Skip to content

Research Allowance in the Automotive Trade: How Our Tax-Case Engine Became a Recognized Research Project (FZulG)

Julian Alessio Goßen (Bachelor of Taxation) July 09, 2026 18 min read Research Allowance FZulG Research & Development Automotive Trade VAT Contract Research

In brief

The research allowance under the Forschungszulagengesetz (Research Allowance Act, FZulG) is a tax incentive for research and development, regardless of legal form or industry. It covers 25% of eligible expenses, or 35% for small and medium-sized enterprises. For contract research, 70% of the fee qualifies. The allowance is credited against income tax, and any excess amount is paid out.

On this page (13 sections)

Research Allowance in the Automotive Trade: How Our Tax-Case Engine Became a Recognized Research Project (FZulG)

Some of the most important problems appear unremarkable from the outside. “Determine the correct VAT treatment for a vehicle sale” is one of them. It sounds like checking a box on a form, but it is actually one of the most difficult tasks that vehicle dealers must solve every day and at scale.

For us at Autaxo, that was reason enough to turn it into a dedicated research and development project. An independent body has now confirmed that we were right: the Bescheinigungsstelle Forschungszulage (Research Allowance Certification Office, BSFZ) has officially recognized the project as research and development.

If you first want to see the most common errors in the overall VAT system that this article addresses from a technical perspective:
→ Common VAT mistakes in the used car trade

This article explains why seemingly dry tax logic conceals a genuine technical problem, how we solved it, and how Germany’s research allowance works, from recognition as R&D to the calculation.

🧰 Related tools and resources for checking and calculating:

Quick Definition: What Is the Research Allowance (FZulG)?

The research allowance under sections 1 et seq. of the Forschungszulagengesetz (Research Allowance Act, FZulG) is a tax incentive for research and development, regardless of legal form or industry. It covers 25% of eligible expenses, or 35% for small and medium-sized enterprises. For contract research, 70% of the fee qualifies. The allowance is credited against income tax; if it exceeds the tax due, the remainder is paid out.

📚 Knowledge base:
Used Car Trade Knowledge Base: Overview and Quick Start

Key Facts

  • Recognition: The BSFZ classified our rules-based tax-case engine as experimental development: novel, involving genuine technical risk, and carried out systematically.
  • The project: Automatically derive the applicable VAT treatment in the automotive trade from transaction data, including margin scheme taxation, intra-Community supplies, third-country exports, and chain transactions.
  • The core: A formal decision and state model with priorities, deliberate hard stops, a traceable rule path, and versioning.
  • Funding amount: 25% of the assessment basis, or 35% for SMEs. For contract research, 70% of the fee qualifies, resulting in effective funding of up to 24.5% of the contract fee (70% × 35%) for SMEs.
  • Payment: The allowance is credited against income tax; any excess amount is refunded, making it attractive even for young companies that are not yet profitable.
  • Two stages: First, obtain the BSFZ certificate (stage 1), then submit the application to the tax office through “Mein ELSTER” (stage 2).

Why the Tax Question Is So Difficult at a Car Dealership

Whether a vehicle transaction is subject to margin scheme taxation, a VAT-exempt intra-Community supply, a third-country export, or standard taxation cannot be determined from a single data point. It depends on an entire set of interdependent characteristics: How was the vehicle purchased? Who is the buyer, and in which country is the buyer established? Is there a valid Umsatzsteuer-Identifikationsnummer (German VAT identification number, USt-IdNr.)? How old is the vehicle, and what is its mileage? And is all required evidence complete?

The real difficulty is not the number of these factors, but how they interact. They influence one another, often in ways that someone without deep specialist knowledge would not expect.

For example, a vehicle purchased under the margin scheme and sold to a business in another EU country with a valid VAT ID simultaneously meets the conditions for two tax treatments that are actually mutually exclusive. The data alone does not determine which one applies. It depends on whether a tax option is exercised. At the same time, the type of purchase may rule out a particular tax treatment for the sale from the outset. Even tax exemptions do not operate symmetrically: what applies to an export to a third country does not necessarily apply to an intra-Community supply.

Dealership staff processing such cases today generally select the applicable treatment manually or rely on preconfigured tax codes and isolated checks. That works as long as the cases are simple. For the special cases that are routine in the used car and cross-border trade, it is error-prone, difficult to trace, and does not scale with the business.
(For more detail, see Margin scheme taxation in the automotive trade under section 25a UStG.)

It Does Not Stop at One Transaction: The Chain Transaction

Things become truly demanding when several sales merge into a single movement of goods. That is precisely what happens in a chain transaction: several businesses enter into successive purchase agreements for the same vehicle, but the vehicle is transported directly from the first supplier to the final customer. There are therefore several supplies, but only one physical movement.

For tax purposes, only one of these supplies can be the “moving” supply, and only that supply can be treated as a VAT-exempt intra-Community supply or export. All others are “stationary” supplies that are taxed where the vehicle begins or ends its journey. Which supply is the moving supply depends on who is responsible for transport and, if a dealer is in the middle of the chain, which VAT ID that dealer uses.

For the dealer, this means that whether its own sale may be treated as VAT exempt or must show VAT, and in which country, is determined not by its transaction alone but by its position in the chain and responsibility for transport. If this is assigned incorrectly, the consequences may include tax being due in the wrong country, an exemption being granted without justification, or, in the worst case, double taxation. For an automated model, this is the most advanced scenario. It must assess not just one transaction, but the interaction of several linked transactions.

The Tax Case Never Stands Still

Even after the correct case has been determined, the work is not finished. The basis for the decision changes repeatedly over the course of a transaction: a buyer turns out to be a business rather than a private individual. Evidence is submitted later or expires. An invoice is canceled and reissued. Each of these changes can make the originally correct tax treatment wrong after the fact.

This creates a deeper problem than it initially appears, a kind of chicken-and-egg relationship. The tax decision determines which evidence is required. Conversely, the status of the evidence determines whether the decision is permissible at all. If one changes, the other must be reassessed without losing the original rationale. This is exactly where simple solutions fail. Changing the tax case without a clean recalculation causes inconsistencies between the decision and everything that depends on it: the invoice, accounting entry, and evidence.

Another practical example involves reimported vehicles, for which the date of first registration is often missing. Available vehicle data sources sometimes provide the manufacturing date instead of the date of first registration. A single gap of this kind can make it impossible to determine whether a vehicle is “new,” potentially changing the entire tax treatment. Assumptions that are too lenient produce incorrect approvals, while assumptions that are too strict block permissible transactions.

What We Developed

Our approach is a rules-based tax-case and evidence engine that automatically derives the applicable VAT treatment from structured transaction data. At its core is a formal decision and state model that we developed in several stages.

First, we formalized the relevant tax cases, their priorities, and their mutual exclusions, translating legal language into an unambiguous structure that a machine can evaluate. Building on this structure, a decision mechanism derives the appropriate case from the available data. Where several regimes compete, it resolves the conflict according to clear precedence rules and evaluates a transaction not only in isolation but also in the context of its position in a supply chain. Where a configuration is impermissible or the data is insufficient, it deliberately stops and requests manual clarification instead of silently making a risky assumption.

Every decision is supported by a traceable rule path. This means that it is possible to see afterward exactly why a particular tax treatment was selected, in a form that the dealer, tax advisor, and tax auditor can all verify. Because data changes, the model is also versioned. It cleanly distinguishes old states from new ones and consistently reassesses a previous decision when the facts change.

This brings together processes that standard systems handle through separate, manually coordinated paths: deriving the tax treatment, managing evidence, and mapping the transaction in financial accounting, all within one consistent and contradiction-free logic.
(For the practical impact on records and evidence, see GoBD at a car dealership and the vehicle trade register.)

Why This Is Research, Not Routine Work

The crucial point is that, at the outset, it was unclear whether this could be achieved at all. The open, technically risky question was whether the full range of special cases could be translated into a complete and consistent formal model, or whether too many cases would remain that could only be resolved through non-generalizable, case-by-case decisions. It was also possible that the precedence rules would form circular dependencies and make an unambiguous sequence impossible.

This very risk, the real possibility of failure, is a hallmark of genuine research. We planned for it from the beginning, with clear termination criteria and a systematic test matrix for detecting and classifying error cases. The BSFZ evaluates projects according to the criteria of novelty, technical uncertainty, and a systematic approach under section 2 FZulG. Passing this assessment independently confirms that we are not merely implementing established knowledge, but have entered genuinely new territory.

What Is the Research Allowance and How Is It Calculated?

The research allowance is the second part of the story and an underused instrument for many small and medium-sized businesses. It is not a project-specific grant that a company must apply for and “win,” but a statutory entitlement for eligible research and development.

Which expenses qualify? Personnel costs for in-house research, the owner’s own work in sole proprietorships and by partners or shareholders, valued at a flat rate, and contract research, meaning R&D commissioned from external service providers under section 3 FZulG.

How much is it? The research allowance equals 25% of the assessment basis, or 35% for small and medium-sized enterprises under section 4 FZulG. For contract research, 70% of the fee paid is treated as an eligible expense. For an SME, this produces effective support of up to 24.5% of the contract fee (70% × 35%). The annual assessment basis is capped, but the cap is in the millions and therefore not a practical limit for the vast majority of SMEs.

How is the money received? The allowance is not a grant. It is assessed by the tax office and credited against corporate or personal income tax under section 8 FZulG. If the tax is lower than the allowance, as is often the case for young companies, the excess is paid out. The research allowance therefore also has an effect when a company is not yet generating profits.

What is the process? It has two stages. First, the company submits an application for certification to the BSFZ under section 6 FZulG; the BSFZ determines only whether a project qualifies as R&D. The company then applies for the actual research allowance to the tax office through “Mein ELSTER,” where the eligible costs are reviewed and assessed.

Note: The specific assessment, particularly for contract research and the allocation of individual costs, depends on the circumstances of each case and should receive professional tax support.

What This Means for Dealers

This is not an abstract topic for our customers. A tax case that is derived automatically and correctly means less manual work, a lower risk of errors, and, perhaps most importantly, a traceable rationale whenever the tax office or tax advisor asks questions. Particularly in trade involving EU and third countries, reimports, chain transactions, and all cases in which the available data changes later, Autaxo removes a layer of complexity that currently costs considerable time and effort and can become expensive if handled incorrectly.

Recognition as research also demonstrates that we are not building just another standard tool. We are developing in depth precisely where daily operations hurt most. Learn more about the team and the BSFZ recognition on About Autaxo.

FAQ About the Research Allowance and the Tax-Case Engine

What is the research allowance and who is eligible?
The research allowance is a tax incentive for research and development under the FZulG. It applies regardless of legal form or industry; in principle, all taxable companies with eligible R&D projects may claim it.

How much is the research allowance?
It equals 25% of eligible expenses, or 35% for small and medium-sized enterprises. For contract research, 70% of the fee qualifies, resulting in effective support of up to 24.5% of the contract fee for SMEs.

Does my company have to make a profit to benefit?
No. The allowance is credited against income tax; if it exceeds the assessed tax, the remainder is paid out. For young companies that are not yet profitable, it can therefore function like a direct refund.

What is the process?
It has two stages: first, obtain certification from the Bescheinigungsstelle Forschungszulage (Research Allowance Certification Office, BSFZ), which only determines whether a project qualifies as R&D. Then apply for the research allowance to the tax office through “Mein ELSTER,” where the costs are reviewed and assessed.

What does tax software have to do with research?
The funded activity is not the software product itself, but the experimental development behind it. At the outset, it was technically uncertain whether the special VAT cases in the automotive trade could be translated into a complete and consistent formal decision model. That risk and the systematic approach make it research within the meaning of the FZulG.

Checklist: Does My Project Qualify for the Research Allowance?

  • ☐ Novelty: Does it go beyond known standard solutions, rather than merely applying existing knowledge?
  • ☐ Technical risk: Was it uncertain at the outset whether and how the objective could be achieved?
  • ☐ Systematic approach: Is there a structured work plan with work packages and intermediate objectives?
  • ☐ Expense categories clarified: Have personnel costs, the owner’s own work, and/or contract research been allocated cleanly?
  • ☐ Ongoing documentation: Are time recorded by work package, supporting documents, notes on edge cases, and failed attempts documented?
  • ☐ Stage 1: Has the application for certification been submitted to the BSFZ under section 6 FZulG?
  • ☐ Stage 2: After certification, has the application for the research allowance been submitted to the tax office through “Mein ELSTER”?
  • ☐ For contract research or related parties: Can arm’s-length fees and actual payments be documented?

Further Articles in the Autaxo Knowledge Base

Legislation and standards:

  • Section 2 FZulG: Eligible research and development projects (basic research, industrial research, and experimental development) and their criteria.
  • Section 3 FZulG: Eligible expenses (personnel costs, the owner’s own work, and contract research at 70% of the fee).
  • Section 4 FZulG: Amount of the research allowance (25%, or 35% for SMEs) and assessment basis.
  • Section 6 FZulG: Application for certification by the Bescheinigungsstelle Forschungszulage (BSFZ).
  • Section 8 FZulG: Assessment and crediting of the research allowance.

Authorities and official sources:

Note: This article provides general information and is not a substitute for professional tax advice on an individual case. The specific assessment of eligibility, cost allocation, and calculation depends on the circumstances of each case.

Put it into practice with Autaxo

Tax case engine for §25a, EU, and exports Automatic tax treatment for margin scheme, intra-community, and export workflows Pricing and free trial Autaxo ONE: €49/month, cancel monthly
Back to the blog