Cyprus IP Box when there is no R&D spending

Three situations need different answers

A founder saying “we have no R&D spending” may mean that development finished last year, that the company bought finished software, or that nobody has assembled the cost records. These are materially different facts. Identify which situation applies before using a calculator or quoting a tax rate.

The nexus calculation connects qualifying profit with the relevant expenditure history of the asset. It is not simply a test of whether the current year’s profit-and-loss statement contains an account called R&D. Accounting classifications can obscure development expenditure, and historical costs may remain relevant.

Conversely, the existence of valuable software does not prove that the claimant incurred qualifying expenditure. Unpaid founder effort, acquired code and development funded by another entity require analysis of the actual taxpayer, transaction and costs.

Compare the numerator and denominator explicitly

The examples below isolate the mathematical issue. They assume that any stated expenditure is correctly classified and belongs to the relevant asset and taxpayer.

No current-year expenditure is not the same as no history

A business may complete the main development of a product and then earn subscription income for several years. If the relevant historic qualifying expenditure is supported and the remaining conditions are met, an absence of new development in a particular year does not by itself erase that history.

Keep the asset schedule available even when the development team has moved to another project. It should explain original development, acquisitions, related-party work and later improvements. Reconcile changes and preserve references to the underlying records.

Do not shift current expenditure from a new product to an older profitable asset merely to improve its fraction. The expenditure must have the required connection with the relevant asset. Historical tracking is valuable precisely because it preserves that connection over time.

Unpaid work and transferred code cannot be filled with invented costs

A founder may have written substantial software before the company existed. The commercial value of that work can be considerable, but value and expenditure are different concepts. A notional salary inserted retrospectively into the company’s spreadsheet does not establish expenditure that the company actually incurred.

If the founder transfers the software to the company, determine what rights were transferred, on what terms and for what consideration. The transaction may raise acquisition, valuation and cross-border questions. It is not automatically equivalent to the company having employed the founder throughout the original development.

Subsequent genuine development by the company may create qualifying expenditure if the conditions are satisfied. Model that work using actual supported amounts and the correct asset history, rather than backdating documents to manufacture a favourable starting position.

What to do when the model shows zero

First, stop treating zero as a default for an unanswered question. Mark unknown fields as unresolved. Obtain accounting records, contracts, payroll and development evidence, and identify who incurred the costs. Determine whether the relevant asset was created, acquired or improved by the claimant.

Second, separate the arithmetic from the legal conclusion. A calculator can show that €0 divided by €100,000 is zero. It cannot establish that the acquisition amount is complete, that the asset qualifies or that a founder’s transfer was correctly treated. Those inputs need their own support.

Finally, document the conclusion and its limitations. If no qualifying expenditure can be established, do not advertise a full-nexus tax rate. If supported historical costs exist, use the appropriate expenditure history and retain the evidence for future periods.

• Ask whether zero means none, none this year, or unknown.
• Identify the claimant and the asset’s development or acquisition history.
• Recover genuine records without presenting later estimates as original documents.
• Calculate uplift only after QE and the other costs are established.
• Use a model that flags a zero denominator instead of assigning 100%.