“The financial system looks calm on the surface right now – but under the hood, powerful forces like AI, stablecoins, and booming private markets are quietly rewriting the rules of global finance.” And this is the part most people miss: the real question is not whether something will break, but where the pressure will show up first.
The rise of AI, stablecoins and private markets: how stable is the financial system?
Global financial risks appear less intense than a few years ago, and markets feel more orderly. Yet trade tensions, shifting US economic policies, geopolitical flashpoints, a possible AI-driven boom (or bubble), and the rapid growth of stablecoins all mean the system is anything but risk-free. The big puzzle is whether the world has actually become safer, or whether new vulnerabilities are simply forming in different corners of the financial landscape.
To unpack these issues, host Paul Gordon sits down with financial stability specialist John Fell, who brings a deep, system-wide perspective to the discussion. Their conversation looks at how these trends interact, not just in isolation but as a network of risks that can amplify one another when stress hits.
As always, the opinions shared in the episode reflect the views of the speakers and do not necessarily represent the official position of the European Central Bank. The discussion was recorded on 24 November 2025 and released to the public on 27 November 2025.
01:25
How stable is the financial system?
After a difficult period, the EU and the United States have reached a trade agreement, easing some of the uncertainty that previously weighed on businesses and markets. The direct economic damage from trade frictions so far has turned out to be less severe than many had feared, which has supported a more positive backdrop for financial stability.
However, the true consequences of these policy changes may only become clear over time, as firms adjust supply chains, investment strategies, and pricing. That raises a key question: does the apparent calm genuinely reflect a more resilient financial system, or are markets underestimating risks that may surface later as new rules and relationships fully take hold?
04:14
Is there an AI bubble?
The idea of an “AI bubble” has become a hot topic in boardrooms, trading floors, and policy circles alike. Yet identifying a bubble in real time is notoriously difficult, especially when a technology may be transforming whole sectors and justifying rapid growth.
Valuations of AI-related firms and projects are a central concern: are current prices anchored in realistic expectations about future profits, or are investors chasing hype and momentum? At the same time, the discussion explores how concentrated the AI market has become, and to what extent funding flows can be circular – for example, when gains from one AI investment are used to fuel further bets in the same ecosystem.
Another big issue is how companies are actually paying for their AI ambitions. Are they relying mainly on retained earnings, raising equity from investors willing to take long-term risks, or taking on significant debt that could become problematic if revenues disappoint? But here’s where it gets controversial: if financing is heavily debt-based, does that make the broader financial system more fragile than current optimism suggests?
06:15
If there is a bubble, what could cause it to burst?
Many AI-related investments are supported by substantial leverage, meaning borrowed money plays a major role in funding expansion. This leverage can amplify returns when things go well, but it can also magnify losses and force abrupt deleveraging if expectations change.
The sustainability of these investments will depend heavily on how widely and deeply AI is adopted across the real economy. If AI becomes a genuine general-purpose technology – like railways, electricity, or the internet – the long-run payoff could justify today’s aggressive spending. But if adoption stalls, or productivity gains fall short, even a small shift in sentiment could trigger sharp repricing, forced asset sales, and spillovers to lenders. Are markets being realistic about the time it takes for transformational technologies to deliver concrete cash flows?
08:15
What about the rise of private credit?
Private credit markets – where non-bank lenders provide loans outside traditional public markets – have expanded rapidly and now play a much larger role in financing companies. Unlike equity financing, which allows firms to raise capital in exchange for ownership stakes and potentially share losses, private credit adds fixed obligations that must be serviced regardless of how profits develop.
This shift raises concerns about euro area financial stability as more risk migrates into less transparent, less regulated corners of the system. When credit is extended through private funds and bilateral deals, information about leverage, asset quality, and investor behavior can be harder to monitor. The discussion looks at how these developments connect to systemic risks in the banking sector, especially when banks are exposed to private credit funds or help fund them.
There is also a link between private credit and the build-up of AI infrastructure: large data centers, specialized chips, and cloud platforms often require significant upfront investment. If a chunk of this spending is financed through private credit, stress in AI-related business models could reverberate through lenders that operate outside the traditional banking spotlight. Should regulators and investors be more worried about what they cannot easily see in these markets?
11:23
Are stablecoins really stable?
Stablecoins are marketed as digital assets designed to maintain a steady value, often pegged to a traditional currency such as the euro or the US dollar. Yet the conversation emphasizes that their stability is not guaranteed, especially under stress, and depends heavily on how reserves are managed and how easily users can redeem their holdings.
One key vulnerability is the risk of a liquidity run, where many holders rush to cash out at the same time. If the underlying reserves are not truly liquid or safe, this can cause fire sales of assets and destabilize parts of the broader financial system. As stablecoins grow quickly and become more intertwined with conventional finance – for example, through links to banks, money market funds, or payment systems – the impact of such a run could spread widely.
This gives rise to a provocative question: should stablecoins be treated more like banks or like lightly regulated tech products? The answer has big implications for how strictly they should be supervised and what safeguards they must maintain.
15:11
What’s the situation for government financing?
Since the pandemic, many governments have managed to bring their debt ratios down from peak levels, easing some immediate concerns about fiscal sustainability. This progress supports financial stability by reducing the perceived risk that public finances could spiral out of control.
Even so, in several euro area countries, government debt remains uncomfortably high. At the same time, political and social pressures to ramp up spending – particularly on defence, but also on areas like climate transition and social support – are intensifying. This combination of high legacy debt and new spending demands stands out as one of the biggest medium-term risks for financial stability.
However, there is a nuanced angle: if increased public investment is well targeted, transparent, and efficiently managed, it could boost long-term growth and productivity, strengthening the economy’s capacity to handle debt. The controversial question is whether current fiscal plans are genuinely focused on growth-enhancing projects or driven more by short-term political considerations.
18:07
What about banks?
Against this complex backdrop – questions about an AI bubble, fiscal sustainability concerns, and the rapid ascent of stablecoins – euro area banks currently appear to be in relatively good shape. Profitability is robust by recent historical standards, giving banks more capacity to absorb potential shocks.
Moreover, non-performing loan (NPL) ratios are at historically low levels, indicating that most borrowers are still managing to meet their obligations. This suggests that, for now, the banking sector is not the epicenter of financial stress.
Yet the conversation highlights that strong current metrics do not guarantee future resilience. If an AI correction, a fiscal shock, or a stablecoin-related disturbance were to hit, banks could still be affected through their exposures to governments, corporates, and financial markets. And this is the part most people miss: the real risk may lie in the connections between sectors, not just in the health of banks themselves.
20:13
Our guest’s hot tip
To wrap up, John Fell offers listeners a “hot tip” – a key insight or recommendation drawn from his work on financial stability. While the details are left for the episode itself, it points listeners toward practical ways of thinking about current risks and preparing for future uncertainty.
For those who want to dive deeper, the conversation is closely linked to the European Central Bank’s Financial Stability Review and its concise overview materials. The discussion also references the book “Chip War” by Chris Miller, which explores the strategic importance of semiconductor technology – a crucial input for AI and digital infrastructure – and how this shapes economic and geopolitical risk.
Now over to you: Do you think AI is driving a genuine productivity revolution, or are markets overhyping its potential and inflating a dangerous bubble? Are stablecoins a useful innovation or an accident waiting to happen? And when it comes to high government debt and the surge in private credit, do you see these as manageable challenges or red flags for the next crisis? Share where you stand – and don’t hesitate to disagree strongly if your view clashes with the mainstream narrative.
