputs are inputs that market participants would use in pricing the asset based on market data obtained from sources independent of the Company. Unobservable inputs are inputs that reflect the Company's assumptions about the assumptions that market participants would use in pricing the asset based on the best information available in the circumstances. The hierarchy is broken down into three levels based on the reliability of inputs as follows:

Level 1-Valuations based on quoted prices in active markets for identical assets or liabilities that the Company has the ability to access. Since valuations are based on quoted prices that are readily and regularly available in an active market, valuation of these instruments does not entail a significant degree of judgment.

Level 2-Valuations based on quoted prices in markets that are not active or for which all significant inputs are observable, directly or indirectly.

Level 3-Valuations based on inputs that are unobservable and significant to the overall fair value measurement.

The following table shows how our investments are categorized.

Fair Value Measurement at September 30, 2008 using:

                                     Fair Value            Quoted Prices        Significant Other          Significant
                                  Measurements at        in Active Markets         Observable          Unobservable Inputs
                                 September 30, 2008          (Level 1)          Inputs (Level 2)            (Level 3)
Description                       ($ in thousands)       ($ in thousands)       ($ in thousands)        ($ in thousands)
Fixed maturity investments                1,753,202                     -               1,753,202                       -
AIG Select Hedge Fund                       173,936                     -                      -                   173,936
Other Equity investments                    181,444                     -                 181,444                       -

The AIG Select Hedge Fund ("Select Hedge") has a monthly reported net asset value - where the fund's holdings can be in various publicly quoted and unquoted investments with observable inputs. Because Select Hedge invests in 30-40 underlying third party funds, there is a one-month delay in its valuation. As a result, the most recently advised net asset value ("NAV") of Select Hedge included in our financial statements at the end of each quarter has historically been based upon the NAVs of the underlying funds at the end of the preceding month. The Select Hedge had been classified as Level 2 as at June 30, 2008 because we were using the final NAV which was an observable input. However, due to the significant market volatility towards the end of the quarter ended September 30, 2008 we have included an estimate of the performance of Select Hedge for the month of September, 2008. The estimate is obtained from the fund's investment manager who derives an estimate of the performance of the Select Hedge based on the month end positions from the underlying third party funds. The use of the estimate in respect of the Select Hedge increases the level of unobservable inputs. Hence, we are classifying the Select Hedge as Level 3. In order to obtain comfort over the reasonableness of this estimate, we assessed the difference between the estimates and final month-end NAVs to ensure that there have been no significant variances in the past. Any movement in the estimated NAV relative to the final NAV of Select Hedge would be recorded in the following reporting period. There is up to a two-month delay prior to receiving funds withdrawn from our Select Hedge investment.

A review of fair value hierarchy classifications is conducted on a quarterly basis. Changes in the observability of valuation inputs may result in a reclassification for certain financial assets and liabilities. Reclassifications impacting Level 3 of the fair value hierarchy are reported as transfers in or out of the Level 3 category as of the beginning of the quarter in which the reclassifications occur.

The following table presents the fair market value of the Company's Level 3 financial assets (and liabilities) as at September 30, 2008.

                                                           Fair value measurements using
                                                             significant unobservable
                                                                 inputs (Level 3)
                                                                   Quarter ended
                                                                September 30, 2008
                                                                 ($ in thousands)
Beginning balance as at July 1, 2008                                                  -
Transfers in and/or out of Level 3                                               192,254
Total gains and losses (realized/unrealized)
included in earnings                                                             (18,318 )
Purchases, issuance, and settlements                                                  -

Ending balance at September 30, 2008                                             173,936

The amount of total gains or losses for the
period included in earnings attributable to the
change in unrealized gains or losses relating to
assets still held at September 30, 2008                                          (18,318 )

Other equity investments represent an investment in two mutual funds. These funds are stated at fair value as determined by the most recently reported net asset value as advised by the fund. These funds have daily reported net asset values - with the funds' holdings predominantly in publicly quoted securities. Due to the funds' values being current net asset values, with no significant lock ups, no delays in withdrawal and not publicly quoted prices, we are classifying other equity investments as Level 2.

Realized gains and losses on sales of investments continue to be determined on a first-in, first-out basis. Net investment income includes interest income on fixed maturity investments, recorded when earned, dividend income on equity investments, recorded when declared, and the amortization of premiums and discounts on investments.

At September 30, 2008 all of our fixed maturity securities were investment grade and were rated, with the exception of two fixed maturity securities issued by Lehman Brothers that were not investment grade.

In respect of our fixed maturity investments, we periodically assess valuation . . .