Radixis at Castro: When Does a Strategic Option Become Another Capital Consumer?

Why Radixis Can No Longer Sit Quietly Inside "Other Unallocated"

The main article argued that Castro bought itself balance-sheet breathing room through Urbanica, but had not yet repaired the core retail margin. This follow-up isolates Radixis because in March 2026 the US operation changed category inside the group. It is no longer only a strategic option on a new geography. It became a capital-allocation decision that dilutes Castro, flattens the subsidiary's capital structure, and adds a governance layer around an activity that still is not disclosed separately.

Radixis itself was never designed as a short-cycle bet. It holds the exclusive right to market, sell, and distribute Yves Rocher products in the United States until December 31, 2045. The agreement also includes a ten-year business plan starting in 2025, with purchase targets that only begin from the third year of activity. In other words, this was always meant to be a long-run platform. That is exactly why the March 2026 transaction matters. Before the market has standalone numbers or a separate operating proof point, shareholders were already asked to approve more capital, more dilution, and more related governance decisions around Ron Rotter.

So the relevant question is not whether Radixis could one day work. The narrower question is whether Castro built a capital structure that protects shareholders until the proof arrives, or whether it has already begun to pay a full equity price for an activity that still sits largely inside promise and partial disclosure.

Control Stayed, but the Economics Changed

The February 2024 founding structure was aggressive but clear. Castro received about 79.9% of the venture's equity through preferred shares. An unrelated strategic partner received another 9.9%, also through preferred shares. Ron Rotter received 9.9% through ordinary shares, subject to a three-year reverse-vesting mechanism. Beyond the ownership split, Castro and the strategic partner also agreed to inject about $6 million and about $745 thousand, respectively, in a way that gave them preferential return rights on their invested capital.

That starting point matters because it clarifies what really changed in March 2026. The shareholder meeting did not only bring in new money. It also changed the quality of Castro's holding. Following the new investment, external investors were allocated about 29% of Radixis's equity and voting rights, a 4% fully diluted option pool was created for TenenGroup and its people, and Ron Rotter remained at 9.9%. Castro itself stayed at about 54% of the equity and voting rights, before full dilution.

That 54% figure can sound comfortable because it still means control. Economically, though, it is a very different position. Castro moved from 79.9% through preferred shares with preferential capital-return rights to about 54% through ordinary shares only. The transaction included a capital flattening so that all shares of the subsidiary were converted into ordinary shares and all holders ended up with identical rights. Put differently, Castro kept control, but gave up both a meaningful part of the economic upside and the preferred layer that had made the original investment more protected.

That is the exact point at which a strategic option starts looking like a heavier capital-allocation commitment. If the venture had stayed under a similar control structure and the same preferred-share protection, one could still argue that Castro retained most of the upside under a partial safety net. After March 2026, that is no longer the right framing.

The chart is intentionally simple. It does not try to describe the entire cap table. It isolates the one number that matters most for Castro shareholders: control stayed, but the economic share dropped sharply.

The Headline Is $4.5 Million, but the Funding Is Less Clean Than It Sounds

The funding layer also looks cleaner in the headline than it does in the details. A disclosed total investment of about $4.5 million gives the impression of a neat external financing round. The actual breakdown tells a more mixed story.

TenenGroup committed about $3 million, but not all at once. The first installment, $1 million, was paid upon completion of the transaction on March 5, 2026. The second installment is due within 60 days of completion, and the third within 120 days. At the same completion date, about $480 thousand was invested by additional private investors, about $550 thousand was invested by Castro itself, beyond the $6 million already approved for investment in April 2024, and the remainder was invested by existing shareholders, including Ron Rotter, who is expected to invest about $450 thousand through his preemptive right.

That breakdown matters because it shows that this is not a clean round in which an outside investor fully takes over the financing burden while Castro simply preserves control. At completion itself, only one third of TenenGroup's commitment had actually entered. Castro and existing holders were still writing checks at the same moment. Radixis therefore did not truly leave the category of capital consumer. It merely shifted to a structure where the burden is shared between new outside capital and continued support from existing holders.

That is a meaningful distinction. If all $3 million from the lead investor had arrived at closing, the deal would have been easier to read as a strong third-party validation and as a reduction of pressure on the parent. That is not what was disclosed. What was disclosed is staged funding, while Castro itself still had to add another $550 thousand on top of its previous commitment.

This chart is not meant to show a cumulative total. It shows the character of the financing. TenenGroup is the lead investor, but at closing Castro and Ron Rotter were still participating in the actual funding. That makes it hard to describe the move as one that already shifts Radixis into a largely externally funded activity.

Governance Is Part of the Price, Not a Footnote

There is another layer the market cannot ignore: the governance layer. The voting form published on January 5, 2026 did not ask shareholders to approve only an investment in a subsidiary. It bundled together the investment agreement, the flattening of Radixis's share capital, the investment by the company and additional current shareholders, Ron Rotter's exercise of his preemptive right, and D&O insurance for the subsidiary. Even at that stage it was clear that this was not just about money. It was a full package of capital, control, and management mechanics.

The March 3, 2026 meeting results then show that all three central resolutions were approved: the investment agreement, the acceleration of Ron Rotter's adaptation grant, and the approval of his terms as CEO of the foreign arm. The related-party note in the annual report states explicitly that on that same date the meeting approved the immediate payment of an adaptation grant equal to four months of salary for Ron Rotter's earlier tenure, and that his terms as CEO of the foreign activity, including as CEO of Radixis, remained unchanged for three years.

None of that proves a governance problem. But it does create a package that needs to be read correctly. When new capital, a management mandate, and an immediate payment that had previously been deferred all move together in the same meeting, the burden of proof rises. Investors are being asked to trust not only the potential of Radixis, but also the corporate framework around it. As long as there is no separate operating disclosure, that means buying more trust and fewer hard numbers.

What the Market Still Does Not Get

This is probably the core of the whole debate. Castro still does not give the market a standalone read of Radixis. Carolina Lemke's foreign activity and Radixis are presented together inside the "other unallocated" bucket because their scale does not meet the quantitative thresholds for a separate reportable segment. In 2025 that combined bucket posted NIS 26.7 million of revenue and an operating loss of NIS 12.4 million.

That is enough to understand that at the consolidated level Radixis is not yet showing up as a result saver. But it is not enough to understand how much of those figures belongs to Radixis itself, how much belongs to Carolina Lemke abroad, or how much cash the US activity is consuming on its own. The market is therefore being asked to absorb Castro's drop from 79.9% to about 54% before full dilution, the loss of the preferred-share layer, and a 4% fully diluted option pool, without getting a basic disclosure of sales pace, required funding, or the point at which the US activity is expected to stop consuming capital.

That is where the long duration of the agreement comes back into the picture. The 2045 exclusivity and the ten-year business plan create a wide runway. That can be an advantage if the platform is built at the right pace. It can also become a disadvantage if what enters before proof is another round of capital, options, and management attention.

Conclusion

Radixis has not lost its strategic potential. There is a long exclusivity agreement, a large addressable market, a new lead external investor, and a structure that still lets Castro retain control. But after March 2026 the activity can no longer be read as if it were just a free option on the United States.

The cleaner thesis now is harsher and simpler: Radixis is first and foremost a capital-allocation test for Castro. The company kept control, but did so at the cost of material dilution, the loss of the preferred layer, another self-funded contribution, and a denser governance package around an activity that still is not being judged on its own operating disclosure.

If separate disclosure eventually arrives, if the remaining investment proceeds are completed according to the disclosed structure, and if Radixis starts showing measurable commercial progress, the market may return to reading this as a strategic move with real upside. Until then, it is more accurate to treat it as another potential capital consumer sitting on top of a group whose core still needs to repair its margin.