CDTI: Public Credit for Innovation
Most companies approach public innovation funding in Spain expecting a grant. What CDTI mostly provides is credit: money that has to be repaid, advanced against a project rather than against collateral, on terms no commercial bank would offer for the same risk. That distinction, credit rather than gift, is the single most useful thing to understand before preparing an application, because it changes what the institution examines, what the file must demonstrate, and what the money will look like on the balance sheet afterwards.
CDTI is the Spanish public entity that finances business research, development and innovation. It does not evaluate companies the way a bank does, and it does not evaluate projects the way an investor does. It applies its own logic, technical first and financial second, and companies that present themselves to it with a bank file or an investor deck tend to fail for reasons they never quite understand. What follows sets out the mechanism, the criteria and the file, so that if an application fails, it at least does not fail on preparation.
Credit against a project, not against collateral
The unit of analysis at CDTI is the project, not the company. A bank asks who you are, what you earn and what security you can offer. CDTI asks what you intend to develop, why the outcome is technically uncertain, and whether your team can carry the work to the end. For the standard R&D project line, CDTI covers up to 85 per cent of the approved budget, and it states the corollary plainly: the company contributes at least 15 per cent of the project from its own resources. The minimum eligible budget is EUR 175,000, projects run between twelve and thirty-six months, and repayment is over ten or fifteen years including a grace period of two to three years. The rate is fixed at one-year Euribor, set on the date the board approves the aid, and where Euribor is negative the rate applied is zero, disbursed against the execution of that budget. Some instruments include a tranche that does not have to be repaid, and this is the figure most often quoted carelessly. CDTI publishes a non-repayable tranche of between 10 and 33 per cent of the aid, but that range is not a negotiating band: it is a table of categories. The 33 per cent belongs to one specific type, oriented projects. Small and medium companies with European co-financing in certain regions reach 20 or 30 per cent. For an ordinary R&D project by a small or medium company with no European co-financing, the realistic figure is up to 17 per cent, and up to 10 for a large company. There is a second precision that matters just as much: the non-repayable tranche is calculated on a maximum coverage of 75 per cent of the approved budget, so the increase in coverage up to 85 per cent generates no additional non-repayable amount. All of it remains subject to available funds.
This structure exists because the risk CDTI accepts is precisely the risk a commercial lender declines. A project whose technical outcome is uncertain cannot show the predictable cash flows that bank credit requires, and no amount of presentation changes that. Public innovation credit fills the gap deliberately. It is not a reward for being innovative. It is a lender of a different kind, one whose mandate allows it to carry technical uncertainty that the private market will not price at any reasonable cost.
It follows that CDTI funding is still debt. It sits on the balance sheet, it has a repayment calendar, and every future lender will read it as an obligation like any other. Companies that treat it as free money discover this at the worst possible moment, when a bank analysing a later financing asks how the public loan will be serviced alongside the new one. The correct frame from the first day is that of a borrower: the money is more patient than the private alternative, and it must still come back.
The eligible budget is also narrower than the project as the company sees it. Personnel dedicated to the development, materials consumed by it, specific subcontracting and certain overheads can qualify; general commercial activity does not. One published limit is worth knowing before a budget is drafted, because it reshapes many projects: external collaborations may not, as a general rule, exceed 65 per cent of the eligible budget. A project designed around subcontracting most of its technical work will meet that ceiling and have to be restructured. Building the budget outward from the eligible categories, rather than trimming a general company budget afterwards, is what produces a file that survives review without rework.
What counts as innovation, and what does not
The threshold question is not whether the result will be new to the company, or even new to the market. It is whether reaching the result requires resolving a technical uncertainty. Buying an advanced machine is investment. Integrating known technologies in a known way is engineering. Developing a product or process whose feasibility is genuinely unresolved at the outset is what this credit exists to fund, and the evaluation is designed to tell the three apart.
Spanish public funding distinguishes between research, development and innovation, and the classification is not decorative. Oriented projects, to take the clearest example, sit apart from the ordinary line: they carry a minimum eligible budget of EUR 2,000,000 and a maximum of EUR 30,000,000, and their rate is calculated loan by loan under the European reference rate communication less fifty basis points. A company that oversells routine development as frontier research does not obtain better terms. It obtains a technical evaluation it cannot survive, because the memory is read by specialists who know what the state of the art in the field actually is, and the gap between claim and substance is exactly what they are trained to find.
Honest classification is therefore a strategy, not a concession. A well-framed development project with real but modest technical risk is fundable. The same project dressed as research is not. The strongest applications state plainly what is already known, what is not, and which of those unknowns the project will resolve. That candour reads as competence, because it is the same structure of thought the evaluator applies, arrived at before the evaluator has to apply it.
The project must also be bounded. A permanent ambition to innovate is not a project; a defined development with a start, an end, work packages and verifiable milestones is. The structure is not bureaucratic decoration. Disbursements and later justification are tied to it, and a project that cannot be described in milestones cannot be administered after approval, whatever its technical merit. Companies that resist this discipline at the application stage pay for it during execution, when every deviation has to be explained.
The file CDTI expects to read
The core of the application is the technical memory, and it is a document with a specific job: to establish the state of the art, describe the advance the project proposes over it, and lay out the plan that gets from one to the other. It is not a brochure and it is not a business plan. The commercial opportunity matters, but it enters as context, as the reason the company is committing its own resources, not as the argument for funding. The argument for funding is technical.
The budget is the second pillar, and it must be real and auditable. Personnel hours by profile, materials, subcontracted work and the basis for every figure, because each of them will have to be justified against the plan when the money is spent. A budget assembled to reach a desired funding amount, rather than built up from the work itself, shows its seams under review. Evaluators have read thousands of budgets and recognise a reverse-engineered one quickly.
The financial dimension comes third but it is not optional. CDTI funds a share of the project, never all of it, and the company must carry at least 15 per cent of the project budget from its own resources and demonstrate that it can sustain the activity while the work runs. There is relief on the cash flow side: CDTI publishes an advance of up to 50 per cent of the aid with a limit of EUR 300,000, and states that it does not require additional guarantees for it. A technically excellent project inside a company that cannot show the capacity to finish it will not be funded, because the institution’s real risk is not that the technology fails but that the project stops halfway.
Finally, capacity: the team, its track record and the means available. Prior projects completed, the technical staff who will actually do the work, the facilities and equipment involved. Where capabilities are missing, the file should show how they are covered, through hiring plans or subcontracting, rather than hoping the gap goes unnoticed. It will not go unnoticed. An evaluator who finds one silent gap starts looking for others.
Where applications fail
The most common failure is tone. Companies write the memory as marketing, full of leadership and disruption, and empty of the one thing the reader is looking for: a precise statement of the technical problem and a credible plan to solve it. The document is being read by an engineer, not a customer. Every adjective that replaces a specification weakens the file, and a memory that could have been written without doing any technical work reads as exactly that.
The second failure is the backward budget, built from the funding the company wants rather than from the work the project needs. It produces inflated personnel allocations, vague subcontracting and materials lines that do not map to the work packages. Even when it passes evaluation, it fails later, at justification, when the spending has to match the plan and does not. The audit stage is where optimistic budgets become repayment problems.
The third failure is financial. The company presents a strong project and weak accounts: little equity, stretched working capital, no explanation of how its share of the budget will be funded. The evaluation does not ignore this in favour of the technology. It reads the whole, and a project the company visibly cannot afford to finish is declined regardless of merit. Strengthening the balance sheet, or documenting the cofinancing, is part of preparing the application, not a separate exercise.
The last failure is timing. Public funding operates on an incentive logic: it exists to make projects happen that would not otherwise happen, and costs incurred before the relevant date can fall outside the eligible budget. The date belongs to each call and to each file, and it is the single most expensive thing to assume. Unlike the regional calls, this line does not open and close: CDTI states that the R&D project call is open all year, which removes the calendar excuse and leaves only the preparation. A company that signs contracts and starts spending, then applies, can find it has converted eligible costs into ineligible ones. The order of operations is part of the mechanism, and it is unforgiving.
The place of public innovation credit in a financing plan
CDTI credit is a layer, not a plan. It funds the innovation component of a company’s activity on terms nothing else matches, and it funds nothing else. The company still needs working capital, still needs investment finance for assets that are not research, and still needs its banking relationships. Treating a public programme as the financing strategy, rather than as one instrument within it, leaves the rest of the balance sheet unattended while attention concentrates on a single application.
The layers also have to know about each other. A bank assessing the company will find the public loan and wants to understand its calendar and obligations; the public institution expects the project’s cofinancing to be real and stable. Presenting each side with a picture that omits the other is not sophistication, it is a defect that surfaces at the worst time. The file that works is the one where the whole structure is visible and the repayment of every layer is explained from the same cash flows, once.
Sequencing matters for the same reason. Because eligibility follows the incentive logic described above, the public application generally anchors the calendar and private commitments are arranged around it. Deciding the order of the two conversations before starting either is a structuring decision, and taking it late usually means taking it badly, with a bank offer expiring while an application is still in evaluation, or eligible costs already committed.
The deeper point is that the discipline is the same. A bounded project, a real budget, a demonstrated capacity to finish, a repayment explained from identifiable flows: this is what CDTI requires, and it is what any serious private lender requires in its own vocabulary. A company that can write this file properly tends to find every other financing conversation easier, because the work of understanding its own project has already been done. The public application is often where that work happens for the first time.
Montclare does not act as a lender. It structures transactions, prepares the file and coordinates financing with authorised institutions.