TRAF: the national reform was decided file by file and canton by canton

On 19 May 2019, Switzerland voted on more than a tax reform. Within months, the vote changed the questions every management team needed to ask before its next year-end: what would the company lose when special tax statuses ended, what could it document in a patent box, which research and development expenditure belonged in the cantonal calculation, and which scenario would remain defensible after the cap on relief? Behind the federal legislation, implementation in twenty-six cantons made any generic answer risky. This market insight reconstructs the commercial window before the rules took effect on 1 January 2020. It explains why a firm could sell a short review before a broader tax engagement, how to distinguish theoretical savings from an evidence-led option, and which events still create legitimate demand. The purpose is not to promise a tax reduction, but to make a tax decision comparable, assigned and verifiable. General analysis updated on 7 August 2026. It is not a substitute for advice on an individual tax position or a decision by the competent cantonal authority.

TRAF abolished the cantonal regimes reserved for companies with special tax status and introduced a framework compatible with international standards. It opened the patent box, allowed cantons to provide an additional deduction for research and development, and imposed a cap on relief. The reform therefore created a new calculation, not an automatic tax reduction.

Tax Reform and AHV Financing was approved on 19 May 2019 and took effect on 1 January 2020. The former privileges for companies operating mainly internationally were abolished. In return, cantons received new instruments and policy discretion. A company therefore had to leave its historical category and reconstruct its treatment from the underlying economic facts.

The patent box allows reduced taxation of part of the profit derived from patents and comparable rights. Cantons may also provide an additional deduction of up to 50% for certain research and development expenditure. A company must, however, remain taxable on at least 30% of the profit it would have reported before the special reliefs. These national limits do not answer the decisive questions: which rule did the canton adopt, and what evidence does the company hold?

The first commercially viable service was a decision map: establish the relevant canton, compare the before-and-after position, inventory the rights and expenditure, then list the options requiring further examination. The firm sold neither a tax favour nor approval in advance; it sold the quality of the analysis delivered to the decision-maker.

A national answer was insufficient because cantons selected different parameters within the discretion left by the reform. The ordinary rate, patent-box reduction, research and development deduction and overall cap could vary. Two comparable companies in different cantons did not therefore necessarily follow the same path.

The reform created a shared grammar, but each cantonal law set several of its values. A useful comparison had to retain the same profit base, distinguish transitional effects from recurring ones and avoid stacking relief beyond the applicable cap. The calculation also had to separate profit derived from an eligible right from general trading profit.

This tax geography created an easily segmented pool of accounts: former special-status companies, patent-owning businesses, structures funding research in Switzerland and groups considering a cantonal location. A public signal was never enough to establish eligibility. It allowed a more precise question: do your year-end and canton make a review of the TRAF instruments worthwhile?

The appropriate first purchase was a focused cantonal review covering the starting position, potentially applicable instruments, available evidence, quantified scenarios and questions requiring confirmation. It could be invoiced promptly because it had a clear scope and outcome. It preceded—but did not require—the preparation of a tax return or ruling request.

The review begins with an interview involving financial management and those responsible for innovation. It then reconciles the accounts, intellectual property, research contracts and location of the teams. If the evidence does not isolate a profit stream or item of expenditure, the scenario remains explicitly unproven. That transparency is more valuable than an attractive number built on hidden assumptions.

The report compares at least a baseline scenario with one supported by evidence. It shows the estimated tax difference, preparation cost, dependencies and owner of each approval. The partner’s fees, start-up time and monthly capacity must be confirmed before any acquisition begins.

To avoid a misleading promise, every saving must be tied to cantonal law, a period, a calculation base and evidence held by the company. The scenario must state what is confirmed, what depends on interpretation and what remains excluded. The theoretical difference is never presented as an amount secured or received.

A robust model starts with profit before relief and ends with an estimated tax charge that can be reconciled. Each measure appears separately between those points. The overall cap is applied after the relevant combination; it must not be overlooked merely because each instrument was calculated in isolation. Effects on capital tax or hidden reserves require an assessment specific to the case.

How to read the diagram. The accounts, evidence and cantonal law precede the model. An unknown is not hidden in the result; it becomes an assigned question.

This approach also protects the commercial relationship. If the option produces no net saving or the documentation cost outweighs the benefit, the firm can recommend retaining the ordinary treatment. A reasoned refusal makes the engagement credible and prevents acquisition from relying solely on a promise to pay less tax.

The priority audience comprised former special-status companies, businesses with patents or research in Switzerland, groups reorganising their functions and management teams planning a cantonal establishment. Priority came from combining an observable event with an imminent decision, not from an undifferentiated list of every company.

In 2019, the announced end of a status was a direct trigger. Today, the signals are different: filing or acquiring a patent, creating a research team, transferring a function, changing canton, undertaking a merger or reaching the first profitable year after an investment phase. Each can reopen the question without artificially recreating the urgency of 2020.

The Accountancy and fiduciary services market page can direct these situations to the relevant articles. Within a campaign, organic search serves management teams already looking for an explanation; referral networks provide trust; and email and telephone follow only after a professional event has been qualified. Individual tax data is never used as prospecting material.

A campaign is measured first by paid reviews, the time from signal to purchase, delivery margin and follow-on work actually invoiced. A visit, response or meeting indicates interest, not revenue. Potential value remains separate from revenue received, and no confidential client result is reconstructed.

The journey may begin with a note explaining the cantonal decision, continue with contact based on a documented event and lead to the review. If the review identifies a viable file, tax-return preparation, annual documentation and review after a change become distinct engagements. If no option is robust, the engagement closes with a reasoned recommendation.

The review becomes recurring when new facts require a fresh decision: annual closing, a new patent, a change to a research project, reorganisation, a group transaction or a cantonal move. Continuity must be triggered and invoiced by an identifiable need. It cannot take the form of a vague subscription to the reform.

The partner can offer an annual collection timetable, checks of expenditure allocation and an alert after a structural event. Every cycle has an evidence list, an owner, a price and an end date. Management retains control of the engagement and understands why further work is necessary.

This continuity requires genuine organisation: monthly capacity, response times, specialist review, confidentiality and an escalation process. getfishnet can support a campaign only when the partner confirms these points, its target margin and its ability to decline out-of-scope cases.

The essential limits are the absence of any tax guarantee, cantonal differences, reliance on the quality of evidence and the need for appropriate professional expertise. TRAF is not a uniform product. Every campaign must stop a file where the canton, period, substance or paying client cannot be established.

The tax authorities retain their discretion. A public article does not replace an assessment of an individual situation, and a firm must not imply that one mechanism is sufficient to validate relief. The acquisition message therefore concerns an evidence-led review, not a minimum rate or promised gain.

The authoritative sources are the Federal Department of Finance, the Federal Tax Administration, TRAF legislation and the applicable cantonal publications. They establish the framework and its limits. They publish neither advisory prices nor conversion rates or expected partner revenue.

The free eligibility test examines your specialism, acquisition challenges, account pool, initial review and delivery capacity. It validates no tax treatment. Its purpose is to decide whether a tailored strategy can connect a documented event with a useful purchase and a durable relationship.