I posted an article in June 2020 on the question of whether Takaful practices risk transfer or risk sharing. I argued that IFRS Standards are based on the “expected” model; the model considers something which has not yet occurred but is expected to occur. And importantly assigns a probability to measure that event. Now a qard is a subordinated loan from the takaful operator to the takaful fund (the Fund) to cover instances when liabilities exceed assets. I argued that if the probability that any qard would be repaid is sufficiently high then there is no risk transfer from the Fund to the takaful operator. Some quarters instead allude to the possibility (without assigning any probability) of the qard being written off by the takaful operator as proof that risk transfer has transpired.
This posting considers only takaful operations where the takaful operator is diligent in its underwriting and pricing, has an appropriate retakaful program in place and practices good Asset Liability management. Such a Fund would be expected to generate annually surpluses rather than losses and is thus likely to be able to repay any qard that it might occasionally incur. This posting does not consider “takaful” operations with takaful funds which are perpetually requiring a qard as such operations are not takaful. Once we have established with an accepted level of probability that the operation is indeed takaful we can say that the takaful fund is underwriting 100% of the risk. This is notwithstanding (like in Malaysia) that the takaful operator may be entitled to a performance bonus which is payable from surplus accruing to the takaful participants.
Sometimes a comparison is made of takaful to a participating life insurance policy (a with-profits policy). For the uninitiated, a participating life insurance policy is where the insurer takes x % of the surplus distributed whilst the participating policyholders takes the remaining 1- x %. Should, however, the funds be insufficient to pay the guaranteed benefits, the insurer is obliged to pay 100% of the deficit arising (sounds familiar?).
The accounting of participating life policies under IFRS 17 is expected to fall under the Variable Fee Approach (VFA). However, even under the VFA, policies at inception would be subject to the General Measurement Model. The only difference is in the subsequent measurement of the Contractual Service Margin (CSM). The IASB has refused to exempt participating life policies from the requirement of establishing annual cohorts, maintaining that this is necessary should any policy guarantee “bite”. In the next paragraph we consider how losses under participating life policies are expected to be treated under IFRS 17.
IFRS 17 applies to accounting for insurance contracts, effectively accounting at the individual policy level. However, understanding that such accounting at a policy level is near impossible, the IASB has ruled that insurance policies should be grouped with the condition that no two policies within the Group can be incepted more than one year apart. This is called establishing a cohort. This need for cohorts is tied to the requirement to amortize any profit via the CSM and over the expected duration of the policies within that cohort. No cohort, no duration over which to amortize the CSM.
The flip side of the CSM is the Loss Component (LC); this is the capital sum of the losses expected to be incurred by the insurer for a Group of policies. Under IFRS 17, all expected losses to the insurer must be charged immediately to retained earnings while all CSMs are amortized over the expected duration of the respective Group of policies. All this is fine, except that a “loss making” Group may subsequently turn profitable in the future. Given that profits and losses are defined within a cohort and for the duration of the policies within that cohort, there is a need to keep track of the LC as it is amortized over the duration of the cohort. Should the cohort turn a profit (defined as when any unamortized LC has been covered) the Group would be re-designated as profitable. The cohort would then establish a CSM mid-duration as any expected profit would be earned through a profit carrier over the unexpired duration of the Group of surviving policies. For those cohorts that are unable to turn a profit, any remaining unamortized LC must be zeroized by the time the last policy in the cohort expires. To summarize, the LC is determined on a Group of policies basis and the LC can only be recovered from any subsequent surpluses emerging from that Group.
Let us now compare the above with how losses arise in a takaful fund and how any qard arising would be repaid. Under the wakala takaful operating model the takaful fund is where the tabarru’ or risk premium is accumulated. Any claim would be paid from the takaful fund. The participants (policyholders) contribute to the takaful fund in the expectation that, should the insured event transpire, the takaful fund would compensate the participant for the loss. The participant also contributes a fee (the wakala fee) to the takaful operator to manage the takaful fund. The takaful contract is therefore a triparty contract between the participant, the takaful operator and the takaful fund. As part of the operating model the takaful operator would, should the need arise, provide a temporary loan (the qard) to the takaful fund. A qard arises when the assets in the takaful fund are not sufficient to cover its liabilities. Under takaful the qard is repayable by the current and future participants of the takaful fund. Thus the qard is “covered” by ALL future surpluses of the takaful fund. This means not just any future surplus from the policies that caused the deficit in the first place, but also the future surplus from the profitable policies in the takaful fund AND surplus from all future new policyholders to the fund.
The table below summarizes the differences between a participating life policy and a takaful certificate:
It is apparent from the above that takaful if properly implemented practices risk sharing. The takaful fund within the takaful operation accepts risk from each participant and pools that risk. The takaful fund is the entity (separate from the participants and the operator) which bears the pooled risks collectively on behalf of the participants. It is also apparent from the above that how takaful treats qard is different from how losses are managed under participating life insurance policies.
The final question is how qard is accounted under IFRS 17. One point of view is the notional (notional meaning undiscounted) qard should be allocated as part of the fulfilment cash flows. However, given that the qard is established at the takaful fund entity level and is not attributable to any specific policy let alone the period of the eventual repayment of that qard, that approach would be questionable even if it is doable.
Zainal Abidin Mohd Kassim