The Progressive Corporation (PGR) 8-K Report on Loss Reserving Practices, dated June 2001

Exhibit No. 99

Report on Loss Reserving Practices, dated June 2001

REPORT ON LOSS RESERVING PRACTICES

The Progressive Corporation June 2001

Preface

This report provides a comprehensive discussion of The Progressive Corporation’s loss reserving practices. We frequently receive requests for information that explains our loss reserving practices as well as the methods and tools used to estimate reserve balances for Progressive. This report is specifically targeted to help Progressive’s investors, employees, and interested parties understand how we reserve and how loss reserves affect our results. Analysts and investors who desire a more in-depth understanding of Progressive’s approach to loss reserving may also find this report helpful.

The report’s objectives are to:

1. Explain the importance of loss reserves and the effect loss reserves have on our balance sheet and on our pricing. Explain why loss reserve activity may be different from what was originally expected and how changes have influenced our results over time.

2. Provide an overview of the methodologies used by Progressive to determine reserve balances for all types of reserves. Explain the difference between reserve strengthening, release, and development, and why these differences matter.

3. Provide an open forum for discussion of reserving methodologies and changes we have made to improve reserving accuracy.

4. Provide educational materials to investors and analysts who want to know more about reserving.

At the end of this report we have provided a section entitled “Terms and Definitions”. This section has been provided for the reader who may not be familiar with some of the terminology used in this report.

We welcome feedback, comments and questions. Contact Al Neis, Chief Actuary, The Progressive Corporation, 6300 Wilson Mills Road, Mayfield Village, Ohio 44143 or email at al_neis@progressive.com. Inquiries may also be directed to Sara Stehlik at sara_stehlik@progressive.com.

Table of Contents:

			Page#
Section 1:	About Progressive
		Our Business		1
		Our Financial Objectives		1
		Pricing and Loss Reserving		1
		Loss Reserving: Definition and Stated Goals		2
		Lodestar: Progressive’s Loss Reporting System		3
Section 2:		About Reserves: How Reserves Work:
		Effect on Balance Sheet and Income Statement		4
		Claims: How Aging Affects Reserve Balances		4
		Types of Reserves
		Case Reserves		7
		Incurred But Not Recorded Reserves (IBNR)		8
		Loss Adjustment Expense Reserves		8
		Supplemental Reserve		9
		Additional Needed Case Reserves (ANCR)		9
		Involuntary Market Operating Loss Reserves		9
		Private Passenger Assigned Risk Plan		9
		Commercial Auto Insurance Plan (CAIP)		9
	Trend			11
	Approach to Loss Reserving: Methods and Tools			13
		Segments		13
		Estimation Methods		14
		The Triangle		15
		Reserve Strengthening, Reserve Releasing
		and Reserve Development		15
		Closure Patterns		18
		Changes in Reserves Versus Inflation Versus ANCR		19
		Salvage and Subrogation		20
Section 3:	Reporting
	Segment Results			21
	Calendar Year vs. Accident Year Loss Ratio			22
	The Decomposition Report			23
Section 4:	A Case Study			24
Terms and Definitions				40

Section 1: About Progressive

Our Business

Progressive’s insurance subsidiaries (The Company) write personal automobile and other specialty property-casualty insurance and related services throughout the United States. The Company’s personal lines business consists predominately of insurance for automobiles, motorcycles, recreational vehicles and travel trailers, and is either generated by an agent or written directly by the Company. The personal lines agent business includes premium written by the Company’s network of more than 30,000 independent insurance agencies. The personal lines-direct business is generated through the telephone (1-800-AUTOPRO) or via the Internet (www.progressive.com).

The Company’s other lines of business include insurance for small fleets of commercial vehicles, collateral protection for lenders, liability insurance for directors and officers of financial institutions, and various non-indemity services.

Our Financial Objectives

Progressive has the following financial objectives:

• Profitability: Our most important goal is to produce a calendar year 4 percent GAAP underwriting profit (96 percent Combined Ratio) .

• Return on Shareholders’ Equity: Our goal is to achieve an after-tax return on shareholders’ 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).

• Growth: Our growth goal is to grow as fast as possible constrained only by our profitability objective and our ability to provide high quality customer service.

We believe that the best way to maximize shareholder value is to consistently achieve our financial objectives. The Loss Reserving Function plays an important role in our ability to routinely deliver excellent financial results to our shareholders. Accurate reserving contributes not only to our ability to price risks accurately, but also to our ability to assess future exposures so that we maintain a strong and stable balance sheet.

Pricing and Loss Reserving

One of the most important functions in the insurance business is setting rates, or “Pricing.” The goal is to determine 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 to our customers.

(3) The amount of premium needed to realize our underwriting profit goal.

1

Our underwriting profit goal remains constant at 4 percent of all premium earned or a 96 percent combined ratio. The role of the Pricing Analyst is to predict accurately the amount of premium needed now to cover future expenses and future losses the organization will incur during the period the premium we charge our customers is earned.

The loss reserving function plays an important role in assisting our Pricing Analysts’ determination of premium rates needed to cover the loss component of our business. Because our Pricing Analysts use a cost-plus pricing strategy, and losses represent the most significant part of any rate making analysis, it is important that our loss reserves are estimated as accurately as possible. If the projected ultimate losses are incorrectly stated, the resulting rates could be either too high, thus constraining growth, or too low, thus affecting profitability.

Loss Reserving: Definition and Stated Goals

Loss Reserves represent our best estimates of Progressive’s ultimate liability for losses and loss adjustment expenses which 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 variables that are difficult to quantify such as price inflation on automobile repair costs and medical costs as well as changes in our own internal claims practices, and changes in state regulatory requirements.

Our goal at Progressive is to ensure 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 our stated goals. In the exhibit, you will note that for accident year 1999, loss reserves as estimated at 12/31/99 were deficient by $75.8 million based on information through 12/31/2000. This was caused by dramatic unanticipated increases in repair and medical costs during the second half of 1999. This will be addressed in more detail later in this document.

Exhibit 1

Loss and LAE Reserves: History of Performance

Goal: Ensure reserves are adequate to cover all loss costs while sustaining minimal variation from the time reserves are initially established until they are fully developed.

Historically, How do we measure up?

Percentage Adequate by Year	1990			1991			1992			1993			1994			1995			1996			1997			1998			1999
Cumulative Adequacy	$	131.20		$	171.00		$	241.80		$	223.80		$	146.00		$	161.10		$	81.90		$	194.40		$	35.20		$	(75.80	)
Percentage Adequate		15.9	%		19.9	%		25.3	%		22.1	%		13.3	%		12.3	%		5.3	%		10.4	%		1.8	%		(3.4	%)
Supplemental Reserve Carried *	$	59.8		$	65.6		$	73.1		$	73.1		$	71.0		$	0.0		$	0.0		$	0.0		$	0.0		$	0.0
Cumulative Adequacy w/o Supp Resv	$	71.4		$	105.4		$	168.7		$	150.7		$	75.0		$	161.1		$	81.9		$	194.4		$	35.2		$	(75.8	)
Percent Adeq/(Def) w/o Supp Resv		9.8	%		13.2	%		19.1	%		16.0	%		7.3	%		12.3	%		5.3	%		10.4	%		1.8	%		(3.4	%)
Total Loss Reserves
Total Loss Reserves	$	791.6		$	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
Total Loss Resv w/o Suppl Reserve	$	731.8		$	795.9		$	883.3		$	939.3		$	1,027.7		$	1,314.4		$	1,532.9		$	1,867.5		$	1,945.8		$	2,200.2

* Progressive discontinued the practice of carrying a supplemental reserve in 1994. More discussion follows in the section entitled “Types of Reserves”.

2

LODESTAR: Progressive’s Loss Reporting System:

Our current data reporting system is known as the LOss DEvelopment STatistical Analysis Reporting System (LODESTAR). This system was implemented in 1987 specifically for loss reserving, but has several side benefits for pricing applications as well. Major features include:

• The data entering LODESTAR is balanced to the data entering the accounting systems on a monthly basis so we are sure that our loss data reconciles to our financial results.

• Significant flexibility in the data configurations which can be grouped on a monthly, quarterly, semiannual, or annual basis.

• Options for segmenting the data by company, state, product (e.g. private passenger automobile, commercial automobile), coverage (e.g. bodily injury, collision) and limit of liability are easy to define.

      Five report configurations are readily available:

1. Period Reports:These reports allow cumulative experience to be grouped by accident period, record period, policy effective period, or rate revision (when prices changed) period.

2. Record Within Accident Period Reports:With this option, data can be organized both by the date of the accident and by the date the loss was recorded following the accident. Data reported in this format allows the user to observe the proportion and development of claims recorded during the accident period, or one period, two periods, etc. after the accident.

3. Size of Loss Reports:These reports segment data by the size of the amount which is paid, incurred, or reserved. This report offers the flexibility of including or excluding losses according to the accident date and record date, and can display the same losses evaluated at different points in time

4. Reserve Runoff Reports:These reports identify the case and IBNR reserves that were open at a given date in the past (the runoff date) and display their development over time.

5. Case Reserve by Age Runoff Reports:In order to produce age-sensitive average reserves, this report separates case reserves at each runoff date according to the number of months between accident date and runoff date.

3

Section 2: About Reserves

How Reserves Work

Effect on Balance Sheet and Income Statement

Following is a theoretical example of the how a claim works from a reserving perspective and its effect on our financial results:

Assume $1000 in earned premium over the policy term. Accident occurred on 5/15/2000 but was not reported to the company until 5/17/2000. The loss was settled for same amount reserved, $700, and paid to the claimant/policyholder on 5/20/2000.

Date: 5/15/2000	Date 5/17/2000	Date: 5/20/2000
Date Accident Occurred	Date Accident Reported	Date Claim Paid
IBNR reserve for $700 previously accrued on balance sheet in anticipation of the accident occurring. This $700 reserve also affects the income statement	Create case reserve on balance sheet in the amount of $700. Claim has now been recorded by the company. Release IBNR reserve, as accident has now been reported to the company.	Pay policyholder $700. Release case reserve on balance sheet since loss has now been paid to claimant.
pure loss ratio* 70%	pure loss ratio 70%	pure loss ratio 70%

* pure loss ratio= (paid losses+change in case reserves+change in IBNR reserves) divided by earned premium.

The pure loss ratio never changes, however, the liability moves on the balance sheet from IBNR (Incurred But Not Recorded) to case reserves once it has been recorded by the company. On 5/20/2000 the loss payment is made to the claimant so we remove the case reserve (liability) from the balance sheet and close the claim.

Claims: How Aging Affects the Reserve Balance

The age of a claim affects the reserving process and ultimately, the cost to the company. The size of a loss payment generally will vary based on the amount of time that it takes to settle the claim following the occurrence. We refer to this as the “age of the claim.” Claims that take longer to settle generally cost the company more. The following model has been repeatedly published and continues to be a good illustration of why the correlation between claim age and payment amount is important. If claim age and payment amount were unrelated, we could simply multiply the number of open claims by the average loss payment over the past twelve months and determine a reserve.

4

The purpose of the model below is to demonstrate that a population of open claims has different properties than a population of closed claims:

For simplicity, assume that there are only two types of claims:

      1. Low-severity, quickly-settled, frequent claims

      2. High-severity, slowly-settled, infrequent claims

	FREQUENCY		SEVERITY		TIME TO SETTLE
Claim Type #1		1 per month		100		1 month
Claim Type #2		1 per year		1,000		12 months

      To Illustrate:

X = Claim is recorded

O = Claim is settled

Claim type #1

Jan 00

Feb 00

Mar 00

Apr 00

May 00

Jun 00

Jul 00

Aug 00

Sep 00

Oct 00

Nov 00

Dec 00

X—

O

X---

O

X---

O

X---

O

X---

O

X---

O

X---

O

X---

O

X---

O

X---

O

X---

O

X---

5

Claim Type #2

Jan 00	Feb 00	Mar 00	Apr 00	May 00	June 00	Jul 00	Aug 00	Sep 00	Oct 00	Nov 00	Dec 00
									O
									X

In reviewing the two tables above, you will note that throughout the example year there is one type #1 claim and one type #2 claim open. Using average cost, our reserves would be $338* however, our reserves should be $1,100 (1 claim at $100 and 1 claim at $1000)

* 12 claims settled at $100 =	$1,200
1 claim settled at $1000 =	$1,000
total claims dollars paid	$2,200
divided by total number of claims paid	13
	$   169	average cost per claim
number of open claims	2
average cost	$   338

The model shown above is fundamental however, it does exhibit how using an average cost method for reserving purposes could misstate our reserve liability. Further discussion about how reserves work, the different types of reserves and tools used to estimate reserves can be found in subsequent sections of this document.

6

Types of Reserves

Progressive has defined six different types of loss reserves. Each needs to be reviewed separately in order to get the total reserve balance as accurate as possible:

1. Loss Case Reservesrepresent 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 and every claim regardless of individual claim characteristics. These reserves are determined by the actuaries. As the claim ages (the length of time from the accident date to the current accounting period), the value of the claim will generally increase as history has proven that severity of a claim generally increases with age. The averages vary by company, state, product, and line coverage.

Generally, an average reserve is based on our estimate of total needed reserves excluding the adjuster set reserves (defined immediately below). 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 will now use average, actuarially-set reserves as opposed to adjuster set reserves. We estimate that between 15 percent to 20 percent of our automobile bodily injury loss reserve dollars will now be based on adjuster set reserves.

Adjuster-Set Reserve: We include the adjuster estimates of how much a claim will cost as the reserve when the adjuster-set reserve provides a more reliable estimate than statistical averages. Adjuster set reserves are primarily used when the claim estimate is greater than or equal to the established threshold discussed above.

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 feel 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, the reserve changes arising from our analyses only reflect shifts in anticipated development patterns. This also contributes to the consistency in data we need when rate making.

Loss case reserves represented approximately two thirds of our total reserve balance as of 12/31/2000.

7

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. We track IBNR emergence to known statistics (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. Given accurate rates, earned premium has advantages over other measures; it reflects changes in expected loss costs due to changes in mix of business and reflects the impact of inflation to the extent that inflation is reflected in our pricing.

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.

Loss IBNR represented about one sixth of our total reserve balance as of 12/31/2000.

3. Loss Adjustment Expense (LAE) Reservesare 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) formerly known as Allocated Loss Adjustment Expense (ALAE) Reserves: These are costs associated with defense, litigation, and cost containment.
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 company, state, product and line coverage.

LAE reserves represented approximately one sixth of our total reserve balance as of 12/31/2000: (about half of our LAE reserves are DCC reserves, while the other half are Adjusting Expense reserves).

8

4. Supplemental Reserve:From 1971 until 1994, Progressive maintained a special supplemental reserve. During the early and mid-1970’s, we experienced unusual fluctuations in frequency and/or severity as the amount of earned premium we wrote wasn’t large enough to produce credible reserve indications. By 1977, the increased size of the Company reduced the need for this reserve. Between 1977 and 1994, the company carried a supplemental reserve to achieve the objective that carried reserves were adequate at a targeted confidence level. In 1994, the supplemental reserve, which totaled $71 million at that time, was released into income. The company has not carried a supplemental reserve since.

5. Additional Needed Case Reserves (ANCR): Additional Needed Case Reserve is a bulk 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). A further explanation of this reserve is included later in this report. Many companies include ANCR in their IBNR estimates.

6. Involuntary Market Operating Loss Reserves (Assigned Risk Reserves): 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.

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

9

cede 100 percent of the business to the state CAIP plans. The state plans then retrocede 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 private passenger auto, 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.

Assigned Risk reserves represented under one half of one percent (.05 percent) of our total reserve balance as of 12/31/2000.

10

Trend

Trend estimates allow us to anticipate future changes in settlement costs due to the passing of time from some historical payment date to a prospective payment date. We compare old data to new data to determine 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 and changes in labor rates and car part prices. 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 period, similar to an inflation adjustment. If the trend factor used to estimate reserves exceeds the actual trend experienced, we could be over reserved. If the trend factor used to estimate reserves is less than the actual trend experienced, risks may be under reserved.

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, 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) in order to understand the trend differences by tier. This is because Progressive’s trend results historically have been more heavily influenced by the nonstandard market. This extra step provides us with a better indication of frequency trend because the preferred and ultra preferred tiers are growing more rapidly and may disguise changes in trend in the other tiers. Pricing analysts 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 then 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

11

increases and decreases) and changes in population density. Severity trend tends to be affected by changes in Consumer Price Index, medical costs, and repair costs.

We are more concerned with BI trend when evaluating reserve adequacy. Injury claims (e.g. BI) are long-tailed and require more data points while property damage (e.g. PD) coverages are short-tailed and require fewer data points. The chart immediately following this paragraph illustrates this fact. BI reserves are a much larger portion of our total reserves than the short-tail coverages such as PD, and historically take longer to fully settle.

      [GRAPH]

12

Approach to Loss Reserving

Segments:Progressive generally starts by reviewing six different estimation methods, described on the next page, to determine needed levels of reserves. 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
Segment #1		New York		Private Passenger		Personal Injury
				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. Currently, there are over 240 independent reserve segments. Segments can be reviewed annually, semiannually, quarterly, or monthly. 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 and may be reviewed using countrywide information. 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’s line coverage 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.

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.

Because we review Loss Reserves at a very refined level, we do not find it necessary to engage a loss reserve committee. Instead, actuarial analysts are responsible, based on the results of their analyses, to either strengthen or release reserves for each segment under review. We then measure our Chief Actuary by the accuracy of our reserve liability balance on an aggregate level. We do this by comparing the accident-year loss ratio to the calendar-year loss ratio looking for minimal variation between the two results. The distinction between calendar-year loss ratio and accident-year loss ratio can be found under Section 3 of this document.

13

Estimation Methods Defined:

We use these estimation methods during a segment review:

1. 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.

2. Average Paid:Paid severity by accident period 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-CWP rate). CWP rate is the percentage of claims recorded but closed without payment. This method uses three triangles of data, paid severity by accident period, CWP rate, and features recorded.

3. 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.

4. Average Incurred:Incurred severity by accident period 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 projected CWP rate). This method uses three triangles: Incurred Severity, CWP rate, and number of recorded feature counts.

5. 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.

6. 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.

14

The Triangle:

The triangle is a tool used by actuaries to show how estimates of incurred or paid-to-date amounts have changed or “developed” over time. Development can be defined as the estimate of losses incurred on business produced in previous years at subsequent evaluation dates. Data from the 1990 accident year has had eleven years to develop while data from the 1998 accident year has only had three years to develop. Exhibit 2 is an example of a triangle that appears in Progressive’s 2000 Annual Report. This triangle represents the loss and LAE reserves for 1990 through 2000. The reserves are re-estimated based on the experience as of the end of each subsequent year, and are increased or decreased as more information becomes available about the frequency or severity of claims reserved for as of each year end. The cumulative deficiency or conservative runoff represents the aggregate change in the estimates over all prior years.

Exhibit 2
Analysis of Loss and Loss Adjustment Expenses (LAE) Development

For Years ending																															(millions)
Dec. 31,	1990			1991			1992			1993			1994			1995			1996			1997			1998			1999			2000
Loss and LAE reserves	$	791.6		$	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
Re-estimated reserves as of:
One Year Later		748.8			810.0			857.9			869.9			1,042.1			1,208.6			1,429.6			1,683.3			1,916.0			2,276.0
Two Years Later		726.5			771.9			765.5			837.8			991.7			1,149.5			1,364.5			1,668.5			1,910.6
Three Years Later		712.7			718.7			737.4			811.3			961.2			1,118.6			1,432.3			1,673.1
Four Years Later		683.7			700.1			725.2			794.6			940.6			1,137.7			1,451.0
Five Years Later		666.3			695.1			717.3			782.9			945.5			1,153.3
Six Years Later		664.8			692.6			711.1			780.1			952.7
Seven Years Later		664.5			688.2			709.2			788.6
Eight Years Later		661.4			687.9			714.6
Nine Years Later		660.4			690.3
Ten Years Later		665.9
Cumulative redundancy/(deficiency)	$	125.7		$	171.2		$	241.8		$	223.8		$	146.0		$	161.1		$	81.9		$	194.4		$	35.2			($75.8	)
Percentage		15.9	%		19.9	%		25.3	%		22.1	%		13.3	%		12.3	%		5.3	%		10.4	%		1.8	%		(3.4	%)

Reserve Development, Reserve Strengthening and Reserve Releasing:

At this point, it may be worthwhile to discuss the difference between reserve development, reserve strengthening and reserve releasing, and why these differences matter. Reserve development is the process of settling a claim for either less or more than the established reserve for that claim. Adverse development will result when a claim is settled for more than the estimated reserve balance. Favorable development will result when a claim is settled for less than the estimated reserve balance.

Reserve strengthening (increasing reserves) or reserve releasing (decreasing reserves) happens when the loss reserving group, after a segment review, determines that the accrued reserves are not enough to pay ultimate losses (requiring an increase in reserves), or the accrued reserves are in excess of what we need (allowing a decrease in reserves). In either case, authority to

15

strengthen or release reserves is an activity limited to the loss reserving staff and is only done after a thorough segment review has been completed.

Exhibit 3, shown on the following page, is an example of an another triangle used to estimate the ultimate paid losses that will occur from a given segment (please note: this triangle is organized differently than exhibit 2. Exhibit 2 is organized from top to bottom and the following triangle is organized from left to right). The complete triangle is shown, including all 14 semiannual periods that are typically reviewed. This is an example of a triangle previously used to project the estimated ultimate costs showing paid losses at the end of each accident period under review. The method used to develop the results is the paid ultimate method. We’ve highlighted several areas on the triangle to show how factors are calculated.

Every insurance company is required by the National Association of Insurance Commissioners (NAIC) to file certain exhibits in the company’s annual statement. Exhibit 4 below displays the Schedule P Part 3 triangle that we reported in the 2000 Statutory Annual Statement for the Progressive Casualty Insurance Company. You’ll notice in exhibit 4 that the column headings are accident year values as of the calendar year end where in exhibit 3, the headings represent evaluation periods at any given point in time. We show this only to emphasize that the construction of a triangle differs depending on its intended use.

Exhibit 4
Schedule P-Part 3B Private Passenger Auto Liability/Medical

Years in																						Number of		Number of
Which																						Claims		Claims
Losses																						Closed with		Closed
Were																						Loss		Without
Incurred																						Payment		Loss
		1991		1992		1993		1994		1995		1996		1997		1998		1999		2000				Payment
Prior			0		40,537		56,709		63,540		67,047		69,340		70,079		70,532		70,687		71,040				1
	1991		90,260		157,397		184,497		194,937		198,898		200,873		201,585		202,043		202,212		202,556		67,625		31,810
	1992		xxx		106,061		182,460		213,330		224,035		228,566		230,820		231,864		232,464		232,232		76,497		35,451
	1993		xxx		xxx		130,055		228,728		261,886		275,119		281,305		283,801		284,997		285,967		94,488		44,564
	1994		xxx		xxx		xxx		192,737		332,662		375,268		395,947		404,645		408,841		409,662		133,752		63,877
	1995		xxx		xxx		xxx		xxx		254,856		412,615		471,115		498,254		512,371		517,048		166,477		80,689
	1996		xxx		xxx		xxx		xxx		xxx		271,334		451,285		518,831		551,107		566,727		187,684		96,336
	1997		xxx		xxx		xxx		xxx		xxx		xxx		341,959		567,699		658,458		694,128		234,158		135,908
	1998		xxx		xxx		xxx		xxx		xxx		xxx		xxx		431,411		716,145		810,644		283,181		166,565
	1999		xxx		xxx		xxx		xxx		xxx		xxx		xxx		xxx		557,472		892,267		335,710		237,982
	2000		xxx		xxx		xxx		xxx		xxx		xxx		xxx		xxx		xxx		627,785		287,513		254,496

16

Exhibit 3
Paid Ultimate Method

Semiannual
Accident	PAID LOSSES (000)
Periods
Ending	1			2			3			4			5			6			7			8			9			10			11			12			13			14
Jul-93		4,323			9,837			11,160			11,662			11,875			11,966			12,026			12,069			12,095			12,112			12,121			12,135			12,155			12,169
Jan-94		4,660			11,648			13,479			14,059			14,342			14,503			14,622			14,688			14,719			14,781			14,812			14,838			14,856
Jul-94		6,480			14,852			17,036			17,710			18,029			18,230			18,323			18,395			18,469			18,528			18,547			18,613
Jan-95		7,216			17,158			19,621			20,346			20,755			20,959			21,128			21,260			21,429			21,477			21,631
Jul-95		7,688			17,606			19,881			20,780			21,289			21,504			21,738			21,859			21,984			22,068
Jan-96		6,439			15,323			17,990			18,892			19,334			19,630			19,790			19,892			20,037
Jul-96		6,413			15,686			17,798			18,642			19,158			19,469			19,675			19,868
Jan-97		6,589			15,109			17,052			17,799			18,249			18,561			18,876
Jul-97		7,145			16,998			19,124			20,078			20,610			21,106
Jan-98		9,703			23,615			27,043			28,516			29,474
Jul-98		12,416			29,964			34,061			35,744
Jan-99		12,918			35,071			39,649
Jul-99		17,223	A		43,214	B
Jan-00		21,487
		1-2			2-3			3-4			4-5			5-6			6-7			7-8			8-9			9-10			10-11			11-12			12-13			13-14
Jul-93		2.275			1.134			1.045			1.018			1.008			1.005			1.004			1.002			1.001			1.001			1.001			1.002			1.001
Jan-94		2.500			1.157			1.043			1.020			1.011			1.008			1.005			1.002			1.004			1.002			1.002			1.001
Jul-94		2.292			1.147			1.040			1.018			1.011			1.005			1.004			1.004			1.003			1.001			1.004
Jan-95		2.378			1.144			1.037			1.020			1.010			1.008			1.006			1.008			1.002			1.007
Jul-95		2.290			1.129			1.045			1.025			1.010			1.011			1.006			1.006			1.004
Jan-96		2.380			1.174			1.050			1.023			1.015			1.008			1.005			1.007
Jul-96		2.446			1.135			1.047			1.028	D1		1.016			1.011			1.010
Jan-97		2.293			1.129			1.044			1.025	D2		1.017			1.017
Jul-97		2.379			1.125			1.050			1.027	D3		1.024
Jan-98		2.434			1.145			1.054			1.034	D4
Jul-98		2.413			1.137			1.049
Jan-99		2.715			1.131
Jul-99		2.509	C
Straight Avg		2.408			1.141			1.046			1.024			1.014			1.009			1.006			1.005			1.003			1.003			1.002			1.001			1.001
Avg x HiLo		2.392			1.139			1.046			1.023			1.013			1.008			1.005			1.005			1.003			1.002			1.002
Wtd Avg All		2.436			1.139			1.047			1.025			1.014			1.009			1.006			1.005			1.003			1.003			1.002			1.001			1.001
Avg Last 4		2.518			1.134			1.049			1.028	E		1.018			1.012			1.007			1.006			1.003			1.003
Wt Avg.4		2.523			1.135			1.050			1.029			1.018			1.012			1.007			1.006			1.003			1.003
Prior Select
Select		2.450			1.125	F		1.045			1.025			1.015			1.011			1.006			1.005			1.003			1.003			1.002			1.001			1.001			Tail
Cumulative		3.094			1.263	H		1.122	G		1.074			1.048			1.032			1.021			1.015			1.010			1.007			1.004			1.002			1.001			1.000
Ultimate Loss		66,473			54,568			44,503	J		38,392			30,886			21,791			19,275			20,167			20,238			22,223			21,718			18,650			14,871			12,169
Ultimate LR		98.0	%		87.7	%		79.3	%		72.3	%		67.0	%		57.4	%		56.8	%		59.2	%		60.0	%		66.6	%		68.4	%		63.3	%		60.1	%		58.8	%
Ultimate PP		165			147			135			125			119			107			107			112			111			120			124			99			60			58
Avg. Ult. Loss		5,848			5,294			4,967			4,770			4,558			4,341			4,267			4,359			4,294			4,403			4,304			4,047			3,967			3,844
Sum all of the values in the Ultimate Loss Row will determine the Ultimate Amount that will be paid for this segment																															$	405,925

Calculations:
      C = B/A: Factor Used to Develop Next Periods Losses
      H = F*G: Cumulative Factor used to multiply losses in most recent period to determine ultimate losses for that Period
      E = (D1+D2+D3+D4)/4: Average of most recent four periods
      J = B*G: Ultimate Loss Dollars for the third Accident Period

17

Closure patterns and their effect on reserves:

During all loss reserving reviews and for each segment(s) under review, we analyze the history of the closure rate (the number of paid claims closed with payment at each evaluation date divided by the estimated ultimate number of incurred claims), as well as the closed without payment rate for the data set (segment) under review.

Closure Rate:When closure rates change, we may pay closer attention to the results generated from the three incurred estimation methods previously discussed. If we see the closure rates decreasing, the results of the paid methods may be understated. The converse is also true: An increase in the closure rate may indicate that the recommended reserve balances generated by the paid methods are overstated.

Closed Without Payment (CWP): There are various reasons a claim is reported and recorded on the company’s records, but is ultimately closed without a claim payment made:

1. The policy was not in force at the time of the accident.
2. Coverage for the claim does not exist.
3. Damage caused to the vehicle is less than the deductible amount listed on the policy.
4. The company has determined that no liability exists.

During a segment review, we run a triangle reporting the CWP rate showing 14 semiannual accident periods to determine what percentage of the ultimate number of reported losses have been closed without payment. When we see significant differences in the CWP rate from what we expected, we further analyze the results to determine if the number of claims recorded has declined, if claims are getting recorded more quickly than they have in the past, or whether the claims organization is identifying those claims that we will ultimately close without payment more quickly than they have in the past. An example of a CWP triangle with actual data can be reviewed in Exhibit 5 along with an explanation.

18

Sample State: Sample Coverage

Exhibit 5

Semiannual
Accident	CLOSED WITHOUT PAYMENT RATE (Features closed without Payment / # Recorded) - ACCIDENT PERIOD ANALYSIS
Periods
Ending	1			2			3			4			5			6			7			8			9			10			11			12			13			14			Ultimate
Oct-94		21.3	%		24.5	%		25.6	%		26.2	%		26.2	%		26.3	%		26.7	%		26.7	%		26.7	%		26.7	%		26.7	%		26.7	%		26.8	%		26.8	%		26.8	%
Apr-95		19.4	%		23.3	%		23.9	%		24.0	%		24.3	%		24.5	%		24.5	%		24.5	%		24.6	%		24.6	%		24.6	%		24.6	%		24.6	%					24.6	%
Oct-95		21.3	%		25.0	%		25.7	%		26.1	%		26.6	%		26.8	%		26.9	%		27.0	%		27.1	%		27.1	%		27.1	%		27.1	%								27.2	%
Apr-96		18.9	%		23.2	%		24.1	%		24.5	%		24.6	%		24.7	%		24.9	%		24.9	%		25.1	%		25.1	%		25.1	%											25.1	%
Oct-96		20.8	%		24.9	%		25.8	%		26.5	%		26.9	%		27.0	%		27.1	%		27.2	%		27.2	%		27.2	%														27.3	%
Apr-97		19.4	%		23.4	%		24.4	%		24.8	%		25.0	%		25.2	%		25.4	%		25.5	%		25.5	%																	25.6	%
Oct-97		22.7	%		25.6	%		26.7	%		27.0	%		27.1	%		27.2	%		27.3	%		27.3	%																				27.5	%
Apr-98		19.9	%		23.6	%		25.0	%		25.3	%		25.6	%		25.7	%		25.7	%																							25.9	%
Oct-98		18.1	%		22.2	%		22.8	%		23.2	%		23.8	%		24.0	%																										24.3	%
Apr-99		20.8	%		24.4	%		25.4	%		26.1	%		26.3	%																													26.8	%
Oct-99		24.6	%		28.2	%		29.0	%		29.3	%																																30.2	%
Apr-00		27.8	%		31.7	%		32.6	%																																			34.1	%
Oct-00		29.3	%		32.3	%																																						34.8	%
Apr-01		27.5	%																																									33.2	%
	1-2			2-3			3-4			4-5			5-6			6-7			7-8			8-9			9-10			10-11			11-12			12-13			13-14
Oct-94		1.150			1.044			1.024			0.999			1.003			1.015			1.000			1.000			1.002			1.000			1.000			1.005			1.000
Apr-95		1.197			1.025			1.008			1.012			1.006			1.000			1.002			1.005			1.000			1.000			1.000			1.000
Oct-95		1.174			1.029			1.015			1.018			1.009			1.002			1.004			1.002			1.001			1.000			1.000
Apr-96		1.229			1.037			1.014			1.008			1.003			1.006			1.002			1.006			1.000			1.000
Oct-96		1.193			1.039			1.025			1.016			1.006			1.004			1.001			1.003			1.000
Apr-97		1.210			1.039			1.020			1.007			1.006			1.008			1.005			1.000
Oct-97		1.129			1.041			1.012			1.005			1.003			1.005			1.000
Apr-98		1.187			1.059			1.010			1.012			1.005			1.001
Oct-98		1.228			1.027			1.016			1.027			1.007
Apr-99		1.173			1.040			1.028			1.009
Oct-99		1.149			1.027			1.011
Apr-00		1.138			1.029
Oct-00		1.102
Straight Avg		1.174			1.036			1.017			1.011			1.005			1.005			1.002			1.003			1.001			1.000			1.000			1.002			1.000
Avg x HiLo		1.175			1.035			1.016			1.011			1.005			1.004			1.002			1.002			1.000			1.000			1.000
Wtd Avg All		1.169			1.036			1.017			1.011			1.005			1.005			1.002			1.003			1.001			1.000			1.000			1.003			1.000
Avg Last 4		1.140			1.031			1.016			1.013			1.005			1.004			1.002			1.003			1.000			1.000
Wt Avg.4		1.137			1.031			1.016			1.013			1.005			1.004			1.002			1.003			1.000			1.000
Fitted		1.120			1.029			1.014			1.001			1.005			1.004			1.002			1.001			1.000			1.000			1.000			1.006			1.000
Prior Select
Select		1.120			1.031			1.016			1.013			1.005			1.004			1.002			1.003			1.000			1.000			1.000			1.003			1.000			Tail
Cumulative		1.210			1.080			1.048			1.031			1.018			1.012			1.008			1.006			1.003			1.003			1.003			1.003			1.000			1.000
	Apr-01			Oct-00			Apr-00			Oct-99			Apr-99			Oct-98			Apr-98			Oct-97			Apr-97			Oct-96			Apr-96			Oct-95			Apr-95			Oct-94
Ult CWP Rate		33.2	%		34.8	%		34.1	%		30.2	%		26.8	%		24.3	%		25.9	%		27.5	%		25.6	%		27.3	%		25.1	%		27.2	%		24.6	%		26.8	%
Ult Counts		2,828			3,079			2,635			2,467			2,028			2,022			1,551			1,560			1,477			1,691			1,902			2,251			2,046			1,669
Ult. Frequency		2.49	%		2.94	%		2.86	%		3.14	%		2.95	%		3.30	%		2.84	%		3.27	%		2.97	%		2.85	%		3.00	%		3.30	%		3.28	%		3.55	%

In exhibit 5 the CWP rate for recent accident periods is up sharply compared to historical figures. This could be due to several possible factors. Most likely, there has been a change in the way claims have been reported: The increased use of a 24-hour call centers for claims reporting may result in more claims being recorded/opened and then the closed without payment more quickly. The recorded count triangle (not exhibited here) shows an increase in frequency, the closed without payment count triangle also shows an increase in frequency, but the incurred count triangle (which is equal to recorded counts - CWP) shows relatively stable frequency. Therefore, we would conclude that there is no significant change to our true incurred count pattern, which is the more important measure in terms of our results.

Changes in Reserves versus Inflation versus ANCR:

Reserve balances change for various reasons. As previously discussed, we will initiate a change to a reserve balance at the completion of a segment review based on the data. We do not review each segment every month. We base our review schedule in part on the potential effect reserve changes may have on our balance sheet (each segment is reviewed at least annually). Because each segment is not reviewed monthly, in an effort to keep reserve balances at or close to their appropriate levels, we apply an inflation factor to the existing average reserves during those

19

months that a segment is not reviewed. We do this because we know that, in general, the claims that remain unpaid longer are more costly to the company.

Additional Needed Case Reserve (ANCR) is a bulk reserve intended to cover further development of the adjusters estimates for large losses, particularly for losses associated with policies that have higher limits (in excess of the threshold). Increases to the incurred loss estimates over the life of a portfolio of large losses will frequently occur as new facts emerge during an investigation of the claims necessitating an ANCR reserve. The ANCR reserve generally will diminish as the pool of claims ages because we become more educated on the exact nature and extent of the cost of these claims. In the industry, the tendency for case reserves to develop may be treated as part of IBNR. Progressive prefers to call the further case development “Additional Needed Case Reserve” since the claims in question are known to the company.

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.

We periodically study the ratio of salvage and subrogation recoverable to total loss reserves to identify any changes in process that may have occurred.

20

Section 3: Reporting

Reporting Segment Review Results All segment reviews are reported internally monthly using a standard format. Completed results of each review are provided to our pricing organization, general managers, and others in the organization as needed. A summary of each review is also completed, highlighting the results of each method used to review reserves, the threshold of each segment, the ANCR factor (if applicable), as well as the selected ultimate reserve need. A subsection of that report is shown below in Exhibit 6 along with definitions for each field reported.

Exhibit 6

This report provides a summary of the results generated after a segment review has been completed by the Actuaries. Once all the indications have been prepared by the Actuaries, a “selected ultimate” indication is chosen which drives the change or impact to that segment’s carried reserve levels for the month.

Below is an example of the results generated during a segment review and definitions of the corresponding fields reported.

Example:

ACCIDENT ANALYSIS
All Dollars in Thousands (000)

Example:

ACCIDENT ANALYSIS

All Dollars in Thousands (000)

Review Date

Month/Year

(A)

(B)

(C)

(D)

Nxt

Bodily

Inflation

$ Resv

Review

Coverage Type

Injury Auto

Threshold

$XX.XX

Factor

7.00%

held before

$4,345

Date

Jul-2001

State

State Name

ANCR

$ Resv

% > Threshold

5.3%

Factor

1.064

held after

$3,955

(E)

(F)

(G)

Indication

Results (000)

1

2

3

4

5

6

Amt Paid

as of

$ Reserve

ULT Inc

BF Incurred

ULTIMATE

ULT paid

BF Paid

Ultimate

Selected

Review

Adequacy

$ IMPACT

% to

Avg Losses

Ultimate

Amt Inc.

Avg. Losses

Ultimate

Amt Paid

Ultimate

date

+/(-)

+/(-)

Indicated

$

112,163

$

112,097

$

112,161

$

110,906

$

110,826

$

110,807

$

111,994

$

108,596

$

947

($390

)

40

%

(H

)

(I

)

(J

)

(K

)

(L

)

(M

)

(N

)

(O

)

(P

)

(Q

)

(R

)

Definition of Fields

(A) Threshold	The point at which the adjuster’s estimate is used as our carried financial reserve for P&L purposes.
(B) Inflation Factor	The annual percentage rate that the financial reserves are inflated monthly.
(C) $ Resv held before	Total reserves held for the segment prior to the review and based on prior month end balance.
(D) Nxt Review Dt	Date next review of this segment will be performed
(E) % > Threshold	The total dollar amount of reserves that exceed the threshold divided by the total reserve balance for the entire segment.
(F) ANCR Factor	A factor used to spread the bulk reserve (also referred to as ANCR). ANCR is a reserve to cover future expected case reserve development. on large losses. ANCR reserves generally diminish as the claim ages and we become more educated on the exact nature and extent of the claim. ANCR factors vary by age of claim and by state. The factor reported above is a weighted average for the state. ANCR factors are generally only reported for the BI segments as claims in excess of the Threshold are more infrequent in the other line coverage’s.
(G) $ Resv held after	Total reserves held for the segment after change is implemented by the actuarial department.
INDICATION RESULTS
(H) ULT INC Avg Losses	Incurred severity by accident period projected to an ultimate level multiplied by the projected ultimate incurred feature counts for the same accident period.
(I) BF Incurred Ultimate	Bornhuetter-Ferguson Incurred Ultimate. We use the incurred loss development pattern to determine the percent (%) unreported. We then apply the % unreported to the expected ultimate loss amount to arrive at the expected unreported amount. The unreported amount is added to the actual incurred to date.
(J) Ult Amt Inc.	Total dollars incurred by age and accident period projected to an ultimate level.
(K) Ult Paid Avg Losses	Paid severity by accident period is projected to ultimate level and multiplied by ultimate paid feature counts. This is the same as the incurred average projection using paid data.
(L) BF Paid Ultimate	Bornhuetter-Ferguson Paid Ultimate. We use the paid loss development pattern to determine the percent (%) unpaid. We then apply the expected unpaid percentage to the expected ultimate loss amount to arrive at the expected unpaid amount. The unpaid amount is added to the actual paid to date.
(M) Ultimate Amt Paid	Total loss dollars paid by age and accident period are projected to an estimated ultimate level.
(N) Selected Ultimate	The resulting indicated ultimate loss value for the accident periods reviewed based upon the selections made by the actuarial staff.
(O) Amt Paid as of Rev date	The total dollars already paid to claimants for the accident periods reviewed.
(P) $ Reserve Adeq	(4)-(15)=(17) or $ Resv held before minus Selected Ultimate
(Q) $ Impact	Amount of Reserve Increase or Decrease to segment under review
(R) % to indicated	% of the total indicated change selected by the actuaries that was actually implemented.

21

Calendar Year vs. Accident Year Loss Ratio:

Understanding the difference between a calendar year loss ratio and an accident year loss ratio is important: 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 internally to our pricing organization and our management team both accident year and calendar year results by state each month. What follows is an explanation of each.

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.

Accident Year Loss Ratio:We 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 therefore change over time as our estimates of loss costs improve or deteriorate. To determine the Accident Year Loss Ratio, we add estimated 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.

22

Decomposition Report:The Decomposition Report is used frequently to review activity of our reserves both on a state level and on a consolidated level. This report is prepared monthly and provides an inventory status of the impact on all claims features and their reserve changes by type of change. A sample of this report can be seen in Exhibit 7. The Decomposition Report is a useful tool to:

1. Provide an understanding of how reserve changes will impact state results
2. Determine the impact our claims adjusters had on the overall reserve change verses changes implemented by our reserving actuaries.
3. Compare impact of loss reserve changes to expected change results.

Exhibit 7

		TYPE OF		AMOUNT OF			NUMBER OF		PRIOR CASE		CURRENT
STATE/GROUP	PRODUCT	CHANGE		CHANGE			FEATURES		RESERVES		RESERVES
STATE NAME	AUTO		NO CHANGE		0			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 < $75,000).

23

Section 4: A Case Study

We will conclude the commentary on the reserving process here but have included a case study immediately following. This case study 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).

Total Needed Reserves Review

There are several key parameters that we consider prior to selecting an ultimate level of needed loss reserves. We will use exhibit 8 to explain considerations we make before making changes to our reserves. Exhibit 8 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 AIn 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.

24

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 8, Section Bdisplays 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.

25

Exhibit 8, Section Cshows 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 8, 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 thecalendar 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.

Exhibit 8, Section Econtains 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 8, Section Fis 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 8 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.

26

• 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.

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 9 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 9 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 9, column 5). Historical development suggests that downward development will occur. Finally, review of the retrospective comparison (exhibit 9, 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 10 through exhibit 15 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.

In defining a lag group please refer to the attached exhibit 10; 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

28

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 10, 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 11 and 12 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 11, 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 12. 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 13, “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 11 and 12 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

29

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 13. 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 14. Once the pure premium is calculated, an average pure premium is selected for each lag-group.

In exhibit 14 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) raised to the power of .25) 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 14 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 15, 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 IBNR 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 15 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.

30

Reconciliation of the Ultimate Value

Exhibit 15 develops an indicated neededIBNRamount of $34,584,463. The record period analysis generated an indicated total neededcase reservesof $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 case study here. We trust that the reader has gained insight and a full appreciation of the level of knowledge and detail that is applied to the reserving process at Progressive.

31

Countrywide X Auto as of February 28, 2001

Exhibit 8

ACCIDENT PERIOD ANALYSIS — ZERO RUNOFF

Section A

Column #	(1)		(2)			(3)			(4)			(5)			(6)			(7)			(8)		(9)		(10)
	Semiannual					Avg			B-F						Avg			B-F			Case Inc’d		Paid
	Accident		Paid			Paid			Paid			Incurred			Incurred			Incurred			Loss as of		Loss as of
	Periods		Method			Method			Method			Method			Method			Method			02/28/2001		02/28/2001		Indicated
	Ending		Ult ($000)			Ult ($000)			Ult ($000)			Ult ($000)			Ult ($000)			Ult ($000)			($000)		($000)		Ultimate
		Aug-94		50,160			50,155		��	50,160			50,160			50,155			50,160			50,160		50,160		50,160
		Feb-95		66,227			66,228			66,227			66,227			66,228			66,227			66,227		66,227		66,228
		Aug-95		76,527			76,517			76,527			76,527			76,517			76,527			76,527		76,527		76,527
		Feb-96		84,899			84,891			84,899			84,928			84,937			84,928			84,928		84,881		84,928
		Aug-96		83,102			83,100			83,102			83,095			83,100			83,095			83,101		83,077		83,101
		Feb-97		93,967			93,954			93,967			93,973			93,954			93,973			93,991		93,934		93,991
		Aug-97		108,819			108,796			108,819			108,814			108,796			108,814			108,834		108,762		108,834
		Feb-98		126,612			126,594			126,612			126,553			126,529			126,553			126,583		126,481		126,583
		Aug-98		134,107			133,909			134,106			133,826			133,838			133,826			133,788		133,582		133,935
		Feb-99		149,979			149,734			149,979			149,685			149,659			149,685			149,555		149,165		149,787
		Aug-99		173,364			173,129			173,363			172,944			172,871			172,945			172,381		171,756		173,285
		Feb-2000		202,912			202,621			202,910			202,412			202,427			202,416			200,448		199,136		202,814
		Aug-2000		214,407			214,599			214,412			213,924			214,095			213,940			208,102		204,735		214,473
		Feb-2001		229,231			227,244			229,015			228,393			227,244			228,379			207,642		179,418		228,251
		Total		1,794,313			1,791,472			1,794,097			1,791,462			1,790,350			1,791,468			1,762,267		1,727,840		1,792,897
		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			63,622			62,510			63,628							65,057
Held Reserve				58,616			58,616			58,616			58,616			58,616			58,616							58,616
Reserve Adequacy				(7,857	)		(5,015	)		(7,641	)		(5,006	)		(3,894	)		(5,012	)						(6,441	)

Section B

(1)			(2)			(3)			(4)		(5)		(6)		(7)		(8)
															Recorded
						Indicated					Closed				*(1-CWP
						Ultimate			Paid		w/Pay		Incurred		Rate)		Indicated
CWP Rate			CWP Rate			CWP			Counts		Counts		Counts		Counts		Ultimate
@6 months			@6 months			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

Countrywide X Auto as of February 28, 2001

Exhibit 8

ACCIDENT PERIOD ANALYSIS — ZERO RUNOFF — Continued

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		1,702		1,702		1,702		1,702		1,702		1,702
		Feb-95		1,744		1,744		1,744		1,744		1,744		1,744
		Aug-95		1,789		1,789		1,789		1,789		1,789		1,789
		Feb-96		1,824		1,824		1,824		1,825		1,825		1,825
		Aug-96		1,877		1,877		1,877		1,877		1,877		1,877
		Feb-97		1,890		1,890		1,890		1,890		1,890		1,890
		Aug-97		1,914		1,914		1,914		1,914		1,914		1,914
		Feb-98		1,932		1,932		1,932		1,931		1,931		1,931
		Aug-98		1,899		1,896		1,899		1,895		1,895		1,895
		Feb-99		1,987		1,984		1,987		1,983		1,983		1,983
		Aug-99		2,013		2,010		2,013		2,008		2,007		2,008
		Feb-2000		2,101		2,098		2,101		2,096		2,096		2,096
		Aug-2000		2,125		2,127		2,125		2,120		2,122		2,120
		Feb-2001		2,185		2,166		2,183		2,177		2,166		2,177
		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		2.5%		2.5%		2.5%		2.5%		2.5%		2.5%
	Aug-95		2.6%		2.6%		2.6%		2.6%		2.6%		2.6%
	Feb-96		2.0%		2.0%		2.0%		2.0%		2.0%		2.0%
	Aug-96		2.9%		2.9%		2.9%		2.9%		2.8%		2.9%
	Feb-97		0.7%		0.7%		0.7%		0.7%		0.7%		0.7%
	Aug-97		1.3%		1.3%		1.3%		1.3%		1.3%		1.3%
	Feb-98		0.9%		0.9%		0.9%		0.9%		0.9%		0.9%
	Aug-98		-1.7%		-1.9%		-1.7%		-1.9%		-1.9%		-1.9%
	Feb-99		4.7%		4.6%		4.7%		4.7%		4.6%		4.7%
	Aug-99		1.3%		1.3%		1.3%		1.2%		1.2%		1.2%
	Feb-2000		4.4%		4.4%		4.4%		4.4%		4.4%		4.4%
	Aug-2000		1.1%		1.4%		1.1%		1.2%		1.2%		1.2%
	Feb-2001		2.8%		1.8%		2.7%		2.7%		2.1%		2.7%
	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-2000		200,606		202,814		2,208
	Apr-2000		207,579		Aug-2000		213,901		214,473		572
	Oct-2000		217,061		Feb-2001		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 9

RECORD PERIOD ANALYSIS — ZERO RUNOFF

Section A

Column #	(1)		(2)			(3)			(4)			(5)			(6)			(7)			(8)		(9)		(10)
																					Adjuster
	Semiannual					Avg			B-F						Avg			B-F			Case Inc’d		Paid
	Accident		Paid			Paid			Paid			Incurred			Incurred			Incurred			Loss as of		Loss as of
	Periods		Method			Method			Method			Method			Method			Method			02/28/2001		02/28/2001		Indicated
	Ending		Ult ($000)			Ult ($000)			Ult ($000)			Ult ($000)			Ult ($000)			Ult ($000)			($000)		($000)		Ultimate
		Aug-94		49,101			49,101			49,101			49,101			49,101			49,101			49,101		49,101		49,101
		Feb-95		64,709			64,709			64,709			64,709			64,709			64,709			64,709		64,709		64,709
		Aug-95		74,983			74,981			74,983			74,983			74,981			74,983			74,983		74,983		74,983
		Feb-96		83,804			83,808			83,804			83,834			83,834			83,834			83,834		83,804		83,834
		Aug-96		81,391			81,392			81,391			81,391			81,391			81,391			81,392		81,390		81,391
		Feb-97		92,678			92,671			92,678			92,676			92,671			92,676			92,678		92,676		92,676
		Aug-97		105,495			105,495			105,495			105,521			105,518			105,521			105,529		105,492		105,507
		Feb-98		126,978			126,973			126,978			127,014			127,009			127,014			127,034		126,970		126,994
		Aug-98		131,328			131,306			131,328			131,273			131,260			131,273			131,284		131,190		131,295
		Feb-99		148,597			148,553			148,597			148,491			148,471			148,491			148,518		148,400		148,533
		Aug-99		169,402			169,321			169,402			169,255			169,210			169,255			169,353		169,073		169,307
		Feb-2000		200,929			200,854			200,929			200,726			200,711			200,726			200,891		200,026		200,812
		Aug-2000		210,924			210,838			210,924			210,480			210,606			210,480			210,490		207,637		210,709
		Feb-2001		223,481			222,934			223,447			223,213			223,192			223,213			226,302		196,223		223,247
		Total		1,763,799			1,762,936			1,763,765			1,762,667			1,762,663			1,762,667			1,766,100		1,731,673		1,763,099
		Paid Loss		1,731,673			1,731,673			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			30,994							31,425
Held Reserve				29,833			29,833			29,833			29,833			29,833			29,833							29,833
Reserve Adequacy				(2,293	)		(1,430	)		(2,258	)		(1,161	)		(1,156	)		(1,160	)						(1,592	)

Section B

(1)			(2)			(3)		(4)		(5)		(6)		(7)
												Recorded
			Indicated					Closed				*(1-CWP
			Ultimate			Paid		w/Pay		Incurred		Rate)		Indicated
CWP Rate			CWP			Counts		Counts		Counts		Counts		Ultimate
@6 months			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

Countrywide X Auto as of February 28, 2001

Exhibit 9

RECORD PERIOD ANALYSIS — ZERO RUNOFF -- Continued

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-2000		2,084		2,084		2,084		2,082		2,082		2,082
		Aug-2000		2,100		2,099		2,100		2,095		2,096		2,095
		Feb-2001		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-2000			5.5%		5.5%		5.5%		5.5%		5.5%		5.5%
	Aug-2000			0.7%		0.7%		0.7%		0.6%		0.7%		0.6%
	Feb-2001			1.7%		1.5%		1.7%		1.8%		1.7%		1.8%
	Total

Countrywide X Auto as of February 28, 2001

Exhibit 9

RECORD PERIOD ANALYSIS — ZERO RUNOFF -- Continued

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)
		34102		40,318		Aug-94		50,426		49,101		(1,325	)
		34286		55,479		Feb-95		66,328		64,709		(1,619	)
		34470		71,753		Aug-95		76,902		74,983		(1,919	)
		34654		79,477		Feb-96		80,616		83,834		3,218
		34838		81,186		Aug-96		80,711		81,391		680
		35022		80,473		Feb-97		91,932		92,676		744
		35206		97,662		Aug-97		104,420		105,507		1,087
		35390		107,800		Feb-98		115,202		126,994		11,792
		35574		118,903		Aug-98		138,689		131,295		(7,394	)
		35758		148,582		Feb-99		148,019		148,533		514
		35942		147,738		Aug-99		173,687		169,307		(4,380	)
		36126		186,662		Feb-2000		200,606		200,812		206
		36310		207,579		Aug-2000		206,665		210,709		4,043
		36494		206,208		Feb-2001		NA		223,247		NA
Total		13 sem		1,629,818				1,534,203		1,763,099		5,648

33

Exhibit 10

Countrywide X Auto as of February 28, 2001

Semiannual
Accident
Periods	RECORDED FEATURE COUNT - ACCIDENT PERIOD ANALYSIS
Ending	1		2		3		4		5		6		7		8		9		10
May-96		28,339		30,243		30,464		30,555		30,629		30,665		30,693		30,711		30,725		30,731
Aug-96		29,928		32,283		32,520		32,637		32,699		32,736		32,759		32,773		32,780		32,784
Nov-96		32,117		35,074		35,423		35,555		35,633		35,669		35,688		35,711		35,729		35,737
Feb-97		34,419		36,722		37,038		37,169		37,237		37,271		37,291		37,302		37,312		37,319
May-97		37,378		40,364		40,649	��	40,801		40,878		40,933		40,968		40,991		41,004		41,016
Aug-97		40,113		43,398		43,722		43,871		43,958		43,993		44,023		44,042		44,059		44,070
Nov-97		43,120		47,048		47,400		47,551		47,619		47,667		47,698		47,717		47,732		47,744
Feb-98		48,987		50,621		50,941		51,075		51,147		51,193		51,225		51,241		51,256		51,268
May-98		49,835		51,629		51,958		52,121		52,227		52,273		52,298		52,316		52,338		52,352
Aug-98		50,931		52,740		53,083		53,254		53,350		53,397		53,439		53,455		53,474		53,481
Nov-98		53,942		55,915		56,320		56,493		56,606		56,651		56,687		56,713		56,728		56,739
Feb-99		57,242		59,305		59,688		59,881		59,974		60,045		60,092		60,125		60,146
May-99		62,839		65,398		65,813		66,008		66,127		66,196		66,239		66,267
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-2000		84,910		87,708		88,155		88,370		88,488
May-2000		86,802		90,012		90,496		90,683
Aug-2000		88,351		91,655		92,117
Nov-2000		88,763		92,402
Feb-2001		93,218

Semiannual
Accident	RECORDED INCREMENTAL FEATURE COUNT - ACCIDENT PERIOD ANALYSIS
Periods	Lag Grp 0		Lag Grp 1		Lag Grp 2		Lag Grp 3		Lag Grp 4		Lag Grp 5		Lag Grp 6		Lag Grp 7		Lag Grp 8		Lag Grp 9
Ending	1		2		3		4		5		6		7		8		9		10
May-96		28,339		1,904		221		91		74		36		28		18		14		6
Aug-96		29,928		2,355		237		117		62		37		23		14		7		4
Nov-96		32,117		2,957		349		132		78		36		19		23		18		8
Feb-97		34,419		2,303		316		131		68		34		20		11		10		7
May-97		37,378		2,986		285		152		77		55		35		23		13		12
Aug-97		40,113		3,285		324		149		87		35		30		19		17		11
Nov-97		43,120		3,928		352		151		68		48		31		19		15		12
Feb-98		48,987		1,634		320		134		72		46		32		16		15		12
May-98		49,835		1,794		329		163		106		46		25		18		22		14
Aug-98		50,931		1,809		343		171		96		47		42		16		19		7
Nov-98		53,942		1,973		405		173		113		45		36		26		15		11
Feb-99		57,242		2,063		383		193		93		71		47		33		21
May-99		62,839		2,559		415		195		119		69		43		28
Aug-99		72,684		2,573		460		230		127		58		30
Nov-99		79,905		2,907		480		208		118		65
Feb-2000		84,910		2,798		447		215		118
May-2000		86,802		3,210		484		187
Aug-2000		88,351		3,304		462
Nov-2000		88,763		3,639
Feb-2001		93,218

34

Exhibit 11

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		8		9		10		Ultimate
May-96		3,565,510		3,501,278		3,470,132		3,454,204		3,416,983		3,415,004		3,424,862		3,442,903		3,439,493		3,437,569		3,437,569
Aug-96		4,409,344		4,286,490		4,239,093		4,189,919		4,185,440		4,184,583		4,182,448		4,179,948		4,185,870		4,185,870		4,184,110
Nov-96		5,493,238		5,241,111		5,199,503		5,190,654		5,191,756		5,186,402		5,186,252		5,186,152		5,186,127		5,186,127		5,182,672
Feb-97		4,506,067		4,279,284		4,239,422		4,255,017		4,234,543		4,232,148		4,230,220		4,230,324		4,227,966		4,225,943		4,222,662
May-97		5,559,617		5,308,146		5,286,386		5,286,034		5,279,481		5,275,988		5,249,070		5,253,604		5,243,604		5,243,604		5,242,462
Aug-97		6,046,692		5,807,174		5,785,475		5,788,650		5,783,227		5,778,288		5,783,675		5,785,740		5,785,849		5,778,530		5,776,046
Nov-97		7,375,863		7,254,442		7,238,616		7,206,675		7,189,864		7,190,336		7,200,237		7,182,011		7,181,496		7,178,855		7,178,631
Feb-98		4,524,973		4,382,121		4,351,051		4,341,804		4,345,496		4,337,016		4,335,225		4,340,802		4,340,802		4,340,802		4,340,389
May-98		4,792,251		4,650,335		4,621,616		4,619,637		4,606,569		4,605,889		4,602,679		4,602,702		4,599,227		4,597,525		4,596,326
Aug-98		5,319,925		5,171,438		5,153,475		5,152,848		5,132,983		5,125,691		5,136,179		5,135,288		5,132,882		5,132,882		5,131,221
Nov-98		5,717,023		5,541,913		5,507,431		5,485,960		5,487,093		5,481,759		5,479,126		5,481,413		5,480,413				5,476,940
Feb-99		5,289,611		5,164,677		5,158,963		5,157,648		5,154,384		5,149,230		5,156,941		5,156,345						5,150,553
May-99		6,136,849		5,952,047		5,944,489		5,923,950		5,933,121		5,921,047		5,927,235								5,920,658
Aug-99		7,526,664		7,260,942		7,254,458		7,257,132		7,242,577		7,228,749										7,225,518
Nov-99		8,659,826		8,526,548		8,480,201		8,489,165		8,476,791												8,463,078
Feb-2000		7,273,274		7,052,120		7,076,164		7,081,966														7,063,284
May-2000		8,236,681		7,919,480		7,927,062																7,885,687
Aug-2000		8,703,268		8,489,973																		8,396,780
Nov-2000		9,391,582																				8,959,569

	1-2		2-3		3-4		4-5		5-6		6-7		7-8		8-9		9-10		10-11
May-96		0.982		0.991		0.995		0.989		0.999		1.003		1.005		0.999		0.999		1.000
Aug-96		0.972		0.989		0.988		0.999		1.000		0.999		0.999		1.001		1.000		1.000
Nov-96		0.954		0.992		0.998		1.000		0.999		1.000		1.000		1.000		1.000		1.000
Feb-97		0.950		0.991		1.004		0.995		0.999		1.000		1.000		0.999		1.000		1.000
May-97		0.955		0.996		1.000		0.999		0.999		0.995		1.001		0.998		1.000		1.000
Aug-97		0.960		0.996		1.001		0.999		0.999		1.001		1.000		1.000		0.999		1.000
Nov-97		0.984		0.998		0.996		0.998		1.000		1.001		0.997		1.000		1.000		1.000
Feb-98		0.968		0.993		0.998		1.001		0.998		1.000		1.001		1.000		1.000		1.000
May-98		0.970		0.994		1.000		0.997		1.000		0.999		1.000		0.999		1.000		1.000
Aug-98		0.972		0.997		1.000		0.996		0.999		1.002		1.000		1.000		1.000
Nov-98		0.969		0.994		0.996		1.000		0.999		1.000		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-2000		0.970		1.003		1.001
May-2000		0.961		1.001
Aug-2000		0.975
Straight Avg		0.964		0.994		0.997		0.998		0.999		1.000		1.000		1.000		1.000		1.000
Avg x HiLo		0.965		0.994		0.997		0.998		0.999		1.000		1.000		1.000		1.000		1.000
Wtd Avg All		0.967		0.995		0.998		0.998		0.999		1.000		1.000		1.000		1.000		1.000
Avg Last 8		0.971		0.998		0.999		0.999		0.999		1.001		1.000		1.000		1.000		1.000
Wt Avg.8		0.972		0.998		0.999		0.999		0.999		1.001		1.000		1.000		1.000		1.000
Avg Last 4		0.973		1.000		1.000		0.999		0.999		1.001		1.000		1.000		1.000		1.000
Wt Avg.4		0.973		0.999		1.000		0.999		0.998		1.001		1.000		1.000		1.000		1.000
Select		0.964		0.994		0.997		0.999		0.999		1.001		1.000		1.000		1.000		1.000
Cumulative		0.954		0.989		0.995		0.997		0.998		1.000		0.999		0.999		0.999		1.000
Ultimate Loss		8,959,569		8,396,780		7,885,687		7,063,284		8,463,078		7,225,518		5,920,658		5,150,553		5,476,940		5,131,221

35

Exhibit 12

Countrywide X Auto as of February 28, 2001

Quarterly
Record
Periods	INCURRED LOSSES QUARTERLY LAG 2 - IBNR ANALYSIS
Ending	1		2		3		4		5		6		7		8		9		10		Ultimate
May-96		770,365		712,299		701,209		687,354		684,115		679,995		679,995		684,711		684,711		684,711		684,711
Aug-96		821,122		745,490		713,522		711,267		707,486		707,439		708,965		713,143		713,143		713,143		710,977
Nov-96		992,289		920,766		911,317		911,196		905,653		900,155		900,155		897,165		895,501		895,501		895,501
Feb-97		912,187		827,755		809,508		803,427		806,676		804,814		796,888		808,030		805,959		805,959		803,896
May-97		1,276,548		1,229,040		1,203,503		1,190,933		1,193,473		1,193,473		1,193,473		1,192,206		1,192,206		1,192,206		1,191,753
Aug-97		1,387,450		1,332,813		1,298,705		1,300,189		1,300,960		1,301,022		1,301,048		1,290,086		1,290,130		1,290,177		1,291,414
Nov-97		1,118,549		1,074,498		1,078,631		1,065,791		1,059,375		1,059,445		1,057,469		1,057,266		1,057,126		1,052,690		1,051,954
Feb-98		1,451,549		1,374,085		1,358,483		1,344,354		1,344,054		1,337,197		1,337,197		1,337,197		1,337,197		1,337,197		1,336,623
May-98		1,422,243		1,381,906		1,358,692		1,359,011		1,353,652		1,351,820		1,353,605		1,353,787		1,353,787		1,353,787		1,353,121
Aug-98		1,386,713		1,324,331		1,325,596		1,321,086		1,316,327		1,316,376		1,316,426		1,316,476		1,316,527				1,315,195
Nov-98		1,536,161		1,432,661		1,423,991		1,413,938		1,406,669		1,405,615		1,405,139		1,399,906						1,397,707
Feb-99		1,492,908		1,392,167		1,365,904		1,362,954		1,366,536		1,365,119		1,361,913								1,359,901
May-99		1,691,912		1,619,294		1,594,488		1,592,653		1,588,513		1,586,402										1,583,430
Aug-99		1,880,903		1,792,471		1,769,981		1,755,897		1,753,214												1,747,869
Nov-99		2,148,501		2,074,814		2,046,275		2,034,381														2,023,000
Feb-2000		1,876,480		1,771,150		1,748,531																1,723,949
May-2000		1,978,122		1,945,414																		1,895,251
Aug-2000		2,045,116																				1,880,756

	1-2		2-3		3-4		4-5		5-6		6-7		7-8		8-9		9-10		10-11
May-96		0.925		0.984		0.980		0.995		0.994		1.000		1.007		1.000		1.000		1.000
Aug-96		0.908		0.957		0.997		0.995		1.000		1.002		1.006		1.000		1.000		1.000
Nov-96		0.928		0.990		1.000		0.994		0.994		1.000		0.997		0.998		1.000		1.000
Feb-97		0.907		0.978		0.992		1.004		0.998		0.990		1.014		0.997		1.000		1.000
May-97		0.963		0.979		0.990		1.002		1.000		1.000		0.999		1.000		1.000		1.000
Aug-97		0.961		0.974		1.001		1.001		1.000		1.000		0.992		1.000		1.000		0.999
Nov-97		0.961		1.004		0.988		0.994		1.000		0.998		1.000		1.000		0.996		1.000
Feb-98		0.947		0.989		0.990		1.000		0.995		1.000		1.000		1.000		1.000		1.000
May-98		0.972		0.983		1.000		0.996		0.999		1.001		1.000		1.000		1.000
Aug-98		0.955		1.001		0.997		0.996		1.000		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-2000		0.944		0.987
May-2000		0.983
Straight Avg		0.944		0.985		0.991		0.997		0.997		1.001		1.001		0.999		1.008		0.998
Avg x HiLo		0.945		0.986		0.992		0.997		0.997		0.999		1.001		1.000		0.999		0.999
Wtd Avg All		0.949		0.986		0.993		0.997		0.998		1.000		1.000		1.000		1.006		0.999
Avg Last 8		0.953		0.988		0.995		0.997		0.999		1.000		1.000		0.999		0.999		1.000
Wt Avg.8		0.954		0.988		0.995		0.998		0.999		1.000		0.999		1.000		0.999		1.000
Avg Last 4		0.962		0.986		0.996		0.998		0.999		1.000		0.999		1.000		0.999		1.000
Wt Avg.4		0.962		0.986		0.995		0.998		0.999		1.000		0.999		1.000		0.999		1.000
Select		0.944		0.988		0.991		0.997		0.999		1.000		1.000		0.999		0.999		1.000
Cumulative		0.920		0.974		0.986		0.994		0.997		0.998		0.999		0.998		0.999		1.000
Ultimate Loss		1,880,756		1,895,251		1,723,949		2,023,000		1,747,869		1,583,430		1,359,901		1,397,707		1,315,195		1,353,121

36

Exhibit 13

Countrywide X Auto as of February 28, 2001

Quarterly
Record
Periods	INCURRED LOSSES QUARTERLY LAG 0-7 EMERGENCE - IBNR ANALYSIS
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-2000		90,173,735		7,063,284		1,723,949		1,234,355		584,067
May-2000		92,079,844		7,885,687		1,895,251		1,168,689
Aug-2000		94,547,077		8,396,780		1,880,756
Nov-2000		95,991,072		8,959,569
Feb-2001		98,816,838
Emergence		96,451,662		8,414,012		1,833,319		1,175,362		660,728		357,894		229,741		182,470
Quarterly
Record
Periods	INFLATED INCURRED LOSSES QUARTERLY LAG 0-7 EMERGENCE - IBNR ANALYSIS
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-2000		94,231,553		7,300,353		1,762,311		1,248,013		584,067
May-2000		95,170,378		8,061,163		1,916,222		1,168,689
Aug-2000		96,650,978		8,489,690		1,880,756
Nov-2000		97,053,212		8,959,569
Feb-2001		98,816,838

37

Exhibit 14

Countrywide X Auto as of February 28, 2001

Quarterly
Record	INFLATED INCURRED LOSSES QUARTERLY LAG 0-7 PURE
Periods	PREMIUMS - IBNR ANALYSIS
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-2000		134.826		10.445		2.522		1.786		0.836
May-2000		127.890		10.833		2.575		1.570
Aug-2000		124.043		10.896		2.414
Nov-2000		120.595		11.133
Feb-2001		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-2000											0.713		0.437		0.344		2.438
May-2000									1.049		0.721		0.442		0.348		3.514
Aug-2000							1.660		1.061		0.729		0.446		0.351		5.213
Nov-2000					2.787		1.679		1.073		0.737		0.451		0.355		8.058
Feb-2001			11.379		2.818		1.697		1.085		0.746		0.456		0.359		19.527

38

Exhibit 15

Countrywide X Auto as of February 28, 2001

SUMMARY OF ESTIMATED IBNR

	(1)		(2)		(3)		(5)		(4)			(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-2000		Factors			Aug-2000		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-2000		2.438		698,911		330,437,470		1,703,804		0.5	%		1,818,422		1.1	%		1,836,089		1.1	%
May-2000		3.514		744,159		347,321,145		2,615,156		0.8	%		3,063,940		1.6	%		3,008,680		1.6	%
Aug-2000		5.213		779,176		358,752,241		4,062,189		1.1	%		10,277,536		4.0	%		10,247,331		4.0	%
Nov-2000		8.058		804,785		368,144,959		6,485,294		1.8	%
Feb-2001		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	)

39

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.

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 included in Calendar Year Results:

1. Average Reserve: These are determined by the actuaries and are generally limited losses estimated to be below a specified threshold.

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.

Closed Without Payment (CWP):A claim which was originally set up with a reserve where no claim payment was necessary and is now closed.

Closure Rate:The number of claims closed with payment at each evaluation date divided by the ultimate number of incurred claims.

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. Generally a claim will involve multiple features such as a Bodily Injury (BI) loss and another for Property Damage (PD).

Financial Case Reserve:See definition for Case Reserves.

Frequency:The number of claims divided by the number of exposure counts. This provides a measure of the proportion of insureds who have claims.

40

High Indication:The highest reasonable indicated reserve estimate generated during a review of a segment by any one of the estimation methods used during a review. If all of the estimation methods show adequacy, the value here is the one that represents the smallest reasonable indicated reserve decrease.

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 but claim has not yet been paid. Incurred Losses = Paid Losses + Case Reserves

Lag Groups:Losses separated into subsets according to the calendar period the claims were recorded following the designated accident period. For example, the claims with accident dates during the first quarter of 1999, that were recorded during the third quarter of 1999, would belong to quarterly record lag two.

Loss Ratio(Calendar Year Loss Ratio): (Paid Losses + Paid LAE + Change in Case Reserves + Change in IBNR Reserves + Change in LAE Reserves + Change in Assigned Risk Reserves) divided by Earned Premium.

Low Indication:The lowest reasonable indicated reserve estimate generated during the review of a segment by any one of the estimation methods used. If more than one of the estimation methods show redundancy, the value here is the one that represents the largest reasonable indicated reserve decrease.

Net Loss Reserves:“Net” indicates that we have deducted the expected reinsurance recoverable from the sum of case and IBNR reserves.

Percent Selected:Percent of difference in selected versus carried reserve. Current carried is based on prior month end open reserve.

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 Loss Ratio:(Paid Losses+Change in Case Reserves+Change in IBNR Reserves) divided by Earned Premium

41

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.

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 and Subrogation:This reserve is the amount of future recoveries we expect to collect. Treated as IBNR, a bulk reserve because there is no feature open.

Segment:A given combination of states, products, and coverage’s with reasonably similar loss development 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.

Taken Dollars Total:The dollar amount of reserve increase or decrease that was actually applied to the reserve once a segment review has been completed.

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

42

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.

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.

43