Actuarial Partners Consulting Hassan Scott Odierno discusses the use of actuarial science to ensure the financial statements of an insurer provide a fair and accurate picture of its financial health under IFRS17.

Most countries in Asia have pledged to follow International Financial Reporting Standards (IFRS) in order to provide a way to standardise financial statements. One major exception to this standardisation has been insurance companies. Insurers have followed a standard called IFRS4, which can vary significantly by country. This has made it difficult for an investor to compare the financial statements of insurance companies in different countries. Although we have been aware of the problems with IFRS4 for many years, implementation challenges have kept us from improving on it until now.

This process of ensuring the financial statements are fair and accurate means stating all items of the financial statements at their market value. For assets this has required fine tuning in some cases, which is the realm of IFRS9. Obtaining the market value of liabilities however is more challenging and is where IFRS17 comes in. We need to use actuarial science to ensure the financial statements of an insurer provide a fair and accurate picture of its financial health.

In the early days of actuarial science we calculated the value of liabilities using a concept called net premium valuation (NPV). The idea was generally to calculate reserves using conservative assumptions. This use of conservative assumptions meant we effectively held back profits. As an example, suppose we had a product where we expect to pay benefits of 80 this year and 80 next year. Perhaps we charge a premium of 100 per year which means we have a profit of 20 this year and 20 next year (ignoring discounting and expenses). If we have used conservative assumptions perhaps, we will assume we pay benefits of 120 this year and 120 next year. Just after the policy starts we have received the first premium of 100 but have not paid any benefits yet. Our reserves would be 120 + 120 – 100 = 140. Thus we have received a premium of 100 but have set up reserves of 140. More than holding back profits we have a strain of 40 (100 premium income – 140 reserves set up).

Over the last decade many countries have moved to a gross premium valuation basis (GPV). In our example above this means that we use a best estimate assumption for benefits, i.e. 80 per year, plus padding. This padding is sometimes defined as a confidence level while at other times is defined as a percentage loading on our best estimate. This padding accounts for the fact that our assumptions are an estimation which could be wrong. For this example assume the padding is 5 per year. Thus we assume that we pay out benefits of 85 each year. Thus the GPV just after the policy is issued would be 85 + 85 – 100 = 70. Considering that we received a premium of 100 and set up reserves of 70 we have made a profit of 30 i.e. (100 – 70). This represents part of the profits we will receive over the next two years. Does this give a fair and accurate picture of the company?

A good way to think of this is that an insurance company does work to earn its profit, similar to any other company. Thus it should earn its profit as it does work rather than receive its profit upfront. In IFRS17 terms we would consider the initial profit in the example above of 30 as the contractual service margin (CSM) which will be released over the lifetime of the contract. The padding of 5 is called Risk Adjustment (RA).

A further challenge in understanding the financial statements of an insurance company relates to savings portions of premiums. Take the example above, but instead of a premium of 100 we have 500, where the additional 400 is for savings. The premium income when the policy first starts is now 500 rather than 100, which we could think of as 100 for benefits (and profit) and 400 for savings (like a deposit). A problem arises if we were to compare the financial statements of an insurance company to a bank. Bank deposits would not be considered as income to the bank, but deposits in an insurance company is considered income to the insurer, making it difficult to compare the two. Therefore premiums will no longer be considered income to the insurer; instead we calculate the expected benefit payouts (and expenses) and release of CSM and use this instead of premium income.

A third challenge to understanding the financial statements of an insurance company is understanding how the insurer makes profits. Thus we will perform our calculations gross of reinsurance and explicitly show the cost (or value) of reinsurance. This will show the reader of the financial statements the extent the insurance company is relying on reinsurance for profit or on the flip side how expensive the reinsurance programme is. Above we mentioned the use of expected benefits (and expenses) instead of premiums, we will compare this with actual benefits and expenses to show the reader underwriting profit. Investment income will be shown separately and compared to the investment income required in our actuarial calculations. Thus it will be clear how much profit is emerging from investment income.

In future the finance team of an insurer will depend on the actuary to put the financial statements together. This is the first challenge that must be overcome. The second challenge is that the CSM must be calculated separately for each year of issue (at a minimum), with potentially different assumptions used in each year (called a cohort).

**Game plan**

The first question shareholders will want to know is how will this affect me? Firstly what is the size of the CSM that must be held for all policies which have been sold and is still on the books compared to the capital currently available? Will injections be required when we move to IFRS17? Secondly, once IFRS17 is up and running, how will the yearly profit and transfer to shareholders compare to pre-IFRS17? This requires an initial study. The initial study will need to account for the specifics of the products sold, the history of the pricing methodology and basis and an understanding of the workings of the underlying models of operations.

Different types of participating plans may also have different methods of modelling. There is significant freedom given in the IFRS17 rules, so the various choices need to be explained carefully so shareholders understand the potential effects of each choice. Publicly listed companies in particular will not appreciate surprises as IFRS17 is implemented, so this study should be performed as soon as possible.

The second issue will be the accounting system itself. The accounting system will need to keep track of many CSM calculations as well as the various actuarial inputs such as expected benefits, expenses and investment income. The actuarial inputs will need to come from actuarial software. Normally a gap study will be performed to see the extent of changes needed in the accounting system. The first set of calculations may be required by December 2019, so during 2018 changes will be made to the system and during 2019 test runs will be performed.

The third issue will be training. In some countries a large increase in actuarial staff will be required, which will require advanced planning and training. Finance staff such as the CFO will need to understand the basics of actuarial science to be able to interpret the data given to them and reasonableness of the accounts which has been put together.

**Conclusion**

IFRS17 will finally allow the financial statements of insurers to be understood by investors and analysts comparing insurers in different countries and different industries. Implementation will require extensive technical work. Insurers will need to put together the initial study along with the gap analysis as soon as possible. Training on IFRS17 will be necessary.