Actuarial Audit Report

City of Chandler — Regular Meeting (2022-02-03)

View PDF Meeting page

Extracted text (via ocr_local) 41179 characters

@ OliverWyman

ACTUARIAL ANALYSIS OF
WORKERS’ COMPENSATION
UNPAID LOSSES & ALAE AS OF
JUNE 30, 2021

CITY OF CHANDLER

September 29, 2021

Actuarial Analysis

CONTENTS

1. Executive Summary..

Purpose and Scope
Actuarial Findings

Relevant COMMENLS.....cccccsccssesecessssetesssesvsesnsscssssssessessssessscesessseeseasaesesesseseuseenenesesecseeenereneasenseeneeserentesens 5
2. Background......ccscescssessseresssessssvsnecsssseseseesseesensaeesseneseseanesssesseesaeanenenesenee ceneeeeneee 6
3. Data....ccccssccscscnscnssseeeceresereeceseenssssnanensensenecesesensnsnsneenensneenscrsnreaaaagansorssees vaenaceners 7
4. Methodology .....

Loss Development Method.
Bornhuetter-Ferguson Method.
Frequency / Severity Method v...ccssessesssessesseessessssensessessesseesneessecanesneseeeneseetansenisssenesniesensasersaneaneensees 9
Loss Cost Method
Estimated Ultimate Number of Claims
Estimated Unpaid Losses and ALAE at Retained Limits ..

Estimates of Reserves at Various Probability Levels.......cccccesssssssesescseerereseessseeseseesnensisseeenessenenennees

Projection of Ultimate Losses and ALAE for Future Policy Periods ..........sssssessssesesesiesseseeseneensess 11
5. Distribution and Use .......csscsccssseesreceseecsseeesteeneeeeesseeee secaesnavenecesenceeeeseeeneeenees 12
6. Considerations and Limitations ..........:csscessereeeeeee sescaaveseeeeeaeeesserenas seceeeveenneees 13
7. Acknowledgement of Qualifications wl7

Appendix A. Glossary

Actuarial Analysis Executive Summary

1. EXECUTIVE SUMMARY

Purpose and Scope

Oliver Wyman Actuarial Consulting, Inc. (Oliver Wyman) has been retained by the City of Chandler (the
City) to provide actuarial consulting services. The purpose of this analysis is to estimate unpaid losses and
allocated loss adjustment expense (ALAE) as well as claim cost projections for the City’s workers’
compensation (WC) exposures. The conclusions in this study are related to its stated purpose and may
not be applicable for other purposes.

The analysis provides the following:
e Unpaid losses and ALAE as of June 30, 2021 for claims incurred through that same date.

e Projected ultimate losses and ALAE for policy period July 1, 2021-22.

e Estimates of the 70%, 80% and 90% probability levels of the unpaid losses and ALAE as well as the
projected ultimate losses as of June 30, 2021.

The estimates are based on data evaluated as of June 30, 2021 and additional information provided to
Oliver Wyman (and only that information provided) through September 29, 2021.

The estimates are prepared on a nominal basis only; that is, there is no consideration of the time value of
money.

Estimates are presented as actuarial central estimates. An actuarial central estimate is defined as an
expected value over the range of reasonably possible, as opposed to all conceivable, outcomes.

Additionally, estimates are provided at the 70%, 80% and 90% probability levels, meaning the estimates
include a margin for potential adverse development over the long term average. These estimates
represent the amount for which there is a 70%, 80% or 90% chance that the actual unpaid losses and ALAE
or projected ultimate losses and ALAE will be less than or equal to this value.

Unless otherwise stated, all references to loss should be understood to include ALAE and be net of salvage
and subrogation. Please note that ALAE is an insurance industry term which generally refers to costs that
can be attributed to the settlement of individual claims. Fees paid to outside defense attorneys and
medical management costs represent a major portion of this expense item. Also, please note that the
terms unpaid losses, reserves and liabilities are used interchangeably throughout this report.

The scope of this project did not include the estimation of any expenses other than ALAE that are
sometimes associated with insurance programs. Such expenses include unallocated loss adjustment
expense (ULAE), reinsurance premiums, the costs of trustee, legal, administrative, risk management and
actuarial services, taxes, fees and assessments, and costs for surety bonds or letters of credit pertaining
to outstanding liabilities or capital requirements. ULAE is the insurance industry term generally used to
refer to any claims-handling costs that cannot be attributed to individual claims. Fees paid to third party
administrators (TPAs) and salaries paid to in-house claim adjusters and in-house defense attorneys
represent examples of this type of expense.

Actuarial Analysis Executive Summary

This report sets forth the results of Oliver Wyman’s analysis and is accompanied by exhibits which should
be considered integral parts of this report.

Actuarial Findings

Table 1 below summarizes the nominal unpaid losses & ALAE, at the appropriate self-insured retention
(SIR) levels as of June 30, 2021. Estimates are presented both at an actuarial central estimate and the 70%,
80% and 90% probability levels.

Table 1: Estimated Unpaid Losses & ALAE

Evaluation Actuarial Probability Levels
Date Central
Estimate 70th 80th 90th
As of June 30, 2021 10,458,761 11,975,281 13,878,776 17,016,404
As of June 30, 2020 8,795,715 10,079,889 11,707,097 14,416,177
Change 1,663,046 1,895,392 2,171,679 2,600,227

The estimate of approximately $10.5 million as of June 30, 2021 represents an increase of approximately
$1.7 million since the actuarial analysis performed as of June 30, 2020. The increase is driven by the
inclusion of the experience in policy year 7/1/20 — 6/30/21, which already has seven claims above the
$75,000 incurred threshold, including claim number CN-21-005148, which has approximately $1.1 million
in case reserves and is one of three police officer COVID claims the City has seen with incurred amounts
over $75,000. Total incurred to date for this policy year, approximately $2.6 million, has already surpassed
our initial projected ultimate loss for the period and is notably higher than the incurred values of prior
policy years 12 months from inception.

Projected Ultimate Losses and ALAE

The nominal losses and ALAE expected to be incurred during policy period 7/1/21— 6/30/22 by the City,
with a self-insured retention of $3.0 million for police and fire employees and $1.0 million for all other
employees, are as shown in Table 2 below.

Table 2: Projected Ultimate Losses and ALAE for Policy Period 21/22

Actuarial Probability Levels
Self-Insured Central
Retention Estimate 70th 80th 90th

1M /3M 2,700,000 | 3,096,900 3,626,100 4,514,400

This forecast is based on Oliver Wyman’s estimated loss costs (losses per unit of exposure) and the City’s
forecasted exposures. It should be noted that since the projected losses and ALAE are directly related to
the forecasted exposures, if actual exposures differ from the forecast, the projected losses and ALAE
should be adjusted accordingly.

Actuarial Analysis Executive Summary

Relevant Comments

Unpaid losses and ALAE are composed of claims examiners’ estimates and IBNR amounts. The claims
examiners’ estimates are established by individual claim adjusters on open claims. Claims examiners’
estimates are based on the facts of a claim as they are known today; these estimates do not anticipate
that a claim will change. The IBNR amounts are estimated by Oliver Wyman and include amounts for late
reported and re-opened claims as well as development on known open claims. The fact that Oliver
Wyman’s IBNR amounts include a provision for development on known claims does not necessarily imply
that there is a problem with the claim examiner estimates. The tendency is that, in aggregate, damages
or injuries are worse than originally thought. Thus, an adequate provision for unpaid losses and ALAE
should include a provision for upward development. Please note that Oliver Wyman’s IBNR amounts are
appropriate for all historical policy periods in aggregate but not necessarily for any one particular policy
period.

Actuarial Analysis

2. BACKGROUND

Founded on May 17, 1912, the City of Chandler is located in the Greater Phoenix Valley of Arizona, located
20 miles from the Phoenix Sky Harbor Airport.

The City began self-insuring its Workers’ Compensation exposures on July 1, 2002 and has maintained a
self-insured retention since inception. Prior to that date, the City’s Workers’ Compensation liabilities were
covered through a guaranteed cost program.

The City’s self-insured retentions since July 1, 2002 are as displayed below. All self-insured retentions are

assumed to be ona per occurrence basis and eroded by loss and ALAE.

Self-Insured Retention

Policy Years
Police / Fire All Other
7/1/02-03 to 7/1/11-12 500,000
7/1/12-13 to 7/1/14-15 600,000
7/1/15-16 to 7/1/17-18 1,000,000 600,000
7/1/18-19 to 7/1/19-20 1,250,000 600,000
7/1/20-21 2,250,000 750,000
7/1/21-22 3,000,000 1,000,000

Actuarial Analysis Data

3. DATA

For this study, Oliver Wyman relied on the accuracy and completeness of data and information provided
by the City without independent audit. This actuarial analysis is based upon the review of this data and
also insurance industry data, along with practical knowledge and training in actuarial theory, which
together enable us to interpret the results of calculations. Throughout the process, the data played an
important role. The accuracy and validity of the actuarial analysis are dependent on, among other things,
the quality of the data used. If the underlying data are inaccurate or incomplete, the results of Oliver
Wyman’s analysis may likewise be inaccurate or incomplete.

Oliver Wyman has attempted to review the data used directly in this analysis for the purpose of identifying
values that are questionable or relationships that are materially inconsistent. In cases where the data are
incomplete, inaccurate, or not as appropriate as desired, Oliver Wyman has considered whether the use
of such data may produce material biases in the results of this study, or whether the data are so
inadequate that the data cannot be used to satisfy the purpose of this study.

Oliver Wyman has not, however, audited or verified this data since this is beyond the scope of the
assignment. Also, it should be noted that Oliver Wyman’s review of data may not always reveal
imperfections and that there may be situations where it is impossible or impracticable to perform a
sufficient review of the data. Neither has Oliver Wyman examined any insurance contracts covering the
years applicable to this analysis.

Specifically, the data relied upon are as follows:
* Historical ground-up and unlimited detailed claims listings as June 30, 2021 provided by the City
* Historical estimated and audited (where applicable) payrolls by period provided by the City

* Oliver Wyman’s prior study Actuarial Analysis of Workers’ Compensation Unpaid Loss and ALAE as of
June 30, 2020 dated September 29, 2020

* Industry loss development factors, increased limit factors and benefit level changes from the 2019
and 2020 Annual Statistical Bulletins published by National Council on Compensation Insurance (NCCI)

Actuarial Analysis Methodology

4. METHODOLOGY

The estimates in this report are developed in accordance with of the Casualty Actuarial Society and the
applicable standards of the American Academy of Actuaries. Below is a discussion of the details underlying
the actuarial methodologies and assumptions used by Oliver Wyman in the analysis. Oliver Wyman’s
projections for the City assumed that the patterns observed in historical policy periods will be applicable
to future policy periods.

Oliver Wyman first selected ultimate losses limited to $75,000, which provides a more stable base for
projections than the City’s historical SIRs (or losses at statutory limits). Oliver Wyman then used the City’s
data supplemented with insurance industry data to bring the $75,000 limit estimates to the actual SIR by
policy period.

To develop the estimates of limited ultimate losses, Oliver Wyman applied the following generally
accepted actuarial techniques.

* Paid loss development method

* Incurred loss development method

* Paid Bornhuetter-Ferguson method

* Incurred Bornhuetter-Ferguson method

* Frequency / severity method

* Loss cost method

The following describes these methods in more detail:

Loss Development Method

This method involves estimating how losses will develop in the future. Historical patterns are identified
by creating a loss triangle. The triangle presents paid or incurred losses over time at successive 12-month
valuations. The triangle reveals how paid or incurred losses have changed over time. From this historical
loss development experience, loss development factors (LDFs) are selected to represent Oliver Wyman’s
expectations of future loss development. If the Company’s exposure is small and there is not enough
historical data to determine its own patterns, or if the Company’s own patterns are volatile, then
insurance industry data may be used to help select appropriate loss development factors.

An accumulation of the selected LDFs provides a means of estimating how losses will develop up to the
time that all losses have been settled. The calculated accumulated LDFs are applied directly to the latest
paid or incurred losses to project ultimate losses.

An advantage of the paid loss development method is that its predictive accuracy is independent of case
reserving practices and philosophies.

The advantages of the incurred loss development method are that incurred losses provide a larger
statistical database than paid losses (since claims examiners’ estimates are used in addition to payments)
and that incurred losses provide more current information (since payments lag behind the establishment
of claims examiners’ estimates).

Oliver Wyman

Actuarial Analysis Methodology

Bornhuetter-Ferguson Method

The Bornhuetter-Ferguson (B-F) method is a blending of the results of the loss development method with
an a priori expectation of ultimate losses. Expected unpaid or unreported losses are added to cumulative
paid or incurred losses to derive an estimate of ultimate losses. Expected unpaid or unreported losses are
derived by applying the portion of ultimate losses that are estimated to be unpaid or unreported to the
expected ultimate losses.

Please note that the preliminary ultimate losses are selected based on the loss development methods and
the loss cost method (described below) for the more recent years. The loss development factors are the
same as those used in the loss development method.

An advantage of the B-F method is that estimates of ultimate losses tend to be more stable than the usual
loss development method. This is because estimates based on the B-F method are generally a weighted
average between the estimate based on a loss development methodology and the preliminary estimates
of ultimate losses.

Frequency / Severity Method

Under the frequency / severity method, the ultimate loss for each policy period is the product of the
selected ultimate severity (i.e., average cost per claim) multiplied by the ultimate number of reported
claims. The preliminary ultimate losses are based on the loss development methods.

The ultimate severity for the most recent policy period is judgmentally selected based on the historical
severities trended to the current period’s cost level. The ultimate severity for prior policy periods equals
the selected current severity, detrended to the appropriate cost level.

Loss Cost Method

Under the loss cost method, an ultimate loss cost is judgmentally selected and multiplied by the exposure
to obtain the ultimate loss estimate. Loss cost equals ultimate losses divided by exposure. Ultimate loss
costs are judgmentally selected based on the City’s historical loss cost, adjusted for inflation. The
preliminary losses are based on the loss development methods.

This method can be useful for estimating ultimate losses for the most recent and for future policy periods.
The advantages of using a loss cost method for the more recent policy periods are that it accounts for
changes in exposure and is less affected by changes in closing and reserving practices than are the incurred
and paid development methods.

Estimated Ultimate Number of Claims

Similar to the paid and incurred loss development factors, reported claim count development factors are
calculated based on historical claims triangles created from the City’s own experience. The selected count
development factors are applied to the City’s reported counts to determine an estimate of the ultimate
number of claims.

Asecond estimate of the ultimate number of claims is developed based on a combination of the frequency
and B-F methods. Oliver Wyman first trends the above-estimated ultimate counts by using a judgmentally

Actuarial Analysis Methodology

selected frequency trend. Using the selected count development factors, the percentage of unreported
claim counts is determined. Lastly, these unreported counts are then added to the reported counts to
date to create a second estimate of ultimate number of claims. Oliver Wyman then judgmentally selects
the ultimate number of claims based on the two estimates.

Estimated Unpaid Losses and ALAE at Retained Limits

To develop the estimates of retained ultimate losses, Oliver Wyman applied three generally accepted
actuarial techniques: the Increased Limit Factor (ILF) Method, the LDF method, and the Excess Method.

In the ILF Method, Oliver Wyman used the City’s data supplemented with insurance industry data to
create ILFs, which is a ratio of expected losses at higher limits to expected losses the selected limit of
$75,000. These selected ILFs were multiplied by the $75,000 limited ultimate loss estimates to bring these
estimates to the actual SIR by policy period. This method can be useful for estimating ultimate losses for
policy periods with a larger unknown. Typically, this is most useful for more recent or future policy periods
and for policy periods with a significantly high SIR, where actual losses have not yet had time to develop
into that layer.

In the LDF Method, Oliver Wyman used internal data to adjust the selected limited loss development
factors to reflect the incurred loss development pattern appropriate for the actual retention limit by policy
period. These LDFs were then applied to the retained incurred losses. This method is useful when it is
appropriate to assume that historical development patterns are representative of future development
patterns, regardless of what has been reported so far in a given policy period.

In the Excess Method, Oliver Wyman used the aforementioned ILFs, LDFs at the $75,000 limits, and LDFs
at the actual retention limit to calculate an estimate of IBNR excess of the $75,000 limit. This amount was
then added to the limited ultimate losses and the excess incurred losses to estimate ultimate losses at the
actual retention limit. This method is a blend of the ILF and LDF methods.

To estimate the unpaid losses at the City’s retention limits, Oliver Wyman selected ultimate losses at the
City’s retention limits based on the three methods described above, and then subtracted the City’s actual
payments at the retention limits.

Estimates of Reserves at Various Probability Levels

Oliver Wyman provides estimates of loss and ALAE reserves as of June 30, 2021 at the actuarial central
estimate and at various probability levels. A compound Poisson model was developed to gauge the
volatility of potential losses. The parameters for the stochastic model were determined based on Oliver
Wyman’s analysis of the claim detail.

Oliver Wyman has assumed the loss and ALAE reserves follow a lognormal distribution with a coefficient
of variation (CV) of 50%. The frequency and severity for the lognormal distribution were estimated based
on the number of open and IBNR claims and based on the CV of the City’s loss experience, respectively.
Oliver Wyman has judgmentally introduced a variable to calculate any additional parameter risk involved.

Oliver Wyman 10

Actuarial Analysis Methodology

These calculations assume Oliver Wyman has correctly measured the parameters of the model and do not
include estimates of the additional variation which will be caused by misestimating the expected number
of claims and the average claim cost.

Projection of Ultimate Losses and ALAE for Future Policy Periods

To project future losses at the SIR (assuming SIR remains unchanged in the upcoming period), Oliver
Wyman adjusted historical ultimate losses limited to $75,000 to the projected cost and benefit levels and
adjusted historical ultimate claim counts to the projected frequency levels.

Using the loss cost method, Oliver Wyman selected a projected loss cost, the result of which was
multiplied by projected payroll. The results are estimated ultimate loss and ALAE, limited to $75,000, for
calendar accident period 7/1/2021 — 6/30/2022. Using the frequency/severity method, Oliver Wyman
selected a projected frequency and severity. The selected frequency was multiplied by projected payroll
to arrive at an estimate of projected ultimate claim counts, which was then multiplied by the selected
severity to arrive at a second indication of the estimate ultimate loss and ALAE, limited to $75,000, for
calendar period 7/1/2021 — 6/30/2022. Oliver Wyman then judgmentally selects a projected ultimate loss
and ALAE, limited to $75,000, based on these two indications and shows the resulting selected limited
forecasted loss cost, limited to $75,000.

This selected limited forecasted loss cost was then multiplied by the projected payroll and an increased
limit factor to bring the estimate up to the SIR.

Actuarial Analysis Distribution and Use

5. DISTRIBUTION AND USE

Usage and Responsibility of Client — Oliver Wyman prepared this report for the sole use of the client
named herein for the stated purpose. This report includes important considerations, assumptions, and
limitations and, as a result, is intended to be read and used only as a whole. This report may not be
separated into, or distributed, in parts other than by the client to whom this report was issued, as needed,
in the case of distribution to such client’s directors, officers, or employees. All decisions in connection
with the implementation or use of advice or recommendations contained in this report are the sole
responsibility of the client named herein.

Third Party Reliance and Due Diligence — Oliver Wyman’s consent to any distribution of this report
(whether herein or in the written agreement pursuant to which we issued this report) to parties other
than of the client named herein does not constitute advice by Oliver Wyman to any such third parties.
Any distribution to third parties shall be solely for informational purposes and not for purposes of reliance
by any such parties. Oliver Wyman assumes no liability related to third party use of this report or any
actions taken or decisions made as a consequence of the results, advice or recommendations set forth
herein. This report should not replace the due diligence on behalf of any such third party.

Actuarial Analysis Considerations and Limitations

6. CONSIDERATIONS AND LIMITATIONS

Data Verification — For our analysis, we relied on data and information provided by the client named
herein without independent audit. Though we have reviewed the data for reasonableness and consistency,
we have not audited or otherwise verified this data. Our review of data may not always reveal
imperfections. We have assumed that the data provided is both accurate and complete. The results of our
analysis are dependent on this assumption. If this data or information is inaccurate or incomplete, our
findings and conclusions might therefore be unreliable.

Prospective Policy / Accident Period Estimates — We estimated the prospective policy/accident period
estimates developed in this analysis using estimated loss costs and the projected exposures. Prospective
period loss and ALAE estimates are directly related to the projected exposures. Therefore, if actual
exposures differ from the projection, we would need to adjust the prospective policy/accident period
estimates accordingly.

Supplemental Data — Where historical data of the client named herein was either (i) not available, (ii) not
appropriate or (iii) not sufficiently credible to develop our actuarial assumptions, we supplemented it with
external information, as we deemed appropriate. Although we believe these external sources may be
more predictive of future experience of the client named herein than any other data of which we are
aware, the use of external data adds to the uncertainty associated with our projections.

Exclusion of Other Program Costs — The scope of the project does not include the estimation of any costs
other than those described herein. Such ancillary costs may include unallocated loss adjustment expenses
(ULAE); excess insurance premiums; the costs of trustee, legal, administrative, risk management and
actuarial services; fees and assessments; and costs for surety bonds or letters of credit pertaining to claim
liabilities.

Assumption of Valid Insurance / Reinsurance ~ We assumed that all insurance/reinsurance is valid and
fully collectible. We made no assessment, and do not express any opinion, concerning the viability or
collectability of any insurance or reinsurance. We have not evaluated the financial strength, claims-paying
ability or any other factors with regard to the past, current, and prospective insurers/reinsurers of the
client named herein.

Funding of Claim Payments — We have not examined any assets that may be supporting the liabilities,
and have made no assumptions regarding the maturities and liquidity of these assets, should they exist.
This examination is beyond the scope of our review.

Rounding and Accuracy — Our models may retain more digits than those displayed. Also, the results of
certain calculations may be presented in the exhibits with more or fewer digits than would be considered
significant. As a result, there may be rounding differences between the results of calculations presented
in the exhibits and replications of those calculations based on displayed underlying amounts. Also,
calculation results may not have been adjusted to reflect the precision of the calculation.

Unanticipated Changes — We developed our conclusions based on an analysis of the data of the client
named herein and on the estimation of the outcome of many contingent events. We developed our
estimates from the historical claim experience and covered exposure, with adjustments for anticipated
changes. Our estimates make no provision for extraordinary future emergence of new types of losses not
sufficiently represented in historical databases or which are not yet quantifiable. Also, we assumed that

Actuarial Analysis Considerations and Limitations

the client named herein will remain a going concern, and we have not anticipated any impacts of potential
insolvency, bankruptcy, or any similar event.

Internal / External Changes — The sources of uncertainty affecting our estimates are numerous and
include factors internal and external to the client named herein. Internal factors include items such as
changes in claim reserving or settlement practices. The most significant external influences include, but
are not limited to, changes in the legal, social, or regulatory environment surrounding the claims process.
Uncontrollable factors such as general economic conditions also contribute to the variability.

Uncertainty Inherent in Projections — While this analysis complies with applicable Actuarial Standards of
Practice, users of this analysis should recognize that our projections involve estimates of future events
and are subject to economic and statistical variations from expected values. We have not anticipated any
extraordinary changes to the legal, social, or economic environment that might affect the frequency or
severity of claims. For these reasons, we do not guarantee that the emergence of actual losses will
correspond to the projections in this analysis.

Probability Levels — We have calculated estimates of the statistical uncertainty associated with the
process risk inherent in our estimates. However, unless otherwise indicated, our probability level
estimates do not address parameter or model risk. To the extent that the probability estimates do not
address parameter risk or model risk, the true variability of results is greater than the range of outcomes
presented. The client named herein may wish to consider this additional uncertainty in evaluating the
estimates presented in this report.

Obamacare - The Patient Protection and Affordable Care Act (PPACA or “Obamacare”) was signed into
law in 2010 with the goals of increasing coverage, affordability, quality, and efficiency of the current health
care system. Provisions to accomplish these intended goals were implemented in phases, with some
provisions taking effect in 2014 and beyond. Our estimates do not account for any potential impact,
including potential future revisions or repeal/replacement, the PPACA will have on any lines of business
included in this analysis.

Workers’ Compensation - The Medicare, Medicaid and SCHIP Extension Act of 2007 (the Act) was enacted
in part to ensure that Medicare does not pay medical benefits for beneficiaries whose medical treatments
arise out of a liability or work-related accident. While Medicare has always been a secondary payer of
these benefits, Medicare’s rights to reimbursement have not been frequently exercised in the past. New
reporting requirements under the Act enable Medicare to enforce its liens against primary payers and
effectuate the reimbursement process. The ability of liability and workers compensation insurers to
procure full and final medical settlements may be diminished under this Act. The passage of this Act
increases the uncertainty in any estimate of unpaid losses and we have not attempted to quantify the
impact of this Act.

COVID-19 Pandemic We have included no explicit adjustments in this report for the effect of the COVID-
19 pandemic on loss experience except as specifically noted in this report. The impact of this event on loss
experience is highly uncertain and generally unquantifiable at this time.

The potential impact of the COVID-19 pandemic on workers compensation experience is highly uncertain.
The unprecedented nature of the pandemic, the numerous potential impacts on claims experience, and
the uncertainty associated with activities and actions intended to temper the impact (such as client-
specific actions or more general governmental interventions) increase the uncertainty underlying the

Actuarial Analysis Considerations and Limitations

estimates contained in this report. There are numerous known and unknown potential impacts. We
describe several of these uncertainties below:

Oliver Wyman

Compensability of COVID-19 Claims Businesses may incur workers compensation claims due to
employees becoming infected with COVID-19. However, claims directly related to COVID-19 may or
may not be determined compensable as an occupational disease, depending on statutes, regulations,
and court interpretations specific to an individual jurisdiction.

— Some jurisdictions have different standards for different diseases.
— Some jurisdictions exclude ordinary diseases of life, such as influenza, from coverage.

— Specific employee classifications, most notably those involved in healthcare with direct patient
contact, and public safety officers, may have lower thresholds of compensability than other
employee classifications.

— Some jurisdictions, such as Washington State, have implemented material policy changes with
respect to the compensability of COVID-19 related claims. Other jurisdictions may follow.

Claim Frequency and Severity Considerations for potential impacts of the COVID-19 pandemic and
related societal and economic impacts on workers compensation claim frequency and severity
include:

— Health Care Coverage: Some states have mandated that health insurance must cover testing and
emergency room/ urgent care visits related to COVID-19 without deductibles or copays. These
types of measures may serve to temper workers compensation claim activity related directly to
the illness. Actions taken by individual employers may have the same impact.

— High Demand Industries and Essential Businesses: The hiring of new employees to meet growing
demand in some industries could lead to an increase in claims, as claim incidence rates associated
with new employees are typically higher than those associated with longer-tenured employees.
Additionally, employees performing work to which they are not accustomed may lead to higher
average severity of claims. The new demand for these workers is likely leading to longer working
hours, which may increase the number of claims. Essential businesses may also see an increase in
mental stress claims. Compensability of mental stress claims varies by jurisdiction.

— Economic Impacts: Sectors most affected by sudden job loss and furloughs may see an increase
in workers compensation claims as employees seek to supplement their lost income. Potentially
offsetting this increase, remaining employees may file fewer claims due to concerns over job
security. The overall impact is unknown at this time. An economic downturn may also lead to an
increase in claim severity. For example, employees may be more likely to involve attorneys in their
claims, and claimants who are already receiving temporary disability benefits may attempt to
extend those benefits.

Employers Liability Employers liability may serve as a remedy for an employee who believes the
infection was the result of employer negligence. If an employee’s family member becomes infected,
that could also lead to an employers liability claim.

Teleworking Companies with a material shift to a teleworking environment may see an increase in
claims for injuries related to non-ergonomic work environments stations and other potential safety
hazards in employees’ homes. Compensability of incidents related to these issues will vary by
jurisdiction. Less exposure to other work-place hazards may have a counter effect.

Actuarial Analysis Considerations and Limitations

* Claims Handling Considerations The claims process is likely to lengthen during this pandemic, leading
to an increase in claim severity as workers compensation hearings and mediations are potentially
postponed, and claimant access to doctors, nurse case managers and medical services is limited.
Reduced availability of transitional duty jobs may also serve to increase claim costs for claimants who
would have otherwise been able to return to work.

Potential impacts of the pandemic are many and the above only represents a sample of anticipated effects.

Oliver Wyman 16

Actuarial Analysis Acknowledgement of Qualifications

7. ACKNOWLEDGEMENT OF QUALIFICATIONS

Thank you for providing Oliver Wyman with the opportunity to develop this analysis for the City. We are
prepared to discuss the results at your request.

|, Adam Hirsch, am a Principal with Oliver Wyman Actuarial Consulting, Inc. | am a Fellow of the Casualty
Actuarial Society (FCAS), a Member of the American Academy of Actuaries (MAAA), and meet the
Qualification Standards of the American Academy of Actuaries to counsel on the results contained herein.

Oliver Wyman Actuarial Consulting, Inc.:

Adam Hirsch, FCAS, MAAA

633 West Fifth Street, Suite 1200
Los Angeles, CA 90071

(213) 346-5352

Actuarial Analysis Appendix A

APPENDIX A. GLOSSARY

Accident Period

The period in which the event giving rise to a claim occurred, regardless of when the claim is actually
reported.

Accounting Date
The point in time at which the estimate of unpaid claims and ALAE is evaluated.

Actuarial Central Estimate

An estimate that represents an expected value over the range of reasonably possible outcomes. Such
range of reasonably possible outcomes may not include all conceivable outcomes.

Allocated Loss Adjustment Expense (ALAE)

Expense costs associated with the handling and settling of an individual claim that can be directly
attributed to the particular claim. Fees paid to outside defense attorneys and investigation firms are
examples of this expense cost.

Case Reserves
The unpaid claim estimates established by adjusters on an individual claim basis.

Claim

A demand for payment under the coverage provided by a plan or contract. As used throughout this
Glossary, it also includes suits, potentially compensable events, notifications, and unasserted claims.

Claim Frequency
The number of claims that occur over a period of time per unit of exposure.

Claim Reporting Pattern
The rate at which claims are assumed to be reported over time.

Claim Severity
The average cost per claim.

Coefficient of Variation

A statistical measure of dispersion. The coefficient of variation is calculated as the standard deviation of
the random process divided by the expected value (mean).

Confidence Level

The probability that the outcome of a random process will not exceed an associated estimate. For
example, a 75% confidence level for an unpaid claim estimate of $10 million would indicate that there is
a 75% probability that the actual claim payments will be less than or equal to $10 million. The estimate
is defined in the context of the risks modeled in our analysis and may not consider all factors
contributing to variability of outcomes.

Credibility
A measure of the predictive value of a body of data.

Oliver Wyman 18

Actuarial Analysis Appendix A

Development

The change between valuation dates in the observed values of certain fundamental quantities that may
be used in the unpaid claim estimation process.

For example, the number of reported claims associated with events occurring within a particular period
will change from one valuation date to the next until all claims have been reported. In a similar fashion,
the paid claim amounts for events occurring within a particular period will change from one valuation
date to the next until all claims have been reported and closed. The change in the number of reported
claims or the change in the paid claim amounts is referred to as development. The concept of
development also applies to reported incurred losses.

Estimated Ultimate Incurred Losses
The estimated cost of claims during a period. Ultimate incurred losses represent the total of paid claim
amounts, case reserves, and IBNR.

Event
The incident or activity that triggers potential for claim or allocated claim adjustment expense payment.

Exposure
A measure of the underlying potential for claim costs.

IBNR

The unpaid claim estimate for: (a) events that have occurred for which claims have not been reported as
of the accounting date, (b) future development of the case reserves, (c) claims that have been reported
but not yet recorded in the loss listing, and (d) claims that have been closed but that will be reopened.
Loss

The cost associated with a claim. The cost may or may not include loss adjustment expenses.

Loss Adjustment Expenses

The costs of administering, determining coverage for, settling, or defending claims. Loss adjustment
expenses include allocated loss adjustment expenses and unallocated loss adjustment expenses.

Loss Cost
The loss amount per exposure unit.

Method
The systematic procedure for developing an actuarial estimate.

Model
A mathematical or empirical representation of a specified phenomenon.

Model Risk
The risk that the methods are not appropriate to the circumstances or the models are not
representative of the specified phenomenon.

Paid Losses
The total aggregate dollar amount paid on all reported claims as of a certain date.

Oliver Wyman 19

Actuarial Analysis Appendix A

Parameter Risk

The risk that the assumptions or parameters used in the methods or models are not representative of
future outcomes.

Payment Pattern
The rate at which claims are paid over time.

Process Risk

The uncertainty associated with the projection of future contingencies that are inherently variable, even
when the parameters are known with certainty.

Reported Incurred Loss Amount
The total of paid claim amounts and case reserves.

Subrogation
Recoveries from a third party responsible for the event for which a claim has already been paid.

Unallocated Loss Adjustment Expense (ULAE)

Loss adjustment expenses that cannot be attributed to an individual claim. Typically includes salaries,
utilities, and rent apportioned to the claim adjustment expense function but not readily assignable to
specific claims.

Unpaid Loss Estimate
The estimate of the obligation for future payments resulting from losses due to past events.

Valuation Date
The date through which transactions are included in the data used in the unpaid claim estimate analysis.