For this interview, Olena Gryniuk, CEO of WOA Digital, speaks with Gašper Rant, President of the Management Board of Borza terjatev (Invoice Exchange), one of the pioneers of marketplace-based SME finance in Central and Eastern Europe. Together they explore how the Slovenian fintech has financed more than €718 million in receivables, built a unique investor-funded alternative to traditional factoring, and achieved a high level of automation in working capital finance. The conversation also examines the SME financing gap in Slovenia and Croatia, the role of technology and risk management, and the lessons learned from a decade of operating a two-sided finance marketplace.
Olena Gryniuk: How does Borza terjatev work, and how is it different from classic factoring?
Gašper Rant: Borza terjatev / Invoice Exchange is an online marketplace for short-term business finance. On one side are SMEs seeking working capital. On the other side are investors with surplus cash looking for short-term, diversified investment in the real economy. Since launching in November 2015, we have financed more than €718 million in Slovenia across more than 40,000 individual deals, as well as more than €40 million in Croatia through our subsidiary, P2P Finance.
The main difference from classic factoring lies in where the funding comes from. A factor funds receivables from its own balance sheet, while we act as the central counterparty to both sides. We purchase each receivable, split it into fractions, and sell those fractions on to investors. We track them off-balance-sheet until they are fully collected. Our investors include companies with surplus cash, institutional investors and, through a dedicated alternative investment fund, professional investors. Because each deal is split, even a small portfolio can achieve the same diversification as a large one, and no SME depends on a single investor for funding.
For the SME, onboarding is fully online, with identification carried out through a qualified digital certificate. Before the first deal, we put collateral in place, such as enforceable promissory notes, mortgages, or guarantees. This typically takes a few days, depending mainly on how quickly the client responds. It’s the one step we can’t fully automate because it involves notaries, but it only needs to be completed once.
After that, financing becomes an operational decision. The client enters an invoice, and the system rates the deal and shows a binding offer with all costs shown in real time. Once the client accepts, the funds are transferred directly to their bank account. At maturity, the debtor pays into our dedicated fiduciary account. That is the same account where investors’ funds are held and from which all payouts are made.
Pricing is based on the deal’s credit rating, which the system calculates automatically. Investors choose which rating classes they are willing to accept, but they do not determine the prices.
A typical invoice is worth €15,000–20,000, with maturities ranging from a few days up to 180 days. In addition to purchasing receivables, we offer short-term loans and supplier financing. Our purpose-specific loans can also finance export receivables or orders where an invoice has not yet been issued.
Automation holds it all together. We are connected via API to a range of data sources, allowing both the credit decision and the execution of each deal to be largely automated. That is what makes the service fast, efficient, and simple enough for a wide range of companies.
OG: Why is this model rare in CEE?
GR: In our part of Europe, we haven’t found another platform operating in quite the same way: financing SME receivables and short-term loans with funds provided directly by other businesses. The only comparable platform in the region is P2P Finance in Croatia, which is our own subsidiary.
The usual explanations are regulation and banks. I don’t think either is the main reason. Banks aren’t our direct competitors, and in Slovenia factoring is not subject to a specific licensing regime. Borza terjatev therefore does not need a separate licence to provide factoring services. As a financial institution, we are subject to the applicable anti-money-laundering and counter-terrorist-financing legislation, which requires us to carry out the relevant checks on our clients. That’s where the real challenge begins. Without a licensing regime, trust must be built through the way the business is structured and through a proven track record. Beyond the statutory AML requirements, the additional safeguards we apply are voluntary. We keep all client funds in a dedicated fiduciary account, and our business is independently audited every year.
The hardest part is building a two-sided marketplace. Volumes need to stay balanced on both sides: investors’ money sitting idle is just as problematic as a good invoice that goes unfunded. That means that we have had to build both sides of the marketplace in parallel, while the investor side naturally takes more time because the sales cycle is much longer. A company that invests also needs a significantly higher level of trust before placing its surplus cash with us than an SME does before selling us an invoice.
Our investors are companies and institutions. Individuals cannot invest directly, as the platform is not designed for retail investors. Individuals who qualify as professional investors can invest through our dedicated alternative investment fund.
Then there’s the size of each market. Entering a new country essentially means building the business again, as the legal framework, the collateral structures, the investor base, and the language all change across the border. When we entered Croatia in 2020, we established P2P Finance as a separate company with its own platform and its own investor base. This was a deliberate choice, as it was the easiest and fastest way to develop the business there. We plan to connect the two platforms, although this will require further work on both the technological and legal side.
The relatively small size of each deal makes scale especially important. Our average deal is around €17,000, so reaching €718 million in financing in Slovenia required more than 40,000 individual transactions. With such volumes, the model only works if key processes such as credit rating, pricing, and payouts are highly automated. This means the technology has to be built and invested in before the business reaches the scale needed to fully support that investment.
Taken together, all of this means the model takes years to build and develop. I think that is the main reason you don’t see more comparable platforms in the CEE market.
OG: What is the SME financing gap in Slovenia and Croatia?
GR: According to EUF estimates, factoring turnover in Slovenia was just under €3 billion in 2025, or 4.2% of GDP, compared with an EU average of 11.5%. If Slovenia reached the EU-average level of factoring penetration, companies would finance approximately €8 billion of receivables each year. This suggests a potential gap of roughly €5 billion in annual factoring volume.
A low factoring-to-GDP ratio on its own does not necessarily indicate unmet demand. Sweden and Ireland, for example, are also below 5%. In Slovenia, however, the need becomes more apparent when looking at the payment practices. According to a 2025 Atradius survey, average B2B payment terms are around 50 days, while overdue invoices account for approximately 44% of B2B sales.
The issue is not that there is a lack of liquidity in the banking system. Slovenian banks’ loans amount to only 69% of their non-bank deposits, well below the euro area average. For SMEs, however, bank financing can take several weeks, and even after completing the full process, approval is not guaranteed. Banks have also become more cautious with smaller companies, which means those clients often turn to us as an alternative. Somewhat larger companies typically use our financing alongside their existing bank credit lines.
Bank credit is generally cheaper, and when a prospective client can access suitable bank financing, we tell them so directly. We do not compete with banks on price. Instead, we serve the segment they are unable or unwilling to finance and compete on speed, automation, and quality of service.
The other side of the gap is the available capital itself. Slovenian companies are now net creditors of their banks. They keep about three-quarters of their deposits on sight, at rates close to zero. A significant share of that cash could instead be used to finance other companies’ receivables. Yet companies and individuals with capital to invest still tend to perceive short-term business finance as less secure than traditional asset classes, and I think that is not justified. An investor on our platform finances a transaction between two local companies in the real economy, while the nominal value of the claim does not fluctuate with financial markets. The key exposure is credit risk related to the debtor and the seller, which can be assessed and secured, rather than market risk.
SMEs face a perception problem of their own. In Slovenia, factoring is still seen as a last resort, used for problematic receivables or by companies already under financial pressure. Factoring is a standard tool for financing working capital, and it is most effective when used proactively, before liquidity becomes a problem, rather than after it does.
Croatia shows a similar pattern, only more pronounced. Factoring turnover there amounted to around 2.1% of GDP in 2025, half the Slovenian level, even though the market grew by about a quarter that year.
So, the financing gap is not really about the lack of money. It’s about connecting companies waiting to be paid and companies whose cash is earning very little. Building that connection is exactly what we do.
OG: How automated is receivables financing on your platform today, and how close are you to real-time decisions?
GR: The credit decision is already made in real time. The system automatically calculates an offer as soon as a client submits an invoice. Unless we’re waiting for confirmation from the debtor, the funds usually reach the client’s account within one or two hours of accepting the offer.
The process begins with onboarding, which can be completed within a single working day. The client is identified with a qualified digital certificate, and the system generates the user agreement and sends it for signing by email. It also automatically retrieves company data from Dun & Bradstreet and other sources, including registration details, financial statements, and bank account blocks, and calculates the client’s credit rating. We review the beneficial ownership and PEP information ourselves, and once the client is approved, they gain access to the platform. Collateral is arranged once before the first deal, and because it usually involves a notary, this is the one step that cannot be automated.
From that point on, the client submits orders through the platform, with a single order able to include several invoices issued to the same debtor. The system rates each deal immediately. Most of the data comes through our API connection to Dun & Bradstreet, which aggregates information from sources such as account blocks and payment indices. Where available, we also include our own payment history, and after more than 40,000 deals, that data has become a valuable signal. The rating determines the price, so the client sees a binding offer with all costs immediately.
How we verify the debtor depends on the type of assignment. With a notified assignment, we contact the debtor, who confirms that the invoice exists and commits to paying into our fiduciary account. This provides the strongest protection against fictitious invoices. With silent assignment, there is no such confirmation, so we rely more heavily on the seller’s credit rating and, above all, on the collateral in place.
Funding is automatic as well. The system splits each receivable into fractions and allocates them across all investors whose portfolios accept that rating class and who have available funds. If a deal cannot be fully funded, the client would be notified. In practice, this happens rarely, because liquidity planning is one of our core processes.
Before any funds are transferred, our back office performs a final review of each order, usually within one or two hours. Once the order is confirmed, the payout is executed as a SEPA instant payment through the Halcom B2B system.
Collection follows the same logic. The system imports the daily bank statement from the fiduciary account and matches incoming payments to open receivables based on the amount and payment reference. Where there is a match, the payment is booked, and the investors’ trading accounts are credited automatically. Any remaining payments are reconciled daily. Payment reminders are also sent automatically, while a few days after maturity our collection team takes over personally.
We built the entire platform in-house, from credit rating to settlement. That lets us add new data sources, adjust the rating model, or modify individual steps in the process ourselves, without having to rely on an external vendor.
OG: What is still surprisingly manual?
GR: What surprises people most is that a person still reviews every order before money leaves our fiduciary account. That is a deliberate choice. Once SEPA instant payment is sent, it cannot be reversed, and a rating model is only as reliable as the data it receives. Our back office checks whether the attached invoice makes sense and that the client has entered the data correctly. The due date is especially important, because it affects both the price and the investors’ expected cash flow. They also check whether the invoice fits the client’s usual business activity, customers, and transaction sizes. A few minutes of human attention at the point where the payment becomes irreversible is a strong safeguard, and it adds no more than an hour or two for the client.
That highlights the real gap, both in our business and across the industry. Company data is rich and largely automated, while Invoice data is not. Through Dun & Bradstreet and our own payment history, we can quickly build a detailed picture of both the seller and the debtor. The invoice itself, though, still reaches us as information entered by the client, with a copy attached. Neither proves that the goods were delivered or the service was performed.
With a notified assignment, that proof comes directly from the debtor. The system automatically drafts the confirmation email, which we usually follow up with a phone call. With a silent assignment, there is no such confirmation. This is why silent deals rely more on the seller’s credit rating and the collateral in place, and why they are priced higher.
Collateral is the other manual step, and the reason it is legal, not technical. It usually involves a notary, so it cannot be completed online. Because collateral is arranged once per client rather than for each individual deal, it has very little impact on the overall speed of financing.
After payout, most of the manual work is concentrated in exceptions. Most debtor payments are matched automatically, but some arrive without a reference, with an incorrect one, or as a single transfer covering several invoices. These cases are reconciled each day manually. Once an invoice is a few days overdue, our collection team takes over directly. At that point, the key question is no longer whether a reminder was sent but why the debtor hasn’t paid: whether there is a dispute, a cash-flow issue, or simply an administrative error. A phone call usually resolves that faster than another automated reminder. Because all our deals are with recourse, the seller is often part of that conversation as well.
From what I see, manual work across the industry tends to remain in the same areas: invoice verification, debtor confirmation, payment reconciliation, and dispute handling. None of these processes is difficult to automate from the technical perspective. They remain manual because the underlying data is often unstructured, cannot be verified directly at the source, or is tied to a physical legal process.
OG: Which signals do you watch when assessing risk, and how do you balance speed with prudence?
GR: For us, speed and prudence aren’t really a trade-off, because most of the risk assessment takes place before the first deal. By the time a client submits an invoice, we have already rated the company, put the required collateral in place, and set its financing limits. The system can then assess each order within those pre-approved limits, which is why a decision can be made within seconds.
Every seller, borrower, and debtor receives an automatic credit rating from 1 to 10, with 10 being the highest. Which rating carries the most weight depends on which party we are ultimately really relying on for repayment. With notified assignment, the debtor has confirmed the invoice and pays into our fiduciary account, so the deal relies primarily on the debtor’s rating. With silent assignment, where the debtor is not notified, the deal relies heavily on the seller’s rating. For loans, the borrower’s rating is the key factor. As said, every deal is with recourse, meaning the seller remains liable for the receivable in all cases. The deal rating determines the price and decides which investors’ portfolios are eligible to fund it.
The rating draws on several types of data, including financial statements, bank account blocks and payment indices, court notices and our own payment history.
After more than 40,000 deals, we often have a strong view of how a debtor pays across multiple clients, not just what its financial statements show. We have updated the rating model several times in recent years, both by adding new data sources and by refining the methodology.
Some signals override the rating completely. A deal is stopped automatically, regardless of how strong the other indicators may be, if there is a blocked bank account, an entry on the tax authority’s list of tax debtors, an insolvency proceeding, or another comparable warning signal.
Limits operate on three levels: seller, debtor and investor. The seller and debtor limits are entered directly into the trading platform, which automatically blocks any order that would exceed them. These checks add no extra time, as they are performed the moment an order is submitted.
The risk monitoring also continues after payout. We monitor sellers and debtors for as long as a deal remains open, using daily information on bank account blocks and alerts from Dun & Bradstreet. Here too, speed is part of prudent risk management. Learning about a blocked account the next day is far more valuable than discovering it only at maturity.
Automation therefore does not lower our standards. It ensures that the same rules are applied consistently to every order, every time. People step in where judgment can add something that rules alone cannot.
OG: What do investors get wrong about this kind of risk?
GR: Investors who are new to receivables finance usually misjudge the risk in one of two ways. Some think recourse and collateral make the investment effectively risk-free. Others see a single default and conclude that the model doesn’t work. Neither view is accurate.
Recourse and collateral provide important protection, but they don’t make the investment risk-free. Every deal is with recourse and is also secured with an izvršnica, a Slovenian enforceable promissory instrument that allows us to pursue collection directly from the client’s bank accounts without first obtaining a court judgment. Depending on the client, we may also require mortgages, guarantees, or other collateral. However, there is no principal guarantee and no reserve fund covering losses. If the debtor, the seller and the collateral all prove insufficient, the investors holding that deal ultimately bear the shortfall.
That’s why concentration matters more than any single deal. Each investor sets a maximum exposure to any one counterparty, and we recommend a limit of 4% of the portfolio. There is no minimum or maximum fraction size, so the system automatically diversifies even a relatively small portfolio within those limits.
The second mistake is treating late payment, a default, and a loss as the same thing. Late payment is a normal part of B2B trade. On our platform, the average agreed payment term is 59 days, while invoices are actually paid after 66 days on average. A default is a status, not necessarily a loss. The system automatically classifies a deal as defaulted 90 days after maturity, while our collection team may classify it earlier if they learn that full repayment is unlikely. These are the same two criteria used by banks.
From there, we pursue both the debtor and, under recourse, the seller, while also enforcing the available collateral. How long recovery takes depends on the case, the debtor, and the type of collateral involved. The izvršnica is generally the fastest route, while mortgages and other forms of collateral are less liquid and therefore take longer to enforce. To date, no deal has resulted in a complete loss with no recovery. While payment remains overdue, we charge default interest to the client, most of which is passed to investors. For an investor, the first cost of a default is therefore time: capital remains tied up longer than planned instead of being reinvested. An actual loss only arises if the recovery is insufficient.
What ultimately matters is the net return on the portfolio. Our target is six-month Euribor plus 3 percentage points per year, net of all write-offs and fees. We measure this as the average across all investors on their entire portfolio, including funds that are not invested at a given time. Idle cash therefore reduces the measured return rather than improving it. We achieved the target in both 2024 and 2025 and are on track to do so again in 2026. Individual results differ, depending on the rating classes and exposure limits selected by each investor.
So, as an investor, I would not ask whether an individual deal can default, because some inevitably will. The more relevant question is what the portfolio delivers over time, net of defaults, write-offs, and fees.
OG: What have ten years taught you?
GR: We went live in November 2015. Looking back, three lessons stand out. Our annual volume peaked in 2021 at more than €110 million. It is now around €60 million, and we expect 2026 to remain in line with last year. The main reason is on the investor side. One of our larger institutional investors invested through its own fund, and after several years, that fund had to be liquidated. As the fund was wound down, its capital gradually left the platform. In a marketplace, when investor funding leaves, fewer invoices can be financed. Replacing that capital takes time, because the investor sales cycle is much longer than the SME one. This was not a credit-quality issue. Average investor returns remain positive, including in years when six-month Euribor was well below zero. The lesson is one we already give our investors. We recommend that they limit their exposure to any single counterparty; the funding side of the platform also needs to be well diversified.
The second lesson is that a simpler process for the client comes from lower risk, not from fewer checks. We launched loans in July 2019, and by 2022 they made up about half our volume. Today they represent roughly a third. The share has declined mainly because notified assignment now makes up a larger part of our business. When a debtor is strong, a notified assignment carries less risk, which means less collateral, such as mortgages, may be required and the process becomes simpler for the client. Loans still play a clear role, particularly for exporters. They allow a company to finance receivables from foreign buyers, and in an export-driven economy like Slovenia’s, that is highly relevant.
The third lesson came from Croatia. When we entered the market in 2020, the plan was to replicate the platform and start trading. We assumed the legal framework would be almost identical, given that many laws in both countries originate from the former Yugoslav system, and both are EU members. Building the Croatian platform showed us that there are, in fact, quite a few differences. The two platforms are very similar at their core, but they differ in many important details. That lesson will be even more valuable as we enter new markets in the future.
OG: If we’re having this conversation again in three years, what headline would you like to see?
GR: “Borza terjatev passes €1 billion and becomes the go-to platform for non-bank SME finance in Slovenia and Croatia.” The number is the easier part. Becoming the “go-to” part matters more to me. When a micro, small, or medium-sized company in Slovenia or Croatia needs finance outside its bank, I want Borza terjatev to be the first name that comes to mind. That does not mean replacing the bank. It means companies no longer see factoring as a last resort, but a normal way to finance working capital – something they use when business is going well, not only when they are already facing difficulties.
