Will AI revolutionise the Securitisation Market?
Dr. Christian Thun, CEO, European DataWareHouse
The securitisation industry has witnessed numerous technological promises over the past decade, from blockchain to tokenisation to artificial intelligence. We spoke with Dr Christian Thun, CEO of European DataWarehouse, to separate hype from reality and understand what lies ahead for Europe’s market.
European DataWarehouse was set up in 2012, creating unprecedented transparency in securitisation. How significant is this competitive advantage to the market?
It’s unmatched anywhere on the planet. There is no other region in the world with that level of transparency when it comes to the underlying data of public securitisations. If you compare it to the US, for example, there simply isn’t the same level of disclosure and data availability that exists in Europe. European DataWarehouse has a giant, standardised, centralised data pool that stretches from Lithuania to Portugal. The Americans market lack this infrastructure—they rather rely on specialised providers to collect information themselves and then build their models.
The irony is that Europe is not using it to the extent that the Americans use their fragmented data. If Europeans leveraged what we have, our market could be miles ahead of where it is now.
Many technologies once presented as transformative, didn’t really deliver. What happened to blockchain and tokenisation?
Everybody was talking about the tokenisation of loan contracts, smart contracts replacing normal processing, everything on the blockchain or Ethereum. These technologies were expected to create a landslide in securitisation. It hasn’t happened yet. We’ve seen some promising pioneers—there was an Italian tokenised securitisation issued by Zenith in 2023 and a few others—but they all got stuck in some kind of prototype stage. Nothing has reached mainstream.
The potential benefits are unquestionable. But there seem to be either regulatory questions or issues related to technical implementation, and data security concerns. The market still hasn’t really tackled these challenges. Setting up a securitisation in the blockchain, creating smart contracts, etc. can be quite elaborate.
So today everyone’s talking about AI. Is this just a hype cycle, or is artificial intelligence genuinely applicable to securitisation?
AI absolutely can help. It can massively improve data and document processing, increase the velocity of the whole issuance process, bring costs down, and make even smaller securitisations more feasible and more economical. It can tick all the boxes. AI has the potential to be very disruptive, specifically on the data side. How securitisation can be set up, how patterns in the underlying data can be recognised, data gaps identified, data quality issues addressed—all these things can be tackled very quickly with AI.
But here’s the critical difference: AI is much more accessible than blockchain. We all use ChatGPT, Grok, or Gemini. For normal day-to-day use, AI is easy to access. Even building your own agent can be done with relative ease. It’s becoming part of daily life: on your iPhone, in text replacement or content creation. I believe AI will be more easily adopted in the securitisation industry than smart contracts and tokenisation ever were.
What are the actual use cases you envision?
Data analysis is the obvious starting point. Pool selection. Data validation. Identification of missing or inconsistent fields. Pattern recognition across large loan datasets. Documentation review. Comparison of prospectus language. Detection of subtle divergences in definitions. All of these are functions where AI has a future. If you could train AI models on European data, these models would learn what is specific for European loans, European behaviour, delinquencies. They would learn how a European mortgage behaves or a European Auto loan.
This could help us select pools of underlying mortgages or Auto loans for securitisation. AI could be used to identify a strong-performing Dutch mortgage or German Auto loan, what data should be available, and could quickly highlight missing information or unusual values. Imagine bringing down the time needed from the decision to securitise to pricing and placing in the market by a third, while reducing costs by a significant percentage. This would allow smaller players to enter the market who currently can’t afford the significant setup costs of a securitisation.
You sound cautious despite the enthusiasm. What’s the catch?
AI can only be as good as the underlying data. I’ve built my own AI models years ago—you can train AIs in multiple ways, but the AI is only as good as the data it uses and the intelligence that went into its design.
At EDW we’ve seen companies use our data and come up with interesting but ultimately misleading findings—like elevated prepayment levels in Spanish mortgages that turned out to be just an anomaly in the data, not a real pattern. You need to understand the underlying data and the economic reality it reflects. Otherwise, you create an AI that gives you misleading information.
Is the European Data Warehouse itself experimenting with AI?
We have a tool on our website using a GPT model to analyse the data. We call it EDWARD. It’s a natural-language-to-SQL AI Assistant so you can ask it questions in any language: like “what’s the most popular car financed by French pensioners?” Or “provide the delinquency rate of Dutch mortgages in the last five years”. However, we don’t use it to build predictive models because that’s beyond our mandate. We’re here to provide good quality data to market players who then build their own predictive analytics.
How does Europe’s regulatory approach compare to the US when it comes to technology adoption?
There’s a fundamental difference in mindset. The US in general has shown more willingness to work with data, more eagerness to build models and predictive analytics. They see the opportunities of combining large data sets and tend to be more focused on what’s possible, sometimes at the expense of individual data rights.
Europe has always been more conservative about data. We rather see the downside of potential data violations, leaks, confidentiality issues, personal rights. The GDPR discussion epitomises this conflict. The EU is more concerned about the rights of the individual and privacy. It prefers a “regulate first, innovate later” approach. There is nothing wrong with this approach but there are obviously trade-offs.
Does this put Europe at a disadvantage?
Not necessarily. The steps taken by European legislators following the Great Financial Crisis was understandable, aiming to guide the market and prevent a repeat of past failures. Tight regulatory boundaries around capital requirements, disclosure, transparency were essential safeguards.
Thanks to reports from Lagarde, Draghi, and others in 2023-2024, it is now recognised securitisation is essential to fund Europe’s transition to a more competitive, and resilient economy. Regulators acknowledge the need to relax some of the very restrictive legislation giving the market breathing space to better bridge banking and capital markets, fund the economy, and drive the kind of growth seen in other parts of the world.
The current wave of regulatory reform is the right approach. There is a shared understanding that getting it right this time is crucial, because if they fail, reopening the securitisation regulation won’t happen again for some time. With today’s geopolitical challenges, this is critical.
Where do you see the market in five to ten years?
I believe the securitisation market of 2030 or 2035 will be very different from what we saw in 2005 or 2015. If applied effectively, technology could make securitisations smaller, faster and cheaper.
There’s massive room for innovation. Leveraging technology, new tools, and data has the potential to be very disruptive, but AI is still in a very nascent state when it comes to securitisation. So far, we’re only just talking about what could be done conceptually, where it could be applied.
The market will be shaped by the generation now in their 30s, young talent who will define how securitisation will develop over the next 10-15 years. Humans will remain essential—to build AI, teach it what’s right and wrong, and negotiate terms. Meanwhile the repetitive tasks like number crunching, paperwork, and routine processes will increasingly be outsourced to AI.
What gives you optimism?
That the industry is actually talking about this. The Amsterdam Securitisation Event is jumping into this discussion rather than just complaining about regulation and wanting things to remain as they were. They’re asking: how should the securitisation market look in 5,10, 15 years? Who will be the players? That forward-thinking approach is what we need. The potential for Europe’s securitisation market to grow is there—undoubtedly. Now we need to realize it.
Register now
At the Securitisation Event 2026, market leaders explore how growth, regulation, new asset classes, AI, technology and data centres are reshaping Europe’s securitisation market. Expect concrete cases, practical insights and open debate across the full value chain. Join the largest annual Dutch securitisation platform and connect with the entire structured finance community.
Securitisation Event
The Securitisation Event, where more than 250 participants come together annually to discuss developments...
Download the brochure
Share
