JV or Contract Manufacturing in Ukraine: Which Structure Fits

Both routes end with your product being made in Ukraine. They differ in what you own when it is.

A third arrangement sits between them and is worth naming because it is often what people mean when they say “JV”: a wholly-owned local subsidiary. It gives full control without a partner, and it also gives you none of a partner's existing plant, workforce, security clearances or customer relationships — which in a wartime defense market is most of what makes a footprint work quickly.

One route runs the other way, and is worth naming so these two are not read as the whole menu: instead of making your product in Ukraine, a foreign manufacturer can license a Ukrainian product and build it abroad. Ukraine's Ministry of Defence licenses items, so a proven Ukrainian design can be produced under licence in your own facility — and even sold back into Ukraine. A lighter version is white-label: a Ukrainian maker produces to your brand as a vendor. Both keep the counterparty at arm's length and neither gives you a footprint inside Ukraine, which is the trade-off the rest of this piece is about.

Read the last two rows first. Capital and reversibility are what most programmes actually decide on, and the rows above them explain why.

Under a supply agreement your influence is whatever the contract says: specifications, acceptance criteria, audit rights, penalties and termination. That is real leverage over the units you have ordered, and no leverage at all over what the plant does the rest of the time — its other customers, its investment decisions, whether it expands the line you depend on, or who else it licenses.

A joint venture converts that into governance. Board composition, reserved matters, the annual budget, the appointment of the managing director and the deadlock mechanism are where a JV is really negotiated, and a shareholding percentage on its own tells you less than the reserved-matters list. Where the majority sits still matters: in the sector's most-cited case, Rheinmetall holds 51% of its venture with the state-owned Ukrainian Defense Industry, which put the foreign partner in control of a manufacturing operation opposite a Ukrainian state counterparty.

The instinct is that contract manufacturing protects IP and a JV exposes it. The reality is that both routes put a technology package in the hands of a Ukrainian entity, because neither produces anything otherwise. What differs is the governance of that transfer and the ownership of what comes next.

• Under a contract, the licence is narrow and time-boxed, tied to the contracted output, with the design remaining yours. The weakness is enforcement distance: you are policing use through a commercial agreement with a company you do not sit inside.
• Under a JV, core IP normally stays with the parent and is licensed in, while improvements made on the line — process know-how, tooling design, adaptations to local components — vest in the venture unless the agreement says otherwise. In a fast-iterating defense product, those improvements can be the most valuable thing produced.
• In both cases, the IP boundary has to be drawn before production starts. Deciding afterwards means negotiating with a counterparty who already holds the know-how.

This is where the two routes are least alike. A supply agreement with a plant that already exists needs a specification, a technology-transfer package, a quality regime and a price — all of which can be negotiated in the time it takes to draft a shareholders' agreement, and none of which requires construction. A joint venture needs incorporation, an SHA, capital contributions, permits and, if the plan involves a new site, a building.

Construction risk deserves to be modelled rather than assumed. Rheinmetall's Ukrainian programme had grown to four announced factories by the end of 2025 — armoured vehicles, artillery ammunition, air defence and propellants — with the repair and vehicle site starting first. The ammunition plant, announced in February 2024 and pointed at a 2026 start, had not broken ground as of May 2026. That is a well-capitalised prime with state-level backing, which is a useful calibration for anyone planning a first footprint on a two-year horizon. The programme is walked through facility by facility in the Rheinmetall model, including which layer moved in months and why.

Reversibility is the dimension that gets discussed last and matters most to a board. A supply agreement ends when its term ends, or earlier on the terms you negotiated; the cost of exit is commercial, and the residual question is what the counterparty retains of your technology.

A shareholding needs a buyer. Exit provisions — tag and drag rights, pre-emption, put and call options, valuation mechanics and deadlock resolution — are what turn a stake into something you can actually leave, and they are worth more attention than the headline percentage. Where the partner is a state-owned enterprise, transfer of a shareholding in a defense asset also engages approvals that a commercial buyer cannot supply on its own, which is a reason to write the mechanics down while everyone is still enthusiastic.

Presented as a binary, the two routes look like alternatives. In practice they are frequently stages, and the market is visibly running the sequence: partnerships and supply relationships are being signed with a stated intent to convert into ventures once the work proves out.

1. 1Contract a limited run.A defined volume to a defined specification, with a narrow technology licence. You learn the plant's real capability, its quality discipline and its management — none of which is visible in a data room.
2. 2Expand the contract and the transfer. More volume, more of the process localised, quality regime tightened to the standard your buyers audit against.
3. 3Convert when the economics justify equity. When the volume is durable, the counterparty is known and the local operation is generating value you would rather own than buy, the venture becomes the cheaper structure.

Running the stages in this order also prices the risk correctly: the expensive, hard-to-reverse commitment is made after the cheap, reversible one has produced evidence.

Four questions usually settle it, and they are worth answering in writing before the first partner meeting.

• How long is the demand? A programme with a defined end-point rarely justifies equity. A capability you expect to sell for a decade does.
• How much of the value is in the process? If the hard part is the design and the plant is assembly, contract. If the hard part is manufacturing at rate and quality, the operation is the asset and you should own part of it.
• Do you need to be a Ukrainian producer? Some outcomes — eligibility, standing with the domestic customer, access to the local regime — attach to the local entity, and a supply contract does not confer them.
• What does your board need to be able to undo?If the answer is “everything, within a year,” the structural choice has already been made for you.

Localizing is one of six routes into Ukraine's defense market, and it is worth confirming it is the right one before optimising the structure — the operator's guide to market entry sets the routes against each other, and how a foreign supplier gets paid covers the funding side that determines whether the volume you are structuring around will actually be purchased.