Malam Team: The Pension-Platform Reset, the Write-Down, and What Has to Happen by the End of 2026
The NIS 27.2 million impairment in Malam Team's pension and provident-fund software was not a stray accounting event. It was management admitting that the bank route failed, and that Mango now has to prove by year-end 2026 that one platform, one migration path, and one efficiency plan can finally close the gap.
The main article argued that Malam Team enters 2026 larger, but still not fully clean. This follow-up isolates only the pension and provident-fund layer, because this is where a quick read can go wrong: the year-end write-down did not hit an otherwise healthy business. It hit an activity that was already failing to reach balance, and that also lost the external shortcut management had hoped might solve scale and economics faster.
That matters precisely because the activity is small inside the segment. In 2025 it generated about NIS 29.3 million of revenue against NIS 295.6 million for the full payroll, HR, and long-term savings segment. But size alone misses the point. Management estimates only about 2% share in the long-term savings solutions market, and the banks are still named as the main competitors in provident-fund processing. The bank negotiation was therefore not just another commercial lead. It was an attempted shortcut in a market that is hard to scale organically.
Three sharp findings stand out from the selected materials:
First: the activity did not fall from zero to negative in the fourth quarter. Even before the write-down it was already behind. In both 2024 and 2025, operating loss stayed close to NIS 9.6 million.
Second: the year-end hit sat on top of a business that still was not balanced. In Q4 2025 the activity already showed positive EBITDA of about NIS 0.7 million, yet the company tied the failed bank talks to a loss of about NIS 30 million and to a NIS 27.2 million impairment of software-development assets.
Third: from here the story is no longer mainly about a transaction. It is about execution. Mango is supposed to unify three systems, customer migrations began in March 2026, and completion is expected to approach year-end 2026.
What The Reset Actually Exposed
The easiest mistake in reading this activity is to collapse the ongoing operating loss into the write-down. The presentation shows 2025 revenue of NIS 29.3 million versus NIS 26.9 million in 2024, yet operating loss stayed at about NIS 9.6 million in both years, and EBITDA also stayed negative at about NIS 1.8 million. In the fourth quarter there was even partial improvement: revenue rose to NIS 7.7 million, operating loss narrowed to about NIS 1.4 million from NIS 2.8 million, and EBITDA turned positive at NIS 0.7 million.
The implication is straightforward. The write-down did not create the problem. It exposed that the problem was already there. If the core activity had already been profitable, year-end 2025 could have been read as a one-off event only. That is not the picture. What appears instead is a business that improved the quarter somewhat, but still had not proved it could carry the cost base and the software layer built around it.
There is one more implication that matters. EBITDA being close to break-even while operating loss still sits at NIS 9.6 million means the 2026 reset is not only a cost-cutting story. The activity also needs enough revenue density and enough margin to carry the software layer already built. That is why operating break-even in 2026 is a real target, but not yet the end of the story.
What The Write-Down Really Says
Note 12(c) leaves little room for soft interpretations. The carrying value of the activity stood at NIS 63.26 million, the recoverable amount was set at NIS 36.06 million, and the result was a NIS 27.2 million impairment. Put differently, almost 43% of the value carried on the balance sheet was cut in one move.
But the number itself is only half the story. The more interesting half sits inside the valuation logic. The recoverable amount was not built on a closed transaction. It was built on two weighted paths: sale of most of the activity, or continuing the business through systems merger and efficiency measures. The sale scenario fell to only 10% in 2025 from 20% in 2024, while the internal continuation path rose to 90%.
That is the core of this continuation thesis. The company has not closed the door on financial-institution cooperation, but it has demoted it from a central thesis to option value. The 2026 business goals still talk about strategic partnerships with the banking system, yet the valuation logic already assumes that the base case has to come from inside. Even the assumptions do not look promotional: a 14.45% discount rate and only 1% long-term growth.
So the impairment is not just an acknowledgment that the bank talks failed. It is an acknowledgment that the value of the activity now has to be rebuilt primarily through self-help, not through an external shortcut.
Why Mango Is The Real 2026 Test
It is worth stopping on the word Mango, because it is easy to read it as just another product name and move on. In practice, this is the move that now carries the entire repair thesis. The group's pension-processing unit is nearing the end of a consolidation of three IT systems into one platform based on its newer pension system. That single platform is supposed to handle new pension, provident funds, study funds, and IRA. Customer migrations started in March 2026, and completion is expected to approach year-end 2026.
This matters not only because one system is cleaner than three. Management explicitly says the move should improve operating, computing, and regulatory efficiency. The risk section adds another layer: provident-fund management is expected to move during the year to Mango, and after that the group should have no mainframe left at all. In other words, year-end 2026 is not just a marketing milestone. It is also an operating deadline.
That leads directly to the real investor test. Banks are still named as the main competitors in provident-fund processing, while the company estimates only about 2% share in long-term savings solutions. Mango is therefore not a strategy on its own. It is the enabling infrastructure that is supposed to make two things more economical: a lower operating burden, and a deeper existing-client business or future financial-institution cooperation built on a more credible platform.
The presentation makes this even sharper. Completing the IT consolidation alongside deepening partnerships with existing clients is expected to bring the activity to operating break-even in 2026 and to profitability from 2027. That is a very precise promise, and a demanding one. It does not promise explosive growth, and it does not promise a transformative deal. It promises, first of all, a repair of the basic economics.
| Checkpoint | What Is Already Known | Why It Matters |
|---|---|---|
| Customer migrations | Began in March 2026 and are expected to approach completion by year-end 2026 | Any delay pushes out both cost efficiency and the 2027 test |
| Platform consolidation | One system is supposed to handle new pension, provident, study-fund, and IRA activity | Without one platform, the write-down and reset do not have a full operating logic |
| Mainframe exit | After the migration, the group should have no mainframe left at all | That makes 2026 a forced execution year, not a comfortable option |
| Operating break-even | This is the explicit target for 2026 | Without break-even, the 2027 profitability target will look too early |
The Real Test Is Balance Versus Delay
This is where the balance-versus-delay test in the title really sits. The broader Malam Team group can absorb this activity even if 2026 turns into another transition year. The core payroll and HR engine remains highly profitable without the long-term savings layer, and the problematic activity is small in revenue terms. So this is not a group-stability question.
But it is very much a repair-credibility question. After the company has already taken a NIS 27.2 million impairment and cut the recoverable value to NIS 36.06 million, it becomes much harder to ask the market for another year of patience without visible operating proof. If by year-end 2026 the company shows advanced migrations, a unified platform, no material operating disruption, and operating break-even, the 2025 write-down will look like a painful but rational reset. If one of those steps slips, the risk is that the impairment will be read not as the end of the cleanup, but as a stop halfway through it.
Conclusion
The main article marked pension and provident-fund processing as one of the areas that still needed repair. This follow-up sharpens the point. The question is no longer whether there is an interesting option here. It is whether the 2025 reset really closed the gap between software, operations, and economics. For now, the answer is still open.
Current thesis: the impairment was justified, but it did not close the story. It only moved the story from a possible-deal path to a mandatory-execution path.
What changed versus the main read? It is now clearer that the activity itself was weak even before the impairment, that the valuation rests mainly on internal repair, and that year-end 2026 is the real line of proof.
Counter-thesis: one can argue that this reading is too severe, because revenue in the activity still grew, Q4 EBITDA already turned positive, and the strong payroll and HR core gives the company enough time to complete the migration without material pressure on the wider group.
What could change the market reading in the short to medium term? Orderly Mango migrations, proof of operating break-even during 2026, no further impairments, and perhaps an early sign that deeper work with existing clients or renewed cooperation with a financial institution is actually being built on a more stable platform.
Why this matters: in regulated software and processing businesses, the gap between a platform that has been built and a platform that actually works economically is enormous. Malam Team has already paid the accounting-recognition price. Now it has to prove that the reset will also become an operating repair.