The claim reserves reported in the Company’s annual report are a combination of case reserves and Incurred But Not Reported (‘‘IBNR’’) claims. The process for establishing claim reserves is not based solely on information from the cedent but includes significant actuarial analysis. Our process is to first establish case reserves on reported claims based on information from cedents (this is a similar process for both insurance and reinsurance claims). The rest of the reserves are for IBNR claims which are determined statistically from actuarial analyses, which includes assessment of available historical data and other approaches in establishing a range of appropriate loss ratios for management to select its best point estimate. For reinsurance lines of business, a higher proportion of initial reserves is statistically based IBNR due to longer reporting delays between the original loss and the time when reinsurers are aware of the extent of the losses. For insurance business, notifications are received earlier and closer to the date of original loss notification. This delay in reinsurance lines explains why there is a higher degree of uncertainty in reinsurance than direct loss reserves.
The higher degree of uncertainty associated with reserves for reinsurance business is associated with the assumptions used in deriving the IBNR. Where it is based on an analysis of past loss experience, then the principal assumption is that past patterns of development will be repeated on current business. The process of extrapolation is by necessity one involving subjective judgment because the actuary has to take into account the impact of the changing business mix as well as changes in legislative conditions, changes in judicial interpretation of legal liability policy coverages and inflation. These factors are incorporated in the recommended reserve range from which management selects its best point estimate.
Full cedent policy information is recorded at the time of underwriting and includes inception and expiry dates, attachment points and limits and our share of the risk. Claims information includes the loss date, details of the claim, the recommended reserve and the loss adjusters dealing with the claim. In respect of pro-rata treaties we receive regular statements (bordereau) which detail for each cedent the earned and unearned premium, summary paid and loss reserves, with large losses separately identified.
| | |
| c. | We intend to supplement our existing disclosures in our Annual Report on Form 10-K in the appropriate place within our ‘‘Critical Accounting Policies — Reserves for Losses and Loss Expenses’’ section of the MD&A, as follows: |
Reinsurance claims are subject to a longer time lag both in their reporting and in their time to final settlement. The time lag is a factor which is included in the projections to ultimate claims within the actuarial analyses and helps to explain why in general a higher proportion of the initial reinsurance reserves are represented by IBNR than for insurance reserves.
| | |
| d. | We intend to supplement our existing disclosures in our Annual Report on Form 10-K in the appropriate place within our ‘‘Critical Accounting Policies — Reserves for Losses and Loss Expenses’’ section of the MD&A, as follows: |
Information from cedents is principally factored into the setting of reported claim reserves. In addition, for significant events such as the hurricanes in 2004 (i.e. Hurricanes Charley, Frances, Ivan and Jeanne and Typhoon Songda), the detailed analysis of our potential exposures included information obtained directly from cedents which had yet to be processed through market systems enabling us to reduce the time lag between a significant event occurring and establishing a case reserve. This additional information is incorporated into the analysis used to determine the actuarial IBNR. Reinsurance intermediaries are used to assist in obtaining information from cedents but we establish all reserves.
| | |
| e. | We intend to supplement our existing disclosures in our Annual Report on Form 10-K in the appropriate place within our ‘‘Critical Accounting Policies — Reserves for Losses and Loss Expenses’’ section of the MD&A, as follows: |
Delays in receiving information from cedents are an expected part of normal business operations and are included within the statistical estimate of IBNR to the extent that current levels of backlog are consistent with historical data. Currently, there are no processing backlogs which would materially affect our financial statements.
| | |
| f. | We intend to supplement our existing disclosures in our Annual Report on Form 10-K in the appropriate place within our ‘‘Critical Accounting Policies — Reserves for Losses and Loss Expenses’’ section of the MD&A, as follows: |
We use reinsurance intermediaries to assist in communicating with and validating information from cedents. In addition, we engage loss adjusters and perform on site cedent audits to validate the information provided.
| | |
| g. | We intend to supplement our existing disclosures in our Annual Report on Form 10-K in the appropriate place within our ‘‘Critical Accounting Policies — Reserves for Losses and Loss Expenses’’ section of the MD&A, as follows: |
Disputes do occur with cedents, but the number and frequency are generally low. In the event of a dispute, intermediaries are used to try to resolve the dispute. If a resolution cannot be reached, then the contracts typically provide for binding arbitration. To date, there are no material disputes with cedents.
| | |
| h. | We intend to supplement our existing disclosures in our Annual Report on Form 10-K in the appropriate place within our ‘‘Critical Accounting Policies — Reserves for Losses and Loss Expenses’’ section of the MD&A, as follows: |
Historical loss information is included in selecting our best estimates of the reserves for each line of business. We take into account the quality of the historical information available and where appropriate historical trends are used to validate information received from cedents.
5
| |
4. | We note that you provide disclosure for and an analysis around your total reserve balances on a net basis. To help us evaluate the adequacy of your disclosure, please provide us, in disclosure-type format, disclosure around the reserve balances for each line of business on a gross basis. Specifically, where you have provided an actuarial range, provide this range for each line of business shown in the table on page 69. In addition, to aid an investor in understanding the variability that is reasonably likely from your estimated gross loss reserve for each line, disclose the key assumptions used to arrive at management’s best estimate and quantify the effect on operations and financial positions of reasonably likely changes in those assumptions. Estimates of reinsurance recoverable should be discussed separately. |
For the Annual Report on Form 10-K for the year ended December 31, 2005, we propose to include disclosure similar to the following tabular disclosure and narrative, though it will be updated to reflect our reserve position as at December 31, 2005:
The following table sets out the selected actuarial gross reserves for each of our segments and compares it to the actuarial mid-point and actuarial upper-point estimates. While the reported reserves make a reasonable provision for unpaid loss and loss adjustment expense obligations, we note that the process of estimating required reserves does, by its very nature, involve uncertainty and therefore the ultimate claims may fall outside this range. The level of uncertainty can be influenced by such factors as the existence of coverage with long duration payment patterns and changes in claims handling practices, as well as the factors described above. Reasonably likely changes in actuarial assumptions could increase gross loss reserves as at December 31, 2004 by $15.4 million to $1,293.3 million or reduce gross loss reserves by $162.1 million to $1,115.8 million. The net actuarial range for reserves for losses and loss expenses established at December 31, 2004 was between $926.8 million and $1,092.9 million. In determining the range of net reserves, we estimate recoveries due under our proportional and excess of loss reinsurance programmes. For the period ended December 31, 2004, there was a strong correlation between the gross and net reserve ranges as the majority of recoveries were from our proportional reinsurance programme and so movements in gross reserves resulted in an equivalent movement in the net reserves. For excess of loss recoveries, if the upper and lower estimates of gross losses are contained within our reinsurance programme then there will be minimal movement in the net losses, if gross losses were to reduce to a level below the attachment point of the excess of loss programme then movements in gross losses would be replicated as movements in the net loss.
Gross Reserves
| | | | | | | | | | | | | | |
| | Management’s selected reserve ($ in millions) | | Mid-point actuarial estimates ($ in millions) | | Upper-point actuarial estimates ($ in millions) |
Property Reinsurance | | | 341.2 | | | | 333.7 | | | | 341.2 | |
Casualty Reinsurance | | | 377.8 | | | | 300.4 | | | | 377.8 | |
Specialty Reinsurance | | | 168.8 | | | | 140.9 | | | | 168.8 | |
Total Reinsurance | | | 887.8 | | | | 775.0 | | | | 887.8 | |
Property Insurance | | | 77.3 | | | | 71.4 | | | | 77.3 | |
Casualty Insurance | | | 294.5 | | | | 253.5 | | | | 309.9 | |
Specialty Insurance | | | 18.3 | | | | 15.9 | | | | 18.3 | |
Total Insurance | | | 390.1 | | | | 340.8 | | | | 405.5 | |
Total Losses and loss expense reserves | | | 1,277.9 | | | | 1,115.8 | | | | 1,293.3 | |
|
| |
5. | Because IBNR reserve estimates are more imprecise, please provide us, in disclosure-type format, the amount of IBNR separately from case reserves for each line of business listed in the table on page 69. |
For the Annual Report on Form 10-K for the year ended December 31, 2005, we propose to include similar tabular and narrative disclosure as provided below, though it will be updated to reflect 2005 year-end results.
6
| | | | | | | | | | | | | | | | | | |
| | Gross Outstandings ($ in millions) | | Gross IBNR ($ in millions) | | Gross Reserve ($ in millions) | | % IBNR |
Property Reinsurance | | | 181.1 | | | | 160.1 | | | | 341.2 | | | 47% |
Casualty Reinsurance | | | 58.6 | | | | 319.2 | | | | 377.8 | | | 84% |
Specialty Reinsurance | | | 32.8 | | | | 136.0 | | | | 168.8 | | | 81% |
Total Reinsurance | | | 272.6 | | | | 615.2 | | | | 887.8 | | | 69% |
Property Insurance | | | 52.1 | | | | 25.2 | | | | 77.3 | | | 33% |
Casualty Insurance | | | 142.0 | | | | 152.5 | | | | 294.5 | | | 52% |
Specialty Insurance | | | 3.0 | | | | 15.3 | | | | 18.3 | | | 84% |
Total Insurance | | | 197.1 | | | | 193.0 | | | | 390.1 | | | 49% |
Total Losses and loss expense reserves | | | 469.7 | | | | 808.2 | | | | 1,277.9 | | | 63% |
|
Gross reserves of the reinsurance segment were $887.8 million, in 2004 as compared to $352.6 million in 2003, of which 69% were IBNR. The increase in reserves was due to the reinsurance portfolio maturing as a book of business. 2004 was only the second full year of operation for the Company, and an increase in reserves would be expected as reserves build up over time, particularly in the longer tail business lines such as Casualty Reinsurance. In addition to the increase in reserves as a result of the second full year of trading, the Property Reinsurance segment was affected by material claims incurred during 2004 from Hurricanes Charley, Frances, Ivan and Jeanne and by Typhoon Songda totalling $298.7 million. The gross reserves of the specialty reinsurance class of business were also impacted by Hurricanes Charley, Frances, Ivan and Jeanne and Typhoon Songda, totalling $23.8 million.
Gross reserves of the insurance segment were $390.1 million in 2004, compared to $173.2 million in 2003, of which 49% were IBNR. Similar to the reinsurance lines, the increase in reserves was due to the insurance portfolio maturing as a book of business. 2004 was only the second full year of operation for the Company, and an increase in reserves would be expected as reserves build up over time, particularly in the longer tail liability business lines. We do not receive notice of liability insurance claims until a considerable time has passed, resulting in a high 52% of gross reserves being IBNR. Property insurance reserves are particularly sensitive to large claims, and the impact a factory fire ($14.4 million) in Suffolk (U.K.), U.S. Hurricanes and Typhoon Songda losses totalling $17.5 million, have greatly increased the claims reserve. The high proportion of IBNR for specialty insurance was due to the class only commencing underwriting in the final quarter of 2004.
Notes to the Consolidated Financial Statements
Note 21. Unaudited Quarterly Financial Data, page F-34
| |
6. | We note that you disclose various non-GAAP financial measures including ‘‘underwriting income,’’ ‘‘total other operating revenue,’’ and ‘‘operating income before tax.’’ We note that these measures exclude realized investment gains and losses and net exchange gains and losses. Elimination of these items from the most comparable GAAP measure appears to have the effect of smoothing earnings. While the acceptability of a non-GAAP financial measure that eliminates recurring items from the most comparable GAAP measure depends on all facts and circumstances, we do not believe that a non-GAAP measure that has the effect of smoothing earnings is appropriate. In addition, we note that the gains and losses you excluded have the following attributes: |
| | |
| • | There is a past pattern of these items occurring in each reporting period ; |
| | |
| • | The financial impact of these items will not disappear or become immaterial in the future; and |
| | |
| • | There is no unusual reason that the company can substantiate to identify the special nature of these items. |
These attributes raise significant questions about managements’ assertions as to the usefulness of these measures for investors and the appropriateness of their presentation in accordance with Item 10 of Regulation S-K. Please refer to ‘‘Frequently Asked Questions Regarding the Use of Non-GAAP Financial Measures’’ on our website at www.sec.gov/divisions/corpfin/faqs/nongaapfaq.htm that we issued on June 13, 2003. Please confirm that you will eliminate these measures from your future filings or tell us, in disclosure-type format, why you believe these measures are useful and appropriate.
7
The quarterly unaudited financial information disclosed in Note 21 in our financial statements in our Annual Report on Form 10-K was a representation of information contained in the Company’s quarterly earnings supplements. The disclosure of these non-GAAP measures was considered appropriate in the quarterly supplements as it was comparable with information provided by the Company’s peers and included a reconciliation to the most comparable GAAP measure. The Company agrees that it will eliminate these non- GAAP measures from future SEC filings.
In response to your comments, we acknowledge that:
| | |
| • | the Company is responsible for the adequacy and accuracy of the disclosure in the filing; |
| | |
| • | staff comments or changes to disclosure in response to staff comments do not foreclose the Commission from taking any action with respect to the filing; and |
| | |
| • | the Company may not assert staff comments as a defense in any proceeding initiated by the Commission or any person under the federal laws of the United States. |
Thank you for your consideration of the responses. If you have any further questions or comments, please contact me at 1-441-297-9371, Michael Groll of LeBoeuf, Lamb, Greene & MacRae at 1-212-424-8616 or Joe Ferraro of LeBoeuf, Lamb, Greene & MacRae at 011-44-207-459-5125.
Yours sincerely,
/s/ Julian Cusack
Julian Cusack
Chief Financial Officer
Aspen Insurance Holdings Limited
| |
Cc: | Dana Hartz Lisa Vanjoske U.S. Securities and Exchange Commission |
Ian Campbell
David Curtin
Patricia Roufca
Aspen Insurance Holdings Limited
Richard Bennison
Robert Lewis
David Greenfield
KPMG Audit Plc
Michael Groll
Joseph Ferraro
LeBoeuf, Lamb, Greene & MacRae
8