How (and why) to read a Statement of Actuarial Opinion (Property/Casualty)

Loss Reserves are generally the largest liability listed on an insurance company’s balance sheet. Loss Reserves are amounts set aside to pay for all claims that have already occurred, whether reported or not, that the company is obligated to pay in the future.  In the past, a high percentage of property/casualty insurance company insolvencies have been due to the inadequacy of these reserves.  For that reason and others, each state now requires that there be an attestation, i.e. opinion, by a qualified professional, as to the reasonableness of these reserves.

Formally called the “Statement of Actuarial Opinion,” (SAO), it is commonly referred to as “the opinion.” 

“Is it a clean opinion?”

One of the first things asked of the opining actuary is whether or not it is a “clean” opinion.  “Clean” is the conversational term for an “unqualified” opinion.  The answer appears in a section labeled “Opinion” about a fourth of the way through the document.  A clean opinion states that the loss reserves:

  1. Meet the requirements of the domiciliary state,
  2. Are consistent with reserves computed with accepted loss reserving standards and principles and,
  3. Make a reasonable provision in the aggregate for all unpaid loss and loss adjustment expense obligations of the Company under the terms of its contracts and agreements.

The first two items are rarely an issue: it is the third point that is usually the reason for the question.  If the booked loss and loss expense reserves equal the actuary’s reserve estimate or fall within a certain range of that estimate, the actuary can say that the booked reserves make a reasonable provision for the reserve liability and can issue a “clean” opinion.

The alternatives to a “clean” opinion are generally not good:

A qualified opinion – This happens when, in the actuary’s opinion, the reserves for a certain item or items are in question because they cannot be reasonably estimated, or the actuary is unable to render an opinion on those items.  Note:  if there is some sort of loss for which the actuary has no means of determining the ultimate value, but the perceived value is not material, the opinion can still be a “clean” one.

A Deficient or Inadequate Provision.  This is the one you do not want to see.  When the carried reserve amount is less than the minimum amount that the actuary believes is reasonable, the actuary will issue a Statement of Actuarial Opinion that the carried reserve amount does not make a reasonable provision for the liabilities associated with the specified reserves.

A Redundant or Excessive Provision.  These are very rare.  When the carried reserve amount is greater than the maximum amount that the actuary believes is reasonable, the actuary issues a redundant opinion.

A No Opinion is another rare rendering.  The actuary’s ability to give an opinion is dependent upon data, analyses, assumptions, and related information that are sufficient to support a conclusion. If the actuary cannot reach a conclusion due to deficiencies or limitations in the data, analyses, assumptions, or related information, then the actuary may issue a statement of no opinion. A statement of no opinion should include a description of the reasons why no opinion could be given.

Some selected sections of the SAO are also of importance.

The “Relevant Comments” section starts with “Risk of Material Adverse Deviation.”  This section answers three important questions: 

  1. Is there a risk of a material deviation (i.e. change) from the loss and loss expense reserves booked?
  2. How large does a possible deviation need to be to be considered “material?”
  3. What is the source of such a deviation(s)?

If the actuary indicates there is one or more items that are a potential source of risk of material deviation, it might or might not be a cause for concern by the reader. 

The absence of a risk of material adverse deviation does not imply that additional factors will not be identified in the future as having a significant influence on the company’s reserves; it is a statement concerning a point in time.

While the interest in Asbestos Exposures and Environmental Exposures has waned in recent years, they still can represent a significant exposure to the company.  They are addressed in the section bearing that as a title.

Most companies do not write policies with durations over thirteen months, but the section titled Disclosure of Unearned Premium Reserves for Long Duration Contracts will provide information to that effect. 

The Reinsurance Collectability section provides information about the likelihood of the company’s ability to collect on its reinsurance.  While the actuary’s comments do not imply an opinion as to the financial stability of any of the reinsurers, it should include information from management regarding any collectability problems the company has encountered with its reinsurers.  The actuary usually uses publicly available information as well as the reinsurers’ ratings, provided by a rating agency, as a reason for whatever conclusion he/she reaches on this topic.

The section on “NAIC IRIS Tests” requires a little explaining.  The National Association of Insurance Commissioners (NAIC) has a series of mathematical formulae that are applied to various Annual Statement entries.  Of the thirteen formulae, the last three, tests 11, 12 and 13, concern the booked loss and loss expense reserves.  The actuary is required to look at these tests and comment on any that fall outside the range of acceptance: i.e. they “fail” the test.  

If the company fails any of these tests, the actuary will provide a narrative as to whether or not their failure is important. You can get more detailed information at http://www.naic.org/prod_serv/UIR-ZB-16_UIR_2016.pdf.

The final pages are the signature page and some exhibits providing selected values. 

Recapping:

  • Is the opinion an unqualified (i.e. a “Clean”) one or not?
  • How financially large does something have to be in order to be a material risk to the finances of the company, in the actuary’s opinion, and are there any such risks?
  • Did the company fail any IRIS Tests?  And if so, why?

Please contact the author for more information:

Grover M. Edie, MBA, FCAS, MAAA, CPCU, ARP, CERA, ARM

grover.edie@hugginsactuarial.com

Explaining IBNR to the Board

“What is IBNR” is likely the most asked question of actuaries at a Board of Directors meeting. Huggins’ actuaries often are asked the question, so we decided to put together this short, printable explanation. Even if you already understand IBNR, this short narrative might enable you to save some time. Feel free to share it with others, including your Board of Directors.

Like most companies, insurance companies are required to hold liabilities on their books for the unpaid costs of claims that have occurred as of the accounting date.  Unlike other companies, there are claims that have not been finalized, and some claims that have not even been presented to the company.  Statutory accounting requires the insurer reflect the amount the insurer expects to pay for losses incurred but not yet reported or reported claims that have not yet been paid as of the accounting date. 

Because an ultimate value of the claim and claim expenses liabilities cannot be precisely determined in advance, the provisions for them at any balance sheet date is usually estimated or tested using actuarial and statistical techniques.

Usually, most of the claims reserve is in the form of claims adjuster estimates.  The total of all the claim adjusters’ reserves is referred to as the case reserve.  But as of the accounting date, not all the information regarding all the claims has been presented to the claim adjusters.  There needs to be an estimate of the value of the cost of the “unknown” information on claims being adjusted.  Some call it “Incurred But Not Enough (information) Reported, or IBNER, to represent the cost of additional unknown information on known claims. 

For known outstanding claims, the IBNER can be calculated as the difference between the current case reserve and the estimate amount ultimately needed to settle the reported claims (provision for future development on known claims), including payments post closure for reopened claims.

As of an accounting date, not all the claims have been presented to the insurance company.  These “unknown” claims are still costs as of the accounting date, and need to be reflected.  Some sources call this “Incurred But Not Yet Reported” claim costs, or IBNYR.  For unreported claims, the IBNYR is sometimes separately calculated.

Often, the IBNER and IBNYR amounts are calculated together, and are referred to combined as IBNR, which stands for “Incurred But Not Reported,” and refers to estimates of the both IBNER and IBNYR.

As a group of claims age, more claims are settled and paid, more information becomes known, and more previously unreported claims are reported; thus, the need for IBNR diminishes.  Therefore, the amount of IBNR for a given accident year generally decreases over time. The declines for all prior years are often more than compensated for by the IBNR needed for the new accident period, and thus overall IBNR increases.

Please contact the author for more information:

Grover M. Edie, MBA, FCAS, MAAA, CPCU, ARP, CERA, ARM

grover.edie@hugginsactuarial.com

Analysis of Economic Scenarios Used by Huggins Actuarial Services, Inc. 2016 through 2018

OVERVIEW

When Huggins’ actuaries create an Economic Capital Model, they use a state-of-the-art modeling platform designed by actuaries and other insurance professionals that is widely known as an industry standard.  These models are used to predict profits, policyholder surplus, income and other financial metrics over a pre-determined timeframe. In order to reflect uncertainty in the future economic activity which affects balance sheet assets (stocks and bonds), Huggins loads Economic Scenarios into the Economic Capital Model. The Economic Scenarios used by Huggins are generated by a top flight firm that specializes in the analysis of industry trends. In the tables below, Huggins actuaries analyze how these economic scenarios for four classes of assets have been changing over the period 2016 to 2018:

  • Corporate A-Rated Bonds;
  • Blue Chip Stocks;
  • Master Limited Partnerships (“MLP Equities”); and
  • Treasury Bonds.

In the analysis for each category of asset, the actuaries consider yields for bonds and total returns for stocks and MLPs using the future forecasts given by one-year time-step (one year in the future), three-year time-step (three years in the future) and five-year time-step (five years in the future) scenarios. Each scenario consists of 10,000 random values generated by the Economic Scenario Generator. The yield and total return data are fit to a normal distribution and the following are derived:

  • Distribution Mean and
  • Distribution Standard Deviation.

 

SUMMARY OF RESULTS

Overall, the yields for bonds and total returns for blue chips stocks and MLP equities increase over time, both as time-step increases and going from the 2016 economic scenarios to the 2018 economic scenarios, reflecting both increasing bond yields and increasing inflation.

What do these increases in bond yields and total returns mean for an insurance company?  The impact depends on the asset mix of the current portfolio and the result is likely to be mixed.  Given that most property casualty companies invest heavily in bonds, the embedded bonds are likely to decrease in value as bonds are marked to market. The increase in bond yields for newly purchased bonds will offset this downward adjustment somewhat. The non-bond portion of the portfolio should have a larger contribution to the bottom line than in the recent past in both dividends and enhanced asset value. Overall, Huggins would expect to see increased investment returns, possibly returning to pre-recession levels.

ANALYSIS BY ASSET CLASS

For two-year “A” rated corporate bonds, the yields increase from 1.66% for one-year time-step to 3.50% for five-year time-step for the 2016 economic scenarios, to 3.32% for one-year time-step to 4.29% for five-year time-step for the 2018 economic scenarios. For five-year “A” rated corporate bonds, the yields increase from 2.13% for one-year time-step to 3.97% for five-year time-step for the 2016 economic scenarios, to 3.59% for one-year time-step to 4.71% for five-year time-step for the 2018 economic scenarios. For ten-year “A” rated corporate bonds, the yields increase from 2.78% for one-year time-step to 4.53% for five-year time-step for the 2016 economic scenarios, to 4.02% for one-year time-step to 5.21% for five-year time-step for the 2018 economic scenarios. See the three tables below.

 

Economic Scenarios – Two-Year “A” Rated Corporate Bonds

 10,000 Yield Scenarios

2016 Two Year

Corporate Bonds

2017 Two Year

Corporate Bonds

2018 Two Year

Corporate Bonds

 

Time-Step

 

Mean

Standard Deviation  

Mean

Standard Deviation  

Mean

Standard Deviation
1 Year 1.66% 0.63% 2.45% 0.76% 3.32% 0.90%
3 Year 2.67% 1.07% 3.18% 1.24% 3.79% 1.40%
5 Year 3.50% 1.45% 3.84% 1.56% 4.29% 1.74%

 

Economic Scenarios – Five-Year “A” Rated Corporate Bonds

 10,000 Yield Scenarios

2016 Five Year

Corporate Bonds

2017 Five Year

Corporate Bonds

2018 Five Year

Corporate Bonds

 

Time-Step

 

Mean

Standard Deviation  

Mean

Standard Deviation  

Mean

Standard Deviation
1 Year 2.13% 0.46% 2.81% 0.61% 3.59% 0.75%
3 Year 3.14% 0.92% 3.60% 1.09% 4.17% 1.26%
5 Year 3.97% 1.33% 4.28% 1.44% 4.71% 1.59%

 

Economic Scenarios – Ten-Year “A” Rated Corporate Bonds

 10,000 Yield Scenarios

2016 Ten Year

Corporate Bonds

2017 Ten Year

Corporate Bonds

2018 Ten Year

Corporate Bonds

 

Time-Step

 

Mean

Standard Deviation  

Mean

Standard Deviation  

Mean

Standard Deviation
1 Year 2.78% 0.37% 3.33% 0.52% 4.02% 0.67%
3 Year 3.74% 0.83% 4.14% 1.00% 4.66% 1.16%
5 Year 4.53% 1.24% 4.81% 1.34% 5.21% 1.49%

 

For Blue Chip Stocks, the total returns increase from 5.04% for one-year time-step to 6.42% for five-year time-step for the 2016 economic scenarios, to 7.06% for one-year time-step to 8.02% for five-year time-step for the 2018 economic scenarios. See table below.

 

Economic Scenarios – Blue Chip Stocks

 10,000 Total Return Scenarios

2016 Blue

Chip Stocks

2017 Blue

Chip Stocks

2018 Blue

Chip Stocks

 

Time-Step

 

Mean

Standard Deviation  

Mean

Standard Deviation  

Mean

Standard Deviation
1 Year 5.04% 16.94% 6.27% 16.50% 7.06% 16.85%
3 Year 5.60% 17.14% 6.91% 16.96% 7.49% 17.09%
5 Year 6.42% 17.12% 7.68% 17.53% 8.02% 17.72%

 

For MLP Equities, the total returns increase from 9.46% for one-year time-step to 10.26% for five-year time-step for the 2016 economic scenarios, to 9.75% for one-year time-step to 11.02% for five-year time-step for the 2018 economic scenarios. See table below.

 

Economic Scenarios – MLP Equities

 10,000 Total Return Scenarios

2016 MLP

Equities

2017 MLP

Equities

2018 MLP

Equities

 

Time-Step

 

Mean

Standard Deviation  

Mean

Standard Deviation  

Mean

Standard Deviation
1 Year 9.46% 18.23% 9.44% 18.12% 9.75% 18.16%
3 Year 9.73% 18.64% 9.88% 18.63% 10.36% 18.76%
5 Year 10.26% 18.59% 10.63% 18.80% 11.02% 18.98%

 

For two-year Treasury bonds, the yields increase from 1.15% for one-year time-step to 2.89% for five-year time-step for the 2016 economic scenarios, to 2.69% for one-year time-step to 3.60% for five-year time-step for the 2018 economic scenarios. For five-year Treasury bonds, the yields increase from 1.50% for one-year time-step to 3.25% for five-year time-step for the 2016 economic scenarios, to 2.85% for one-year time-step to 3.92% for five-year time-step for the 2018 economic scenarios. For ten-year Treasury bonds, the yields increase from 1.94% for one-year time-step to 3.62% for five-year time-step for the 2016 economic scenarios, to 3.09% for one-year time-step to 4.23% for five-year time-step for the 2018 economic scenarios. See the three tables below.

 

 

Economic Scenarios – Two-Year Treasury Bonds

 10,000 Yield Scenarios

2016 Two Year

Treasury Bonds

2017 Two Year

Treasury Bonds

2018 Two Year

Treasury Bonds

 

Time-Step

 

Mean

Standard Deviation  

Mean

Standard Deviation  

Mean

Standard Deviation
1 Year 1.15% 0.62% 1.93% 0.75% 2.69% 0.89%
3 Year 2.10% 1.05% 2.61% 1.22% 3.13% 1.38%
5 Year 2.89% 1.43% 3.22% 1.54% 3.60% 1.68%

 

Economic Scenarios – Five-Year Treasury Bonds

 10,000 Yield Scenarios

2016 Five Year

Treasury Bonds

2017 Five Year

Treasury Bonds

2018 Five Year

Treasury Bonds

 

Time-Step

 

Mean

Standard Deviation  

Mean

Standard Deviation  

Mean

Standard Deviation
1 Year 1.50% 0.44% 2.17% 0.60% 2.85% 0.74%
3 Year 2.46% 0.91% 2.91% 1.08% 3.40% 1.25%
5 Year 3.25% 1.31% 3.56% 1.42% 3.92% 1.57%

 

Economic Scenarios – Ten-Year Treasury Bonds

 10,000 Yield Scenarios

2016 Ten Year

Treasury Bonds

2017 Ten Year

Treasury Bonds

2018 Ten Year

Treasury Bonds

 

Time-Step

 

Mean

Standard Deviation  

Mean

Standard Deviation  

Mean

Standard Deviation
1 Year 1.94% 0.35% 2.49% 0.51% 3.09% 0.65%
3 Year 2.86% 0.81% 3.25% 0.98% 3.70% 1.15%
5 Year 3.62% 1.22% 3.90% 1.33% 4.23% 1.47%

 

Authored by

GROVER M. EDIE

MBA, FCAS, MAAA, CPCU, ARP, CERA, ARM

and

FRANK SULLIVAN, PH.D.

Driverless Cars and Insurance

How Driverless Cars Affect the Auto Insurance Industry

by James Chang, FCAS, CPCU, ARe, MAAA and Kim E. Piersol, FCAS, MAAA

The American public has mixed feelings about giving up control of their cars’ steering wheels.  Despite the enthusiasm with which autonomous vehicles are being developed by auto manufacturers and technology companies, recent polls show that few drivers are interested in autonomous vehicle technology despite the potential safety and time-saving benefits.

 

Auto insurers are already thinking about how self-driving cars will affect them even though it might be a long time before we see them regularly on the road.  From the invention of the horseless carriage in 1890’s, car insurance has evolved from simple handwritten contracts to the high-tech global industry that it is today.

 

The transition to driverless cars will create the need for more personal coverage and to increase the need for more commercial insurance as car manufacturers will assume much of the risk for this new techology.  However, it is too soon to completely know what the insurance strategy will be now since it could be 25 to 30 years before we transition to driverless cars.

 

Currently, technology companies and auto manufacturers are capturing data, and that data can be used for actuarial models.  Insurers will also need to form partnerships with the winners in the autonomous vehicle space.  Insurers are having conversations with the autonomous vehicle providers to solve problems points and lot of benefits are to be gained through these partnerships.

 

We can speculate where the commercial liability responsibility lies for autonomous vehicle accidents.  The responsibility for liability becomes complicated when multiple parties are involved in an accident.  The fault can be divided between consumer, manufacturer, and software maker.  Manufacturers may be required to purchase insurance for their vehicles in the future.  There may be a battle between the manufacturer and insurance company to see who pays for the accident.  Other trends driverless cars can have on the insurance industry include dropping of insurance premium as accidents decline and more drivers could reduce or drop coverage.

 

There could be new lines of business created as a result from driverless cars.  Cybersecurity, such as protection against remote vehicle theft, ransomware, coverage for identity theft, and theft or misuse of personal data will emerge as well as product liability for software and sensors and public infrastructure insurance for cloud server systems that manage traffic and road network.

 

Industry experts conclude that insurers can take key steps in preparation of a driverless future.

 

  • Build expertise in big data and analytics; insurers must be able to harness the data generated by autonomous vehicles and systems that support them.
  • Develop the actuarial modeling framework by utilizing advanced actuarial modeling techniques as they adapt to the autonomous vehicle features.
  • Collaborate with automakers, software developers, governmental entities, manufacturers and the like to share data and expertise.

 

The business model will change as the auto insurance risks change followed by insurance policy changes.

“Stuff”

Too much clutter in your life?  Read the latest column from Huggins’ consulting actuary Grover Edie, MBA, FCAS, MAAA, CERA, CPCU, ARM, ARP in the latest edition of Actuarial Review.

 

 

NAIC Draft Model Law On Cybersecurity by Chandu C. Patel, FCAS, MAAA, Rusty Kuehn, FCAS, MAAA, and Kim Piersol, FCAS, MAAA