Takaful is 1.4% of Islamic finance, its profits are 87% Saudi, and in Malaysia family cover has stopped growing. The reason may be buried in how the money moves between two funds.
A takaful operator's finance team sits down after year end to work out whether there is a surplus to give back.
The participants' fund took in more than it paid out. On the face of it, money should go back to the people who contributed it, which is the entire point of the structure.
Then somebody opens the ledger and finds an interest-free loan the shareholders made to that same fund four years ago, still outstanding, which has to be repaid before anybody gets anything. The surplus is real. It just is not theirs yet.
That sequence is the most important thing to understand about this market, and it explains a set of numbers that otherwise look contradictory.
The Islamic Financial Services Board publishes the closest thing the sector has to official statistics. Its 2025 stability report put total Islamic financial services assets at USD 3.88 trillion at the end of 2024, up 14.9% in a year.
Takaful accounted for 1.4% of that.
Islamic banking held 71.6% and ṣukūk 23.3%. So the insurance arm of a near-four-trillion-dollar industry is roughly fifty billion in assets, which is smaller than several individual European insurers. Growth is respectable at 16.9% for the year, though ṣukūk grew faster at 25.6%.
Small is not the same as unpromising, and at this size it cuts the other way: the sector is compact enough that one operator getting something structurally right is visible across the whole market rather than lost in it.
Profitability has been good. S&P Global Ratings put aggregate net profit for GCC takaful operators at about USD 1.1bn in 2024, up from USD 940m, and expects the region's Islamic insurers to hold roughly 10% annual growth through 2026, with earnings recovering again in the first half of this year.
One number in that set deserves more attention than it usually gets. Saudi Arabia accounted for 87.3% of the GCC total.
This is not a regional market. It is one market with satellites, and the satellites are under pressure: S&P notes rising compliance costs, minimum capital requirements and hard competition in motor and health squeezing smaller operators into consolidation. Walaa's absorption of SABB Takaful, taking its capital to SAR 851m with a quarter of it held by former SABB Takaful shareholders, is the shape of things.
If the GCC shows what concentration looks like, Malaysia shows what maturity looks like, and the picture there is more sobering than the growth rates suggest.
Malaysian Takaful Association figures put industry assets at RM62.54bn by mid-2025. General takaful had a good year, with gross written contributions up 12.38% to RM6.64bn.
Family takaful, which is the part that protects households, did this: new business contributions of RM9.74bn, an increase of RM10m on the previous year.
That is growth of about a tenth of a percent.
Penetration tells the same story. It sat at 19.63% in 2025 against 19.57% in 2024, a move of six hundredths of a percentage point in the most developed takaful market in the world. The longer view is genuinely impressive, with active family certificates reaching 6.69 million by August 2025 against 4.55 million a decade earlier. The recent view is a plateau.
Operators know it. Syarikat Takaful Malaysia Keluarga has been pushing regular-contribution family business and leaning on bancatakaful partnerships specifically to get away from credit-linked products, and the industry's stated priorities for 2026 are medical and health cover, wider use of hibah, and digital distribution. Those are the moves of people who can see a flat line.
To see why the plateau might be structural rather than commercial, you have to look at how money actually moves through a takaful operator.
Conventional insurance has one pool. Premium comes in, claims and expenses go out, whatever remains is the shareholders' profit. Takaful splits that in two and adds obligations between the halves.
Every one of those is an accounting relationship a conventional core system has no field for.
The last two interact in a way that matters more than either does alone.
The IFSB has published its own view of how this works in practice, and it is blunter than anything a vendor would write.
In a working paper on the takaful model, it records that the continued deficits in many operators' participants' risk funds are “attributable to the magnitude of the wakālah fees and mudarabah profit shares paid to the TO”, producing a situation in which operators “routinely rely on qard to meet the obligations of takāful participants”.
Read that slowly. The deficit is not primarily bad luck on claims. It is a consequence of the fee structure the model runs on.
That is worth separating from any suggestion of avarice, because operators are boxed in here. The wakala fee is how the operator is paid at all, since it takes no underwriting profit from the participants' fund. Regulators require a separate solvency margin on the shareholders' side precisely because the qard obligation exists. Competition on contribution rates pushes the other way. Most of these fee levels were set years ago, under different loss experience, by people who are no longer in the building.
The mechanics are circular in a way that is easy to miss when the two halves are described separately. The fee comes out of the participants' fund, so if the fee is large enough the fund cannot cover its own claims from what remains; the shareholders then lend it the shortfall, interest free, as they are obliged to; that loan sits as a receivable against the fund, and it has to be repaid out of future surplus before any surplus can be distributed to the people whose contributions generated it. Each step is individually defensible and the sequence is self-reinforcing.
The effect, whoever set the numbers, is that the mechanism which makes takaful distinctive gets deferred by the mechanism that funds the operator. The IFSB adds that repayment terms are “often unclear and sometimes not stipulated”, and asks the obvious question: if the qard is never recovered, is the operator willing to write it off while the deficit grows?
What makes this worth an operator's attention is that it is fixable, and most of the things holding the sector back are not. Nobody can quickly change how many people want cover, or how much they can pay. A fee basis, a surplus calculation and the order in which a qard is repaid are all decisions inside the building.
Now put that next to Malaysia. Family takaful is sold on a promise conventional insurance cannot make, which is that if the fund does well you share in it. A promise like that only does commercial work once buyers can see it operating, and a sector-wide pattern of surplus being absorbed by outstanding qard makes it hard for any individual operator to point at evidence. Penetration moving six hundredths of a point in a year is roughly what you would expect from a differentiator the market cannot yet observe.
None of which makes it the only reason. Affordability, distribution and product design all matter, and probably matter more. It is simply the reason written into the structure rather than into the market.
All of which lands on something unglamorous: whether your platform can express these relationships, or whether it is a conventional core with the words changed.
Relabelling "premium" as "contribution" in a user interface is trivial and worth nothing. What matters is whether the ledger genuinely holds two funds, whether the operator's fee is computed as a defined charge against one of them, whether surplus is a periodic calculation that knows about outstanding qard and its precedence, and whether every step of that is inspectable by a Sharia board months later.
That is a configuration problem more than a coding one, which is where Openkoda's takaful setup starts. Participant and operator funds are modelled separately rather than as a reporting view over a single pool. Wakala and mudarabah remuneration are defined in the rating engine alongside contribution rates, so a change to the fee basis is an edit and an effective date rather than a release. Surplus and qard movements run as scheduled calculations in the policy administration layer, and the Product Builder lets a business user reshape a product when a regulator changes what is permitted.
The governance side carries as much weight here as the arithmetic. A Sharia board reviewing a surplus decision needs to see the inputs, the rule applied and who approved it, which is what a real audit trail is for, and running the platform in your own environment tends to matter more in jurisdictions with data residency rules than it does elsewhere.
If takaful mechanics are new to you, our guide to the model covers the principles, and the core administration piece goes through the system requirements in more detail.
Takaful has spent a decade being described as an underpenetrated opportunity, and the asset growth keeps supporting that description. The contribution figures from the most mature market in the sector do not.
Which leaves an opening rather than a verdict. The first operator that can show a participant where the surplus went, and that it arrived, holds a claim no conventional insurer is able to copy, in a sector small enough for that to be noticed. Getting there is an operational question rather than a theological one, which is the encouraging part: operational questions have answers.
US MGA premium grew 12% in 2025 against the wider market's 5%. In the same window AM Best cut its outlook on the MGA channel and on E&S from positive to stable. Both things are true, and they have the same cause.
Insurers spent 2025 and 2026 writing AI out of standard policies and building a new market to replace it. Here is what Armilla, Munich Re, AIUC, Klaimee, HSB and Testudo actually sell, and what it takes to launch a product like theirs.
US cyber premium fell 7% in 2024, the first decline in the market's history, while policy count held flat and loss ratios stayed strong. A price correction, not a retreat.
Book a live, personalized demo with our product team - tell us your use case and see the platform work with your data. No commitment.