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