This Thing Called Qard

Qard is the most misunderstood term in takaful. How Qard is utilized in takaful varies by jurisdiction and, for the purpose of this note, we will assume that Qard is defined as the amount that the takaful operator is required to inject into the Participants Risk Fund (PRF) in order to ensure that the admissible assets in the PRF exceed the PRF total liabilities. Again for the purpose of this note, we will assume that the accounting for liabilities and assets for the purpose of IFRS and regulatory reporting is the same.

First and foremost Qard does not automatically arise when a contract is onerous. If we go back to the example given in the July note:

Even when by definition the contract is not onerous, if the wakala fee is set as 50% of the contribution, then from the PRF perspective the contract is onerous as:

To add to the complications, even if there is a liability of 10cu, this does not automatically trigger a Qard in the PRF as under IFRS17 paragraphs B67 to B70, it is possible to allocate this liability to other profitable contracts within the PRF.

However, in order to make progress to understand the role of Qard in takaful, we need for now to leave such details behind. The point I wanted to make at the outset is that Qard is not necessarily directly related to whether a takaful contract or group of contracts are onerous. It is important to realize that Qard is instead only triggered should the PRF be in a position of valuation deficit. Furthermore we also cannot talk about Qard without also simultaneously talking about surplus, as Qard is essentially negative surplus.

So what is Qard? Qard is essentially the mechanism by which the sharing of risks is extended forward, beyond the accounting period being considered. Let us consider the following simple example:

An insured event has a 1-in-10-year chance of happening. Should the event occur the pay-out is 1000cu. The tabarru’ (that part of the contribution paid to the PRF) payable in each year (the accounting period) is 100cu. The following would be true:

If the event occurs in the 10th year, then in the first nine years there is expected to be a surplus in the first nine years whilst a deficit would arise in the final year (see diagram below):

In the above scenario, there is expected to be a surplus of 100cu for nine years and a Qard of 900cu in the 10th year.

However, if the accounting period is extended from one year to ten years, a different picture would arise:

Clearly if an accounting period of ten years is instead considered, then no Qard arises as the total tabarru’ of 1000cu over the ten years is sufficient to pay the claim of 1000 in the 10th year.

Thus, in this example the Qard arising in the 10th year simply arises due to the timing of the actual occurrence of the event. The event could occur in any of the ten years and in this example it happened in the 10th year. The Qard could also have been prevented if the PRF did not distribute the surplus that arose in the first 9 years but instead set aside the yearly surplus in a provision to meet the eventual claim. What is clear here is that the Qard did not arise due to mispricing as the tabarru’ of 100cu per year is sufficient to cover the expected claim. This example shows the use of Qard as a mechanism to extend the pool of risks. In this case clearly having a pool of one policy does not work. For the given probability of a claim, a minimum pool of 10 policies is needed as the event is expected to occur once in every ten years. By using Qard, however, the pool is extended to ten policy years instead of 10 policies in one year. So the conclusion here is that Qard need not reflect a real loss but instead is a mechanism used to expand the pool of risks across time.

Of course, Qard can also arise when experience differs from the expected, whether it is because of a higher incidence of the insured event, the greater severity of the claim or simply because of mispricing. The concept of pooling over extended accounting periods still applies though. It would just entail the loss arising from this accounting period being offset against surpluses from future accounting periods.  The issue of equity between different generations of participants is not of primary importance in takaful as the premium paid into the PRF is treated as a donation (the literal meaning of tabarru’).

Qard is an important feature for takaful as the PRF does not necessarily have any capital to absorb underwriting losses arising either from timing issues or otherwise. Well established PRFs (those which have been in existence for many years) may indeed have retained surpluses that can be used to absorb losses. However, many other PRFs would not. Each takaful operator may indeed have multiple PRFs, each PRF housing takaful contracts written with a certain attribute which makes it necessary to group them in a separate PRF. For takaful operators which write long term contracts the valuation methodology may also result in “virtual” capital. By virtual I mean negative reserves which represent future expected surplus yet to be realised and not yet allocated to the policies currently in hand.

The intention of this note is to refute the assumption that the existence of the possibility of the operator having to inject Qard into the PRF implies that there is a risk transfer from the takaful participants to the operator. If we were to for a moment assume indeed that any Qard paid reflects a loss that the Operator has to account for, then the takaful business model is unsustainable, as such a premise would imply that the operator is only entitled to 50% of any surplus in the PRF (50% being the maximum that the operator can take of any surplus in the PRF in Malaysia) but has to carry 100% of any loss (as quantified by the Qard paid). The conventional insurance business model only works because, although the insurer bears all of the underwriting losses, it likewise takes all of the underwriting profits. Clearly all is not what it is assumed to be. Fake news perhaps?

Zainal Abidin Mohd Kassim

September, 2020.