We sit in Limassol; we get the IP Box question monthly. Here is our plain summary for tech companies weighing Cyprus—not legal advice, just what we watch in models.
What the regime tries to do
Qualifying intellectual property profit can attract a much lower effective rate than standard corporate tax—often discussed around 2.5% effect on that slice when the substance and nexus tests are met. The spread matters only if you actually generate qualifying IP income with documentation and people where you say they are.
When it helps
- You have capitalised development with clear ownership and compliance.
- Decision-makers and development substance live in the story you tell—and can evidence.
- Profit is not entirely passive licensing to yourself with no real function in Cyprus.
Four cases where we have seen it not pay off
- Pure contract dev shops reselling hours with no durable IP asset to ring-fence.
- US parents expecting IP Box to erase global structure questions—it does not replace transfer pricing work.
- Teams with no local hiring plan beyond a mailbox. Auditors and banks both ask.
- Early pre-revenue labs spending more on compliance than the tax delta for years.
Tiny model (illustrative)
| Item | Standard slice | Qualifying IP slice (illustrative effective) |
|---|---|---|
| Profit before IP split | €1.00m | — |
| Qualifying IP profit | — | €0.35m |
| Blended effect | 12.5% on CoC story | Depends on assets, caps, rules in force |
Swap in your lawyer’s rates and caps—ours is the habit: model the scenario with and without substance cost.
Our read
IP Box is a structuring tool, not a coupon. We recommend it when IP, people, and forecasts already justify Cyprus. We do not recommend it when the only driver is a headline rate on a slide.