CENTRAL OHIO RISK MANAGEMENT ASSOCIATION (CORMA) ACTUARIAL REPORT ON UNPAID LOSS AND LOSS ADJUSTMENT EXPENSES AS OF SEPTEMBER 30, 2017

Transcription

1 CENTRAL OHIO RISK MANAGEMENT ASSOCIATION (CORMA) ACTUARIAL REPORT ON UNPAID LOSS AND LOSS ADJUSTMENT EXPENSES AS OF SEPTEMBER 30, 2017 October 25, 2017

2 October 25, 2017 Sent Via Ms. Angel Mumma Director of Finance City of Dublin 5200 Emerald Parkway Dublin, Ohio On behalf of CORMA Dear Angel: Enclosed please find our report regarding indicated liabilities as of September 30, 2017 for CORMA. The report presents our summary and conclusions as well as the detailed supporting exhibits. We are available to discuss our analysis or answer any questions you may have upon review. We thank you for the opportunity to continue to provide service to CORMA. Sincerely, Robert W. Van Epps, FCAS, MAAA Managing Principal David S. Benusa Actuarial Analyst c: Tom Wichert Rich Studenic

3 TABLE OF CONTENTS Page INTRODUCTION...2 CONDITIONS & LIMITATIONS...4 DISTRIBUTION & USE...8 SUMMARY OF FINDINGS...9 ANALYSIS...11 EXHIBITS

4 INTRODUCTION Financial Risk Analysts, LLC has been retained by the Central Ohio Risk Management Association (CORMA) to provide indicated loss and allocated loss adjustment expense (ALAE) liabilities as of September 30, CORMA is a pool of eight Ohio municipalities that began providing property and liability coverage to its members on October 1, Coverages provided are Automobile Liability, General Liability, Police & Public Officials Liability, Employee Benefits Liability, Medical Professional Liability, Property, Crime, and Automobile Physical Damage. Coverage is provided on an occurrence basis, except for Public Officials, Medical Professional, and Employee Benefits Liability, which are on a claims made basis. For the latest year, CORMA purchased specific excess reinsurance which limits its retention to $150,000 per occurrence for liability losses and $25,000 for property losses. CORMA carries deductibles of $50,000 and $250,000 per occurrence for auto physical damage and flood losses, respectively. An overall aggregate retention limits CORMA s total annual liability losses to a maximum amount. The specific excess retentions, deductibles, and aggregate retentions vary by year and are displayed on Exhibit 9. The municipalities comprising CORMA are the cities of Dublin, Westerville, Upper Arlington, Pickerington, Grove City, Powell, Groveport, Canal Winchester, and Grandview Heights. CORMA did not provide liability coverage to Upper Arlington prior to October 1, 1999 since the city maintained a $100,000 self insured retention for liability losses. The cities of Grove City and Powell joined the pool on October 1, 2009 and February 1, 2010, respectively. The city of Groveport joined on October 1, The city of Canal Winchester joined on October 1, The city of Grandview Heights joined on June 6, Administrative services including claims administrative services are provided by Wichert Insurance Services, Inc. 2

5 The estimates in this report were developed in accordance with the principals of the Casualty Actuarial Society and the applicable standards of the American Academy of Actuaries. I, Robert W. Van Epps, am a Principal for Financial Risk Analysts, LLC. I am an independent consultant to CORMA. I am a Member of the American Academy of Actuaries and a Fellow of the Casualty Actuarial Society. I meet the qualification standards of the American Academy of Actuaries to issue this report. 3

6 CONDITIONS & LIMITATIONS GENERAL Throughout this report and its exhibits the term "loss" is used to refer to both losses and allocated loss adjustment expense. Allocated loss adjustment expense or ALAE provides for all expenses associated with the handling and settling of claims that can be directly attributable to a particular claim. Such expenses typically include legal fees, investigatory expenses, expert witness fees, etc. Unallocated loss adjustment expenses or ULAE are expenses associated with the handling and settling of claims that cannot be specifically assigned to a particular claim. Such expenses are generally in the nature of overhead expenses and would include fees paid to CORMA s third party administrator, Wichert Insurance Services, Inc. DATA We used historical data and other information provided to us by CORMA and Wichert Insurance Services, administrator for CORMA. This information consists of premium, loss, exposure, and reinsurance information. In addition, we have relied on various oral representations made to us by the responsible employees of CORMA and Wichert Insurance Services. Certain adjustments to the loss data were made to eliminate Upper Arlington liability losses prior to October 1, We also made adjustments to include certain NC claims that were not included on the ASC loss runs provided to us. CORMA has changed claims administrators several times since 2005, most recently transitioning to Wichert Insurance Services in Prior administrators were unable to provide data valued as of September 30, 2006 that appear reasonable. We have therefore not included this valuation date in our loss triangles. The result is a missing diagonal and a lack of development factors based on CORMA s experience that would have used data with this valuation date. 4

7 Additionally, because of the change in claims administrators noted above, we are unable to reconcile two data items: 1) Net ratable expenditures, which are used as an exposure base for general liability, increased significantly as of the September 2010 evaluation. However, based on discussions with Wichert Insurance Services, we believe that this increase was driven by a change in the way exposures are reported. To account for this, we used the reported ratable exposure for 10/1/08-09 and estimated exposures for previous years based on an assumed trend. 2) Cumulative net losses by year do not match data that was provided historically. To maintain the integrity of historical triangles, we have relied upon the incremental paid during the previous year as well as current case outstanding losses. To the extent that either of these data items has an effect on the estimation of historical exposures or historical payment patterns, our estimation of outstanding losses and estimated pure premiums could be affected, perhaps materially. Although the data supplied have been reviewed for purposes of reasonability we have not independently audited or verified this information and we assume it to be accurate and complete. The results of our analysis will be contingent upon the reliability of the information supplied to us and such reliability is the responsibility of CORMA and Wichert Insurance Services. Should CORMA or Wichert Insurance Services become aware of any significant discrepancies in the data reported to us we should be notified of such discrepancies and this report will be amended, if necessary. In performing our analysis we have also relied on loss development data and information obtained from insurance industry sources to supplement the actual historical data provided to us. 5

8 UNCERTAINTY Actuarial projections, by their nature, are estimates of future contingent events which cannot be known with certainty. The ultimate liability of CORMA for its retained losses will be subject to events that have yet to occur such as the size of future liability awards, medical cost inflation, judicial interpretations as to liability, as well as the propensity for claimants to bring suit. No assumptions have been made in this report as to any extraordinary changes in the legal, social, or judicial environment that might affect future losses. In particular it should be recognized that CORMA has undergone changes to their claims administrator and claims reserving practices. They also have relatively little pooled experience available at this time. These factors will tend to increase the potential variability of any actuarial estimate. While we believe the results presented in this report are reasonable and reflect the use of accepted actuarial principles and standards of practice it is possible that actual future loss results of CORMA will differ, perhaps materially, from those projected herein. Nothing in this report should be construed as a warranty or guarantee as to the adequacy of the funding estimates contained herein. UNINSURED EXPOSURES The estimates contained in this report are intended to cover only those exposures contained within the terms and conditions of the commercial excess insurance policies purchased by CORMA. Therefore, we have not included a specific provision for any losses that might arise as a result of exposures that fall within any exclusions of CORMA s commercially purchased policies. Such exposures often include, but are not limited to environmental impairment liability and asbestos liability. 6

9 EXCESS INSURANCE The funding estimates contained in this report are net of all excess insurance purchased by CORMA. We have made no attempt to analyze the security of CORMA's excess insurers nor have we made provision for any additional funding that would be required by CORMA should any of its excess insurers be unable to perform on their obligations. 7

10 DISTRIBUTION & USE This report is intended solely for the use of CORMA in evaluating its loss and allocated loss adjustment expense liabilities as of September 30, A copy of this report may also be provided to CORMA's administrators and external auditor with the proviso that the report is copied in its entirety and each party receiving a copy of this report agrees to not distribute the report to any other third party. Distribution of this report to parties other than those referenced above is not authorized without the express written consent of Financial Risk Analysts, LLC 8

11 SUMMARY OF FINDINGS Indicated Liability as of September 30, 2017 Based on the analysis contained herein and as displayed on Exhibit 1.1 the following presents our indicated CORMA liabilities (unpaid losses and loss expenses as discussed below) as of September 30, 2017: CORMA Indicated Liabilities as of September 30, 2017 Loss & ALAE $1,226,699 Claim Service Fees $35,000 TOTAL $1,261,699 The liability indication is intended to provide for all losses incurred by CORMA during the period October 1, 1997 through September 30, 2017 under the terms and conditions of its policy provisions, but which are unpaid as of September 30, The liability includes a provision for the following: For occurrence coverages, losses and loss adjustment expenses on claims occurring during the period October 1, 1997-September 30, For claims made coverages, losses and loss adjustment expenses on claims reported during the period October 1, September 30, 2017 provided that the loss occurred subsequent to the retroactive date for the coverage provided. The liabilities presented above do not include a provision for future investment income to be earned on the invested assets supporting the liabilities nor do they contain a provision for any expenses associated with CORMA s operations other than loss adjustment expense. 9

12 The provision for claim service fees is intended to provide for estimated future third party administrative claim costs involved in the handling and settling of all claims incurred as of September 30, GASB-10 requires public entity pools to accrue a liability for these costs. The estimate of these costs has been judgmentally determined based on discussions with Wichert Insurance Services, CORMA s current third party administrator, and our understanding of their fee structure. We recommend that CORMA consult with its external auditors as to the reasonability of booking a provision for this. 10

13 ANALYSIS Loss and Allocated Loss Adjustment Expense Liabilities Exhibit 1.1 presents a summary of the results of our analysis by coverage. We used historical loss experience of the members of CORMA combined with internal municipal pool data to project ultimate losses for each coverage in the pool. We used four commonly applied actuarial methodologies, the Incurred Loss Development, Paid Loss Development, Incurred Bornhuetter- Ferguson and Paid Bornhuetter-Ferguson methods to estimate CORMA s ultimate losses for General Liability and Automobile Liability. For Automobile Physical Damage and Property we used only the loss development methods. Incurred Loss Development Method The incurred loss development method is a method in which incurred losses are projected to an ultimate level based on historical development patterns. An analysis of the changes in accident year incurred losses between various valuation points provides a basis for estimating future changes. Assuming that reporting patterns and reserving practices have remained reasonably stable, this method can be used to provide reasonable estimates of ultimate losses. Since there is little historical loss development experience available for CORMA, we relied on a combination of CORMA s available historical experience and insurance industry loss development data in our selection of loss development factors. Results of the Incurred Loss Development method are displayed on Exhibit 6. Paid Loss Development Method This method is similar to the incurred loss development method except that paid losses are used instead of incurred losses. Paid losses are projected to an ultimate level based on historical paid loss development patterns. This method assumes that loss settlement and 11

14 payment patterns have remained reasonably stable over time. Unlike the incurred method, the paid loss development method is not dependent on reasonably consistent reserving practices. As with the incurred loss development method, paid loss development factors have been derived from a combination of CORMA s available historical experience and insurance industry loss development data. Results of the Paid Loss Development method are also displayed on Exhibit 6. Bornhuetter-Ferguson Incurred Method This approach is essentially a credibility weighted average of two other reserving methods; the incurred loss development method and the expected loss method. The expected loss method simply sets the indicated reserve equal to the difference between expected losses and paid losses. It produces very stable results but fails to incorporate actual results as they emerge. The Bornhuetter-Ferguson Incurred method blends the incurred loss development method and the expected loss method by splitting expected losses into two distinct pieces; expected reported losses and expected unreported losses. As an accident year matures, the expected incurred losses are replaced with actual reported losses plus expected unreported losses. Thus, as the year matures, the initial expected incurred loss estimate becomes less important while the actual reported loss experience increases in importance. In order to use this method, one must estimate both the initial expected losses and the expected loss reporting pattern. Initial expected losses have been estimated based on CORMA s historical loss costs adjusted to current cost levels. This is displayed on Exhibit 5. Expected loss reporting patterns are the reciprocal of the incurred loss development factors from the incurred loss development method. Results of the Incurred Bornhuetter-Ferguson method are displayed on Exhibit 4. 12

15 Bornhuetter-Ferguson Paid Method The Bornhuetter-Ferguson paid loss method is similar to the Bornhuetter-Ferguson incurred loss method except that expected losses are split between expected paid and unpaid losses. Expected loss payment patterns are derived from the paid loss development factors. Initial expected losses are the same as those used in the Bornhuetter-Ferguson Incurred method. Ultimate loss projections for Liability coverages are summarized on Exhibits 3.2 and 3.3. Ultimate losses are selected from the results of all four methods. For Property and Automobile Physical Damage we used the results from the loss development methods as displayed on Exhibits 6.3 and 6.4. The ultimate loss selections by coverage are summarized on Exhibit 3.1 and limited to the aggregate stop loss amount. On Exhibit 1.1, losses paid to date are subtracted from selected ultimate losses to produce the indicated outstanding losses net of all excess insurance. For five of CORMA s first six years paid losses have exceeded the aggregate so there are no outstanding losses on a net of reinsurance basis. For older years where the aggregates are not exceeded but for which all claims are closed, ultimate losses are selected to equal paid losses so there are no outstanding losses. Exhibit 2 shows changes in selected ultimate losses from our September 30, 2016 analysis. The selected ultimates for all lines of business have increased slightly in some pool years and decreased slightly in others. Overall, this indicates that ultimate losses have not changed materially from the prior year. 13

16 Exhibit 1.1 INDICATED OUTSTANDING LOSSES NET OF REINSURANCE AS OF SEPTEMBER 30, 2017 Selected Indicated Ultimate Outstanding Aggregate Losses Net Losses Case Accident Stop Net Of Paid Net Of Outstanding Period Loss Reinsurance Losses Reinsurance Losses IBNR (1) (2) (3) (4) (5) = (3) - (4) (6) (7) = (5) - (6) 10/1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , , /1/ , , ,399 19, ,318 10/1/ , , ,208 82,763 46,390 36,373 10/1/ , , , ,822 55, ,322 10/1/ , , , , ,821 87,907 10/1/ , , , , , ,804 TOTAL 6,716,250 5,489,550 1,226, , ,723 (2) Exhibit 9 (3) Exhibit 3.1 (4) Exhibits limited to aggregate (6) Exhibits Outstanding

17 Exhibit 2 COMPARISON OF SELECTED ULTIMATE LOSSES TO PRIOR ANALYSIS AS OF SEPTEMBER 30, 2017 Accident General Liability Automobile Liability Period Current Prior Change Current Prior Change (1) (2) (3) (4)=(2)-(3) (5) (6) (7)=(5)-(6) 10/1/ , , ,842 13, /1/ , , ,142 1, /1/ , , ,452 35, /1/ , , ,840 8, /1/ , , ,901 9, /1/ , , ,132 8, /1/ , , ,901 18, /1/ , , ,830 86, /1/ ,588 13, ,987 7, /1/ ,485 71, ,431 17, /1/ ,477 66, ,507 28, /1/ ,295 36, ,921 23, /1/ , , , , /1/ , , ,300 20, /1/ , ,185-7,899 47,906 47, /1/ ,967 45,222-10,255 5,597 5, /1/ , ,488-24,557 68,011 69,294-1,283 10/1/ , ,853-10,894 66,101 26,603 39,498 10/1/ , ,349 54,198 28,602 46,615-18,013 TOTAL 3,488,048 3,487, , ,198 20,202 Accident Property Automobile Physical Damage Period Current Prior Change Current Prior Change (1) (8) (9) (10)=(8)-(9) (11) (12) (13)=(11)-(12) 10/1/ ,153 29, ,354 18, /1/ ,586 26, ,981 30, /1/ ,785 17, ,559 17, /1/ ,721 41, ,319 30, /1/ ,879 38, ,279 9, /1/ , , ,111 41, /1/ ,425 58, ,467 47, /1/ ,727 84, ,876 80, /1/ ,651 22, , , /1/ ,599 16, ,518 8, /1/ ,462 74, ,779 83, /1/ ,628 36, , , /1/ ,991 83, ,562 65, /1/ , , ,735 19, /1/ , , ,509 47, /1/ ,992 30, , , /1/ ,263 56, ,766 30, /1/ ,499 49,173-2,674 50,077 49,076 1,001 10/1/ , ,516-36,358 91,988 89,711 2,277 TOTAL 1,109,694 1,148,725-39,031 1,095,970 1,092,692 3,278 (2) Exhibit 3.2 (11) Exhibit 6.4 (5) Exhibit 3.3 (3),(6),(9),(12) From previous analysis (8) Exhibit Comp to Prior

18 Exhibit 3.1 INDICATED OUTSTANDING LOSSES NET OF REINSURANCE AS OF SEPTEMBER 30, 2017 Selected Selected Selected Ultimate Losses Ultimate Ultimate Automobile Aggregate Losses Losses Accident General Automobile Physical Stop Subject To Net Of Period Liability Liability Property Damage TOTAL Loss Aggregate Reinsurance (1) (2) (3) (4) (5) (6) (7) (8) (9) 10/1/ ,449 13,842 29,153 18, , , , ,000 10/1/ ,737 1,142 26,586 30, , , , ,000 10/1/ ,100 35,452 17,785 17, , , , ,750 10/1/ ,397 8,840 41,721 30, , , , ,000 10/1/ ,345 9,901 38,879 9, , , , ,404 10/1/ ,418 8, ,172 41, , , , ,283 10/1/ ,237 18,901 58,425 47, , , , ,030 10/1/ ,317 86,830 84,727 80, , , , ,750 10/1/ ,588 7,987 22, , , ,000 21, ,749 10/1/ ,485 17,431 16,599 8, , ,000 88, ,033 10/1/ ,477 28,507 74,462 83, , ,000 94, ,225 10/1/ ,295 23,921 36, , , ,000 60, ,250 10/1/ , ,997 83,991 65, , , , ,400 10/1/ ,663 20, ,228 19, , , , ,926 10/1/ ,286 47, ,775 47, , , , ,477 10/1/ ,967 5,597 30, , , ,000 40, ,717 10/1/ ,931 68,011 56,263 30, , , , ,971 10/1/ ,959 66,101 46,499 50, , , , ,636 10/1/ ,547 28,602 67,158 91, , , , ,295 10/1/ ,811 62, , , , , , ,354 TOTAL 3,869, ,671 1,349,042 1,212,894 7,215,466 4,914,146 6,716,250 (2) Exhibit 3.2 (7) Exhibit 9 (3) Exhibit 3.3 (8) 10/1/02-03 and subsequent: Aggregate only applies to liability losses (4) Exhibit 6.3 (9) Losses from column (6) limited to aggregate (5) Exhibit 6.4 (6) (2)+(3)+(4)+(5) 3.1 Ult Summary

19 Exhibit 3.2 SUMMARY OF METHODOLOGY RESULTS AND SELECTION OF ULTIMATE LOSSES AS OF SEPTEMBER 30, 2017 GENERAL LIABILITY Projected Ultimate Losses by Method Selected Accident Incurred Paid Incurred Paid Incurred Paid Ultimate Period Losses Losses LDF LDF Born-Ferg Born-Ferg Losses (1) (2) (3) (4) (5) (6) (7) (8) 10/1/ , , , , , , ,449 10/1/ , , , , , , ,737 10/1/ , , , , , , ,100 10/1/ , , , , , , ,397 10/1/ , , , , , , ,345 10/1/ , , , , , , ,418 10/1/ , , , , , , ,237 10/1/ , , , , , , ,317 10/1/ ,588 13,588 13,588 13,588 13,588 13,588 13,588 10/1/ ,485 71,485 71,485 71,485 71,485 71,485 71,485 10/1/ ,477 66,477 66,809 66,809 67,522 67,522 66,477 10/1/ ,295 36,295 36,767 37,021 39,258 40,854 36,295 10/1/ , , , , , , ,850 10/1/ , , , , , , ,663 10/1/ , , , , , , ,286 10/1/ ,649 15,649 17,026 18,560 34,967 53,093 34,967 10/1/ ,559 83, , , , , ,931 10/1/ , , , , , , ,959 10/1/ , , , , , , ,547 10/1/ ,762 9, , , , , ,811 TOTAL 3,381,788 2,973,787 4,139,426 3,812,327 3,951,427 3,890,342 3,869,859 (2),(3),(4),(5) Exhibit 6.1 (6) Exhibit 4.1 (7) Exhibit Ultimate GL

20 Exhibit 3.3 SUMMARY OF METHODOLOGY RESULTS AND SELECTION OF ULTIMATE LOSSES AS OF SEPTEMBER 30, 2017 AUTOMOBILE LIABILITY Projected Ultimate Losses by Method Selected Accident Incurred Paid Incurred Paid Incurred Paid Ultimate Period Losses Losses LDF LDF Born-Ferg Born-Ferg Losses (1) (2) (3) (4) (5) (6) (7) (8) 10/1/ ,842 13,842 13,842 13,842 13,842 13,842 13,842 10/1/ ,142 1,142 1,142 1,142 1,142 1,142 1,142 10/1/ ,452 35,452 35,452 35,452 35,452 35,452 35,452 10/1/ ,840 8,840 8,840 8,840 8,840 8,840 8,840 10/1/ ,901 9,901 9,901 9,901 9,901 9,901 9,901 10/1/ ,132 8,132 8,132 8,132 8,132 8,132 8,132 10/1/ ,901 18,901 18,901 18,901 18,901 18,901 18,901 10/1/ ,830 86,830 86,830 86,830 86,830 86,830 86,830 10/1/ ,987 7,987 7,987 7,987 7,987 7,987 7,987 10/1/ ,431 17,431 17,431 17,431 17,431 17,431 17,431 10/1/ ,507 28,507 28,507 28,507 28,507 28,507 28,507 10/1/ ,921 23,921 23,921 23,921 23,921 23,921 23,921 10/1/ , , , , , , ,997 10/1/ ,300 20,300 20,300 20,300 20,300 20,300 20,300 10/1/ ,906 47,906 47,906 47,906 47,906 47,906 47,906 10/1/ ,597 5,597 5,597 5,597 5,597 5,597 5,597 10/1/ ,011 68,011 68,011 68,011 68,011 68,011 68,011 10/1/ ,805 9,305 66,101 9,770 66,096 12,403 66,101 10/1/ ,653 9,653 11,815 14,701 22,838 34,366 28,602 10/1/ ,801 12,301 49,005 41,222 58,960 65,582 62,271 TOTAL 731, , , , , , ,671 (2),(3),(4),(5) Exhibit 6.2 (6) Exhibit 4.3 (7) Exhibit /25/201710:42 AM 3.3 Ultimate AutoLiab

21 Exhibit 4.1 GENERAL LIABILITY PROJECTED ULTIMATE LOSSES AS OF SEPTEMBER 30, 2017 BORNHUETTER-FERGUSON INCURRED LOSS METHOD Operating Initial Initial Expected % Expected % Expected Projected Accident Budget Expected Expected Of Losses Losses Reported Of Losses Unreported Ultimate Period (000) Pure Pemium Losses Reported Actual Expected Unreported Losses Losses (1) (2) (3) (4)=(2)x(3) (5) (6) (7)=(4)x(5) (8)=1.0-(5) (9)=(4)x(8) (10)=(6)+(9) 10/1/ , , % 101, , % 0 101,449 10/1/ , , % 185, , % 0 185,737 10/1/ , , % 163, , % 0 163,100 10/1/ , , % 248, , % 0 248,397 10/1/ , , % 112, , % 0 112,345 10/1/ , , % 407, , % 0 407,418 10/1/ , , % 189, , % 0 189,237 10/1/ , , % 155, , % 0 155,317 10/1/ , , % 13, , % 0 13,588 10/1/ , , % 71, , % 0 71,485 10/1/ , , % 66, , % 1,045 67,522 10/1/ , , % 36, , % 2,963 39,258 10/1/ , , % 230, , % 5, ,698 10/1/ , , % 136, , % 9, ,733 10/1/ , , % 328, , % 13, ,324 10/1/ , , % 15, , % 19,318 34,967 10/1/ , , % 129, , % 35, ,648 10/1/ , , % 221, , % 75, ,618 10/1/ , , % 436, , % 131, ,511 10/1/ , , % 131,762 94, % 276, ,074 TOTAL 1,571,086 7,398,557 3,381,789 6,828, ,638 3,951,427 (2) Exhibit 8 (3) Exhibit 5.1. Prior to 10/1/14 from previous analysis (5) Reciprocal of development factors from Exhibit 6.1 (6) Exhibit GLBFIncurred

22 Exhibit 4.2 GENERAL LIABILITY PROJECTED ULTIMATE LOSSES AS OF SEPTEMBER 30, 2017 BORNHUETTER-FERGUSON PAID LOSS METHOD Operating Initial Initial Expected % Expected % Expected Projected Accident Budget Expected Expected Of Losses Losses Paid Of Losses Unpaid Ultimate Period (000) Pure Pemium Losses Paid Actual Expected Unpaid Losses Losses (1) (2) (3) (4)=(2)x(3) (5) (6) (7)=(4)x(5) (8)=1.0-(5) (9)=(4)x(8) (10)=(6)+(9) 10/1/ , , % 101, , % 0 101,449 10/1/ , , % 185, , % 0 185,737 10/1/ , , % 163, , % 0 163,100 10/1/ , , % 248, , % 0 248,397 10/1/ , , % 112, , % 0 112,345 10/1/ , , % 407, , % 0 407,418 10/1/ , , % 189, , % 0 189,237 10/1/ , , % 155, , % 0 155,317 10/1/ , , % 13, , % 0 13,588 10/1/ , , % 71, , % 0 71,485 10/1/ , , % 66, , % 1,045 67,522 10/1/ , , % 36, , % 4,559 40,854 10/1/ , , % 230, , % 9, ,109 10/1/ , , % 136, , % 15, ,868 10/1/ , , % 328, , % 21, ,180 10/1/ , , % 15, , % 37,444 53,093 10/1/ , , % 83, , % 84, ,213 10/1/ , , % 221, , % 158, ,299 10/1/ , , % 196, , % 239, ,583 10/1/ , , % 9,972 24, % 345, ,548 TOTAL 1,571,086 7,398,557 2,973,787 6,482, ,555 3,890,342 (2) Exhibit 8 (3) Exhibit 5.1. Prior to 10/1/14 from previous analysis (5) Reciprocal of development factors from Exhibit 6.1 (6) Exhibit GLBFPaid

23 Exhibit 4.3 AUTOMOBILE LIABILITY PROJECTED ULTIMATE LOSSES AS OF SEPTEMBER 30, 2017 BORNHUETTER-FERGUSON INCURRED LOSS METHOD Number Initial Initial Expected % Expected % Expected Projected Accident Of Expected Expected Of Losses Losses Reported Of Losses Unreported Ultimate Period Vehicles Pure Pemium Losses Reported Actual Expected Unreported Losses Losses (1) (2) (3) (4)=(2)x(3) (5) (6) (7)=(4)x(5) (8)=1.0-(5) (9)=(4)x(8) (10)=(6)+(9) 10/1/ , % 13,842 32, % 0 13,842 10/1/ , % 1,142 19, % 0 1,142 10/1/ , % 35,452 35, % 0 35,452 10/1/ , % 8,840 32, % 0 8,840 10/1/ , % 9,901 35, % 0 9,901 10/1/ , % 8,132 38, % 0 8,132 10/1/ , % 18,901 43, % 0 18,901 10/1/ , % 86,830 44, % 0 86,830 10/1/ , % 7,987 41, % 0 7,987 10/1/ , % 17,431 47, % 0 17,431 10/1/ , % 28,507 43, % 0 28,507 10/1/ , % 23,921 54, % 0 23,921 10/1/ , % 223,997 56, % 0 223,997 10/1/ , % 20,300 62, % 0 20,300 10/1/ , % 47,906 56, % 0 47,906 10/1/ , % 5,597 63, % 0 5,597 10/1/ , % 68,011 60, % 0 68,011 10/1/ , % 64,805 63, % 1,291 66,096 10/1/ , % 9,653 58, % 13,185 22,838 10/1/ , , % 30,801 47, % 28,159 58,960 TOTAL 14, , , ,259 42, ,590 (2) Exhibit 8 (3) Exhibit 5.2. Prior to 10/1/14 from previous analysis (5) Reciprocal of development factors from Exhibit 6.2 (6) Exhibit AutoBFIncurred

24 Exhibit 4.4 AUTOMOBILE LIABILITY PROJECTED ULTIMATE LOSSES AS OF SEPTEMBER 30, 2017 BORNHUETTER-FERGUSON PAID LOSS METHOD Number Initial Initial Expected % Expected % Expected Projected Accident Of Expected Expected Of Losses Losses Paid Of Losses Unpaid Ultimate Period Vehicles Pure Pemium Losses Paid Actual Expected Unpaid Losses Losses (1) (2) (3) (4)=(2)x(3) (5) (6) (7)=(4)x(5) (8)=1.0-(5) (9)=(4)x(8) (10)=(6)+(9) 10/1/ , % 13,842 32, % 0 13,842 10/1/ , % 1,142 19, % 0 1,142 10/1/ , % 35,452 35, % 0 35,452 10/1/ , % 8,840 32, % 0 8,840 10/1/ , % 9,901 35, % 0 9,901 10/1/ , % 8,132 38, % 0 8,132 10/1/ , % 18,901 43, % 0 18,901 10/1/ , % 86,830 44, % 0 86,830 10/1/ , % 7,987 41, % 0 7,987 10/1/ , % 17,431 47, % 0 17,431 10/1/ , % 28,507 43, % 0 28,507 10/1/ , % 23,921 54, % 0 23,921 10/1/ , % 223,997 56, % 0 223,997 10/1/ , % 20,300 62, % 0 20,300 10/1/ , % 47,906 56, % 0 47,906 10/1/ , % 5,597 63, % 0 5,597 10/1/ , % 68,011 60, % 0 68,011 10/1/ , % 9,305 61, % 3,098 12,403 10/1/ , % 9,653 47, % 24,713 34,366 10/1/ , , % 12,301 22, % 53,281 65,582 TOTAL 14, , , ,802 81, ,047 (2) Exhibit 8 (3) Exhibit 5.2. Prior to 10/1/14 from previous analysis (5) Reciprocal of development factors from Exhibit 6.2 (6) Exhibit AutoBFPaid

25 Exhibit 5.1 GENERAL LIABILITY PROJECTED PURE PREMIUM Initial Selected Trended Accident Ultimate Pure Trend To Pure Year Retention Losses Exposure Premium April 1, 2017 Premium (1) (2) (3) (4) (5)=(3)/(4) (6) (7)=(5)x(6) 10/1/ , ,793 63, /1/ , ,662 65, /1/ , ,345 67, /1/ , ,418 68, /1/ , ,237 70, /1/ , ,317 72, /1/ ,000 13,588 74, /1/ ,000 71,485 76, /1/ ,000 66,477 78, /1/ ,000 36,295 80, /1/ , ,850 82, /1/ , ,663 88, /1/ , ,602 83, /1/ ,000 17,793 83, /1/ , ,718 94, /1/ , ,861 95, /1/ , , , (7) MEAN 3.71 (8) LATEST (9) LATEST (10) LATEST 5 EX. HI/LOW 3.84 (11) WTD AVG ALL YEARS 3.63 (12) SELECTED 3.65 FORECAST PURE PREMIUM Retention Detrend Forecast Retention Factor Factor Pure Premium (13) (14) (15)=1.0 / (6) (16) 10/1/ , /1/ , /1/ , (2),(13) Exhibit 9 (3) Exhibit 6.1 adjusted to 150K SIR (4) Exhibit 8 (6) +2.0% annual trend based on industry (14) Based on industry increased limits factors (16) (12) x (14) x (15) 5.1 GLPP

26 Exhibit 5.2 AUTOMOBILE LIABILITY PROJECTED PURE PREMIUM Initial Selected Trended Accident Ultimate Pure Trend To Pure Year Retention Losses Exposure Premium April 1, 2017 Premium (1) (2) (3) (4) (5)=(3)/(4) (6) (7)=(5)x(6) 10/1/ ,000 35, /1/ ,000 8, /1/ ,000 9, /1/ ,000 8, /1/ ,000 18, /1/ ,000 86, /1/ ,000 7, /1/ ,000 17, /1/ ,000 28, /1/ ,000 23, /1/ , , /1/ ,000 20, /1/ ,000 47, /1/ ,000 5, /1/ ,000 68, /1/ ,000 66, /1/ ,000 13, (7) MEAN 69 (8) MEAN POST '03 SIR CHANGE 76 (9) LATEST 3 55 (10) LATEST 5 46 (11) LATEST 5 EX. HI/LOW 48 (12) WTD AVG ALL YEARS POST SIR CHANGE 76 (13) SELECTED 75 FORECAST PURE PREMIUM Retention Detrend Forecast Retention Factor Factor Pure Premium (14) (15) (16)=1.0 / (6) (17) 10/1/ , /1/ , /1/ , (2),(13) Exhibit 9 (3) Exhibit 6.2 adjusted to $150K SIR (4) Exhibit 8 (6) +2.93% annual trend based on industry (15) Based on industry increased limits factors (17) (13) x (15) x (16) 5.2 AutoPP

27 Exhibit 6.1 GENERAL LIABILITY PROJECTED ULTIMATE LOSSES AS OF SEPTEMBER 30, 2017 INCURRED AND PAID LOSS DEVELOPMENT METHODS Initial Incurred Projected Paid Projected Selected Accident Incurred Development Ultimate Paid Development Ultimate Ultimate Year Losses Factor Losses Losses Factor Losses Losses (1) (2) (3) (4)=(2)x(3) (5) (6) (7)=(5)x(6) (8) 10/1/ , , , , ,449 10/1/ , , , , ,737 10/1/ , , , , ,100 10/1/ , , , , ,397 10/1/ , , , , ,345 10/1/ , , , , ,418 10/1/ , , , , ,237 10/1/ , , , , ,317 10/1/ , ,588 13, ,588 13,588 10/1/ , ,485 71, ,485 71,485 10/1/ , ,809 66, ,809 66,477 10/1/ , ,767 36, ,021 36,295 10/1/ , , , , ,850 10/1/ , , , , ,663 10/1/ , , , , ,286 10/1/ , ,026 15, ,560 17,793 10/1/ , ,957 83, , ,153 10/1/ , , , , ,745 10/1/ , , , , ,209 10/1/ , ,142 9, , ,062 TOTAL 3,381,789 4,139,426 2,973,787 3,812,327 3,935,605 (2),(3) Exhibit 7.1 (5),(6) Exhibit GLLDF

28 Exhibit 6.2 AUTOMOBILE LIABILITY PROJECTED ULTIMATE LOSSES AS OF SEPTEMBER 30, 2017 INCURRED AND PAID LOSS DEVELOPMENT METHODS Initial Incurred Projected Paid Projected Selected Accident Incurred Development Ultimate Paid Development Ultimate Ultimate Year Losses Factor Losses Losses Factor Losses Losses (1) (2) (3) (4)=(2)x(3) (5) (6) (7)=(5)x(6) (8) 10/1/ , ,842 13, ,842 13,842 10/1/ , ,142 1, ,142 1,142 10/1/ , ,452 35, ,452 35,452 10/1/ , ,840 8, ,840 8,840 10/1/ , ,901 9, ,901 9,901 10/1/ , ,132 8, ,132 8,132 10/1/ , ,901 18, ,901 18,901 10/1/ , ,830 86, ,830 86,830 10/1/ , ,987 7, ,987 7,987 10/1/ , ,431 17, ,431 17,431 10/1/ , ,507 28, ,507 28,507 10/1/ , ,921 23, ,921 23,921 10/1/ , , , , ,997 10/1/ , ,300 20, ,300 20,300 10/1/ , ,906 47, ,906 47,906 10/1/ , ,597 5, ,597 5,597 10/1/ , ,011 68, ,011 68,011 10/1/ , ,101 9, ,770 66,101 10/1/ , ,815 9, ,701 13,258 10/1/ , ,005 12, ,222 45,114 TOTAL 731, , , , ,169 (2),(3) Exhibit 7.3 (5),(6) Exhibit AutoLDF

29 Exhibit 6.3 PROPERTY PROJECTED ULTIMATE LOSSES AS OF SEPTEMBER 30, 2017 INCURRED AND PAID LOSS DEVELOPMENT METHODS Incurred Projected Paid Projected Selected Accident Incurred Development Ultimate Paid Development Ultimate Ultimate Year Losses Factor Losses Losses Factor Losses Losses (1) (2) (3) (4)=(2)x(3) (5) (6) (7)=(5)x(6) (8) 10/1/ , ,153 29, ,153 29,153 10/1/ , ,586 26, ,586 26,586 10/1/ , ,785 17, ,785 17,785 10/1/ , ,721 41, ,721 41,721 10/1/ , ,879 38, ,879 38,879 10/1/ , , , , ,172 10/1/ , ,425 58, ,425 58,425 10/1/ , ,727 84, ,727 84,727 10/1/ , ,651 22, ,651 22,651 10/1/ , ,599 16, ,599 16,599 10/1/ , ,462 74, ,462 74,462 10/1/ , ,628 36, ,628 36,628 10/1/ , ,991 83, ,991 83,991 10/1/ , , , , ,228 10/1/ , , , , ,775 10/1/ , ,992 30, ,992 30,992 10/1/ , ,263 56, ,263 56,263 10/1/ , ,499 46, ,499 46,499 10/1/ , ,729 63, ,587 67,158 10/1/ , ,348 60, , ,348 TOTAL 1,326,234 1,347,613 1,166,234 1,228,135 1,349,042 (2),(3) Exhibit 7.5 (5),(6) Exhibit PropLDF

30 Exhibit 6.4 AUTOMOBILE PHYSICAL DAMAGE PROJECTED ULTIMATE LOSSES AS OF SEPTEMBER 30, 2017 INCURRED AND PAID LOSS DEVELOPMENT METHODS Incurred Projected Paid Projected Selected Accident Incurred Development Ultimate Paid Development Ultimate Ultimate Year Losses Factor Losses Losses Factor Losses Losses (1) (2) (3) (4)=(2)x(3) (5) (6) (7)=(5)x(6) (8) 10/1/ , ,354 18, ,354 18,354 10/1/ , ,981 30, ,981 30,981 10/1/ , ,559 17, ,559 17,559 10/1/ , ,319 30, ,319 30,319 10/1/ , ,279 9, ,279 9,279 10/1/ , ,111 41, ,111 41,111 10/1/ , ,467 47, ,467 47,467 10/1/ , ,876 80, ,876 80,876 10/1/ , , , , ,523 10/1/ , ,518 8, ,518 8,518 10/1/ , ,779 83, ,779 83,779 10/1/ , , , , ,406 10/1/ , ,562 65, ,562 65,562 10/1/ , ,735 19, ,735 19,735 10/1/ , ,509 47, ,509 47,509 10/1/ , , , , ,160 10/1/ , ,766 30, ,766 30,766 10/1/ , ,077 50, ,077 50,077 10/1/ , ,148 91, ,988 91,988 10/1/ , ,924 94, , ,924 TOTAL 1,217,765 1,211,053 1,190,790 1,198,375 1,212,894 (2),(3) Exhibit 7.7 (5),(6) Exhibit PhysLDF

31 Exhibit 7.1 GENERAL LIABILITY INCURRED LOSSES AND ALAE NET OF RETENTION AND DEDUCTIBLES Accident Months Of Development Period /1/ ,594 54, ,309 92, , ,450 10/1/ , , , , , , ,738 10/1/ , ,170 99,500 99,500 99, , ,100 10/1/ , , , , , , ,397 10/1/ , , , , , , ,345 10/1/ , , , , , , ,418 10/1/ , , , , , , ,237 10/1/ ,113 95, , , , , ,317 10/1/ ,432 13,588 13,588 13,588 13,588 13,588 13,588 10/1/ , , ,408 98,908 71,485 71,485 71,485 71,485 10/1/ , , , , , ,236 66,254 66,477 10/1/ ,204 7,704 7,704 42,704 36,295 36,295 36,295 36,295 10/1/ , , , , , , , ,850 10/1/ , , , , , , ,663 10/1/ , , , , , ,286 10/1/ ,601 19,522 14,522 15,649 15,649 10/1/ ,984 92, , ,559 10/1/ , , ,933 10/1/ , ,241 10/1/ ,762 Accident Loss Development Factors Period ult. 10/1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ Mean Latest Latest ex. hi/low Industry Prior Selected Cumulative GLIncurred

32 Exhibit 7.2 GENERAL LIABILITY PAID LOSSES AND ALAE NET OF RETENTION AND DEDUCTIBLES Accident Months Of Development Period /1/ ,694 25,028 72,827 92, , ,450 10/1/ , , , , , , ,738 10/1/ , ,170 99,500 99,500 99, , ,100 10/1/ ,596 38,220 86,732 90, , , ,397 10/1/ ,415 68,653 89, , , , ,345 10/1/ ,936 75,893 94, , , , ,418 10/1/ , , , , , , ,237 10/1/ ,611 72,041 98, , , , ,317 10/1/ ,432 13,588 13,588 13,588 13,588 13,588 13,588 10/1/ ,093 36,383 36,408 70,515 71,485 71,485 71,485 71,485 10/1/ ,206 12,164 27,756 46,205 57,755 65,342 66,254 66,477 10/1/ ,704 7,704 7,704 25,355 36,295 36,295 36,295 36,295 10/1/ ,463 84, , , , , , ,850 10/1/ ,757 61, , , , , ,663 10/1/ , , , , , ,286 10/1/ ,489 14,522 14,522 15,649 15,649 10/1/ ,431 35,528 43,146 83,169 10/1/ , , ,933 10/1/ , ,420 10/1/ ,972 Accident Loss Development Factors Period ult. 10/1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ Mean Latest Latest ex. hi/low Industry Prior Selected Cumulative GLPaid

33 Exhibit 7.3 AUTOMOBILE LIABILITY INCURRED LOSSES AND ALAE NET OF RETENTION AND DEDUCTIBLES Accident Months Of Development Period /1/ ,169 18,312 13,843 13,843 13,843 13,843 13,843 13,843 10/1/ ,189 1,142 1,142 1,142 1,142 1,142 1,142 10/1/ ,131 11,099 11,099 11,099 11,099 35,452 35,452 10/1/ ,340 9,340 9,340 8,840 8,840 8,840 8,840 10/1/ ,996 20,894 19,394 10,546 9,901 9,901 9,901 10/1/ ,191 14,141 7,747 8,132 8,132 8,132 8,132 10/1/ ,584 19,144 18,901 18,901 18,901 18,901 18,901 10/1/ ,222 33,536 86,830 86,830 86,830 86,830 86,830 10/1/ ,987 7,987 7,987 7,987 7,987 7,987 7,987 10/1/ ,583 19,701 37,201 17,431 17,431 17,431 17,431 17,431 10/1/ , ,507 28,507 28,507 28,507 28,507 28,507 28,507 10/1/ ,920 34,921 23,921 23,921 23,921 23,921 23,921 23,921 10/1/ , , , , , , , ,997 10/1/ ,381 20,300 20,300 20,300 20,300 20,300 20,300 10/1/ ,141 28,499 47,999 47,999 47,906 47,906 10/1/ ,597 5,597 5,597 5,597 5,597 10/1/ , ,847 68,011 68,011 10/1/ ,295 14,055 64,805 10/1/ ,403 9,653 10/1/ ,801 Accident Loss Development Factors Period ult. 10/1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ /1/ Mean Latest Latest ex. hi/low Industry Prior Selected Cumulative AutoIncurred