Exhibit No. 99 -------------- Report on Loss Reserving Practices, dated June 2002 REPORT ON [PHOTO] LOSS RESERVING PRACTICES [PHOTO] THE [PHOTO] PROGRESSIVE CORPORATION JUNE 2002 [PHOTO] - ------------------------------------------------------ PREFACE --------- Progressive's report on loss reserving practices describes how we establish and evaluate reserves for policy related claims activity. Our report should help Progressive's investors, employees, and interested parties understand how loss reserves affect our financial results. Specifically, this report seeks to: - - Explain the importance of loss reserves and the effect loss reserves have on our operating results, our balance sheet and our pricing. - - Explain how loss reserve estimates can change over time and how these changes affect earnings. - - Provide an overview of the loss reserve methodologies and processes used by Progressive to determine reserve balances for varied types of risks. - - Discuss reserving methodologies and changes we made in 2001 to improve reserve accuracy. - - Provide a case study, complete with results, to allow readers to test their comprehension of this material after reading this report. The 2002 report improves upon last year's report in several ways: - - We've improved and expanded our section on prior year loss development by providing multiple ways to look at loss development and how prior year development can and will influence current year results. - - Included is a section outlining a few questions actuaries may ask themselves when selecting an ultimate reserve. - - Process changes made during 2001 to improve the accuracy of our reserve estimates are discussed. We feel this report provides a fairly comprehensive explanation of how we look at reserves, will provide insight into why reserve balances change and may lessen the frequency of questions that we are asked to address. Terms used herein are defined in the section entitled "Terms and Definitions" at the end of this report. As always, we welcome feedback, comments and questions. Contact Sara Stehlik, The Progressive Corporation, 6300 Wilson Mills Road, Mayfield Village, Ohio 44143 or e-mail sara_stehlik@progressive.com. Copyright(C)2002 Progressive Casualty Insurance Company. All rights reserved. - ------------------------------------------------------ TABLE OF CONTENTS ------ SECTION I: ABOUT PROGRESSIVE Our Business ................................................. 1 2001 Highlights .............................................. 1 Our Financial Policies and Objectives ........................ 1 Pricing and Loss Reserving ................................... 2 SECTION II: ABOUT RESERVES Loss Reserving: Definition and Stated Goals .................. 4 Reserve Development .......................................... 4 Development Patterns ......................................... 9 Accident vs. Calendar Year Loss Ratio ........................ 9 Trend ........................................................ 10 SECTION III: TYPES OF RESERVES Case ......................................................... 12 Incurred But Not Recorded (IBNR) ............................. 12 Loss Adjustment Expense (LAE) ................................ 13 Involuntary Market Operating Loss Reserves ................... 13 Salvage and Subrogation ...................................... 14 SECTION IV: ESTIMATING LOSS RESERVES Segments ..................................................... 16 Methods ...................................................... 16 Triangle ..................................................... 18 Decomposition Report ......................................... 20 Selecting the Ultimate Reserve ............................... 21 SECTION V: PROCESS CHANGES INTRODUCED IN 2001 Uninsured Motorist Bodily Injury (UMBI) ...................... 22 Flood, Hail and Wind ......................................... 22 Commercial Vehicle ........................................... 22 SECTION VI: CASE STUDY - LOSS RESERVES .............................. 23 SECTION VII: CASE STUDY - LOSS ADJUSTMENT EXPENSE ................... 37 TERMS AND DEFINITIONS ............................................... 44 APPENDIX A .......................................................... 49 SECTION I: - ---------- ABOUT PROGRESSIVE ------------------------------------------------ OUR BUSINESS Since the Progressive insurance organization began business in 1937, we have been innovators growing into new markets and pioneering new ways to meet consumers needs. In 1956, Progressive Casualty Insurance Company was founded to be among the first specialty underwriters of nonstandard auto insurance. Today, The Progressive group of companies (collectively referred to as "Progressive" or "The Company") provide drivers throughout the United States with competitive rates and 24-hour, in-person and online services, through 73 subsidiary companies and two affiliates of the Progressive Corporation. We aspire to be recognized as the preeminent consumer franchise in auto insurance and other specialty property-casualty insurance throughout the United States. The Company's personal lines consists predominately of insurance for automobiles, motorcycles, recreational vehicles and travel trailers. Business is either generated by the Company's network of more than 30,000 independent agents or is written directly by the Company via the telephone (1-800-PROGRESSIVE) or via the Internet (progressive.com). 2001 BUSINESS HIGHLIGHTS 2001 marked our return to profitability. The GAAP Combined Ratio (CR) was 95.2% in 2001, meeting our financial goals, compared to 104.4% in 2000. During 2001, the Company's loss ratios improved reflecting both the Company's intense focus on pricing as well as favorable loss reserve development. During the year, we saw signs of moderating severity trend and experienced $99.0 million, or 1.4% of earned premium, favorable loss reserve development compared to ($75.8) million, or -1.2% of earned premium, adverse development in 2000. We also executed 81 rate revisions in 2001 responding to changes in both severity and frequency trends. While our growth in policies in force was less than we would have liked, we ended 2001 with an underwriting profit of 4.8%. By the end of the year, we reported stronger growth rates and were prepared to manage profitable growth in 2002. Expense management also remains critical. Our Expense Ratio of 21.7% remained generally flat over the prior year and we continue to focus on keeping our expenses low. We constantly push ourselves to identify and implement processing and electronic advancements making it easier and more efficient for our agents and customers to do business with us. OUR FINANCIAL OBJECTIVES We believe that Progressive will be a more successful company if we establish and measure ourselves against goals that are internally consistent, understandable and capable of driving behavior. Transparency about our goals and performance is a critical part of our corporate culture which we believe attracts better investors and increases valuation. Our financial goals are to write business at a 96 combined ratio, grow as fast as possible while still maintaining our profitability and customer service standards and earn an after-tax return on equity over a five year period that is at least 15 percentage points greater than the rate of inflation (as measured by the Consumer Price Index). 1 Through Enterprise Risk Management initiatives, we have identified and manage three separate risk categories that we believe could impair our Company's ability to grow profitably. Those major risk categories include: 1. Underwriting Risk: Failure to grow earned premium and earn a GAAP 96 Combined Ratio; 2. Investment Risk: Failure to maintain a liquid diversified investment portfolio. 3. Financing Risk: Failure to maintain flexibility and earn the cost of equity capital. We believe that the best way to maximize shareholder value is consistently to achieve our financial objectives by actively managing identifiable risks. The loss reserving function, a component of Underwriting Risk, plays an important role in our ability routinely to deliver underwriting profitability, since accurate reserving contributes not only to our ability to price risks accurately, but also contributes to reducing variability in the earnings that we report. Two important statements produced by our accountants are (1) the balance sheet documenting the Company's financial position and (2) the income statement documenting the performance of the Company over a specific period of time. Both statements rely heavily on the Actuary's ability accurately to estimate reserve liabilities. Since loss and loss adjustment expense reserves represent over 40% of our total liability balance it is important that our reserves are as precise as possible. It is also important to note that loss reserves can only be established for events that have already occurred. P&C companies cannot establish a reserve for a hurricane, hail, flood or any other catastrophic event that "may" occur later in the year. With Progressive adopting monthly external reporting of operating results, analysts, shareholders, and employees should expect the Company to experience substantial fluctuations in monthly results when catastrophic events happen because we cannot "slowly build" reserve liabilities for anticipated future events. PRICING AND LOSS RESERVING The goal of pricing or "setting rates" is to determine the premium rates that are adequate to achieve our profitability goals without being excessive, inadequate or unfairly discriminatory to our customers. Premium rates are made up of three components: (1) The amount of premium needed to cover losses and loss adjustment expenses (2) The amount of premium needed to cover organizational expenses which allow us to provide insurance coverage and service our customers (3) The amount of premium needed to realize our underwriting profit goal Our underwriting profit goal is 4 percent of all premium we earn or a 96% combined ratio. Ultimate losses and loss adjustment expenses represent the largest expense component of the combined ratio. If these estimates are inaccurate, insurance rates could be too high, which restricts growth, or too low, which impacts profitability. We segregate the duties of the pricing analysts and loss reserving analysts to review independently expected losses and expected loss adjustment expenses. If these estimates generate materially different values, assumptions are reviewed and revised. Sensitivity analysis and credibility ratings are considered to develop a single projected amount for losses and loss adjustment expenses. Both groups must consider changes in trends which impact car prices, medical costs and litigation expenses. Standard actuarial techniques are used to project losses and to evaluate trend. 2 The Product Management Group (Pricers, Product Managers and General Managers) use a cost-plus strategy, beginning with the projected ultimate loss and loss adjustment expense amount. Future organizational expenses are then forecasted. Future premium rates are the sum of an estimate of future ultimate losses and loss adjustment expenses added to forecasted organizational expenses and a 4 percent profit load. The comparison of future premium rates to current rates indicates the rate change required. Pricers, and ultimately Product and General Managers, are responsible for detailed rate level analysis and investigating frequency and severity of losses to determine areas where rate adjustments are necessary. The Product Management Group is also involved in the review of underwriting rules which determine premium rates. Lastly, this Group is required to work with each State's Department of Insurance to ensure our business practices align and are compliant with the state laws. Immediately following is a chart illustrating expense components expressed as a percent of earned premium. Loss and loss adjustment expenses represent approximately 74% of every premium dollar earned and underwriting expenses represent approximately 22%, leaving us with a 4% underwriting profit. [GRAPH] EXPENSE COMPONENTS Percent of Earned Premium Losses (61%) LAE (13%) Underwriting Exp. (22%) Underwriting Profit (4%) 3 SECTION II: - ----------- ABOUT RESERVES ---------------------------------------------------- LOSS RESERVING: DEFINITION AND STATED GOALS Loss and LAE reserves represent our best estimate of Progressive's ultimate liability for losses and loss adjustment expenses that occurred prior to the end of any given accounting period but have not yet been paid by the Company. These estimates will be influenced by many variables that are difficult to quantify, such as price inflation on automobile repair costs and medical costs which will influence the final amount of the claims settlement. That, coupled with changes to internal claims practices, changes in the legal environment and changes in state regulatory requirements, makes reserve setting as much an art as a science. Loss and LAE reserves include: 1. Case reserves that are assigned to a specific claim 2. IBNR reserves for claims that have occurred but have not been recorded (Incurred But Not Recorded) 3. LAE reserves to cover the costs the company incurs to settle claims, other than loss payments (Loss Adjustment Expense) 4. Involuntary Market Operating Loss Reserves for the expected operating losses from state assigned risk plans for private passenger and commercial auto insurance. RESERVE DEVELOPMENT Our goal is to ensure total reserves are adequate to cover all loss costs, while sustaining minimal variation from the time reserves are initially established until losses are fully developed. Exhibit 1 illustrates our performance against this goal. EXHIBIT 1 ANALYSIS OF LOSS AND LOSS ADJUSTMENT EXPENSES (LAE) DEVELOPMENT (NOT COVERED BY REPORT OF INDEPENDENT ACCOUNTANTS) FOR THE YEARS ENDED DECEMBER 31, 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 - ------------ Original Loss and LAE estimated Reserves liability----------------- from claims on business written and earned during that year. $861.5 $956.4 $1,012.4 $1,098.7 $1,314.4 $1,532.9 $1,867.5 $1,945.8 $2,200.2 $2,785.3 $3,069.7 - ------------ - ------------ Re-estimated reserve estimates reveals accuracy of original----- estimation. Re-estimated - ------------reserves as of: One year later 810.0 857.9 869.9 1,042.1 1,208.6 1,429.6 1,683.3 1,916.0 2,276.0 2,686.3 Two years later 771.9 765.5 837.8 991.7 1,149.5 1,364.5 1,668.5 1,910.6 2,285.4 Three years later 718.7 737.4 811.3 961.2 1,118.6 1,432.3 1,673.1 1,917.3 Four years later 700.1 725.2 794.6 940.6 1,137.7 1,451.0 1,669.2 Five years later 695.1 717.3 782.9 945.5 1,153.3 1,445.1 Six years later 692.6 711.1 780.1 952.7 1,150.1 Seven years later 688.2 709.2 788.6 952.6 Eight years later 687.9 714.6 787.5 Nine years later 690.3 713.3 Ten years later 691.1 Additions to or Cumulative releases development: of original conservative/ reserve.----(deficient) Additions $170.4 $243.1 $224.9 $146.1 $164.3 $87.8 $198.3 $28.5 ($85.2) $99.0 are shown as negative. Percentage 19.8% 25.4% 22.2% 13.3% 12.5% 5.7% 10.6% 1.5% -3.9% 3.6% -------------- Percent overstated/ understated when compared to original estimate ---------------- 4 For accident years 1999 and prior, loss reserves as estimated at 12/31/99, were deficient (will cost us more than originally estimated) by $85.2 million based on the information we had through 12/31/2001. We initially established reserves for 2000 and prior accident years as of 12/31/2000 at $2,785.3 million. By the end of 2001 we estimated the same claims during that accident year will now cost the Company $2,686.3 million, somewhat less, or a favorable development of $99 million. Unlike manufacturers, insurers do not have the advantage of being able to account accurately for the cost of goods sold prior to the sale of the product. Insurance is a promise or commitment to cover the cost of future claims that may or may not occur. We are required to forecast how much claims will ultimately cost us and establish or set a reserve liability for claims that have happened until we ultimately release payment to the claimant and close the claim. During any given accident year, we will review the reserves that were originally estimated and determine if they are developing consistently with our original predictions. Often, the amount we ultimately need to cover outstanding claims during an accident year will differ from the estimate made at the time we originally established the reserve. This is called reserve development. Reserve development will have an effect on calendar year financial results. In exhibit 2, we show how. We refer to this exhibit as the "if we knew then what we know now" exhibit since it shows the contribution of prior year reserve development on the current earnings per share. In other words, if a company under-reserves, it will report a higher income level. Conversely, if a company over-reserves, it will report a lower income level. EXHIBIT 2 [GRAPH] CONTRIBUTION OF PRIOR YEARS DEVELOPMENT ON EARNINGS PER SHARE 1994 1995 1996 1997 1998 1999 2000 Reported EPS $1.20 $1.09 $1.38 $1.77 $2.04 $1.32 $0.21 EPS after Developed $1.21 $1.40 $1.33 $2.04 $1.59 $0.99 $0.72 represents net income per share on a diluted basis adjusted to reflect 3:1 stock split April 2002 5 The nature of the insurance business is to estimate loss costs over time as premium is earned and losses develop. Estimated loss costs are considered to be a liability of the Company since they will eventually be paid to claimants. Changes are often made to our original estimated loss costs because the claim was closed for more or less than we originally anticipated. Inflation will also change our ultimate liability estimate. Reserve development is an important concept to understand. As shown in exhibit 2, reserve development created from claims that happened in prior years can have an influence on the current year earnings. Progressive feels it is such an important concept that this year we began quarterly disclosure of current and prior year actuarial adjustments. We also began disclosing prior year reserve development, which represents the sum of actuarial adjustments, adjustments made to claims reserves by claims adjusters, and other adjustments (predominately closed claims). At this point, it may be appropriate to provide a model, showing the life of multiple claims and how reserve development works. We will exhibit three separate claims, pointing out prior period reserve development by year. 6 LIFE OF A CLAIM AND ITS RESERVE CLAIM #1 DATE ACTIVITY RESERVE PAID PRIOR YEAR RESERVE DEVELOPMENT --------------------------------------------------------------------------- NOTE: CLAIM INCURRED IN 02/00 Open Claim ACCIDENT YEAR 2000. --------------------------------------------------------------------------- --------------------------------------------------------------------------- Reserve established using 02/00 segment average set by Actuaries $100 --------------------------------------------------------------------------- --------------------------------------------------------------------------- NOTE: ADDITIONAL RESERVE Actuarial review of segment NEED DETERMINED IN 2001 FOR A 04/01 average indicates an additional $105 +$5 CLAIM THAT HAPPENED IN PRIOR reserve need of $5 YEAR (2000). --------------------------------------------------------------------------- --------------------------------------------------------------------------- NOTE: CLAIM PAID IN CALENDAR Claim closed by claims adjuster YEAR 2001 AND CLAIM COST 08/01 for less than the amount of $104 COMPANY LESS THAN THE reserve. ACTUARIAL AVERAGE --------------------------------------------------------------------------- --------------------------------------------------------------------------- 08/01 Reserve taken down $0 -$1 --------------------------------------------------------------------------- ============ TOTAL PRIOR YEAR RESERVE DEVELOPMENT IN 2001 FOR CLAIM #1 $4 ============ CLAIM #2 DATE ACTIVITY RESERVE PAID PRIOR YEAR RESERVE DEVELOPMENT --------------------------------------------------------------------------- NOTE: CLAIM INCURRED IN 03/00 Claim opened ACCIDENT YEAR 2000. --------------------------------------------------------------------------- --------------------------------------------------------------------------- Reserve established at segment 03/00 average $100 --------------------------------------------------------------------------- --------------------------------------------------------------------------- NOTE: CLAIM WAS SETTLED FOR Claim closed based on the merits THE EXACT AMOUNT ORIGINALLY 03/01 of the claim at $100 $100 RESERVED. --------------------------------------------------------------------------- --------------------------------------------------------------------------- Reserve taken down 03/01 0 --------------------------------------------------------------------------- ============ TOTAL PRIOR YEAR RESERVE DEVELOPMENT IN 2001 FOR CLAIM #2 $0 ============ CLAIM #3 DATE ACTIVITY RESERVE PAID PRIOR YEAR RESERVE DEVELOPMENT --------------------------------------------------------------------------- NOTE: CLAIM INCURRED IN 06/01 Claim opened ACCIDENT YEAR 2001. --------------------------------------------------------------------------- --------------------------------------------------------------------------- Reserve established at segment 06/01 average $150 --------------------------------------------------------------------------- --------------------------------------------------------------------------- NOTE: ACTUARIES REVIEWED Actuarial review of segment SEGMENT AND CHANGED RESERVE indicates segment average IN SAME YEAR THAT CLAIM 11/01 reserve should be adjusted to $160 HAPPENED. THEREFORE, NO $160 PRIOR YEAR DEVELOPMENT HAS OCCURRED UP TO THIS POINT. --------------------------------------------------------------------------- --------------------------------------------------------------------------- Inflation factor applied to open 12/01 reserves within the segment as a $170 result of the claim aging. --------------------------------------------------------------------------- --------------------------------------------------------------------------- NOTE: CLAIM CLOSED FOR LESS Claim closed by claims adjuster THAN THE AMOUNT RESERVED IN 03/02 $163 SUBSEQUENT YEAR. PRIOR YEAR ADJUSTMENTS TOTAL $7. --------------------------------------------------------------------------- --------------------------------------------------------------------------- 03/02 Reserve taken down $0 -$7 --------------------------------------------------------------------------- ============ TOTAL PRIOR YEAR RESERVE DEVELOPMENT IN 2002 FOR CLAIM #3 -$7 ============ Summary: In 2001, we had prior year reserve adjustments of $4. In 2002, we had prior year reserve adjustments of ($7) Therefore, calendar year 2000 estimated losses were understated by $4 and calendar year 2001 losses were overstated by $11 ($7+$4). 7 Starting in the first quarter of 2002, Progressive began quarterly disclosure of current and prior year reserve development. Below in exhibit 4 is a copy of that portion of the schedule that reports this new information. Immediately following, we run through a few calculations on interpreting the information. For those readers interested in reviewing a full copy of the new supplemental information released, we have provided a copy of the entire schedule as appendix A. exhibit 4 THREE MONTHS ENDED MARCH 31, -------------------- 2002 2001 -------- -------- NET PREMIUMS EARNED Personal Lines-Agent $1,241.5 $1,132.5 Personal Lines-Direct 517.0 398.4 -------- -------- Total Personal Lines 1,758.5 1,530.9 Other Lines 209.0 143.0 -------- -------- Companywide $1,967.5 $1,673.9 ======== ======== COMPANYWIDE GAAP CR Loss & loss adjustment expense ratio 69.0 76.2 Underwriting expense ratio 21.5 21.1 -------- -------- 90.5 97.3 ======== ======== COMPANYWIDE ACCIDENT YEAR Loss & loss adjustment expense ratio 69.2 76.7 ======== ======== ACTUARIAL ADJUSTMENTS-Favorable/(Unfavorable) Prior accident years $ 13.9 Current accident year 0.2 * -------- -------- Calendar year actuarial adjustment $ 14.1 ======== ======== PRIOR ACCIDENT YEARS DEVELOPMENT-Favorable/(Unfavorable) Actuarial adjustment $ 13.9 All Other development (10.8) * -------- -------- Total Development $ 3.1 $ 8.9 ======== ======== By looking at the report, we can easily determine that earned premiums have grown and our profitability has also improved over the periods shown. We also know that prior year reserves were originally established at a level slightly higher than we now expect them to be paid and releasing these reserves into calendar year 2002 results provided a favorable 0.2 impact to this years calendar year GAAP loss and LAE ratio. 2002 Accident Year Loss and LAE ratio 69.2 less 2002 Calendar Year Loss and LAE ratio 69.0 ----- Favorable/(Unfavorable) Development: 0.2% influence on GAAP CR due to prior year adj.'s When you multiply the .2% (rounded) favorable development by the reported earned premium of $1,967.5 million, you will arrive at the $3.1 million total development for prior accident years reported in exhibit 4. This reconciles to the amount of reserves that Progressive released into current year earnings from claims that occurred prior to the beginning of 2002 but were not yet paid. 8 DEVELOPMENT PATTERNS The longer a claim stays open, the more difficult it is to establish an accurate reserve. Injury claims (e.g. bodily injury or "BI") are long-tailed, require more data points and are more sensitive to inflation when reserving than short-tailed property damage claims. The "tail" refers to the length of time it takes to settle a claim once it has occurred. The shorter the tail, the less time there is for unanticipated occurrences to impact further the size of the claim and thus, should show relatively stable loss development statistics over the years. In other words, short tailed claims should develop close to original reserve projections. Below, we have provided an updated exhibit to the one we provided in last year's report comparing the number of months it takes to settle bodily injury claims vs. the short tailed property damage claims. exhibit 5 BODILY INJURY AND PROPERTY DAMAGE PAID DEVELOPMENT PATTERNS [GRAPH] ACCIDENT VS. CALENDAR YEAR LOSS RATIO Understanding the meaning of the accident year loss ratio vs. the calendar year loss ratio becomes much clearer once the reader understands reserve development. Actuaries study claim development on an accident year basis while our accountants report results on a calendar year basis. Therefore, we feel that it is important to report both accident year and calendar year results by quarter to our pricing organization, management team and shareholders. We also produce the same information monthly for internal use by state, channel and product to enable management to understand trend and price better. CALENDAR YEAR LOSS RATIO. When losses are paid or reserve changes are made, we account for those payments or reserve changes during the calendar year in which the payment or reserve change took place, regardless of when the accident happened. Prior calendar year results are not restated as new reserving information comes to light. ACCIDENT YEAR LOSS RATIO. Our actuaries estimate reserves on an accident year basis. Losses which occur in any period are assigned to that accident period. A period may be defined as monthly, quarterly, semiannual or annual. Reserves, regardless of when they are established or changed are assigned to that same accident period. Accident period results will change over time as future information develops and as our estimates of loss costs change accordingly. To determine accident year loss ratio, we add estimated ultimate costs of all losses that occurred in a specific accident period and divide that sum by the total amount of premium earned for that same period of time. 9 Exhibit 6 shows how accident year loss trends impact calendar year results. Unfavorable loss trends came out of the 1999 accident year causing the 2000 calendar year loss ratio to increase by 1.2%. In the subsequent year, 2001, Progressive reduced the 2000 accident year reserves by $99.0 million causing the 2001 calendar year results to improve by a full 1.4%. EXHIBIT 6 ANALYSIS OF CALENDAR YEAR DEVELOPMENT (based on one year development of prior year) 2000 2001 Premiums Earned $6,348.4 $7,161.8 Accident Year Incurred Loss $5,203.6 $5,363.1 --------------------- Prior Year Development** ($75.8) $ 99.0 ------------------------ --------------------- Calendar Year Incurred Loss $5,279.4 $5,264.1 Accident Year Loss Ratio 82.0% 74.9% PRIOR PERIOD DEVELOPMENT -1.2% 1.4% ------------------------ --------------------- Calendar Year Loss Ratio 83.2% 73.5% Expense Ratio 21.2% 21.7% ------------------------ Calendar Year Combined Ratio 104.4% 95.2% Accident Year Combined Ratio 103.2% 96.6% ** For year 2000 prior year development of ($75.8), refer to the 2000 annual report TREND Trend is the rate of change in both frequency and severity of claims. One key aspect of trend is inflation. The anticipated effect of inflation is explicitly considered when estimating liabilities for losses and LAE. While anticipated price increases due to inflation are considered in estimating the ultimate claim costs, the increase in the average severity of a group of claims is caused by a number of factors that vary with the individual type of policy written. Future average severities are projected based on historical trends adjusted for anticipated changes in underwriting standards, policy provisions and general economic conditions. These anticipated trends are monitored based on actual development and are modified when necessary. Trend estimates allow us to anticipate settlement costs in the future. We compare old data to new data to project how much future costs will increase or decrease over time. We call the result of this calculation "trend." Settlement costs will be affected by external factors such as inflation, change in medical costs, changes in labor rates and car part prices, and changes in litigation and court decisions. At the same time, internal factors such as changes in claims practices (speeding up or slowing down the settlement process) will impact trend. The purpose of a trend factor is to restate losses from some historic period to levels appropriate in a future payment period, similar to an inflation adjustment. If the trend factor used to estimate reserves exceeds the actual trend experienced, we could over reserve. If the trend factor used to estimate reserves is less than the actual trend experienced, we may under reserve. 10 There are two different parameters that we study regarding trend: severity and frequency. Severity measures the dollar amount of claims during a given accident period divided by the total number of claims. Frequency measures the number of ultimate claims by accident period divided by the number of earned car years during that period. Severity times frequency equals pure premium (loss costs per Earned Car Year). Reviewing the trend of these two parameters allows us to better understand the change impacting our pure premium trend. Industry-wide trend data is distributed by a statistical agent called The National Association of Independent Insurers (NAII). We review the NAII data to identify similarities between what NAII is reporting and what Progressive is experiencing. The NAII reported data lags Progressive's data by a quarter. For comparative purposes and because Progressive's trend results historically have been more heavily influenced by the nonstandard market, Progressive generates trend information not only on a consolidated basis but also separates its preferred and ultra-preferred books from the other tiers (nonstandard, middle market and standard). This extra step provides us with a better indication of frequency trend because the preferred and ultra preferred tiers of our business are growing more rapidly and may disguise changes in trend in the other tiers. In other words, separating the data provides us with a better understanding of trend differences between high risk drivers and standard drivers. Pricing analysts and product managers also prefer to review the data separately. Each quarter, we produce severity, frequency and pure premium trend results for each line coverage (bodily injury, property damage, collision, comprehensive and personal injury protection) by tier and also for those states that represent our largest market share. We compare these results to the same period in the previous year. We review the trend factors for consistent improvement, deterioration, or variability from quarter to quarter. It is common practice for us to review a four-quarter moving average when examining a trend factor. Such a time-series analysis will contain seasonality, cyclicality, randomness and true trend. The moving averages will tend to eliminate and/or flatten both seasonality and randomness. From this analysis, we use the trend factor to bring historical losses to tomorrow's cost. Frequency trend tends to be affected by changes in the weather, changes in laws, changes in gasoline prices (both increases and decreases) and changes in population density, among other factors. Severity trend tends to be principally affected by changes in medical costs and repair costs. 11 SECTION III: - ----------- TYPES OF RESERVES ------------------------------------------------- Reserves are considered an operating liability on our balance sheet. While reported in aggregate, there are actually many different types of reserves, all of which are reviewed independently of each other in order to get the total reserve balance as accurate as possible: 1. CASE RESERVES represent the largest portion of reserves held on the balance sheet for automobile products. Case reserves are accrued to pay claims that have already been reported by the claimant and recorded into Progressive's systems, but have not yet been paid. There are two types of case reserves included in our financial statement results: AVERAGE RESERVES. An average is determined for each claim regardless of individual claim characteristics and is assigned to a claim at the time it is reported to the Company. Averages vary by "segment". A segment combines data by state, product, and line coverage and is determined by the actuaries based on the Company's experience. As the claim ages (the length of time from the accident date to the current accounting period), the reserve for that claim will generally increase until the claim is paid. This is because history has proven that the severity of a claim generally increases with age. We use two different factors to increase the case reserve, inflation and ANCR (additional needed case reserve). ANCR will be discussed later in this section. Generally, an average reserve is based on our estimate of total needed reserves excluding the adjuster set reserves (defined immediately below) for that type of claim (BI, PD, PIP, etc.). During the third quarter of 2000, we redefined the attachment point (or threshold) we used for setting bodily injury (BI)case reserves by formula. The threshold change was initiated because Progressive now writes a substantial percentage of its business on policies with higher limit policies. By increasing the threshold, a larger portion of our BI reserves now use average, actuarially-set reserves as opposed to adjuster set reserves. ADJUSTER-SET RESERVES. We use adjuster estimates of how much a claim will cost when the reserve estimate is above the threshold vs. using the statistical average set by the actuaries. We include the adjuster estimates of how much a claim will cost when the adjuster-set reserve provides more reliability than the estimates based on statistical averages as the occurrence of larger claims happens less routinely. Our goal is to use the average reserve approach for as many claims as possible, since using average reserves allows claims personnel to concentrate their efforts on adjusting claims rather than accounting for them. However, for claims in excess of our threshold, we have determined adjuster-set reserves provide a better estimate of our ultimate liability because the adjuster has typically spent a great deal of time on these large claims and understands the individual claim characteristics. Our average reserves are adjusted for inflation monthly as claims age while our adjuster-set reserves are not. Thus, reserve changes arising from analyses by our actuaries which includes both average and adjuster set reserves, will reflect shifts in anticipated development patterns. This also contributes to the consistency in data we need for rate making. 2. INCURRED BUT NOT RECORDED RESERVES (Loss IBNR Reserves) are estimates of amounts needed to settle losses which have already occurred but have not been recorded by the Company. In other words, we estimate an unknown group of claims and determine the number (frequency) and dollar amount (severity) of the claims that have already happened and we are liable for, but which have not yet been reported by claimants. We track IBNR emergence to known statistics 12 (generally earned car years) that measure our past exposure, then apply the value to the current exposure to produce the IBNR reserve. We convert our projected needed IBNR to a percentage of earned premium and then apply that percentage to past earned premium to establish the carried amount for the financials. When setting an IBNR reserve, we identify the lag time between when the premium was earned and when the losses will be reported. Different IBNR factors are assigned to different lag time periods and decline over time. ADDITIONAL NEEDED CASE RESERVES (ANCR). Additional Needed Case Reserve is a reserve intended to cover further development of the adjuster's estimates for large losses, particularly on policies with higher limits (in excess of the threshold). Many companies include ANCR in their IBNR estimates. 3. LOSS ADJUSTMENT EXPENSE (LAE) RESERVES are estimates of the costs that we will incur to settle claims other than the indemnity payments themselves. There are two categories of LAE reserves: 1. Defense and Cost Containment (DCC) (roughly equivalent to the old term Allocated Loss Adjustment Expense (ALAE) Reserves): These are costs associated with defense, litigation, and cost containment and include the following: accident investigation, fixed amounts for medical cost containment, litigation management expenses, fees for appraisers, hearing representatives, attorney fees, etc. 2. Adjusting Expenses, formerly known as Unallocated Loss Adjustment Expenses(ULAE) Reserves: All other loss adjusting costs, including salaries and benefits of claims adjusters, rent and occupancy, systems, and fees paid to outside adjusters. Both DCC and Adjusting Expense reserves are reviewed and estimated by segment (a segment is a subset of our total data and is defined by a combination of state, product, and line coverage) as an average for each open case reserve and a percentage of that segment's total loss IBNR reserves. As a result, LAE reserve averages may vary by state, product and line coverage. 4. INVOLUNTARY MARKET OPERATING LOSS RESERVES (ASSIGNED RISK RESERVES). State insurance regulations require us to participate in various assigned risk plans. Generally, assigned risk plans provide insurance to those who cannot find a carrier that will insure them voluntarily. Participation requirements in an assigned risk plan differ from state to state, but generally each carrier is assigned a certain number of "state plan risks" based on the amount of premium the company writes in that state and type of coverage offered. PRIVATE PASSENGER ASSIGNED RISK PLAN. Current writings obligate us to accept assigned risk business. We generally expect to incur an operating loss on these assignments in most states (premium less losses and expenses). The assigned risk reserve is a function of (1) our voluntary market share (2) the amount of business we voluntarily write (3) the expected size of the assigned risk pool and (4) the expected operating results of the policies we will be assigned. The process of determining the reserve for a state is to multiply our market share by the size of the assigned risk pool in that state to determine our portion of the pool. We reduce this product by any state credits we receive for business that we wrote voluntarily that would otherwise be considered assigned risk type business. We then multiply the estimated premium we expect to receive from the state by the operating loss percentage we expect to incur (generally equal to the expected Combined Ratio-100 percent) to determine the assigned risk reserve. 13 COMMERCIAL AUTO INSURANCE PLAN (CAIP). Progressive is also required to share in the operating results of state commercial auto plans. Due to the more complex nature of commercial business, these plans do not assign policies, rather they operate as a Joint Underwriting Association (JUA) and use a small number of carriers to "service" the business. These companies are known as "servicing carriers." Progressive is a servicing carrier in 26 states. The servicing carriers transfer the insurance risk or "cede" 100 percent of the business to the state CAIP plans. The state plans then retrocede, or further transfer, the operating results to all companies writing commercial automobile business in proportion to their share of the voluntary market for the state in question. If the plan generates a loss, we are assessed for our portion of the loss. If the plan generates a profit, we participate in the profits. As with the private passenger assigned risk plan, there is a two year lag from the time our writings incur the liability until we are assessed. But again, because the amount of participation is assigned to Progressive based on current year writings, we reserve for this liability in the current operating year. SALVAGE AND SUBROGATION. As required by Generally Accepted Accounting Principles (GAAP), our loss reserves are stated net of a deduction for the estimated amounts of salvage and subrogation that we will recover. Field claims offices investigate claims and identify salvage or subrogation potential. When potential salvage and/or subrogation opportunities exist, they are entered into our claims system. Salvage follow up is handled by branch claims offices while subrogation work is handled by specialized units across the country. While salvage and subrogation balances are reviewed by our actuaries, they are not considered an independent type of reserve. Salvage and Subrogation are different transactions. Below is an explanation of each: SALVAGE: Progressive takes title to a vehicle that has been stolen and recovered or has been declared a total loss. We sell the remnants to a salvor. Work is handled by branch claims offices. SUBROGATION: Progressive insured is involved in an accident where the other party is at-fault. The insured submits the loss to Progressive; we pay the claim and obtain the insured's right to recover damages from the at-fault party. Work is handled by a specialized group. TYPICAL SUBROGATION CLAIM ------------------------- (3) ------------------------ ------------------------ ABC Auto Insurance Progressive ---------- Carrier ------------------------ ------------------------ | | | (2) | | | | | | | | (1) | ------------------------ ------------------------ Insured Vehicle ---------- Claimant Vehicle ------------------------ ------------------------ (1) ABC Auto Insurance policyholder rear-ends Progressive insured (2) Progressive pays collision claim to Progressive insured (3) Progressive subrogates ABC Auto Insurance to recover balance owed by claimant. 14 We thought it would be interesting to provide a chart showing Progressive's distribution of net loss reserves through 12/31/01. Case loss reserves represent the largest proportion of the reserves we maintain, followed by LAE (DCC plus Adjusting Expenses) and then IBNR (which includes salvage and subrogation). [PIE CHART] exhibit 7 RESERVE DISTRIBUTION Progressive Insurance (12/31/01) Case (64.2%) IBNR (16.4%) LAE (18.6%) Assigned Risk (0.8%) 15 SECTION IV: - ------ ESTIMATING LOSS RESERVES ----------------------------------------------- SEGMENTS When reviewing the need to increase or decrease our reserve balances, we break the data into "segments." Segments are defined as a combination of state, product, and line coverage. The following chart provides examples of three different segments: - ---------------------------------------------------------------------------------------------------------- STATE PRODUCT LINE COVERAGE - ---------------------------------------------------------------------------------------------------------- New York (down Private Passenger Personal Injury SEGMENT #1 state) Automobile Protection - ---------------------------------------------------------------------------------------------------------- SEGMENT #2 Ohio Motorcycle Bodily Injury - ---------------------------------------------------------------------------------------------------------- SEGMENT #3 California Private Passenger Property Damage Automobile - ---------------------------------------------------------------------------------------------------------- Each business segment is reviewed by our actuarial staff to understand reserving needs at a very detailed level as long as the data under review is large enough to produce credible results. Credibility can be defined as a statistical measure of the reliability of experience data. Currently, there are approximately 250 independent reserve segments. Segments are generally reviewed annually, semiannually or quarterly depending on the size of the segment. Segments with large reserve balances are reviewed more often. Smaller segments (or states) that are not large enough to provide credible results on their own are often consolidated with other segments having similar risk characteristics. When a state begins to write enough premium to independently produce credible results within a given coverage, we will create a new segment and review that state separately. We believe that reserving by segments allows us to reserve more accurately and contributes significantly to better pricing practices. By reviewing reserve needs at such a detailed level, we also have the ability to identify and measure variances in trend by state, product, and line coverage that we would not otherwise see on a consolidated basis. 16 The most important benefits of segmentation are the ability to: a. measure individual segment profitability more accurately leading to more accurate pricing. b. respond more quickly to product mix changes. c. respond to changes in claims handling on a state-by-state basis. d. provide independent state results by product and by line coverage avoiding the potential of adverse selection in the pricing process. Trends for one line coverage may differ substantially from the trend we see in another line coverage. For example, severity and frequency may both be going up for the Bodily Injury (BI) segments we review, while the cost and number of claims for Property Damage (PD) may be going down. For this reason, there may be instances where our actuaries find it necessary to increase reserves one month (the month in which a substantial number of BI segments are reviewed) and then turn around the next month and release reserves (the month in which significantly more PD segments are reviewed). We point this out, so there is an understanding that variability in reserve changes (strengthening vs. releasing) can and will happen from one month to the next. ESTIMATION METHODS DEFINED Our standard procedures require us to review the results of six different estimation methods prior to determining if a reserve change is required. Three of the six methods use paid data and the other three methods use incurred data. There are strengths and weaknesses to both data sets. The three "paid" methods listed below (amount paid, average paid and modified paid) all utilize paid loss data. The paid methods estimate growth and development of an accident year by looking at the development of earlier accident years. This method assumes that past paid loss development is a predictor of future paid loss development. The primary strength of this method is that it removes the potential for distortions that may be created by including estimated data. The drawback to the paid method is that it does not always accurately project ultimate losses in the most recent periods under review since it also depends heavily on consistent claims closure or settlement practices. If these practices change, the results of the paid methods could be distorted. The three incurred methods use paid losses combined with loss reserves in any given accident year and assume historical incurred loss development will be predictive of our future incurred loss development. The primary strength of incurred methods is that we can make use of our Company's ultimate loss estimates for reported claims. This is especially critical when estimating the longer tail "BI" type claims. PROGRESSIVE USES SIX INDICATION METHODS AMOUNT PAID We take the total loss dollars paid by accident period and project them to an estimated ultimate level. We review total loss dollars only and base our future loss development projections largely on the development of prior periods. This method uses only one triangle of data. The "triangle" is a tool used by actuaries to show how estimates of incurred or paid to date amounts have changed or "developed." Further discussion of the "triangle" follows on the next page. AVERAGE PAID Paid severity by accident is projected to an ultimate level then multiplied by the ultimate number of feature counts to be paid ("feature" is defined in the terms and definitions section at the end of this document). Ultimate paid feature counts equals the projected ultimate number of features recorded generally multiplied by (1-ultimate CWP rate). CWP rate is the percentage of claims recorded but closed without payment. This method uses three triangles of data, paid severity, CWP rate, and feature count. 17 AMOUNT INCURRED We take the total dollars incurred by age and accident period and project them to an ultimate level. We review total dollars incurred only and base our future loss development projections largely on the development of prior periods. This method uses only one triangle of data. AVERAGE INCURRED Incurred severity by accident is projected to an ultimate level then multiplied by the ultimate number of features to be paid. Ultimate incurred feature count equals the projected ultimate number recorded multiplied by (1-ultimate CWP rate). This method uses three triangles: incurred severity, CWP rate, and recorded feature counts. MODIFIED PAID This method uses the paid-loss development pattern. We use the development of paid losses to determine the percent unpaid. Then we apply the percent unpaid to the expected ultimate loss to arrive at the expected unpaid amount. The unpaid amount is added to actual paid-to-date. MODIFIED INCURRED This method uses the incurred loss development pattern. We use the development of incurred losses to determine the percent unreported. Then we apply the percent unreported to the expected ultimate losses to arrive at the expected unreported amount. The unreported amount is added to actual losses incurred to date. In the event we see wide variation between results reported using the six methods noted above, we will further analyze the data with additional techniques. THE TRIANGLE Contained within the definitions of the methods we use to estimate reserve needs, we refer to the use of a "triangle." The triangle is a tool used by actuaries to show how loss reserve estimates have changed or "developed" over time. Development can be defined as the estimate of losses paid or incurred on business produced in previous years at subsequent evaluation dates. During a routine segment review, we run a triangle showing 14 semiannual accident periods. The last number along the diagonal represents the most recent semiannual period result. The triangle shown in exhibit 8 displays average incurred data. We compare development from one semiannual accident period to the next to develop our loss reserve factors. An improvement over last year, we are now including prior period selects on all triangles to determine whether what we saw in our prior review has changed, is the same, or highlights the fact that we are experiencing something very different from what we expected. This helps in the consistency of our reviews by highlighting changes in the data and suggests if we need to conduct more analytical rigor on the review prior to making our ultimate reserve selection. 18 exhibit 8 ANYSTATE-ANY LINE COVERAGE AS OF FEBRUARY 28, 2002 Semiannual Accident AVERAGE INCURRED LOSSES - ACCIDENT PERIOD ANALYSIS Periods Ending 1 2 3 4 5 6 ------ - - - - - - Aug-95 9,160 11,455 11,210 10,917 10,513 10,187 Feb-96 9,754 12,450 12,188 11,833 10,649 10,259 Aug-96 10,259 12,687 12,435 12,151 12,107 12,201 Feb-97 10,098 11,846 11,843 11,877 11,876 11,851 Aug-97 8,835 10,130 10,727 10,896 11,075 11,186 Feb-98 8,289 10,638 11,374 12,126 12,558 12,884 Aug-98 7,926 9,982 11,153 11,917 12,527 12,602 Feb-99 7,506 10,274 11,448 11,956 12,412 12,825 Aug-99 8,596 10,897 11,227 11,738 12,083 11,944 Feb-00 9,373 10,987 11,791 11,946 11,955 Aug-00 9,725 11,541 12,121 12,229 Feb-01 10,847 11,689 11,711 Aug-01 (#3) A 9,773 B 11,254 Feb-02 10,247 1-2 2-3 3-4 4-5 5-6 6-7 --- --- --- --- --- --- Aug-95 1.250 0.979 0.974 0.963 0.969 0.969 Feb-96 1.276 0.979 0.971 0.900 0.963 0.977 Aug-96 1.237 0.980 0.977 0.996 1.008 1.021 Feb-97 1.173 1.000 1.003 1.000 0.998 1.009 Aug-97 1.147 1.059 1.016 1.016 1.010 1.025 Feb-98 1.283 1.069 1.066 1.036 1.026 1.017 Aug-98 1.259 1.117 1.069 1.051 1.006 1.014 Feb-99 1.369 1.114 1.044 1.038 1.033 0.993 Aug-99 1.268 1.030 1.046 1.029 0.988 Feb-00 1.172 1.073 1.013 1.001 Aug-00 1.187 1.050 1.009 Feb-01 1.078 1.002 Aug-01 (#3) C 1.152 Straight Avg 1.219 1.038 1.017 1.003 1.000 1.003 Avg x HiLo 1.219 1.036 1.016 1.010 1.001 1.005 Avg Last 4 1.147 1.039 1.028 1.030 1.013 1.012 Fitted 1.164 1.033 1.012 1.013 1.006 1.006 Special Adj. 0 0 0 0 0 0 Prior Sel. @ 6 Mth (#1) 1.193 1.067 1.043 1.039 1.019 1.016 Prior Sel. @ 3 Mth (#1) 1.160 1.050 1.031 1.025 1.023 1.015 SELECT 1.147 1.039 1.028 1.020 1.013 1.012 Cumulative 1.292 1.126 1.084 1.055 1.034 1.020 ======================================================================================================================= FEB-02 AUG-01 FEB-01 AUG-00 FEB-00 AUG-99 ------ ------ ------ ------ ------ ------ ULT. SEVERITY 13,241 12,677 12,699 12,899 12,363 12,187 ULT. COUNTS 1,646 1,927 2,431 3,047 3,426 3,260 ULTIMATE LOSS (#2) 21,794,686 24,422,241 30,871,269 39,303,253 42,355,638 39,729,620 ULTIMATE LR 68.0% 77.8% 93.6% 104.1% 101.3% 99.2% ULTIMATE PP 499 483 499 518 494 511 ======================================================================================================================= exhibit 8 ANYSTATE-ANY LINE COVERAGE AS OF FEBRUARY 28, 2002 Semiannual Accident AVERAGE INCURRED LOSSES - ACCIDENT PERIOD ANALYSIS Periods Ending 7 8 9 10 11 12 ------ - - - -- -- -- Aug-95 9,869 9,738 9,394 9,254 9,254 9,231 Feb-96 10,027 9,805 9,972 9,914 9,853 9,844 Aug-96 12,462 12,720 12,948 12,950 12,987 12,561 Feb-97 11,955 11,924 11,877 11,846 11,676 Aug-97 11,468 11,595 11,592 11,492 Feb-98 13,104 13,362 13,369 Aug-98 12,773 12,828 Feb-99 12,735 Aug-99 Feb-00 Aug-00 Feb-01 Aug-01 Feb-02 7-8 8-9 9-10 10-11 11-12 12-13 --- --- ---- ----- ----- ----- Aug-95 0.987 0.965 0.985 1.000 0.997 1.000 Feb-96 0.978 1.017 0.994 0.994 0.999 0.997 Aug-96 1.021 1.018 1.000 1.003 0.967 Feb-97 0.997 0.996 0.997 0.986 Aug-97 1.011 1.000 0.991 Feb-98 1.020 1.000 Aug-98 1.004 Feb-99 Aug-99 Feb-00 Aug-00 Feb-01 Aug-01 Straight Avg 1.003 0.999 0.994 0.996 0.988 0.999 Avg x HiLo 1.004 1.003 0.994 0.997 0.997 Avg Last 4 1.008 1.004 0.996 0.996 Fitted 1.014 1.006 1.000 0.996 1.076 0.781 Special Adj. 0 0 0 0 0 0 Prior Sel. @ 6 Mth (#1) 1.012 1.008 1.000 1.000 1.000 1.000 Prior Sel. @ 3 Mth (#1) 1.005 1.005 1.000 1.000 1.000 1.000 SELECT 1.008 1.000 1.000 1.000 1.000 1.000 Cumulative 1.008 1.000 1.000 1.000 1.000 1.000 ========================================================================================================================== FEB-99 AUG-98 FEB-98 AUG-97 FEB-97 AUG-96 ------ ------ ------ ------ ------ ------ ULT. SEVERITY 12,838 12,828 13,369 11,492 11,676 12,561 ULT. COUNTS 2,413 1,947 1,404 1,304 975 730 ULTIMATE LOSS (#2) 30,978,094 24,976,116 18,770,076 14,985,568 11,384,100 9,169,530 ULTIMATE LR 85.4% 76.3% 64.2% 66.7% 61.2% 64.5% ULTIMATE PP 483 471 400 421 394 404 ========================================================================================================================== exhibit 8 ANYSTATE-ANY LINE COVERAGE AS OF FEBRUARY 28, 2002 Semiannual Accident AVERAGE INCURRED LOSSES - ACCIDENT PERIOD ANALYSIS Periods Ending 13 14 ------ -- -- Aug-95 9,234 9,236 Feb-96 9,814 Aug-96 Feb-97 Aug-97 Feb-98 Aug-98 Feb-99 Aug-99 Feb-00 Aug-00 Feb-01 Aug-01 Feb-02 13-14 ----- Aug-95 1.000 Feb-96 Aug-96 Feb-97 Aug-97 Feb-98 Aug-98 Feb-99 Aug-99 Feb-00 Aug-00 Feb-01 Aug-01 Straight Avg 1.000 Avg x HiLo Avg Last 4 Fitted 0.941 Special Adj. 0 0 Prior Sel. @ 6 Mth (#1) 1.000 Prior Sel. @ 3 Mth (#1) 1.000 SELECT 1.000 TAIL Cumulative 1.000 1.000 ========================================================================================================================== FEB-96 AUG-95 ------ ------ ULT. SEVERITY 9,814 9,236 ULT. COUNTS 410 431 ULTIMATE LOSS (#2) 4,023,740 3,980,716 ULTIMATE LR 52.9% 66.0% ULTIMATE PP 271 309 ========================================================================================================================== #1 If you reviewed the triangle presented in last year's report, you will note that we have added two line items to this report, prior select @ 6 months and prior select @ three months. This data was added to enable our actuaries to be able to easily view factor selections which were made during previous reviews and determine if what they are seeing in the current data is reflective of what they expected to see during the prior review. When differences occur, we will drill down into the data to determine what has changed. #2 The sum of all of the values in the Ultimate Loss Row will determine Ultimate Losses Paid $316,744,647 #3 B/A=C Factor used to develop next periods losses THE DECOMPOSITION REPORT Another important tool used by the Actuaries as well as Pricer's, Product Managers and General Managers is called the Decomposition Report. This report allows us to review reserve activity on both a state level as well as on a consolidated level and not only provides information regarding how reserve changes will effect monthly results, but also reports the effect claims adjusters had on the overall reserve changes verses changes implemented by our actuaries. The report is prepared monthly and provides an inventory status of all claims features and their reserve changes by type of change. A sample of this report can be seen in Exhibit 9 immediately followed by the supporting definitions. EXHIBIT 9 TYPE OF AMOUNT OF NUMBER OF PRIOR CASE CURRENT STATE/GROUP PRODUCT CHANGE CHANGE FEATURES RESERVES RESERVES - ------------- ------- ------------------------------------------------------------------------------------------------------- STATE NAME AUTO NO CHANGE 29 CLOSED (16,396,930) 4,833 OPENED 13,907,161 4,651 AGING ON AVG CASES 1,557,429 2,258 INFLATION 399,181 9,767 AVERAGE TO ADJUSTER 447,882 8 ADJ CHANGE ABOVE CAP 15,000 1 RESEGMENTATION 69 5 ---------- ------ ---------- ---------- (70,209) 21,552 64,720,043 64,649,834 DEFINITIONS: NO CHANGE Amount of Financial Reserves (FinR) is the same as prior month. CLOSED Amount of Financial Reserves (FinR) at prior month end on features closed during the month. OPENED Amount of FinR (average reserves and adjuster estimates) on features opened during the month, based on this month's reserve table. This does not include ANCR on these features. Re-opened features are included here. IMPACT ON NEW FEATURES Change in FinR for opened features (see defn. above) due to Loss Reserving changes in average reserve; this month's reserve table vs. prior month end's reserve table. AGING ON AVG CASE Change in FinR on features open at the prior month end carrying an average reserve that are still open and carrying an average reserve at the current valuation due to changing age groupings. LOSS RESERVE IMPACT Change in FinR on features open at the prior month end carrying an average reserve, that are still open and carrying an average reserve at the current valuation, due to change in average reserves by Loss Reserving. INFLATION Change in FinR on features open at the prior month end carrying an average reserve that are still open and carrying an average reserve at the current valuation, due to inflationary change in average reserves. ADJUSTER TO AVERAGE Change in FinR due to change in carried FinR such that carried FinR was Adj estimate + ANCR at prior month end and is now an average reserve. The change = current average reserve -- prior month's Adj estimate. AVERAGE TO ADJUSTER Change in FinR due to change in carried FinR such that carried FinR was an average reserve at prior month end and is now Adj estimate + ANCR. The change = current Adj estimate -- prior month's average reserve. ADJ CHANGE ABOVE CAP Change in FinR due to Adjuster changing an above-the-cap Adj estimate. Included here are FinR which equal Adj estimate + ANCR for both the prior month end and the current month. The change = Adj estimate (current) -- Adj estimate (prior month end). RESEGMENTATION Change in the FinR due to feature changing region, rate manual group, or line coverage group code. *** Average reserve means average financial reserve set by Loss Reserving. (usually for BI features with Adj estimates less than $75,000). 20 SELECTING THE ULTIMATE RESERVE Selection of the ultimate reserve is driven by the results of the six different indication methods calculated. However, underlying variables must be considered prior to reaching our decision with respect to the ultimate reserve selection for any specific segment. Once the results of each indication method are delivered, the actuary may consider these questions: Q. Is the severity for the three "paid" and the three "incurred" methods trending in the same direction? A. We look to determine if the results of paid methods are trending either up or down in the same direction as the incurred methods. Often however, the results will move in different directions causing the actuaries to look further into closure rates and any process changes our claims organization may have made causing the results to trend apart from each other. Q. Are the number of exposures going up and the ultimate loss indications going down? If so, what other activity may be impacting results? A. At this point, our actuaries would look at frequency and severity separately. If the number of exposures is going up, the number of claims should be going up. If not, we would likely look for a shift in the mix of business to determine if this is causing frequency indications to change. Q. Is severity moving up and frequency moving down? If so, why? A. The preferred auto book of business generally (but not always) carries higher limits than the nonstandard book of business. With higher limits, it makes sense that severity will be greater for the preferred business than for the nonstandard business. On the other hand, frequency in the preferred business should be lower than frequency for the non standard business. When severity moves up and frequency moves down, our actuaries will again look for a shift in business mix. Q. Are today's results consistent with what we expected to happen in prior reviews? If not, what has changed? A. When this happens, we will often see change in the most recent accident semester (the last number on the diagonal of a triangle) because the most recent accident semester is the one most susceptible to ultimate reserve selections. This will tell us whether we were too aggressive or not aggressive enough assigning a loss development factor to the reserve segment the last time reserves were reviewed. Q. Are the ultimate loss ratio results and the ultimate pure premium trending in the same direction? A. If not, we will initially look at our pricing actions to explain the disparity. There are many other questions that we may ask ourselves when we see results vary from what we expect. Indications alone may not be sufficient to drive the determination of an ultimate reserve. 21 SECTION V: PROCESS CHANGES INTRODUCED IN 2001 Most approaches to loss reserving and financial reporting deal with the aggregation of data over a period of time. As Progressive continues to grow, the ability further to refine, segregate and aggregate similar data into even more segments becomes possible. Last year, we reported that we reviewed in excess of 240 segments, this year that number has climbed to approximately 250 segments. This is important as the greater the number of credible breakdowns we have, the better we are able to price, reserve, and understand why the same line coverages with the same limits profiles differ from state to state. During 2001, we made a few changes to enhance further our ability to reserve accurately. - - Uninsured Motorist Bodily Injury (UMBI): During 2001, we found that we finally had credible data in the larger states (NY, FL, OH, CA, etc.) to analyze two segments; 1. reserves for policies with limits less than $50,000/$100,000 and 2. reserves for policies with limits equal to or in excess of $50,000/$100,000. We found that reviewing the limits profile separately delivered different results in both development and severity. - - Tropical Storm Allision: We learned a great deal about catastrophic flood losses as a result of Allision. Our process had been to segment by state (or group of states for the smaller states that lacked enough data to provide credibility) comprehensive exposures and group such exposures as flood, wind, and hail together. After Allision, it became apparent that flood losses cost the Company significantly more than other comprehensive losses. As a result, we now segment, review and establish average reserves for flood losses separately from other comprehensive losses, as flood losses have a higher severity than wind and hail. We now carry four different average reserves for comprehensive losses, i.e. flood, wind and hail, theft, and all other. - - Commercial Vehicle: We improved the accuracy of our carried reserves for our Commercial Vehicle business by increasing the threshold. You will recall, when a reported claim is determined to be below the threshold, actuarially established average reserves are assigned to the claim. By increasing the threshold, a significantly larger portion of the total case reserve is now actuarially decided. In addition, we determined we had enough credibility in our data to breakout our specialty commercial vehicle risks into three separate groups to track and trend our experience better as they have differing development and severity characteristics. 22 SECTION VI: A CASE STUDY We will conclude the commentary on the reserving process here but have included several case studies immediately following. In this section, we have included the exact same case study we published in last years report to provide new readers with experience in calculating Loss Reserves. In the following section of this report, Section VII, we have provided a new case study addressing LAE reserves. The case study immediately following centers around discussing the process we use for reviewing the reserve needs in a given segment. In this case study we will review the following: - - an indication for the total needed reserves using accident period data - - an indication of the needed case reserves using record period data - - an indication of the needed IBNR reserve using record within accident period data. We assembled real data which is used in this case study for your review and reference. This particular example shows countrywide data (excluding a few of the states). Note that all exhibits for both case studies is located at the end of each section. TOTAL NEEDED LOSS RESERVES REVIEW There are several key parameters that we consider prior to selecting an ultimate level of needed loss reserves. We will use exhibit 10 to explain considerations we make before making changes to our reserves. Exhibit 10 has been divided into six sections to make it easier for the reader to understand and refer to. Immediately below you will find a description of each section. SECTION A In this section we show the results of the six different projection methods along with other meaningful data. The ending date of the accident period (we are studying semi annual accident periods in this example) is displayed down the far left (column 1). The six columns immediately following are the results of the six estimation methods previously discussed in this document and are used to project ultimate needed reserves. Columns 8, 9, and 10 show the case incurred (paid + outstanding case reserves) as of the review date, paid losses as of the review date and the indicated ultimate. Below is a description of each of the columns in section A. COLUMN 1: Semiannual accident periods ending: The ending date of the accident period under review. COLUMN 2: Paid Method -- triangular projection of the paid loss amounts to ultimate value. COLUMN 3: Average Paid Method -- average paids are displayed in the triangular format then projected to ultimate by accident period. This ultimate severity is then multiplied by the ultimate number of claims to be paid to estimate ultimate losses. COLUMN 4: B-F Paid Method -- this method determines an expected amount of future paid loss development by accident period (age) then adds this amount to the actual paid-to-date future estimated ultimate loss for that accident period. This method was earlier referred to as modified paid method. COLUMN 5: Incurred Method -triangular projection of the incurred loss amounts to ultimate value. COLUMN 6: Average Incurred method -- average incurreds are displayed in the triangular format then projected to ultimate by accident period. This ultimate severity is then multiplied by the ultimate number of claims incurred to estimate ultimate losses. 23 COLUMN 7: B-F incurred method -- this method determines an expected amount of future incurred loss development by accident period (age) then adds this amount to the actual incurred-to-date future estimated ultimate loss for that accident period. This method was earlier referred to as modified incurred method. COLUMN 8: Case Inc'd Losses as of Feb-2001 -- We show here the incurred (paid plus case reserves) losses by accident period as of the evaluation date; in this case Feb-2001. We take these amounts from the last diagonal of the incurred amount triangle. COLUMN 9: Paid Losses as of Feb-2001 -- Paid loss amounts by accident period as of the evaluation date of Feb-2001 are shown here. We pull these amounts from the last diagonal of the paid amount triangle. COLUMN 10: Indicated Ultimate -- These are selected indicated ultimate loss costs by accident period. EXHIBIT 10, SECTION B displays the different projections for the ultimate number of claims to be paid by accident period along with information regarding claims payment patterns and reporting patterns. There are eight columns shown in this section. Column 1 is the closure rate at the first evaluation as taken from the development triangle. You will recall from our prior discussion that closure rate is number of features closed with payment divided by the ultimate number of features to be paid. We use this information to identify whether or not there is a change in the time it takes to settle claims. COLUMN 2 shows the CWP rate (closed without payment) at the first evaluation point in the development triangle. The CWP rate equals the number of features closed without payment divided by the number of features recorded. We use this statistic to understand or identify any change in the reporting of claims to Progressive. In column 3 we show the indicated ultimate CWP rate. The ultimate CWP rate is used when determining the ultimate claim count projection. COLUMNS 4 THROUGH 7 are the results of the four projection methods we use to determine the ultimate number of features to be paid. They are a) paid count method b) closed with payment count method c) incurred count method and d) ultimate number of recorded counts times one minus the ultimate CWP rate method. The last projection, recorded*(1-CWP rate), is completed by projecting the ultimate number of features to be recorded by accident period then multiplying by the complement of the ultimate CWP rate. COLUMN 8 in section B is the selected ultimate number of claims to be paid by accident period. You will notice that in this example the selected ultimate number of claims to be paid column exactly matches the Recorded *(1-CWP) rate count method. This method is often selected (although not always) since the underlying number recorded data generally displays a consistent development pattern. Also, the CWP rate tends to be reasonably consistent with historical values. EXHIBIT 10, SECTION C shows loss severities by accident period resulting from each of the six different estimation methods described in section A. Two indications, average paid (column 3) and average incurred (column 6), are the ultimate from the triangles of average severity values while the others are the projected ultimate loss amounts divided by the projected ultimate counts by accident period. The lower portion of section C shows the change in the indicated ultimate severities from the prior accident period for each projection. EXHIBIT 10, SECTION D, bottom right, is included to show growth statistics, earned premium (EP) and exposures (defined as earned car years (ECY)) for the segment under review. We show the CALENDAR PERIOD earned premium, earned exposures and the average earned premium (EP/ECY's = AEP). We also show the change in exposures from prior periods as well as the change in the average earned premium. 24 EXHIBIT 10, SECTION E contains our indicated ultimate frequency and severity by accident period. The frequency (f) is the ultimate number of claims to be paid divided by the number of earned exposures (# claims / # ECY's = f). Severity (s) is our selected ultimate loss costs by accident period divided by the ultimate number of claims to be paid by accident period ($ loss / # claims = s). The bottom portion of section E includes the indicated loss ratio (Indicated Ultimate $Loss/EP) and the Indicated Pure Premium (Indicated Ultimate$Loss/ECY's). EXHIBIT 10, SECTION F is the "Retrospective Comparison of Indicated Ultimates". This table compares the selected ultimates made during this review to the selected ultimates we made the last time we reviewed the segment. The purpose for the comparison is to identify how much the current selection differs from the prior selections. We then decide if some or all of the accident period statistics need to be reviewed further in order to understand the changes we are seeing within the data. REVIEWING THE INFORMATION INCLUDED ON THE SUMMARY We review the following data from exhibit 10 to help ensure our indication is reasonable: - - The results of the six loss development methods. An ultimate loss for each accident period is then selected. - - The range of the indications resulting from the different projections in section A. In this case the reserve adequacy ranges from a deficiency of $7.8 million (column 2) to a deficiency of $3.9 million (column 6). The range of the indicated ultimate loss costs is also reviewed for the last accident period. The example shows a range from $227.2 million (column 6) to $229.2 million (column 2). - - The results of the four count development methods. An ultimate count for each accident period is then selected. - - The closure rate over time to determine possible influences internal process changes have had on the paid projections. In section B, column 1, the closure rate for the last accident period is 90.0%. This is consistent with the historical values. - - The CWP rates over time to determine possible influences process or reporting pattern changes may have on the incurred projections (section B, column 2). - - The ultimate frequency (f) indication is also reviewed, section E, column 3, along with the trend (change over time, column 4). Does the change in frequency over time appear to be reasonable? - - The indicated ultimate severity (section E, column 1). Changes in the accident period severity (s) are compared to other studies available and also to the expected annual trend. - - The indicated ultimate severity (s) is compared to the projected value from the avg. paid (section C, column 3) and avg. incurred indications (section C, column 6). If there is a question about the ultimate severity, we will look at the change in severity resulting from each projection to give credence to our selected changes. - - The retrospective comparison of indicated ultimates in section F to determine if we are changing our opinion significantly from our last review and if so, we review the underlying data which produced the change in the indication. In studying the reserve indications, other triangles of several different statistics are reviewed including the following: - - a triangle showing the average adjusters case reserves to see if the adequacy of the case reserves is changing - - triangles showing our paid to date amount divided by our selected ultimate loss - - triangles showing our adjuster incurred to date divided by our selected ultimate losses The three triangles mentioned immediately above are not shown in this report. 25 CASE RESERVE REVIEW The data used in the summary discussed above is sorted by accident period. Indications based on data sorted by record date are also reviewed. Record date indications project ultimate costs for the claims that have been recorded to date (i.e. it only includes reserves for know claims) and is needed to project total needed case reserves. Exhibit 11 shows the parameters that are compared and contrasted. Many are similar to the summary for the accident period review. The primary difference between the accident period and record period review summaries is that exposure and premium data are not included: Frequency, loss ratio or pure premium are not included on the record period summary because it is not appropriate to relate record period losses to the exposure and premium statistics. The sections are labeled according to the information included and are consistent with the prior summary page for the accident period analysis. SECTION A - six projections of ultimate cost by record period SECTION B - projection of the ultimate number of claims to be paid by record period SECTION C - indicated severities by record period for each projection method SECTION D - intentionally left blank as mentioned above SECTION E - ultimate severity by record period, again there is no frequency, loss ratio or pure premium statistics included SECTION F - retrospective comparison of indicated ultimates Our approach for reviewing the record period analysis is similar to the review of the accident period projections. The range of indications is studied. The range of the indicated cost for the last record period shown is also reviewed. Here, the range is from $222.9 million (section A, column 3) to $223.5 million (section A, column 2). A selection is then made and document it in the column entitled "indicated ultimate" (section A, column 10). The results in Section B are also reviewed by studying the range of the indication for the ultimate number of claims to be paid by record period. The CWP rate is reviewed to note if any change in pattern appears. The next step is to look at the ultimate severities (section E) by record period and note any unusual trends. The case incurred loss total (Section A column 8) is compared to total indicated ultimate (Section A column 10). In exhibit 11 compare $1.766 billion to the total indicated ultimates of $1.763 billion. The results of this comparison tell us that the case reserves are expected to develop downward by approximately $3.0 million. Next look at the incurred method projection total of $1.763 billion (exhibit 11, column 5). Historical development suggests that downward development will occur. Finally, review of the retrospective comparison (exhibit 11, section F) of indicated values may point toward a particular record period that may need to be reviewed for some unexpected emerging development. In the example shown here, the indicated required reserve is $31.4 million (Section A, column 10). Carried reserves are $29.8 million indicating a deficiency of $1.6 million, or about 5%. IBNR RESERVE EXHIBIT 12 THROUGH EXHIBIT 17 show the calculations for the IBNR reserve projection. The data for IBNR analyses is organized into record lag periods following the accident date. This format, record within accident period reports, separates all losses from a particular accident quarter into lag groups according to the quarter in which the loss is recorded. 26 In defining a lag group please refer to the attached exhibit 12; the development triangle of recorded claim count for the appropriate accident quarters. The numbers in the first column are considered claims reported in lag-group 0. The reason this is labeled as lag 0 is that these claims were reported in the period they occurred, therefore, no lag exists. The difference in the counts shown in column 2 and column 1 is the number of claims belonging to lag-group 1 because these counts are considered to be recorded one quarter late. The number of claims in the lag group 2 column are considered to be the number of claims reported 2 quarters late. The same logic applies for lag groups 3, 4, 5, 6 etc. Ultimate loss cost are then projected for each lag-group by accident quarter. The development of the loss amounts for each lag-group for a particular accident quarter is the changing valuation of a finite set of claims. FROM EXHIBIT 12, the number of claims recorded in the first column for accident quarter ending May, 2000 is 86,802. In projecting the ultimate cost of these lag-group 0 claims the cost of only these 86,802 claims is tracked. The number shown in column 2 is 90,012 which is greater than the column 1 value by 3,210 which are the total lag-group 1 claims for this accident quarter. The cost of these 3,210 claims are developed to ultimates to determine lag group 1 costs. EXHIBITS 13 AND 14 show the development triangles for lag-group 1 and lag-group 2 incurred losses as examples of this projection. You should note that the triangles are truncated at the tenth evaluation column for display purposes. The calculated link or development factors from the history shown are included. At the bottom of the exhibits different averages of the historical development factors are shown, then the selected incremental loss development factors are shown. The selected factor is the expectation of how the amounts will develop during the next period. Finally, the cumulative development factor (i.e. the product of all preceding incremental factors) is shown which when applied to the figures in the last diagonal of the triangle provides the projected ultimate cost for the applicable accident quarter. EXHIBIT 13, lag group 1, shows at the bottom of the first column averages of the historical loss development factors ranging from 0.964 to 0.973. The selected incremental factor is 0.964. When the selected factor is applied to the cumulative factor in the second column of 0.989 it calculates a cumulative factor of 0.954. By multiplying 0.954 by the figure in the last diagonal located in first column (9,391,582) gives us the projected ultimate cost for accident quarter ending Nov-2000 lag-group 1 losses of $8,959,569 shown at the very bottom of the page as well as on the right hand side of the exhibit. If you followed these steps for lag-group 1 loss projections, you can apply the same steps for lag-group 2 loss development as shown in exhibit 14. By doing this, you will determine that the projected ultimate cost for accident quarter ending Aug.-2000 lag-group 2 losses are $1,880,756. Once the ultimate loss cost by accident quarter and lag-group are projected, we then array the numbers as shown in the top triangle on exhibit 15, "Incurred Losses Quarterly Lag 0-7 Emergence IBNR Analysis". Note the amount shown at the bottom of the column labeled 1 (lag-group 1) is $8,959,569. Lag-group 2 is $1,880,756. These results come directly from Exhibits 13 and 14 previously reviewed. Now that past projected IBNR losses for individual accident periods have been obtained, estimates for future emergence of IBNR losses must be projected. To do this, you must inflate past costs to current dollars because these losses were incurred during a prior period and are not stated in current dollars. Inflated incurred losses are shown in the triangle on the bottom portion of exhibit 15. For accident quarter Aug-2000, lag group 1, the ultimate of $8,396,780 inflated forward one quarter at an annual rate of 4.5% equals $8,489,690. Next, calculate pure premiums by dividing the losses by earned car years (ECY's) to relate the losses from each accident period on a consistent basis. These pure premiums are as shown on exhibit 16. Once the pure premium is calculated, an average pure premium is selected for each lag-group. 27 In EXHIBIT 16 the selected pure premium for each lag-group in current dollars are inflated to the future points in time when they are expected to emerge. For the most recent accident quarter ending Feb-2001, we expect $11.379 ($11.255 inflated by the annual rate of 4.5%: = 11.255 * (1.045) to the power of 1/4th) to emerge one quarter late for every earned car year in the quarter ending Feb-2001. Subsequent IBNR loss amounts per earned car year, inflated into the future, are displayed. Note that no IBNR pure premium is required for lag-period one for the quarter ending Nov-2000, as these losses have already been recorded at Feb-2001. The same reasoning applies to the other blank cells shown in the exhibit. The far right column in exhibit 16 is the sum of all inflated expected future pure premiums for late reported claims by accident period. Only the first seven quarters are displayed, but this calculation is carried out up to six years of lag. In the following EXHIBIT 17, the total future pure premiums (PP) for expected IBNR are converted to earned premium factors. The product of PP and earned exposures equals the required IBNR. By dividing the required IBNR by the earned premium it will provide the required IBNR factors (column 5). Next, convert to earned premium factors for the calculation of INBR going forward to accommodate any short-term shifts in product mix that would not otherwise appear in exposure data. Required IBNR factors are then compared to the current factors to assist in determining a new set of selected factors to establish our IBNR reserve. IBNR loss emergence during the last 6 months (since Aug-2000) of $18,130,936, are shown by accident quarter in column 7. These late reported losses plus future expected emergence for accidents prior to Aug-2000, equals $30,236,912. This amount is labeled "Six Mth Runoff Estimate #1". The future expected emergence for accidents prior to and including Aug-2000 equals the total expected emergence minus the expected emergence for accident quarters ending Nov-2000 and Feb-2001, ($34,584,463 - $6,485,294 - $15,993,193 = $12,105,976). In other words, the IBNR estimate of $30,236,912 = $18,130,936 + 12,105,976. This IBNR reserve amount is compared to the carried IBNR reserve six months prior. COLUMNS 8 AND 9 IN EXHIBIT 17 develop estimated required IBNR factors using a similar approach as the one just discussed (i.e. emerged since Aug-2000) only we take an average of the amounts emerging during the last three six month periods. RECONCILIATION OF THE ULTIMATE VALUE EXHIBIT 17 develops an indicated needed IBNR amount of $34,584,463. The record period analysis generated an indicated total needed CASE RESERVES of $31,425,000 which, together indicates total needed reserves of $66.0 million. The result of the accident period review was a total needed reserve amount of $65.1 million. These two approaches are reasonably close, within 1.5% of each other. In the example above, with an difference of 1.5%, we will generally accept a level of carried reserves between the two amounts ($65.1 million and $66.0 million). However, if the difference is greater than 5%, we will typically do additional analyses to explain differences between the accident period indication and the sum of the record and IBNR indication. Part of any discrepancy may be due to open claims included in the record period analysis that are not, by definition, included in the accident period analysis. By increasing the number of years included in the analysis we can reduce this bias. Another reason may be a changing frequency of late reported claims. We will conclude the loss reserve case study here. In the next section, Section VII, we will review Loss Adjustment Expenses. 28 EXHIBIT 10 COUNTRYWIDE X AUTO AS OF FEBRUARY 28, 2001 ACCIDENT PERIOD ANALYSIS - ZERO RUNOFF ==================================================================================================================================== SECTION A COLUMN # (1) (2) (3) (4) (5) Berquist- Least Least Semiannual Avg B-F Sherman Squares Squares Accident Paid Paid Paid Paid Paid PP Paid LR Incurred Periods Method Method Method Method Method Method Method Ending Ult ($000) Ult ($000) Ult ($000) Ult ($000) Ult ($000) Ult ($000) Ult ($000) ------ ---------- ---------- ---------- ---------- ---------- ---------- ---------- Aug-94 50,160 50,155 50,160 50,160 50,160 50,160 50,160 Feb-95 66,227 66,228 66,227 66,227 66,227 66,227 66,227 Aug-95 76,527 76,517 76,527 76,527 76,528 76,525 76,527 Feb-96 84,899 84,891 84,899 84,881 84,895 84,926 84,928 Aug-96 83,102 83,100 83,102 83,047 83,099 83,094 83,095 Feb-97 93,967 93,954 93,967 93,939 93,958 93,980 93,973 Aug-97 108,819 108,796 108,819 108,753 108,808 108,817 108,814 Feb-98 126,612 126,594 126,612 126,479 126,599 126,560 126,553 Aug-98 134,107 133,909 134,106 133,834 133,884 133,809 133,826 Feb-99 149,979 149,734 149,979 149,557 149,736 149,660 149,685 Aug-99 173,364 173,129 173,363 172,622 173,113 172,856 172,944 Feb-00 202,912 202,621 202,910 201,982 202,579 202,166 202,412 Aug-00 214,407 214,599 214,412 213,467 213,376 212,598 213,924 Feb-01 229,231 227,244 229,015 223,492 223,213 223,494 228,393 Total 1,794,313 1,791,472 1,794,097 1,784,968 1,786,174 1,784,871 1,791,462 Paid Loss 1,727,840 1,727,840 1,727,840 1,727,840 1,727,840 1,727,840 1,727,840 Required Reserve 66,473 63,631 66,257 57,127 58,334 57,031 63,622 Held Reserve 58,616 58,616 58,616 58,616 58,616 58,616 58,616 Reserve Adequacy (7,857) (5,015) (7,641) 1,489 283 1,585 (5,006) ==================================================================================================================================== SECTION A COLUMN # (1) (6) (7) Berquist- Least Least Semiannual Avg B-F Sherman Squares Squares Accident Incurred Incurred Incurred Inc'd PP Inc'd LR Periods Method Method Method Method Method Ending Ult ($000) Ult ($000) Ult ($000) Ult ($000) Ult ($000) ------ ---------- ---------- ---------- ---------- ---------- Aug-94 50,155 50,160 50,160 50,160 50,160 Feb-95 66,228 66,227 66,227 66,227 66,227 Aug-95 76,517 76,527 76,527 76,525 76,525 Feb-96 84,937 84,928 84,928 84,926 84,926 Aug-96 83,100 83,095 83,101 83,094 83,094 Feb-97 93,954 93,973 93,991 93,980 93,976 Aug-97 108,796 108,814 108,906 108,817 108,809 Feb-98 126,529 126,553 126,704 126,560 126,555 Aug-98 133,838 133,826 134,081 133,809 133,807 Feb-99 149,659 149,685 149,894 149,660 149,658 Aug-99 172,871 172,945 173,152 172,856 172,797 Feb-00 202,427 202,416 202,532 202,166 201,982 Aug-00 214,095 213,940 213,930 212,598 212,550 Feb-01 227,244 228,379 227,109 223,494 223,466 Total 1,790,350 1,791,468 1,791,242 1,784,871 1,784,530 Paid Loss 1,727,840 1,727,840 1,727,840 1,727,840 1,727,840 Required Reserve 62,510 63,628 63,401 57,031 56,690 Held Reserve 58,616 58,616 58,616 58,616 58,616 Reserve Adequacy (3,894) (5,012) (4,785) 1,585 1,926 ==================================================================================================================================== SECTION A COLUMN # (1) (8) (9) (10) Semiannual Case Inc'd Paid Accident Loss as of Loss as of Periods 02/28/2001 02/28/2001 Indicated Ending ($000) ($000) Ultimate ------ ------ ------ -------- Aug-94 50,160 50,160 50,160 Feb-95 66,227 66,227 66,228 Aug-95 76,527 76,527 76,527 Feb-96 84,928 84,881 84,928 Aug-96 83,101 83,077 83,101 Feb-97 93,991 93,934 93,991 Aug-97 108,834 108,762 108,834 Feb-98 126,583 126,481 126,583 Aug-98 133,788 133,582 133,935 Feb-99 149,555 149,165 149,787 Aug-99 172,381 171,756 173,285 Feb-00 200,448 199,136 202,814 Aug-00 208,102 204,735 214,473 Feb-01 207,642 179,418 228,251 Total 1,762,267 1,727,840 1,792,897 Paid Loss 1,727,840 Required Reserve 65,057 Held Reserve 58,616 Reserve Adequacy (6,441) ================================================================================================================================ SECTION B (1) (2) (3) (4) (5) (6) (7) (8) Recorded Closed *(1-CWP INDICATED Paid w/Pay Incurred Rate) INDICATED Closure Rate CWP Rate ULTIMATE Counts Counts Counts Counts ULTIMATE @6 months @6 months CWP RATE Method Method Method Method COUNTS --------- --------- -------- ------ ------ ------ ------ ------ 88.2% 26.0% 28.5% 29,468 29,468 29,468 29,468 29,468 89.5% 27.3% 29.8% 37,976 37,976 37,976 37,975 37,975 89.4% 26.8% 29.3% 42,772 42,772 42,772 42,771 42,771 89.8% 26.9% 29.5% 46,538 46,538 46,541 46,541 46,541 87.8% 27.5% 30.3% 44,277 44,277 44,276 44,273 44,273 88.1% 29.1% 32.0% 49,714 49,714 49,712 49,711 49,711 87.6% 30.5% 33.2% 56,847 56,848 56,844 56,842 56,842 90.9% 31.9% 33.9% 65,536 65,537 65,530 65,525 65,525 90.6% 31.3% 33.3% 70,639 70,638 70,635 70,627 70,627 90.1% 33.4% 35.5% 75,502 75,504 75,487 75,471 75,471 89.6% 37.2% 39.6% 86,188 86,190 86,161 86,134 86,134 90.8% 42.3% 44.1% 96,637 96,643 96,577 96,578 96,578 89.5% 43.2% 45.2% 100,982 100,927 100,814 100,893 100,893 90.0% 43.4% 45.2% 105,963 105,596 105,722 104,914 104,914 909,039 908,628 908,515 907,723 907,723 ================================================================================================================================ SECTION C x x x COLUMN # (1) (2) (3) (4) (5) Semiannual Berquist- Least Least Accident Avg B-F Sherman Squares Squares Periods Paid Paid Paid Paid Paid PP Paid LR Incurred Ending Severity Severity Severity Severity Severity Severity Severity ------ -------- -------- -------- -------- -------- -------- -------- Aug-94 1,702 1,702 1,702 1,702 1,702 1,702 1,702 Feb-95 1,744 1,744 1,744 1,744 1,744 1,744 1,744 Aug-95 1,789 1,789 1,789 1,789 1,789 1,789 1,789 Feb-96 1,824 1,824 1,824 1,824 1,824 1,825 1,825 Aug-96 1,877 1,877 1,877 1,876 1,877 1,877 1,877 Feb-97 1,890 1,890 1,890 1,890 1,890 1,891 1,890 Aug-97 1,914 1,914 1,914 1,913 1,914 1,914 1,914 Feb-98 1,932 1,932 1,932 1,930 1,932 1,931 1,931 Aug-98 1,899 1,896 1,899 1,895 1,896 1,895 1,895 Feb-99 1,987 1,984 1,987 1,982 1,984 1,983 1,983 Aug-99 2,013 2,010 2,013 2,004 2,010 2,007 2,008 Feb-00 2,101 2,098 2,101 2,091 2,098 2,093 2,096 Aug-00 2,125 2,127 2,125 2,116 2,115 2,107 2,120 Feb-01 2,185 2,166 2,183 2,130 2,128 2,130 2,177 Total SECTION C x x x COLUMN # (1) (6) (7) Semiannual Berquist- Least Least Accident Avg B-F Sherman Squares Squares Periods Incurred Incurred Incurred Inc'd PP Inc'd LR Ending Severity Severity Severity Severity Severity ------ -------- -------- -------- -------- -------- Aug-94 1,702 1,702 1,702 1,702 1,702 Feb-95 1,744 1,744 1,744 1,744 1,744 Aug-95 1,789 1,789 1,789 1,789 1,789 Feb-96 1,825 1,825 1,825 1,825 1,825 Aug-96 1,877 1,877 1,877 1,877 1,877 Feb-97 1,890 1,890 1,891 1,891 1,890 Aug-97 1,914 1,914 1,916 1,914 1,914 Feb-98 1,931 1,931 1,934 1,931 1,931 Aug-98 1,895 1,895 1,898 1,895 1,895 Feb-99 1,983 1,983 1,986 1,983 1,983 Aug-99 2,007 2,008 2,010 2,007 2,006 Feb-00 2,096 2,096 2,097 2,093 2,091 Aug-00 2,122 2,120 2,120 2,107 2,107 Feb-01 2,166 2,177 2,165 2,130 2,130 Total Change in Change in Change in Semiannual Change in Change in Berquist- Least Least Accident Change in Avg B-F Sherman Squares Squares Change in Periods Paid Paid Paid Paid Paid PP Paid LR Incurred Ending Severity Severity Severity Severity Severity Severity Severity ------ -------- -------- -------- -------- -------- -------- -------- Aug-94 NA NA NA NA NA NA NA Feb-95 2.5% 2.5% 2.5% 2.5% 2.5% 2.5% 2.5% Aug-95 2.6% 2.6% 2.6% 2.6% 2.6% 2.6% 2.6% Feb-96 2.0% 2.0% 2.0% 1.9% 1.9% 2.0% 2.0% Aug-96 2.9% 2.9% 2.9% 2.9% 2.9% 2.9% 2.9% Feb-97 0.7% 0.7% 0.7% 0.7% 0.7% 0.7% 0.7% Aug-97 1.3% 1.3% 1.3% 1.2% 1.3% 1.3% 1.3% Feb-98 0.9% 0.9% 0.9% 0.9% 0.9% 0.9% 0.9% Aug-98 -1.7% -1.9% -1.7% -1.8% -1.9% -1.9% -1.9% Feb-99 4.7% 4.6% 4.7% 4.6% 4.7% 4.7% 4.7% Aug-99 1.3% 1.3% 1.3% 1.1% 1.3% 1.2% 1.2% Feb-00 4.4% 4.4% 4.4% 4.4% 4.4% 4.3% 4.4% Aug-00 1.1% 1.4% 1.1% 1.2% 0.8% 0.7% 1.2% Feb-01 2.8% 1.8% 2.7% 0.7% 0.6% 1.1% 2.7% Total =================================================================================================================================== Change in Change in Change in Semiannual Change in Change in Berquist- Least Least Accident Avg B-F Sherman Squares Squares Periods Incurred Incurred Incurred Inc'd PP Inc'd LR Ending Severity Severity Severity Severity Severity ------ -------- -------- -------- -------- -------- Aug-94 NA NA NA NA NA Feb-95 2.5% 2.5% 2.5% 2.5% 2.5% Aug-95 2.6% 2.6% 2.6% 2.6% 2.6% Feb-96 2.0% 2.0% 2.0% 2.0% 2.0% Aug-96 2.8% 2.9% 2.9% 2.9% 2.9% Feb-97 0.7% 0.7% 0.7% 0.7% 0.7% Aug-97 1.3% 1.3% 1.3% 1.3% 1.3% Feb-98 0.9% 0.9% 0.9% 0.9% 0.9% Aug-98 -1.9% -1.9% -1.8% -1.9% -1.9% Feb-99 4.6% 4.7% 4.6% 4.7% 4.7% Aug-99 1.2% 1.2% 1.2% 1.2% 1.2% Feb-00 4.4% 4.4% 4.3% 4.3% 4.2% Aug-00 1.2% 1.2% 1.1% 0.7% 0.7% Feb-01 2.1% 2.7% 2.1% 1.1% 1.1% Total ==================================================================================================================================== =================================================================================================================================== SECTION D EARNED Change in Change Premium Earned Earned In ($000) Exposures Exposures Avg EP Avg EP ------ --------- --------- ------ ------ 89,251 422,904 211 98,904 508,182 20.2% 195 -7.8% 109,715 594,217 16.9% 185 -5.1% 111,725 601,429 1.2% 186 0.6% 119,401 624,412 3.8% 191 2.9% 131,031 671,355 7.5% 195 2.1% 156,037 790,363 17.7% 197 1.2% 183,355 922,102 16.7% 199 0.7% 204,961 1,022,919 10.9% 200 0.8% 217,902 1,082,352 5.8% 201 0.5% 246,091 1,229,677 13.6% 200 -0.6% 274,950 1,403,549 14.1% 196 -2.1% 303,192 1,570,679 11.9% 193 -1.5% 325,232 1,679,639 6.9% 194 0.3% 2,571,746 13,123,779 ================================================================================================================================ SECTION E (1) (2) (3) (4) INDICATED CHANGE INDICATED Change ULTIMATE IN ULTIMATE In SEVERITY SEVERITY FREQUENCY Frequency -------- -------- --------- --------- 1,702 6.97% 1,744 2.5% 7.47% 7.2% 1,789 2.6% 7.20% -3.7% 1,825 2.0% 7.74% 7.5% 1,877 2.9% 7.09% -8.4% 1,891 0.7% 7.40% 4.4% 1,915 1.3% 7.19% -2.9% 1,932 0.9% 7.11% -1.2% 1,896 -1.8% 6.90% -2.8% 1,985 4.7% 6.97% 1.0% 2,012 1.4% 7.00% 0.5% 2,100 4.4% 6.88% -1.8% 2,126 1.2% 6.42% -6.6% 2,176 2.3% 6.25% -2.8% Indicated Indicated Loss Pure Ratio Premium ----- ------- 56.2% 119 67.0% 130 69.8% 129 76.0% 141 69.6% 133 71.7% 140 69.7% 138 69.0% 137 65.3% 131 68.7% 138 70.4% 141 73.8% 145 70.7% 137 70.2% 136 69.7% 137 =============================================================================== =================================================================================================================================== SECTION F RETROSPECTIVE COMPARISON OF INDICATED ULTIMATES Semiannual Ultimate Semiannual Est Ult Ultimate Accident Loss at Accident Loss at Loss at Periods 11/30/1999 Periods 11/30/1999 02/28/2001 Change Ending ($000) Ending ($000) ($000) ($000) ------ ------ ------ ------ ------ ------ Apr-94 40,318 Aug-94 50,426 50,160 (266) Oct-94 55,479 Feb-95 66,328 66,228 (101) Apr-95 71,753 Aug-95 76,902 76,527 (375) Oct-95 79,477 Feb-96 82,871 84,928 2,057 Apr-96 84,568 Aug-96 84,073 83,101 (973) Oct-96 83,826 Feb-97 95,763 93,991 (1,772) Apr-97 101,731 Aug-97 108,771 108,834 63 Oct-97 112,291 Feb-98 120,002 126,583 6,581 Apr-98 123,857 Aug-98 140,340 133,935 (6,405) Oct-98 148,582 Feb-99 148,019 149,787 1,768 Apr-99 147,738 Aug-99 173,687 173,285 (401) Oct-99 186,662 Feb-00 200,606 202,814 2,208 Apr-00 207,579 Aug-00 213,901 214,473 572 Oct-00 217,061 Feb-01 NA 228,251 NA Total 13 sem NA 1,561,690 1,564,647 2,957 Total 1,660,922 NA 1,792,897 NA ================================================================================================================================ COUNTRYWIDE X AUTO AS OF FEBRUARY 28, 2001 EXHIBIT 11 RECORD PERIOD ANALYSIS - ZERO RUNOFF ================================================================================================================================== SECTION A COLUMN # (1) (2) (3) (4) (5) (6) Semiannual Avg B-F Avg Accident Paid Paid Paid Incurred Incurred Periods Method Method Method Method Method Ending Ult ($000) Ult ($000) Ult ($000) Ult ($000) Ult ($000) ------ ---------- ---------- ---------- ---------- --------- Aug-94 49,101 49,101 49,101 49,101 49,101 Feb-95 64,709 64,709 64,709 64,709 64,709 Aug-95 74,983 74,981 74,983 74,983 74,981 Feb-96 83,804 83,808 83,804 83,834 83,834 Aug-96 81,391 81,392 81,391 81,391 81,391 Feb-97 92,678 92,671 92,678 92,676 92,671 Aug-97 105,495 105,495 105,495 105,521 105,518 Feb-98 126,978 126,973 126,978 127,014 127,009 Aug-98 131,328 131,306 131,328 131,273 131,260 Feb-99 148,597 148,553 148,597 148,491 148,471 Aug-99 169,402 169,321 169,402 169,255 169,210 Feb-00 200,929 200,854 200,929 200,726 200,711 Aug-00 210,924 210,838 210,924 210,480 210,606 Feb-01 223,481 222,934 223,447 223,213 223,192 Total 1,763,799 1,762,936 1,763,765 1,762,667 1,762,663 Paid Loss 1,731,673 1,731,673 1,731,673 1,731,673 1,731,673 Required Reserve 32,126 31,263 32,092 30,994 30,989 Held Reserve 29,833 29,833 29,833 29,833 29,833 Reserve Adequacy (2,293) (1,430) (2,258) (1,161) (1,156) ======================================================================================== COLUMN # (7) (8) (9) (10) B-F Case Inc'd Paid Incurred Loss as of Loss as of Method 02/28/2001 02/28/2001 Indicated Ult ($000) ($000) ($000) Ultimate ---------- ------ ------ 49,101 49,101 49,101 49,101 64,709 64,709 64,709 64,709 74,983 74,983 74,983 74,983 83,834 83,834 83,804 83,834 81,391 81,392 81,390 81,391 92,676 92,678 92,676 92,676 105,521 105,529 105,492 105,507 127,014 127,034 126,970 126,994 131,273 131,284 131,190 131,295 148,491 148,518 148,400 148,533 169,255 169,353 169,073 169,307 200,726 200,891 200,026 200,812 210,480 210,490 207,637 210,709 223,213 226,302 196,223 223,247 1,762,667 1,766,100 1,731,673 1,763,099 1,731,673 1,731,673 Required Reserve 30,994 31,425 Held Reserve 29,833 29,833 Reserve Adequacy (1,160) (1,592) ======================================================================================================================== SECTION B (1) (2) (3) (4) (5) (6) (7) Recorded Closed *(1-CWP INDICATED Paid w/Pay Incurred Rate) Indicated CWP Rate ULTIMATE Counts Counts Counts Counts Ultimate @6 months CWP RATE Method Method Method Method Counts --------- -------- ------ ------ ------ ------ 26.2% 28.6% 29,221 29,221 29,221 29,221 29,221 27.4% 29.7% 37,702 37,702 37,702 37,702 37,702 27.0% 29.2% 42,496 42,496 42,496 42,495 42,495 27.1% 29.3% 46,361 46,361 46,363 46,363 46,363 27.8% 30.4% 43,814 43,814 43,815 43,815 43,815 29.2% 31.9% 49,643 49,644 49,642 49,639 49,639 30.6% 33.2% 56,179 56,180 56,181 56,179 56,179 32.0% 33.9% 66,459 66,460 66,459 66,457 66,457 31.3% 33.3% 70,444 70,443 70,439 70,432 70,432 33.4% 35.4% 75,219 75,220 75,208 75,197 75,197 37.2% 39.5% 85,782 85,784 85,764 85,741 85,741 42.3% 44.0% 96,434 96,440 96,405 96,398 96,398 43.2% 45.2% 100,498 100,502 100,397 100,457 100,457 43.3% 45.2% 105,045 104,838 104,669 104,659 104,659 905,297 905,105 904,761 904,755 904,755 ======================================================================================================================== ============================================================================================================================== SECTION C COLUMN # (1) (2) (3) (4) (5) (6) (7) Semiannual Accident Avg B-F Avg B-F Periods Paid Paid Paid Incurred Incurred Incurred Ending Severity Severity Severity Severity Severity Severity ------ -------- -------- -------- -------- -------- Aug-94 1,680 1,680 1,680 1,680 1,680 1,680 Feb-95 1,716 1,716 1,716 1,716 1,716 1,716 Aug-95 1,765 1,764 1,765 1,765 1,764 1,765 Feb-96 1,808 1,808 1,808 1,808 1,808 1,808 Aug-96 1,858 1,858 1,858 1,858 1,858 1,858 Feb-97 1,867 1,867 1,867 1,867 1,867 1,867 Aug-97 1,878 1,878 1,878 1,878 1,878 1,878 Feb-98 1,911 1,911 1,911 1,911 1,911 1,911 Aug-98 1,865 1,864 1,865 1,864 1,864 1,864 Feb-99 1,976 1,976 1,976 1,975 1,974 1,975 Aug-99 1,976 1,975 1,976 1,974 1,974 1,974 Feb-00 2,084 2,084 2,084 2,082 2,082 2,082 Aug-00 2,100 2,099 2,100 2,095 2,096 2,095 Feb-01 2,135 2,130 2,135 2,133 2,133 2,133 Total Semiannual Change in Change in Change in Change in Accident Change in Avg B-F Change in Avg B-F Periods Paid Paid Paid Incurred Incurred Incurred Ending Severity Severity Severity Severity Severity Severity ------ -------- -------- -------- -------- -------- Aug-94 NA NA NA NA NA NA Feb-95 2.1% 2.1% 2.1% 2.1% 2.1% 2.1% Aug-95 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% Feb-96 2.4% 2.4% 2.4% 2.5% 2.5% 2.5% Aug-96 2.8% 2.8% 2.8% 2.7% 2.7% 2.7% Feb-97 0.5% 0.5% 0.5% 0.5% 0.5% 0.5% Aug-97 0.6% 0.6% 0.6% 0.6% 0.6% 0.6% Feb-98 1.7% 1.7% 1.7% 1.8% 1.8% 1.8% Aug-98 -2.4% -2.4% -2.4% -2.5% -2.5% -2.5% Feb-99 6.0% 6.0% 6.0% 5.9% 5.9% 5.9% Aug-99 -0.0% -0.0% -0.0% -0.0% -0.0% -0.0% Feb-00 5.5% 5.5% 5.5% 5.5% 5.5% 5.5% Aug-00 0.7% 0.7% 0.7% 0.6% 0.7% 0.6% Feb-01 1.7% 1.5% 1.7% 1.8% 1.7% 1.8% Total ============================================================================================================================== ================================================================================ SECTION D SECTION D INTENTIONALLY LEFT BLANK ================================================================================ ================================================================================ SECTION E (1) (2) INDICATED CHANGE ULTIMATE IN SEVERITY SEVERITY -------- -------- 1,680 1,716 2.1% 1,765 2.8% 1,808 2.5% 1,858 2.7% 1,867 0.5% 1,878 0.6% 1,911 1.8% 1,864 -2.4% 1,975 6.0% 1,975 -0.0% 2,083 5.5% 2,097 0.7% 2,133 1.7% ================================================================================ ============================================================================================================== SECTION F RETROSPECTIVE COMPARISON OF INDICATED ULTIMATES Semiannual Ultimate Semiannual Ultimate Ultimate Accident Loss at Accident Loss at Loss at Periods 11/30/1999 Periods 11/30/1999 02/28/2001 Change Ending ($000) Ending ($000) ($000) ($000) ------ ------ ------ ------ ------ May-93 40,318 Aug-94 50,426 49,101 (1,325) Nov-93 55,479 Feb-95 66,328 64,709 (1,619) May-94 71,753 Aug-95 76,902 74,983 (1,919) Nov-94 79,477 Feb-96 80,616 83,834 3,218 May-95 81,186 Aug-96 80,711 81,391 680 Nov-95 80,473 Feb-97 91,932 92,676 744 May-96 97,662 Aug-97 104,420 105,507 1,087 Nov-96 107,800 Feb-98 115,202 126,994 11,792 May-97 118,903 Aug-98 138,689 131,295 (7,394) Nov-97 148,582 Feb-99 148,019 148,533 514 May-98 147,738 Aug-99 173,687 169,307 (4,380) Nov-98 186,662 Feb-00 200,606 200,812 206 May-99 207,579 Aug-00 206,665 210,709 4,043 Nov-99 206,208 Feb-01 NA 223,247 NA Total 13 sem 1,629,818 1,534,203 1,763,099 5,648 ============================================================================================================== 29 EXHBIT 12 COUNTRYWIDE X AUTO AS OF FEBRUARY 28, 2001 Semiannual Accident RECORDED FEATURE COUNT - ACCIDENT PERIOD ANALYSIS Periods Ending 1 2 3 4 5 6 7 ------ - - - - - - - May-96 28,339 30,243 30,464 30,555 30,629 30,665 30,693 Aug-96 29,928 32,283 32,520 32,637 32,699 32,736 32,759 Nov-96 32,117 35,074 35,423 35,555 35,633 35,669 35,688 Feb-97 34,419 36,722 37,038 37,169 37,237 37,271 37,291 May-97 37,378 40,364 40,649 40,801 40,878 40,933 40,968 Aug-97 40,113 43,398 43,722 43,871 43,958 43,993 44,023 Nov-97 43,120 47,048 47,400 47,551 47,619 47,667 47,698 Feb-98 48,987 50,621 50,941 51,075 51,147 51,193 51,225 May-98 49,835 51,629 51,958 52,121 52,227 52,273 52,298 Aug-98 50,931 52,740 53,083 53,254 53,350 53,397 53,439 Nov-98 53,942 55,915 56,320 56,493 56,606 56,651 56,687 Feb-99 57,242 59,305 59,688 59,881 59,974 60,045 60,092 May-99 62,839 65,398 65,813 66,008 66,127 66,196 66,239 Aug-99 72,684 75,257 75,717 75,947 76,074 76,132 76,162 Nov-99 79,905 82,812 83,292 83,500 83,618 83,683 Feb-00 84,910 87,708 88,155 88,370 88,488 May-00 86,802 90,012 90,496 90,683 Aug-00 88,351 91,655 92,117 Nov-00 88,763 92,402 Feb-01 93,218 Semiannual Accident RECORDED FEATURE COUNT - ACCIDENT PERIOD ANALYSIS Periods Ending 8 9 10 ------ - - -- May-96 30,711 30,725 30,731 Aug-96 32,773 32,780 32,784 Nov-96 35,711 35,729 35,737 Feb-97 37,302 37,312 37,319 May-97 40,991 41,004 41,016 Aug-97 44,042 44,059 44,070 Nov-97 47,717 47,732 47,744 Feb-98 51,241 51,256 51,268 May-98 52,316 52,338 52,352 Aug-98 53,455 53,474 53,481 Nov-98 56,713 56,728 56,739 Feb-99 60,125 60,146 May-99 66,267 Aug-99 Nov-99 Feb-00 May-00 Aug-00 Nov-00 Feb-01 Semiannual RECORDED INCREMENTAL FEATURE COUNT - ACCIDENT PERIOD ANALYSIS Accident Periods LAG GRP 0 LAG GRP 1 LAG GRP 2 LAG GRP 3 LAG GRP 4 LAG GRP 5 LAG GRP 6 Ending 1 2 3 4 5 6 7 ------ - - - - - - - May-96 28,339 1,904 221 91 74 36 28 Aug-96 29,928 2,355 237 117 62 37 23 Nov-96 32,117 2,957 349 132 78 36 19 Feb-97 34,419 2,303 316 131 68 34 20 May-97 37,378 2,986 285 152 77 55 35 Aug-97 40,113 3,285 324 149 87 35 30 Nov-97 43,120 3,928 352 151 68 48 31 Feb-98 48,987 1,634 320 134 72 46 32 May-98 49,835 1,794 329 163 106 46 25 Aug-98 50,931 1,809 343 171 96 47 42 Nov-98 53,942 1,973 405 173 113 45 36 Feb-99 57,242 2,063 383 193 93 71 47 May-99 62,839 2,559 415 195 119 69 43 Aug-99 72,684 2,573 460 230 127 58 30 Nov-99 79,905 2,907 480 208 118 65 Feb-00 84,910 2,798 447 215 118 May-00 86,802 3,210 484 187 Aug-00 88,351 3,304 462 Nov-00 88,763 3,639 Feb-01 93,218 Semiannual RECORDED INCREMENTAL FEATURE COUNT - ACCIDENT PERIOD ANALYSIS Accident Periods LAG GRP 7 LAG GRP 8 LAG GRP 9 Ending 8 9 10 ------ - - -- May-96 18 14 6 Aug-96 14 7 4 Nov-96 23 18 8 Feb-97 11 10 7 May-97 23 13 12 Aug-97 19 17 11 Nov-97 19 15 12 Feb-98 16 15 12 May-98 18 22 14 Aug-98 16 19 7 Nov-98 26 15 11 Feb-99 33 21 May-99 28 Aug-99 Nov-99 Feb-00 May-00 Aug-00 Nov-00 Feb-01 31 EXHIBIT 13 COUNTRYWIDE X AUTO AS OF FEBRUARY 28, 2001 Quarterly Record INCURRED LOSSES QUARTERLY LAG 1 - IBNR ANALYSIS Periods Ending 1 2 3 4 5 6 7 ------ - - - - - - - May-96 3,565,510 3,501,278 3,470,132 3,454,204 3,416,983 3,415,004 3,424,862 Aug-96 4,409,344 4,286,490 4,239,093 4,189,919 4,185,440 4,184,583 4,182,448 Nov-96 5,493,238 5,241,111 5,199,503 5,190,654 5,191,756 5,186,402 5,186,252 Feb-97 4,506,067 4,279,284 4,239,422 4,255,017 4,234,543 4,232,148 4,230,220 May-97 5,559,617 5,308,146 5,286,386 5,286,034 5,279,481 5,275,988 5,249,070 Aug-97 6,046,692 5,807,174 5,785,475 5,788,650 5,783,227 5,778,288 5,783,675 Nov-97 7,375,863 7,254,442 7,238,616 7,206,675 7,189,864 7,190,336 7,200,237 Feb-98 4,524,973 4,382,121 4,351,051 4,341,804 4,345,496 4,337,016 4,335,225 May-98 4,792,251 4,650,335 4,621,616 4,619,637 4,606,569 4,605,889 4,602,679 Aug-98 5,319,925 5,171,438 5,153,475 5,152,848 5,132,983 5,125,691 5,136,179 Nov-98 5,717,023 5,541,913 5,507,431 5,485,960 5,487,093 5,481,759 5,479,126 Feb-99 5,289,611 5,164,677 5,158,963 5,157,648 5,154,384 5,149,230 5,156,941 May-99 6,136,849 5,952,047 5,944,489 5,923,950 5,933,121 5,921,047 5,927,235 Aug-99 7,526,664 7,260,942 7,254,458 7,257,132 7,242,577 7,228,749 Nov-99 8,659,826 8,526,548 8,480,201 8,489,165 8,476,791 Feb-00 7,273,274 7,052,120 7,076,164 7,081,966 May-00 8,236,681 7,919,480 7,927,062 Aug-00 8,703,268 8,489,973 Nov-00 9,391,582 Quarterly Record INCURRED LOSSES QUARTERLY LAG 1 - IBNR ANALYSIS Periods Ending 8 9 10 ULTIMATE ------ - - -- -------- May-96 3,442,903 3,439,493 3,437,569 3,437,569 Aug-96 4,179,948 4,185,870 4,185,870 4,184,110 Nov-96 5,186,152 5,186,127 5,186,127 5,182,672 Feb-97 4,230,324 4,227,966 4,225,943 4,222,662 May-97 5,253,604 5,243,604 5,243,604 5,242,462 Aug-97 5,785,740 5,785,849 5,778,530 5,776,046 Nov-97 7,182,011 7,181,496 7,178,855 7,178,631 Feb-98 4,340,802 4,340,802 4,340,802 4,340,389 May-98 4,602,702 4,599,227 4,597,525 4,596,326 Aug-98 5,135,288 5,132,882 5,132,882 5,131,221 Nov-98 5,481,413 5,480,413 5,476,940 Feb-99 5,156,345 5,150,553 May-99 5,920,658 Aug-99 7,225,518 Nov-99 8,463,078 Feb-00 7,063,284 May-00 7,885,687 Aug-00 8,396,780 Nov-00 8,959,569 1-2 2-3 3-4 4-5 5-6 6-7 7-8 --- --- --- --- --- --- --- May-96 0.982 0.991 0.995 0.989 0.999 1.003 1.005 Aug-96 0.972 0.989 0.988 0.999 1.000 0.999 0.999 Nov-96 0.954 0.992 0.998 1.000 0.999 1.000 1.000 Feb-97 0.950 0.991 1.004 0.995 0.999 1.000 1.000 May-97 0.955 0.996 1.000 0.999 0.999 0.995 1.001 Aug-97 0.960 0.996 1.001 0.999 0.999 1.001 1.000 Nov-97 0.984 0.998 0.996 0.998 1.000 1.001 0.997 Feb-98 0.968 0.993 0.998 1.001 0.998 1.000 1.001 May-98 0.970 0.994 1.000 0.997 1.000 0.999 1.000 Aug-98 0.972 0.997 1.000 0.996 0.999 1.002 1.000 Nov-98 0.969 0.994 0.996 1.000 0.999 1.000 1.000 Feb-99 0.976 0.999 1.000 0.999 0.999 1.001 1.000 May-99 0.970 0.999 0.997 1.002 0.998 1.001 Aug-99 0.965 0.999 1.000 0.998 0.998 Nov-99 0.985 0.995 1.001 0.999 Feb-00 0.970 1.003 1.001 May-00 0.961 1.001 Aug-00 0.975 8-9 9-10 10-11 --- ---- ----- May-96 0.999 0.999 1.000 Aug-96 1.001 1.000 1.000 Nov-96 1.000 1.000 1.000 Feb-97 0.999 1.000 1.000 May-97 0.998 1.000 1.000 Aug-97 1.000 0.999 1.000 Nov-97 1.000 1.000 1.000 Feb-98 1.000 1.000 1.000 May-98 0.999 1.000 1.000 Aug-98 1.000 1.000 Nov-98 1.000 Feb-99 May-99 Aug-99 Nov-99 Feb-00 May-00 Aug-00 1-2 2-3 3-4 4-5 5-6 6-7 7-8 --- --- --- --- --- --- --- Straight Avg 0.964 0.994 0.997 0.998 0.999 1.000 1.000 Avg x HiLo 0.965 0.994 0.997 0.998 0.999 1.000 1.000 Wtd Avg All 0.967 0.995 0.998 0.998 0.999 1.000 1.000 Avg Last 8 0.971 0.998 0.999 0.999 0.999 1.001 1.000 Wt Avg.8 0.972 0.998 0.999 0.999 0.999 1.001 1.000 Avg Last 4 0.973 1.000 1.000 0.999 0.999 1.001 1.000 Wt Avg.4 0.973 0.999 1.000 0.999 0.998 1.001 1.000 SELECT 0.964 0.994 0.997 0.999 0.999 1.001 1.000 Cumulative 0.954 0.989 0.995 0.997 0.998 1.000 0.999 ==================================================================================================================================== ULTIMATE LOSS 8,959,569 8,396,780 7,885,687 7,063,284 8,463,078 7,225,518 5,920,658 ==================================================================================================================================== 8-9 9-10 10-11 --- ---- ----- Straight Avg 1.000 1.000 1.000 Avg x HiLo 1.000 1.000 1.000 Wtd Avg All 1.000 1.000 1.000 Avg Last 8 1.000 1.000 1.000 Wt Avg.8 1.000 1.000 1.000 Avg Last 4 1.000 1.000 1.000 Wt Avg.4 1.000 1.000 1.000 SELECT 1.000 1.000 1.000 Cumulative 0.999 0.999 1.000 ======================================================================= ULTIMATE LOSS 5,150,553 5,476,940 5,131,221 ======================================================================= 32 EXHIBIT 14 COUNTRYWIDE X AUTO AS OF FEBRUARY 28, 2001 Quarterly Record INCURRED LOSSES QUARTERLY LAG 2 - IBNR ANALYSIS Periods Ending 1 2 3 4 5 6 7 ------ - - - - - - - May-96 770,365 712,299 701,209 687,354 684,115 679,995 679,995 Aug-96 821,122 745,490 713,522 711,267 707,486 707,439 708,965 Nov-96 992,289 920,766 911,317 911,196 905,653 900,155 900,155 Feb-97 912,187 827,755 809,508 803,427 806,676 804,814 796,888 May-97 1,276,548 1,229,040 1,203,503 1,190,933 1,193,473 1,193,473 1,193,473 Aug-97 1,387,450 1,332,813 1,298,705 1,300,189 1,300,960 1,301,022 1,301,048 Nov-97 1,118,549 1,074,498 1,078,631 1,065,791 1,059,375 1,059,445 1,057,469 Feb-98 1,451,549 1,374,085 1,358,483 1,344,354 1,344,054 1,337,197 1,337,197 May-98 1,422,243 1,381,906 1,358,692 1,359,011 1,353,652 1,351,820 1,353,605 Aug-98 1,386,713 1,324,331 1,325,596 1,321,086 1,316,327 1,316,376 1,316,426 Nov-98 1,536,161 1,432,661 1,423,991 1,413,938 1,406,669 1,405,615 1,405,139 Feb-99 1,492,908 1,392,167 1,365,904 1,362,954 1,366,536 1,365,119 1,361,913 May-99 1,691,912 1,619,294 1,594,488 1,592,653 1,588,513 1,586,402 Aug-99 1,880,903 1,792,471 1,769,981 1,755,897 1,753,214 Nov-99 2,148,501 2,074,814 2,046,275 2,034,381 Feb-00 1,876,480 1,771,150 1,748,531 May-00 1,978,122 1,945,414 Aug-00 2,045,116 Quarterly Record INCURRED LOSSES QUARTERLY LAG 2 - IBNR ANALYSIS Periods Ending 8 9 10 ULTIMATE ------ - - -- -------- May-96 684,711 684,711 684,711 684,711 Aug-96 713,143 713,143 713,143 710,977 Nov-96 897,165 895,501 895,501 895,501 Feb-97 808,030 805,959 805,959 803,896 May-97 1,192,206 1,192,206 1,192,206 1,191,753 Aug-97 1,290,086 1,290,130 1,290,177 1,291,414 Nov-97 1,057,266 1,057,126 1,052,690 1,051,954 Feb-98 1,337,197 1,337,197 1,337,197 1,336,623 May-98 1,353,787 1,353,787 1,353,787 1,353,121 Aug-98 1,316,476 1,316,527 1,315,195 Nov-98 1,399,906 1,397,707 Feb-99 1,359,901 May-99 1,583,430 Aug-99 1,747,869 Nov-99 2,023,000 Feb-00 1,723,949 May-00 1,895,251 Aug-00 1,880,756 1-2 2-3 3-4 4-5 5-6 6-7 7-8 --- --- --- --- --- --- --- May-96 0.925 0.984 0.980 0.995 0.994 1.000 1.007 Aug-96 0.908 0.957 0.997 0.995 1.000 1.002 1.006 Nov-96 0.928 0.990 1.000 0.994 0.994 1.000 0.997 Feb-97 0.907 0.978 0.992 1.004 0.998 0.990 1.014 May-97 0.963 0.979 0.990 1.002 1.000 1.000 0.999 Aug-97 0.961 0.974 1.001 1.001 1.000 1.000 0.992 Nov-97 0.961 1.004 0.988 0.994 1.000 0.998 1.000 Feb-98 0.947 0.989 0.990 1.000 0.995 1.000 1.000 May-98 0.972 0.983 1.000 0.996 0.999 1.001 1.000 Aug-98 0.955 1.001 0.997 0.996 1.000 1.000 1.000 Nov-98 0.933 0.994 0.993 0.995 0.999 1.000 0.996 Feb-99 0.933 0.981 0.998 1.003 0.999 0.998 May-99 0.957 0.985 0.999 0.997 0.999 Aug-99 0.953 0.987 0.992 0.998 Nov-99 0.966 0.986 0.994 Feb-00 0.944 0.987 May-00 0.983 Straight Avg 0.944 0.985 0.991 0.997 0.997 1.001 1.001 Avg x HiLo 0.945 0.986 0.992 0.997 0.997 0.999 1.001 Wtd Avg All 0.949 0.986 0.993 0.997 0.998 1.000 1.000 Avg Last 8 0.953 0.988 0.995 0.997 0.999 1.000 1.000 Wt Avg.8 0.954 0.988 0.995 0.998 0.999 1.000 0.999 Avg Last 4 0.962 0.986 0.996 0.998 0.999 1.000 0.999 Wt Avg.4 0.962 0.986 0.995 0.998 0.999 1.000 0.999 SELECT 0.944 0.988 0.991 0.997 0.999 1.000 1.000 Cumulative 0.920 0.974 0.986 0.994 0.997 0.998 0.999 =================================================================================================================================== ULTIMATE LOSS 1,880,756 1,895,251 1,723,949 2,023,000 1,747,869 1,583,430 1,359,901 =================================================================================================================================== 8-9 9-10 10-11 --- ---- ----- May-96 1.000 1.000 1.000 Aug-96 1.000 1.000 1.000 Nov-96 0.998 1.000 1.000 Feb-97 0.997 1.000 1.000 May-97 1.000 1.000 1.000 Aug-97 1.000 1.000 0.999 Nov-97 1.000 0.996 1.000 Feb-98 1.000 1.000 1.000 May-98 1.000 1.000 Aug-98 1.000 Nov-98 Feb-99 May-99 Aug-99 Nov-99 Feb-00 May-00 Straight Avg 0.999 1.008 0.998 Avg x HiLo 1.000 0.999 0.999 Wtd Avg All 1.000 1.006 0.999 Avg Last 8 0.999 0.999 1.000 Wt Avg.8 1.000 0.999 1.000 Avg Last 4 1.000 0.999 1.000 Wt Avg.4 1.000 0.999 1.000 SELECT 0.999 0.999 1.000 Cumulative 0.998 0.999 1.000 ================================================================== ULTIMATE LOSS 1,397,707 1,315,195 1,353,121 ================================================================== 33 EXHIBIT 15 COUNTRYWIDE X AUTO AS OF FEBRUARY 28, 2001 Quarterly Record INCURRED LOSSES QUARTERLY LAG 0-7 EMERGENCE - IBNR ANALYSIS Periods Ending 0 1 2 3 4 5 6 7 ------ - - - - - - - - May-96 35,006,078 3,437,569 684,711 366,055 265,140 165,808 161,717 94,898 Aug-96 36,680,040 4,184,110 710,977 413,830 269,616 182,741 133,911 14,423 Nov-96 39,364,951 5,182,672 895,501 489,966 352,416 189,898 114,419 90,356 Feb-97 40,736,857 4,222,662 803,896 514,647 277,031 180,505 123,907 76,961 May-97 44,084,212 5,242,462 1,191,753 594,678 316,638 192,164 121,463 95,719 Aug-97 47,920,902 5,776,046 1,291,414 553,786 308,604 270,404 209,562 156,660 Nov-97 50,719,745 7,178,631 1,051,954 795,145 390,532 241,180 208,905 125,585 Feb-98 58,004,870 4,340,389 1,336,623 711,203 441,389 402,513 137,374 85,993 May-98 57,469,615 4,596,326 1,353,121 728,303 492,691 293,498 279,203 201,338 Aug-98 59,590,223 5,131,221 1,315,195 785,063 474,482 368,215 170,199 135,172 Nov-98 64,450,176 5,476,940 1,397,707 655,292 614,304 391,829 213,994 176,689 Feb-99 67,182,405 5,150,553 1,359,901 825,705 531,226 411,193 179,697 226,079 May-99 71,911,496 5,920,658 1,583,430 942,834 504,764 421,604 244,247 144,641 Aug-99 79,235,439 7,225,518 1,747,869 1,031,631 670,450 287,054 265,278 Nov-99 86,695,059 8,463,078 2,023,000 1,123,041 727,666 365,023 Feb-00 90,173,735 7,063,284 1,723,949 1,234,355 584,067 May-00 92,079,844 7,885,687 1,895,251 1,168,689 Aug-00 94,547,077 8,396,780 1,880,756 Nov-00 95,991,072 8,959,569 Feb-01 98,816,838 Avg 3 Emergence 96,451,662 8,414,012 1,833,319 1,175,362 660,728 357,894 229,741 182,470 Quarterly Record INFLATED INCURRED LOSSES QUARTERLY LAG 0-7 EMERGENCE - IBNR ANALYSIS Periods Ending 0 1 2 3 4 5 6 7 ------ - - - - - - - - May-96 43,146,526 4,190,586 825,566 436,527 312,724 193,425 186,588 108,294 Aug-96 44,714,987 5,044,839 847,853 488,099 314,523 210,845 152,815 16,279 Nov-96 47,462,865 6,180,433 1,056,215 571,575 406,615 216,705 129,142 100,867 Feb-97 48,579,460 4,980,494 937,793 593,796 316,138 203,732 138,321 84,973 May-97 51,995,909 6,115,646 1,375,036 678,626 357,382 214,518 134,108 104,528 Aug-97 55,902,599 6,664,362 1,473,718 625,046 344,502 298,555 228,847 169,204 Nov-97 58,520,093 8,192,011 1,187,317 887,640 431,190 263,375 225,633 134,157 Feb-98 66,193,193 4,898,899 1,492,106 785,245 482,008 434,744 146,750 90,857 May-98 64,864,651 5,130,995 1,493,992 795,325 532,143 313,530 294,996 210,398 Aug-98 66,522,067 5,665,423 1,436,226 847,926 506,867 389,042 177,858 139,709 Nov-98 71,159,971 5,980,955 1,509,627 700,018 649,051 409,461 221,176 180,621 Feb-99 73,364,866 5,562,978 1,452,719 872,409 555,131 424,994 183,696 228,581 May-99 77,669,737 6,324,765 1,672,993 985,262 521,706 430,986 246,950 144,641 Aug-99 84,643,557 7,634,214 1,826,523 1,066,256 685,369 290,230 265,278 Nov-99 91,598,784 8,843,917 2,090,899 1,148,031 735,718 365,023 Feb-00 94,231,553 7,300,353 1,762,311 1,248,013 584,067 May-00 95,170,378 8,061,163 1,916,222 1,168,689 Aug-00 96,650,978 8,489,690 1,880,756 Nov-00 97,053,212 8,959,569 Feb-01 98,816,838 34 EXHIBIT 16 COUNTRYWIDE X AUTO AS OF FEBRUARY 28, 2001 Quarterly Record INFLATED INCURRED LOSSES QUARTERLY LAG 0-7 PURE PREMIUMS - IBNR ANALYSIS Periods Ending 0 1 2 3 4 5 6 7 ------ - - - - - - - - May-96 141.040 13.698 2.699 1.427 1.022 0.632 0.610 0.354 Aug-96 141.818 16.000 2.689 1.548 0.998 0.669 0.485 0.052 Nov-96 146.080 19.022 3.251 1.759 1.251 0.667 0.397 0.310 Feb-97 142.913 14.652 2.759 1.747 0.930 0.599 0.407 0.250 May-97 138.831 16.329 3.671 1.812 0.954 0.573 0.358 0.279 Aug-97 139.638 16.647 3.681 1.561 0.861 0.746 0.572 0.423 Nov-97 140.531 19.672 2.851 2.132 1.035 0.632 0.542 0.322 Feb-98 137.353 10.165 3.096 1.629 1.000 0.902 0.305 0.189 May-98 131.761 10.423 3.035 1.616 1.081 0.637 0.599 0.427 Aug-98 132.351 11.272 2.857 1.687 1.008 0.774 0.354 0.278 Nov-98 136.649 11.485 2.899 1.344 1.246 0.786 0.425 0.347 Feb-99 136.616 10.359 2.705 1.625 1.034 0.791 0.342 0.426 May-99 133.876 10.902 2.884 1.698 0.899 0.743 0.426 0.249 Aug-99 135.821 12.250 2.931 1.711 1.100 0.466 0.426 Nov-99 136.914 13.219 3.125 1.716 1.100 0.546 Feb-00 134.826 10.445 2.522 1.786 0.836 May-00 127.890 10.833 2.575 1.570 Aug-00 124.043 10.896 2.414 Nov-00 120.595 11.133 Feb-01 120.649 Straight Avg 135.010 13.126 2.925 1.669 1.022 0.678 0.446 0.300 Avg x HiLo 135.196 12.916 2.909 1.659 1.019 0.677 0.444 0.312 Avg Last 8 129.327 11.255 2.757 1.642 1.038 0.706 0.427 0.333 Avg Last 4 123.294 10.827 2.659 1.696 0.984 0.636 0.405 0.325 ========================================================================================================================= SELECT 129.327 11.255 2.757 1.642 1.038 0.706 0.427 0.333 ========================================================================================================================= Quarterly FUTURE PURE PREMIUMS BY QUARTERLY LAG, INFLATED Record Total Periods Future Ending 0 1 2 3 4 5 6 7 Pure Prem ------ - - - - - - - - --------- May-96 0.057 Aug-96 0.063 Nov-96 0.068 Feb-97 0.074 May-97 0.079 Aug-97 0.105 Nov-97 0.132 Feb-98 0.158 May-98 0.185 Aug-98 0.364 Nov-98 0.545 Feb-99 0.728 May-99 0.913 Aug-99 0.336 1.260 Nov-99 0.432 0.340 1.706 Feb-00 0.713 0.437 0.344 2.438 May-00 1.049 0.721 0.442 0.348 3.514 Aug-00 1.660 1.061 0.729 0.446 0.351 5.213 Nov-00 2.787 1.679 1.073 0.737 0.451 0.355 8.058 Feb-01 11.379 2.818 1.697 1.085 0.746 0.456 0.359 19.527 35 EXHIBIT 17 COUNTRYWIDE X AUTO AS OF FEBRUARY 28, 2001 SUMMARY OF ESTIMATED IBNR (1) (2) (3) (4) (5) (6) (7) (8) (9) Quarterly IBNR Avg IBNR Record Total Required Emerged Required Emerged Required Periods Future Earned Earned Required IBNR Since IBNR Since IBNR Ending Pure Prem Exposures Premium IBNR Factors Aug-00 Factors Aug-00 Factors ------ --------- --------- ------- ---- ------- ------ ------- ------ ------- May-96 0.057 305,916 182,707,245 17,581 0.0% 2,966 0.0% 3,749 0.0% Aug-96 0.063 315,298 189,294,752 19,791 0.0% 2,966 0.0% 3,749 0.0% Nov-96 0.068 324,911 196,146,572 22,135 0.0% 2,966 0.0% 3,749 0.0% Feb-97 0.074 339,923 204,804,526 24,999 0.0% 2,966 0.0% 3,749 0.0% May-97 0.079 374,526 224,067,400 29,595 0.0% 21,056 0.0% 22,875 0.0% Aug-97 0.105 400,340 239,126,469 42,105 0.0% 21,056 0.0% 22,875 0.0% Nov-97 0.132 416,420 246,524,368 54,806 0.0% 21,056 0.0% 22,875 0.0% Feb-98 0.158 481,920 278,347,423 76,310 0.0% 21,056 0.0% 22,875 0.0% May-98 0.185 492,289 277,743,993 91,258 0.0% 209,267 0.1% 177,470 0.1% Aug-98 0.364 502,617 278,081,535 183,134 0.1% 209,267 0.1% 177,470 0.1% Nov-98 0.545 520,751 280,693,980 283,981 0.1% 209,267 0.2% 177,470 0.2% Feb-99 0.728 537,014 279,298,219 391,107 0.1% 209,267 0.2% 177,470 0.2% May-99 0.913 580,161 293,396,358 529,858 0.2% 388,888 0.3% 412,210 0.3% Aug-99 1.260 623,200 307,522,334 785,010 0.3% 552,332 0.4% 587,634 0.4% Nov-99 1.706 669,022 321,801,544 1,141,021 0.4% 1,092,689 0.7% 1,018,621 0.7% Feb-00 2.438 698,911 330,437,470 1,703,804 0.5% 1,818,422 1.1% 1,836,089 1.1% May-00 3.514 744,159 347,321,145 2,615,156 0.8% 3,063,940 1.6% 3,008,680 1.6% Aug-00 5.213 779,176 358,752,241 4,062,189 1.1% 10,277,536 4.0% 10,247,331 4.0% Nov-00 8.058 804,785 368,144,959 6,485,294 1.8% Feb-01 19.527 819,043 373,175,036 15,993,193 4.3% 12,854,162 6,820,909,077 34,584,463 18,130,936 17,939,953 Six Mth Runoff Six Mth Runoff Zero Runoff Estimate #1 Estimate #2 ----------- ----------- ----------- Indicated IBNR: 34,584,463 30,236,912 30,045,929 Carried IBNR: 28,782,946 27,252,921 27,252,921 Adequacy: (5,801,517) (2,983,991) (2,793,008) 36 SECTION VII: CASE STUDY -- LOSS ADJUSTMENT EXPENSE RESERVE REVIEW The next case study is a Loss Adjustment Expense (LAE) reserve review. We review the two types of expenses, Defense and Cost Containment (DCC), formerly referred to as allocated loss adjustment expense or ALAE and adjusting expenses, formerly referred to as unallocated loss adjustment expense or ULAE, separately. The review for this case study is a countrywide segment (excluding certain states or "CWX") for "Feel Bad" coverage with data evaluated as of March 31, 2002. When developing ultimate LAE costs several techniques need to be employed that include losses as a base since LAE costs vary relative to changes in loss patterns. Generally, LAE is incurred only when there is claim activity. You will see how we utilize changes in loss patterns as we develop the indications of ultimate costs below. TOTAL NEEDED DCC (ALAE) RESERVE DCC (ALAE) review -- Exhibit 18 is a summary of the different projections of DCC (ALAE) cost by accident semiannual periods. In January 1998 the statutory definition of ALAE was changed by the NAIC to only include expenses for litigation, defense and cost containment. The primary change resulted in outside adjusting expenses moving to be included in ULAE; previously they were included in ALAE. EXHIBIT 18 is a summary document used by Progressive to display consolidated results of needed information to determine reserve needs. We have partitioned Exhibit 18 into sections which we will comment on individually. EXHIBIT 18, SECTION A In this section we show the results of the three different projection methods along with other meaningful data. The ending date of each accident period (we are studying semiannual accident periods in this example) is displayed down the far left (column 1). The three columns immediately following are the results of the three estimation methods consistent with those previously discussed in this document and are used to project ultimate needed DCC (ALAE) reserves. Column 5 and 6 show the losses that have already been paid by accident period as well as the indicated ultimate for those accident periods as well. Below is a description of each of the columns in section A. COLUMN 1: Semiannual accident periods ending: The ending date of the accident period under review. COLUMN 2: Results of Paid Method -- triangular projection of the paid DCC (ALAE) amounts to ultimate value. COLUMN 3: Results of Average Paid ALAE Method -- average paid DCC (ALAE) is displayed in the triangular format then projected to ultimate by accident period. This ultimate severity is then multiplied by the ultimate number of claims which is expected to incur DCC (ALAE) costs to estimate an ultimate amount. COLUMN 4: Results of Paid ALAE to Paid Loss Method -- this method takes the statistic, paid ALAE to paid loss ratio, and arrays it in a triangle of values for each accident semester by age. The ultimate ALAE to Loss ratio is then projected. These ultimate ratios when multiplied by the ultimate projected loss costs generate estimates of ultimate DCC (ALAE) by accident semester. 37 COLUMN 5: Paid Legal ALAE as of March 2002- Paid DCC (ALAE) amounts by accident period as of the evaluation date of March 2002 are shown here. We pull these amounts from the last diagonal of the paid amount triangle. COLUMN 6: Indicated Ultimate -- These are selected indicated ultimate DCC (ALAE) costs by accident period. EXHIBIT 18, SECTION B. displays two different projections for the ultimate number of claims that will incur DCC (ALAE) costs. COLUMN 1: Ultimate Legal ALAE Count -- is a triangular projection of the number of claims that include a paid DCC (ALAE) amount to an ultimate number. COLUMN 2: Ultimate Projected Utilization -- we have defined the ratio of number of claims incurring DCC (ALAE) to total number of claims as a utilization ratio. We array the ratios in a triangle format and project an ultimate ratio by accident semester. COLUMN 3: Projected Counts -- multiplying the Ultimate Projected Utilization ratio (Column 2) by the ultimate loss counts by accident semester gives us a second projection of ultimate counts. COLUMN 4: Selected Legal ALAE Counts -- These are the selected ultimate number of claims that are expected to incur in DCC (ALAE) costs. COLUMN 5: Selected Ultimate Utilization -- this is the ratio of the selected ultimate number of claims to incur DCC (ALAE) costs (Column (4)) divided by the ultimate loss counts. COLUMN 6: Prior Review Ultimate Utilization -- these are the ultimate utilization ratios that were selected for the appropriate accident periods as a result of our prior review. This information is used to identify accident semesters where we may have changed our opinion significantly of the Ultimate Utilization ratio. Significant changes will generally prompt us to review the underlying data and try to identify what is causing us to change our opinion. EXHIBIT 18, SECTION C includes loss, premium and exposure values that are needed to calculate the ratios or relative DCC (ALAE) measures used in the analysis. COLUMN 1: Indicated Ultimate Loss -- These amounts are used to calculate the amounts shown in Section A, column (4) (Paid ALAE to Paid Loss Method Ult.) COLUMN 2: Indicated Ultimate Loss Counts -- These values were needed to determine the projected counts as shown in Section B Column (3) and the Selected Ultimate Utilization (Column (5), Section B) COLUMN 3: Earned Premium by Semester COLUMN 4: Earned Exposures which are shown as a growth statistic for the segment being reviewed. EXHIBIT 18, SECTION D includes the ultimate cost generated by each projection discussed in Section A as a ratio to the ultimate loss amounts, plus the projected average paid ALAE severity. COLUMN 1: Semiannual accident periods COLUMN 2: Paid Ult ALAE/Loss -- the projected amounts from column (2) in Section A divided by the Ultimate Loss (Section C, column (1)) COLUMN 3: Average Paid Ult ALAE/Loss -- the projected amounts from column (3) in Section A divided by the Ultimate Loss amounts (Section C, column (1)) COLUMN 4: Paid to Paid Ult ALAE/Loss -- the projected ultimate ALAE to Loss ratios from the triangular projection that were multiplied by the Indicated Ultimate Loss (Section C, column (1)) to calculate the amount in column (4) in Section A. 38 COLUMN 5: Projected Average Paid ALAE Severity -- ultimates from triangles of average paid DCC (ALAE) amounts. EXHIBIT 18, SECTION E is statistics calculated using selected Indicated Ultimate Legal ALAE amounts shown in column (6) of Section A. COLUMN 1: Selected Ultimate Legal ALAE/Loss -- amounts in column (6) of Section A divided the indicated Ultimate Loss amounts in Section C, Column (1). COLUMN 2: Selected Ultimate Legal ALAE/EP -- the ALAE ratio to Earned Premium. It is calculated by dividing the Indicated Ultimate Legal ALAE amount in column (6) of Section A by the Earned Premium (Section C, column (3)). COLUMN 3: Selected Ultimate Legal ALAE Severity -- Indicated Ultimate Legal ALAE (Section A, column (6)) divided by the Selected Legal ALAE counts shown in column (4) of Section B. COLUMN 4: Prior Review Legal ALAE Severity -- the ultimate DCC (ALAE) severities developed in the prior review of this segment. Again we look at the comparison of our results in this review compared to those of the prior review to note where we may differ and possibly identify changes that we should study more completely. EXHIBIT 16, SECTION F is the "Retrospective Comparison of Indicated Ultimates". This table compares the selected ultimates made during this review to the selected ultimates we made the last time we reviewed the segment. The purpose for the comparison is to identify how much the current selection differs from the prior selections. We then decide if some or all of the accident period statistics need to be reviewed further in order to understand the changes we are seeing within the data. REVIEWING THE INFORMATION INCLUDED ON THE SUMMARY We review the following data from Exhibit 18 to help ensure our indication is reasonable: - - The results of the three DCC (ALAE) development methods. An ultimate cost for each accident period is then selected. - - The range of the indications resulting from the different projections in Section A. In this case the reserve adequacy ranges from an adequacy of $115,000 (column 3) to an adequacy of $320,000 (column 2). The range of indicated ultimate loss costs is also reviewed for the last accident period. The example shows a range from $5,083,000 (column 4) to $5,205,000 (column 3). - - The results of the two count projection methods shown in Section B. An ultimate count for each accident period is then selected. - - The utilization rate over time to determine possible influences internal process changes may have on the ultimate cost indications. In Section B, column 5, the rate for the last accident period is 57.1%. This is consistent with the historical values. - - The ultimate utilization ratios should generally be the same as those selected in the prior review. - - The indicated ultimate DCC (ALAE) severity (Section E, column 3). Changes in the accident period severity(s) are compared to other information or reviews available and to our expected annual trend. - - The ultimate severity(s) by accident period is reviewed for consistency over time and that the annual change is reasonable. If there is a question about the ultimate severity, we will look at the change in severity resulting from the prior review to hopefully identify where we might further analyze the underlying data. - - The retrospective comparison of indicated ultimates in Section F to determine if we are changing our opinion significantly from our last review and if so, we will review the underlying data which produced the change in the indication. 39 EXHIBIT 18, SECTION G is where we calculate the actual change in average reserves and the IBNR factor is made. COLUMN (1) ALAE Case Reserve Calculation -- is the calculation of needed average reserve compared to the present average reserve. In making the calculation of the indicated needed average reserve we subtract the number and amount of reserve for adjuster's estimates greater than or equal to the threshold ($15,000). There are eight claims for $170,040. The indicated needed case reserves of $19.7 million minus the $170,040 for case >= threshold divided by the number of reserves less than the threshold (37,094) equals $525. The present average reserve is $533 and we have decided to select an average of $527. COLUMNS (2) through (9) is the determination of the new average reserves for each age group. (1-6 months, 7-18 months, and 19 months and older). COLUMN (10): ALAE IBNR Reserve Calculation -- is the determination of indicated needed IBNR factor. The LAE IBNR factors are applied to the carried loss IBNR to generate the amount of LAE IBNR reserve we carry. As you can see the indicated needed ratio is 3.9% and our carried factor is 3.9%, therefore only a small change is generated. The small indicated inadequacy is due to rounding to 0.1% In the middle of Section G, we show the reserve changes for the Case and IBNR. The total in this example is a decrease of $203,296. TOTAL NEEDED ADJUSTING EXPENSE (ULAE) RESERVES The Adjusting Expense (ULAE) reserve indications are completed generally by reviewing calendar quarter Adjusting Expense (ULAE) cost statistics. Exhibit 19 is the basic review completed for most segments. Basically the method is to study the paid adjusting expenses (ULAE) to paid loss ratios by calendar quarter, then estimate what we believe to be the expected ratio of adjusting expense (ULAE) to loss going forward, then apply the estimated unpaid portion to our carried loss reserves. COLUMN 1: Adjusted ULAE Counts -- This is a weighted average of the number of new recorded claims and the number of paid claims. This count is used as a surrogate of the proportional number of transactions that require claim staff activity. COLUMN 2: Paid Loss Capped -- total paid loss amounts with claims capped at $50,000. We use this as a basis assuming that the amount of staff involvement does not increase proportionally as the size of the claim increases in dollars greater that $50,000. COLUMN 3: Paid Loss Uncapped -- total paid loss amounts by calendar quater. This statistic is the basis for the measure to be applied to the IBNR as we cannot assume we know the expected late report costs capped at $50,000. COLUMN 4: Total ULAE charged -- total amount of paid adjusting expenses (ULAE) by calendar quater related to the segment being reviewed. COLUMN 5: ULAE charged per ULAE count -- column (4) divided by the number of adjusted ULAE counts (column (1)) -- This measure is a statistic derived to possibly identify changing cost for handling claims. COLUMN 6: Ratio of ULAE to Capped Paid Loss -- column (4) divided by column (2) -- paid-to-paid ratio used to develop the needed adjusting expense (ULAE) reserve for settling open case reserves. 40 COLUMN 7: Ratio of ULAE to Uncapped Paid Loss -- column (4) divided by column (3) -- This statistic is used to determine the expected needed adjusting expense (ULAE) reserve to adjust the IBNR (late reported) claims. You will note that we show moving twelve month values for the statistics shown in columns (5), (6) and (7) (ULAE per Count, per capped paid losses, per uncapped paid losses). Based on these historical statistics we make a selection of our estimated adjusting expense (ULAE) to loss ratios expected to be incurred in adjusting (settling) the open case and IBNR reserves. Columns (8) through (14) are the values used to determine our adjusting expense (ULAE) reserve adequacy. COLUMN 8: Loss Reserves -- open case reserves capped at $50,000 and the total IBNR reserves. COLUMN 9: Incurred ULAE -- The projected total incurred adjusting expense (ULAE) for adjusting all claims reserved. This amount equals the selected adjusting expense (ULAE) to the loss ratio shown at the bottom of column (6) applied to the capped case reserves and the selected ratio shown at the bottom of column (7) applied to the IBNR reserves. COLUMN 10: % Unpaid -- these are the estimates of the portion of the adjusting expenses (ULAE) that are unpaid. An estimate of this portion is determined from Progressive's time tracking studies. The underlying concept is that adjusting expenses (ULAE) are incurred when a claim is opened (actually even prior as demonstrated by our application of 95% for IBNR) and regularly during the life of a claim through the time a claim is settled. COLUMN 11: Indicated ULAE reserve -- Indicated needed adjusting expense (ULAE) reserve which equals column (9) multiplied by column (10). COLUMN 12: Carried ULAE Reserve-- COLUMN 13: Reserve Adequacy -- Indicated needed change in the carried adjusting expense (ULAE) reserves equal to column (12) minus column (11). COLUMN 14: # ULAE Reserve -- This count equals the number of open loss reserves. Column 15: Present ULAE Avgs -- Current average reserve and IBNR factor used to establish the current reserve shown in column (12). COLUMN 16: Indicated ULAE Avgs -- needed average reserve and IBNR factor to calculate the amount of Indicated ULAE Reserve shown in column (11). The indicated average reserve is determined by dividing the case amount shown in column (11) by column (14). The indicated needed IBNR factor is calculated by taking the IBNR amount shown in column (11) divided by the loss IBNR reserve in column (8). COLUMN 17: Selected ULAE Avgs -- The new average reserve and IBNR factor to be used in the future for generating the adjusting expense (ULAE) reserves. COLUMN 18: $ Impact ULAE Reserve -- The change in the carried reserves after adopting the selected average and IBNR factors. The impact is calculated by multiplying the average reserve in column (17) by the number of reserves shown in column (14) minus the carried reserve amount for open cases as shown in column (12). For the IBNR impact we multiply the IBNR factor in column (17) by the loss IBNR reserve in column (8) minus the carried reserve for IBNR as shown in column (12). The adjusting expense (ULAE) cost statistics have been increasing for this segment during the periods prior to March, 2001. They have leveled off to where the selected ratio to capped losses of 24.0%, appears an adequate estimate going forward. The selected average of $315 is close to the indicated average of $314 while increasing the IBNR factor to 21.0% increased the reserve for ULAE IBNR to a level that is reasonably close to the indicated overall needed reserve. 41 Exhibit 18 ================================================================================================================================ SECTION A (1) (2) (3) (4) (5) (6) Semiannual Avg Paid ALAE Paid Accident Paid ALAE Paid ALAE to Paid Loss Legal ALAE INDICATED Periods Method Method Method to date ULTIMATE Ending Ult ($000) Ult ($000) Ult ($000) ($000) LEGAL ALAE ------ ---------- ---------- ---------- ------ ---------- Sep-95 1,543 1,543 1,539 1,536 1,542 Mar-96 1,839 1,841 1,827 1,823 1,836 ---------------------------------------------------------------------------------------------- Sep-96 2,095 2,095 2,095 2,041 2,095 Mar-97 2,254 2,249 2,249 2,146 2,251 Sep-97 2,775 2,769 2,753 2,558 2,766 Mar-98 2,761 2,759 2,763 2,421 2,761 ---------------------------------------------------------------------------------------------- Sep-98 2,524 2,522 2,506 2,053 2,517 Mar-99 2,738 2,739 2,725 1,961 2,734 Sep-99 3,525 3,561 3,519 2,036 3,535 Mar-00 4,255 4,293 4,276 1,819 4,275 ---------------------------------------------------------------------------------------------- Sep-00 4,963 4,987 5,011 1,415 4,987 Mar-01 4,681 4,700 4,720 731 4,700 Sep-01 4,942 4,971 4,997 307 4,970 Mar-02 5,134 5,205 5,083 68 5,144 Total 46,029 46,234 46,065 22,917 46,113 Paid ALAE 22,917 22,917 22,917 22,917 ================================================================================================================================ REQUIRED RESERVE 23,112 23,317 23,147 23,196 HELD RESERVE 23,432 23,432 23,432 23,432 RESERVE ADEQUACY 320 115 285 237 ================================================================================================================================ ================================================================================================================ SECTION B (1) (2) (3) (4) (5) (6) ================================ Ultimate Ultimate SELECTED SELECTED PR REVIEW Legal ALAE Projected Projected LEGAL ALAE ULTIMATE ULTIMATE Counts Utilization Counts COUNTS UTILIZATION UTILIZATION ------ ----------- ------ ------ ----------- ----------- 3,780 50.6% 3,776 3,778 50.6% 52.2% --------------- 3,851 51.9% 3,845 3,848 51.9% 50.4% - --------------------------------------------------------------------------------- 4,241 54.2% 4,235 4,238 54.3% 55.4% 4,460 55.1% 4,453 4,456 55.2% 56.3% 5,120 57.2% 5,111 5,115 57.2% 57.6% --------------- 5,217 54.1% 5,214 5,215 54.1% 53.4% - --------------------------------------------------------------------------------- 5,375 52.1% 5,371 5,373 52.2% 53.9% 5,995 56.2% 5,986 5,990 56.3% 55.7% 6,765 54.3% 6,756 6,760 54.4% 55.4% --------------- 7,606 56.2% 7,623 7,615 56.2% 54.2% - --------------------------------------------------------------------------------- 8,771 58.2% 8,869 8,820 57.9% 60.2% 8,218 56.7% 8,336 8,277 56.3% 56.9% 8,363 56.0% 8,439 8,401 55.8% 59.2% 8,543 57.8% 8,769 8,656 57.1% 86,305 86,782 86,543 55.3% ================================================================================================================ ================================================================= SECTION C (1) (2) (3) (4) Indicated Indicated Ultimate Ultimate Earned Earned Loss Loss Premium Exposures ($000) Counts ($000) ($000) ------ ------ ------ ------ 23,418 7,467 152,210 428,343 22,786 7,410 157,566 436,078 - ----------------------------------------------------------------- 23,477 7,808 170,529 466,125 23,917 8,080 184,019 503,024 27,409 8,937 206,615 559,890 28,100 9,641 230,209 624,568 - ----------------------------------------------------------------- 31,546 10,300 239,906 683,728 32,453 10,643 239,246 735,520 37,786 12,434 254,610 858,595 41,692 13,556 278,198 1,000,884 - ----------------------------------------------------------------- 47,627 15,227 305,199 1,153,169 46,307 14,702 318,530 1,253,556 48,178 15,068 345,089 1,354,756 50,975 15,160 385,876 1,466,823 485,671 156,433 3,467,801 11,525,059 ================================================================= ================================================================================================= SECTION D (1) (2) (3) (4) (5) Semiannual Projected Accident Average Paid Average Periods Paid Ult Paid Ult to Paid Ult Paid ALAE Ending ALAE/Loss ALAE/Loss ALAE/Loss Severity ------ --------- --------- --------- -------- Sep-98 8.00% 8.00% 7.94% 469 Mar-99 8.44% 8.44% 8.40% 457 Sep-99 9.33% 9.42% 9.31% 527 Mar-00 10.21% 10.30% 10.26% 564 - ---------------------------------------------------------------------------------------------------------------- Sep-00 10.42% 10.47% 10.52% 565 Mar-01 10.11% 10.15% 10.19% 568 Sep-01 10.26% 10.32% 10.37% 592 Mar-02 10.07% 10.21% 9.97% 601 ================================================================================================= ================================================================ Section E (1) (2) (3) (4) Selected Selected Selected Ultimate Ultimate Ultimate Prior Review Legal Legal Legal ALAE Legal ALAE ALAE/Loss ALAE/EP Severity Severity 7.98% 1.05% 469 454 8.42% 1.14% 456 455 9.36% 1.39% 523 568 ---------------- 10.25% 1.54% 561 591 - ------------------------------------------------ 10.47% 1.63% 565 544 10.15% 1.48% 568 562 10.32% 1.44% 592 596 10.09% 1.33% 594 ================================================================ ================================================================================================================================ SECTION F RETROSPECTIVE COMPARISON OF INDICATED ULTIMATES Semiannual Ult. Legal Semiannual Ult. Legal Ult. Legal Accident ALAE at Accident ALAE at ALAE at Periods 10/31/2001 Periods 10/31/2001 03/31/2002 Change Ending ($000) Ending ($000) ($000) ($000) ------ ------ ------ ------ ------ ------ Prior 15,842 15,767 (75) Prior 15,842 Mar-99 2,666 2,734 68 Apr-99 2,695 Sep-99 3,712 3,535 (177) Oct-99 3,915 Mar-00 4,273 4,275 2 Apr-00 4,344 Sep-00 4,881 4,987 106 Oct-00 4,988 Mar-01 4,751 4,700 (51) Apr-01 4,704 Sep-01 5,215 4,970 (245) Oct-01 5,317 Mar-02 NA 5,144 NA Total 41,807 41,340 40,969 (372) ================================================================================================================================ ==================================================================================================================================== ALAE CASE BY-AGE AVERAGES ALAE CASE RESERVE CALCULATION (1) (2) (3) (4) (5) (6) (7) Current Age Age # of Current Indicated Case inflation 8.00% Relativities (mos) Rel. Reserves Averages Averages ------ ------------ ----- ---- -------- -------- -------- Number of ALAE Reserves 37,102 1.000 1-6 0.400 17,233 306 210 $ ALAE Reserves >= 15,000 170,040 1.719 7-18 1.055 13,274 516 553 # ALAE Reserves >= 15,000 8 3.869 19+ 2.475 6,587 1,160 1,296 ----- --- ----- ----- ----- ----- # ALAE Reserves < 15,000 37,094 1.767 Tot.(Avg) 1.003 37,094 533 525 Carried ALAE Case Reserves 19,936,235 Indicated Case Reserves 19,660,874 Case Reserve Adequacy 275,361 Indicated Case Average 525 Present LAE Avg < 15000 533 Selected LAE Avg < 15000 527 --------------------------------------------------- $ ALAE CASE IMPACT (217,657) $ ALAE IBNR IMPACT 14,362 --------------------------------------------------- $ ALAE TOTAL IMPACT (203,296) --------------------------------------------------- ==================================================================================================================================== =================================================================================================================================== ALAE CASE BY-AGE AVERAGES ALAE CASE RESERVE CALCULATION (8) (9) New Age Case inflation Averages Dollars -------- ------- Number of ALAE Reserves 210 3,622,356 ALAE IBNR RESERVE CALCULATION $ ALAE Reserves >= 554 7,359,097 (10) # ALAE Reserves >= 1,301 8,567,085 ----- --------- # ALAE Reserves < 527 19,548,538 Carried IBNR Loss reserves 90,013,755 Carried ALAE Case Reserves Carried IBNR ALAE reserves 3,496,175 Indicated Case Reserves Indicated IBNR ALAE reserves 3,536,527 Case Reserve Adequacy Reserve Adequacy (40,352) Indicated Case Average Indicated IBNR ALAE Factor 3.9% Present LAE Avg < 15000 Current IBNR ALAE Factor 3.9% Selected LAE Avg < 15000 Selected IBNR ALAE Factor 3.9% =============================================================================================================================== 42 PROGRESSIVE CORPORATION CWX FEEL BAD ADJUSTING (ULAE) AS OF 03/31/2002 EXHIBIT 19 - ----------------------------------------------------------------------------------------------------------------------------------- (1) (2) (3) (4) (5) (6) (7) ULAE Charged Ratio of ULAE Ratio of ULAE Quaterly Adjusted Total Per ULAE to Capped to Uncapped Calendar ULAE Paid Loss Paid Loss ULAE Count Paid Loss Paid Loss Enddates Counts Capped Uncapped Charged (4) / (1) (4) / (2) (4) / (3) -------- ------ ------ -------- ------- --------- --------- --------- Jun-99 7,467 14,236,816 15,001,601 3,089,622 414 21.7% 20.6% Sep-99 8,226 15,836,481 16,742,747 3,197,987 389 20.2% 19.1% Dec-99 8,749 17,190,541 17,985,754 3,268,595 374 19.0% 18.2% Mar-00 8,786 16,627,381 17,775,641 3,466,173 395 20.8% 19.5% - ----------------------------------------------------------------------------------------------------------------------------------- Jun-00 9,721 16,623,423 17,615,676 3,654,137 376 22.0% 20.7% Sep-00 9,817 17,340,405 18,133,668 3,956,445 403 22.8% 21.8% Dec-00 9,769 19,513,160 20,998,277 3,973,114 407 20.4% 18.9% Mar-01 9,407 19,077,683 20,363,781 4,596,006 489 24.1% 22.6% - ----------------------------------------------------------------------------------------------------------------------------------- Jun-01 9,534 20,101,751 21,527,582 4,792,713 503 23.8% 22.3% Sep-01 9,555 19,700,105 20,902,461 5,028,207 526 25.5% 24.1% Dec-01 9,927 22,658,783 24,386,180 5,185,277 522 22.9% 21.3% Mar-02 9,240 21,249,950 22,971,464 5,624,340 609 26.5% 24.5% Wtd Avg for Apr-99 thru Mar-00 392 20.4% 19.3% Wtd Avg for Apr-00 thru Mar-01 418 22.3% 21.0% Wtd Avg for Apr-01 thru Mar-02 539 24.6% 23.0% Selected @ Mar-02 24.0% 22.5% Selected @ Oct-01 23.5% 22.0% Carried ULAE Paid to Paid Ratio 22.6% Revised ULAE Paid to Paid Ratio 22.5% Note: Adjusted ULAE Counts is a weighted average of Number Recorded and Number Paid. (8) (9) (10) (11) (12) (13) (14) Indicated Carried Reserve # ULAE Loss Reserves Incurred ULAE % Unpaid ULAE Reserve ULAE Reserve Adequacy Reserves ------------- ------------- -------- ------------ ------------ -------- --------- Capped Case 110,400,786 26,496,189 44% 11,658,323 11,822,559 164,237 37,102 IBNR Reserve 22,503,439 5,063,274 95% 4,810,110 4,376,468 (433,642) Total 132,904,224 31,559,462 16,468,433 16,199,028 (269,405) ----------------------------------------- (15) (16) (17) (18) Present Indicated Selected $ Impact ULAE Avgs. ULAE Avgs. ULAE Avgs. ULAE Reserve ---------- ---------- --------------- ------------------ Capped Case 319 314 315 (135,429) IBNR Reserve 19.4% 21.4% 21.0% 349,254 --------------------- Total 213,825 ----------------------------------------- - ----------------------------------------------------------------------------------------------------------------------------------- 43 TERMS AND DEFINITIONS ACCIDENT YEAR LOSSES (number or dollar amount) Losses, which occur in any given 12 month period, are assigned to an accident period. Reserves, regardless of when they are changed or booked, are assigned to that same 12 month period in which the accident occurred. As a result, accident period results will change over time as our estimates of loss costs improve or deteriorate when payments are made or reserves for that accident period are reviewed. ADJUSTMENT EXPENSE Expenses of investigating and settling claims. These include court costs, fees and expenses of adjusters (salaries, wages, overhead, systems), lawyers, witnesses and other expenses that can be charged to a claim. There are two types of adjustment expenses Defense and Cost Containment (DCC) and Adjusting Expenses. ASSUME To accept part or all of a risk transferred by a primary insurer or another reinsurer. BODILY INJURY Physical damage to one's person. The purpose of liability (casualty) insurance is to cover bodily injury to a third party resulting from the negligent or intentional acts and omissions of an insured. CALENDAR YEAR LOSSES (number or dollar amount) Reserve changes and claims payments are recorded in the year the activity takes place without regard to the year in which the accident occurred. CASE RESERVES Estimates of amounts required to settle losses which have already been reported by claimants but have not yet been fully paid. Case reserves represent the largest portion of the reserves for automobile products. There are two types of Case Reserves: 1. Average Reserve: These are determined by the actuaries and are generally limited to losses estimated to be below a specified threshold. Our goal is to use the average reserve approach for as many claims as possible, since using the average reserve allows claims personnel to concentrate on adjusting efforts rather than accounting for them. 2. Adjuster Set: Adjuster Set reserves are the claims adjuster's best estimate of how much a specific loss will cost. Only claims estimates above a specified threshold are included in financial reserves. CATASTROPHE A term applied to an incident, storm or series of related incidents resulting in losses which total more than $25 million in property damage for the insurance industry and affect a significant number of property and casualty insurance policyholders and insurers. CEDE To transfer a risk from an insurance company to a reinsurance company. 44 CLAIM A demand for payment of a policy benefit because of the occurrence of an insured event, such as damage, destruction, injury or death. CLAIMANT One who submits a claim for an incurred loss. CLOSED WITHOUT PAYMENT (CWP) A claim which was originally set up with a reserve that did not require a payment and is now closed. CLOSURE RATE The number of claims closed with payment at each evaluation date divided by the ultimate number of claims to be paid in any given accident period. COLLISION Physical contact of an automobile with another inanimate object resulting in damage to the insured vehicle. COMBINED RATIO A combination of the underwriting expense ratio and the loss and loss adjustment expense ratios. A combined ratio less than 100% indicates an underwriting profit; one in excess of 100% indicates an underwriting loss. COMPREHENSIVE Coverage in automobile insurance providing protection in the event of physical damage (other than collision) or theft of the insured vehicle. For example, fire damage to an insured vehicle would be covered under the comprehensive section of the personal automobile policy. CREDIBILITY A statistical measure of the reliability of experience data, based on the size of the sample. DEFENSE AND COST CONTAINMENT EXPENSE (DCC) A component of LAE, DCC expenses include defense , litigation and medical cost containment whether internal or external. DCC expenses include but are not limited to accident investigation, litigation management expenses, loss adjustment expenses for participation in voluntary and involuntary market pools if reported by accident year, cost of engaging experts, etc. EARNED CAR YEAR Counts of "earned car months" are generated and converted to "earned car years." Earned car months are obtained by taking each vehicle or person insured, and multiplying by the term of the policy. Property damage coverage on one policy for 6 months = 6 car months. (We use the term "car month" even when referring to coverages that protect persons, such as bodily injury or personal injury protection, because coverage is tied to and follows the vehicle, not the person.) Property damage coverage on one policy for 12 months = one earned car year. EARNED PREMIUM The portion of premium paid by an insured that has been allocated to the insurance company's loss experience, expenses and profit year to date. EXPENSE RATIO Expenses associated with the cost of doing business EXCLUDING those expenses associated with losses and LAE, divided by the total earned premium. 45 EXPOSURE The basic rating unit underlying an insurance premium. The unit of exposure will vary based upon the characteristics of the insurance coverage involved. For automobile insurance, one automobile insured for a period of twelve months is a car year or one exposure. A single policy that provides coverage for three vehicles for a six month period is 1.5 car years or 1.5 exposures. FEATURE The smallest divisible part of a claim. This is a loss on one coverage for one person. Often a claim will involve multiple features such as a Bodily Injury (BI) loss and another for Property Damage (PD). FREQUENCY The number of claims divided by the number of exposure counts. This provides a measure of the proportion of insureds who have claims. INCURRED BUT NOT RECORDED (IBNR) These are estimates, during any given reporting period, of amounts that will be needed to settle losses which have already occurred but have not been reported by the claimant. INCURRED LOSSES Those losses which have been recorded and have either been paid or have had reserves established. LOSS ADJUSTMENT EXPENSES (LAE) Costs involved in an insurance company's adjustment of losses. LAE=Defense and Cost Containment + Allocated Loss Adjustment Expenses LOSS RATIO (CALENDAR YEAR LOSS RATIO) (Paid Losses + Paid LAE + Change in Case Reserves + Change in IBNR Reserves + Change in LAE Reserves) divided by Earned Premium. NET LOSS RESERVES "Net" indicates that we have deducted the expected reinsurance recoverable from the sum of case and IBNR reserves. PERSONAL INJURY PROTECTION (PIP) Coverage to pay basic expenses for an insured and his or her family in states with no fault automobile insurance. No fault laws generally require drivers to carry both liability insurance and personal injury protection (PIP) coverage to pay for basic needs of the insured, such as medical expenses, in the event of an accident. POLICY YEAR LOSSES (NUMBER OR DOLLAR AMOUNT) Losses, whenever paid or reserved, are assigned to the effective date of the policy which covers the claim. We also collect rate revision period losses which is a special case of the policy period. For rate revision periods, the losses are assigned according to the policy effective date, but policies written between subsequent rate revisions are grouped together. Because rate revisions are not always completed on an annual basis, the rate revision periods may be longer or shorter than one year. PURE PREMIUM Loss dollars divided by the total exposure count. Pure premium is also equal to frequency times severity. This provides the loss dollar component of the premium per exposure count. 46 RECORD YEAR LOSSES (NUMBER AND DOLLAR AMOUNT) Loss activity, whenever paid or reserved, is assigned to the year the claim is recorded by the company. RELATIVITY The magnitude by which a measure of a subset of data differs (expressed as a quotient) from the same measure for the entire set of data. For example, if the countrywide average paid BI claim was $4,500 and Florida's average BI claim was $5,200, then Florida's paid BI claim relativity would be 1.156. REOPENED CLAIM A claim that was closed (with or without payment) but reopened again at a later date due to the discovery of additional information. We reserve for future reopened claims as IBNR. RESERVES Actuarial estimates of our ultimate liability for losses and loss adjustment expenses which have occurred prior to the end of any given accounting period, but have not yet been paid by the Company. These estimates are influenced by many difficult to quantify variables, such as trend/inflation on automobile costs, changes in claims practices, etc. SALVAGE This reserve is the amount of future recoveries we expect to collect for damaged property previously owned by the insured for which all legal rights have been assigned to the insurance company. Treated as IBNR, salvage is a bulk reserve because there is no feature open. SEGMENT A given combination of states, products, and coverage's with reasonably similar loss characteristics grouped together when assessing reserve adequacy to maximize the data set's credibility. SELECTED The dollar value of the reserve increase or decrease selected by the actuarial staff upon completion of segment review. SEVERITY Loss dollars divided by number of claims. This indicates the dollar amount of the average claim. SUBROGATION Gives an insurer the right to take legal action against a third party responsible for a loss to an insured for which a claim has been paid. THRESHOLD The point at which the dollar value of a claim is large enough to justify having the adjuster set the reserve on the claim, versus an average reserve balance being assigned by the system. TREND FACTOR The purpose of a trend factor is to restate losses from some historic period to levels appropriate in a future period. This is similar to an inflation adjustment. If the trend estimate exceeds the actual trend, we could be over-reserved and will understate earnings in the current period. If the trend estimate is less than the actual trend, risks may be under-reserved and we will overstate earnings in the current period. There are two different trend factors reported and used by Progressive, severity and frequency; the product of which produce a pure premium trend. 47 ULTIMATE The final or farthest possible extent of the analysis. The projected final value of a quantity (e.g. loss, severity, frequency) once development has concluded. UNEARNED PREMIUM RESERVE Written Premium-Earned Premium: That portion of the written premium that has not been exposed to loss and has not been brought into income. These are funds that have been received from the policyholder but for which services have not yet been provided. 48 APPENDIX A THE PROGRESSIVE CORPORATION AND SUBSIDIARIES (SUPPLEMENTAL INFORMATION) (MILLIONS) (UNAUDITED) THREE MONTHS ENDED MARCH 31, 2002 2001 CHANGE -------- -------- --------- NET PREMIUMS WRITTEN(1) Personal Lines-- Agent $1,340.2 $1,075.4 25% Personal Lines-- Direct(1) 585.1 463.9 26% -------- -------- TOTAL PERSONAL LINES(1) 1,925.3 1,539.3 25% Other Lines 261.5 163.5 60% -------- -------- COMPANYWIDE(1) $2,186.8 $1,702.8 28% ======== ======== NET PREMIUMS EARNED Personal Lines-- Agent $1,241.5 $1,132.5 10% Personal Lines-- Direct 517.0 398.4 30% -------- -------- TOTAL PERSONAL LINES 1,758.5 1,530.9 15% Other Lines 209.0 143.0 46% -------- -------- COMPANYWIDE $1,967.5 $1,673.9 18% ======== ======== PERSONAL LINES-- AGENT CR Loss & loss adjustment expense ratio 70.4 77.6 7.2 pts. Underwriting expense ratio 20.4 18.9 (1.5) pts. -------- -------- 90.8 96.5 5.7 pts. ======== ======== PERSONAL LINES-- DIRECT CR Loss & loss adjustment expense ratio 67.2 74.5 7.3 pts. Underwriting expense ratio 23.5 26.7 3.2 pts. -------- -------- 90.7 101.2 10.5 pts. ======== ======== PERSONAL LINES-- TOTAL CR Loss & loss adjustment expense ratio 69.5 76.8 7.3 pts. Underwriting expense ratio 21.3 20.9 (.4) pts. -------- -------- 90.8 97.7 6.9 pts. ======== ======== OTHER LINES-- CR Loss & loss adjustment expense ratio 65.0 69.9 4.9 pts. Underwriting expense ratio 23.3 22.7 (.6) pts. -------- -------- 88.3 92.6 4.3 pts. ======== ======== COMPANYWIDE GAAP CR Loss & loss adjustment expense ratio 69.0 76.2 7.2 pts. Underwriting expense ratio 21.5 21.1 (.4) pts. -------- -------- 90.5 97.3 6.8 pts. ======== ======== 49 COMPANYWIDE ACCIDENT YEAR Loss & loss adjustment expense ratio 69.2 76.7 7.5 pts. ======== ======== ACTUARIAL ADJUSTMENTS-Favorable/(Unfavorable) Prior accident years $ 13.9 Current accident year .2 * -------- -------- CALENDAR YEAR ACTUARIAL ADJUSTMENT $ 14.1 ======== ======== PRIOR ACCIDENT YEARS DEVELOPMENT- Favorable/(Unfavorable) Actuarial adjustment $ 13.9 All other development (10.8) * -------- -------- TOTAL DEVELOPMENT $ 3.1 $ 8.9 NM ======== ======== COMPANYWIDE STATUTORY CR Loss & loss adjustment expense ratio 69.0 76.4 7.4 pts. Underwriting expense ratio 20.3 20.3 -- pts. -------- -------- 89.3 96.7 7.4 pts. ======== ======== MARCH 2002 MARCH 2001 CHANGE ---------- ---------- ------ AUTO POLICIES IN FORCE (IN THOUSANDS) Agent 2,932 2,805 5% Direct 1,302 1,100 18% ----- ----- TOTAL 4,234 3,905 8% ===== ===== (1) Excluding the effect of the commutation of the reinsurance agreement in 2001, net premiums written grew 37% for Direct, 28% for Total Personal Lines and 31% Companywide. * Prior and current accident year actuarial adjustments for 2001 are not provided. 50