menu
close

Author(s):

Jón Danielsson | London School of Economics
Robert Macrae | London School of Economics

Keywords:

financial stability , systemic risk , external shocks , financial regulation , financial supervision , scenario analysis , uncertainty

JEL Codes:

G18 , G28 , H6 , K20

We thank Andreas Uthemann for valuable comments. All errors are ours.

Abstract
Systemic financial risk has both internal and external drivers. So, when we focus too strongly on preventing internal crises, such as the 2008 Global Financial Crisis, we tend to miss out on the more important external risks. Five external risk factors stand out: populism, debt-driven death spirals, manufactured tensions, artificial intelligence and geopolitics. To combat these threats, the macroprudential authorities should reorient away from derisking the financial system and towards resilience and economic growth via deregulation and diversification.

Introduction

Although most systemic risk methodologies indicate that systemic risk is low at present, such risk assessments tend to miss the most important current drivers of systemic risk, those arriving from outside the financial system.

Three fundamental drivers influence the rise in systemic risk. First, while we know conceptually why crises happen, they differ in detail, frustrating analysis. In other words, the lessons we learn from past crises are of limited value.

Second, when the authorities seek to exercise control in one area of the financial system, the forces of instability simply manifest elsewhere. In an infinitely complex financial system, it is axiomatic that crises emerge where the supervisors are not patrolling.

Finally, macroprudential policies aim to secure a stable future with costs borne in the present – an unpopular necessity that provokes significant political resistance.

The fundamental driver of almost all crises is debt, especially real estate and sovereign. Yet high debt levels alone do not fully explain why crises occur because, under normal conditions, there are plenty of safeguards in place. What is further needed is a catalyst – an external shock or internal vulnerability that viciously interacts with elevated debt in such a way as to make what is sustainable one day unsustainable the next. To quote Hemingway (1926): “How did you go bankrupt? Two ways. Gradually, then suddenly.”

Consequently, when assessing the financial system’s stability, we must look for such a catalyst. Many have been mooted, such as shadow banking, private credit, asset bubbles, weak oversight, financial innovation, technological disruptions, climate change and (de)globalisation. While of concern, these are not the most important threats to financial stability. Instead, there are five classes of catastrophic outside events that could cause a financial crisis. In the discussion below, these events are illustrated with representative examples. While each specific case may not be one of the five most likely ones, the class it represents is.

Populism and credibility

The financial authorities’ ability to prevent and fight crises depends crucially on their credibility. Ireland’s sovereign debt crisis was caused by the Irish government issuing a blanket guarantee of troubled banks’ debt that the markets did not believe. Faith in the ability and determination of the European Central Bank and its president, Mario Draghi, personally, made the promise to do “whatever it takes” effective in 2012.

Credibility works both ways. Trust in the Fed made large professional depositors not bother to read Silicon Valley Bank’s financial statements, proving – as Minsky might have suggested – that regulatory credibility has become so high that market participants feel free to act recklessly. The belief that the European authorities will do “whatever it takes”, however loudly they may protest they would never do that, drives the taking of risk.

In other words, both credibility and the lack of it drive systemic financial risk.

Two factors disrupt “whatever it takes”: populism and resource constraints.

Voters are entitled to question the competence and motivation of government organs. When a policy creates identifiable winners and losers, as inevitably happens, it fuels populism. This undermines the authority of the central bank because the losers will cry foul, and trust in the institutions will fall.

Populism limits the ability of central banks to act, as it drags them into the political domain and hence threatens their independence. In turn, that can prevent them from taking unpopular measures, both in preventing future crises and in containing them once they happen.

Furthermore, when we imbue the central banks with responsibilities outside their core monetary and financial stability remit, it erodes their credibility.

Meanwhile, the authorities have limited room to manoeuvre as their fiscal and monetary space is increasingly constrained. We could marshal significant resources in 2008, but that will be hard to repeat in future crises because we have significantly drawn on our reserves since then.

Both factors, rising populism and eroding fiscal and monetary space, make future crises more likely.

Debt death spirals – Europe and Italy

Debt is at the core of every crisis. An indebted country may enter a debt death spiral, where the government has to meet increasingly costly obligations in an environment in which it is difficult to raise taxes. This creates a vicious cycle in which more and more needs to be borrowed to service existing debt and meet demands for public services, ultimately leading to financial collapse or default.

A debt death spiral can happen in any country, but for global systemic consequences, it needs to happen in one of the world’s largest economies – Europe, the USA or China. Are all increasingly vulnerable. Ray Dalio recently used the term “fiscal heart attack” to describe his concerns about a potential US debt crisis. What matters is not the absolute level of debt, but whether the debt can be credibly repaid. When the only plausible reply is “it cannot”, a crisis is not far off.

The European Union is the third-largest economy in the world. It used to keep pace with the USA until the crisis in 2008, but since then, its economic growth has lagged the USA by perhaps 15% to 25%.

Europe displays all the features of an economy heading for a debt death spiral. Growth is long-term stagnant, while debt, interest payments and demands for public services consistently trend upwards. The recent realisation that more needs to be spent on national defence can only make that worse.

What might instigate the debt death spiral?

The weakest large member – Italy – could pose a global threat. Normally, one would not expect a relatively small country to do so. But if Italy defaults on its sovereign debt, which is plausible and even likely given the long-run trend in its fundamentals, how is the rest of Europe going to react? Will member states that have been through a crisis and come out stronger because they have reformed be willing to bail Italy out? If the EU somehow bails out Italy, moral hazard suggests that similar fiscal misbehaviour will continue. The likelihood of a bailout is a key factor in the low CDS spreads on Italian sovereign debt.

If a bailout is not an option, the European Central Bank and the euro will be subject to significant – even unsustainable – strain, which could threaten the existence of the EU itself.

Manufactured tensions – China and Taiwan

An aggressive foreign policy is a tried-and-tested way to restore faith in a failing government. Witness Argentina’s invasion of the Falkland Islands in 1982 and Britain’s response, the Spanish-American War (1898) and Russia’s annexation of the Crimea in 2014.

China faces a multitude of economic and political challenges. These include significant youth unemployment and the potential of a real estate crash, opacity about very high private and public debt, the difficult politics of debt crisis management and a slowdown in growth. This is not a good place to be in for a government whose legitimacy rests on delivering prosperity.

Proposing reunification with Taiwan provides an opportunity to deflect domestic criticism and rally the population around a common goal. But rhetoric can only go so far. If things really come to a head, China may take action and attack Taiwan. Given Taiwan’s pivotal role in global supply chains, this could easily lead to a global systemic crisis.

Artificial intelligence

Technological advancements have always driven both efficiency and instability, and artificial intelligence (AI) is no exception. At the same time, it would be a mistake to consider the risks from AI in the same way as those emanating from other new technology. Existing technology provides input into human-centric decision processes, whereas AI makes decisions. In the language of Norvig and Russell (2021), AI is a highly effective “rational maximising agent”.

Danielsson (2025) discusses the various financial stability threats arising from AI. The most important of these might be the impact of AI on crises. As an autonomous decision-making entity, AI’s behaviour, especially during crises, is an unknown-unknown. The more it gets embedded in the financial system – and that trend is accelerating – the more vulnerable we become to the consequent systemic risk.

In a crisis, the ability of AI engines to analyse the state of the financial system rapidly will speed up the reaction of private firms. Crises that once took days or weeks will be reduced to minutes or hours. That means the central banks’ liquidity interventions might come too late. If the private-sector AI engines collectively conclude that the central banks are not adequately engaged with the new threats emanating from AI, that by itself will make systemic crises more likely.

Geopolitics and trade tensions

Geopolitics has historically been a key driver of systemic crises.

In 1914, globalism was at its peak. Archduke Ferdinand’s assassination on 28 June 1914 caused a global systemic crisis a month before World War I was declared. The financial authorities in the main creditor nations closed down cross-border payments, triggering widespread bankruptcies. This was one of the most serious systemic financial crises ever and it was triggered by geopolitics (Roberts 2014).

Fifteen years later, the Great Depression was the most calamitous economic event we have ever seen. It started as a financial crisis, first in Europe and then in the USA, only to escalate into the Great Depression when the USA passed the Smoot-Hawley Tariff Act in 1930. Retaliation and opportunistic copying of trade restrictions culminated in global trade falling by over two-thirds, spreading the Depression to almost every country in the world.

These two events, with nationalism and geopolitics at their core, led to the most serious global economic crises in the past century. To prevent a repeat, we established a global order founded on international organisations and agreements. But that global order is under growing strain. The lessons learned from these early 20th-century crises are increasingly being disregarded: countries act for narrow nationalistic interests without considering the impact of the counter-reactions of other countries. That makes systemic financial crises more likely.

What can be done?

Over the past few decades, macroprudential policy has focused on preventing internal crises, such as the GFC in 2008. Today’s threats are different, with the financial system on the receiving end of shocks originating elsewhere. That suggests that the authorities should reorient their priorities towards meeting the external threats to financial stability.

To begin with, the central banks need to more actively consider the external political environment and how their involvement in political discourse can deplete their political capital. That is particularly important when populism adversely affects their ability to prevent crises and react to them once they happen.

On a more technical level, the central banks should shift their focus from preventing 2008-style crises to strengthening the system’s resilience. Two ways to achieve that are deregulation and diversification.

A financial system with onerous regulations imposes high costs on financial intermediation, holding back growth. The more the economy grows, the more resources we have to meet demands for public services and strengthen public finances. Growth reduces the need to manufacture tensions and restrict trade. And finally, growth helps to meet the challenges of populism, which are fed by perceptions of personal stagnation, the lack of opportunities and crumbling public services.

We achieve resilience by adopting a basic principle of finance: diversification. Systemic risk increases when financial institutions perceive and react to events in the same way. Regulations that seek to harmonise beliefs and actions inherently undermine resilience. Unfortunately, many of the post-2008 GFC reforms are exactly of that nature.

The more different the institutions in the system, the better the system can absorb shocks. The more tailored financial products are to the economy, the more they stimulate growth. As a bonus, such a system requires less capital and regulation.

The authorities can meet the stability threats from AI by developing their own AI expertise, using AI-to-AI API links to facilitate benchmarking, emphasising triggered facilities, and engaging with local private-sector AI firms, as discussed in Danielsson (2025).

Note: Image generated with Chat GPT

Conclusion

Some of the most disastrous economic events we could suffer are systemic financial crises. They cost tens of per cent of GDP,  trillions of euros for the EU member states, (Barnichon et al. 2021). Preventing and containing systemic crises have always been a core function of the financial authorities, whether they have acknowledged that or not, and we will demand that they do what they can to contain systemic crises when they happen.

Systemic crises can come from inside the financial system, such as the 2008 GFC, or from outside, as has been more common historically. The financial authorities need to consider both possibilities.

This article identifies five outside threats. Debt-driven death spirals, manufactured tensions and geopolitics, amplified by populism that undermines the effectiveness of the authorities, and artificial intelligence that makes financial crises more vicious and faster.

In recent years, the thrust of macroprudential policy has been on inside threats and the prevention of another 2008-style crisis. Meanwhile, the five outside threats discussed in this article have become much more relevant, suggesting that inside threat-dominated macroprudential policy, as practised over the past 15 years or so, may not adequately meet these outside systemic threats.

The financial authorities can mitigate these outside dangers by emphasising resilience and supporting economic growth. That means avoiding onerous regulations and encouraging diversification in the types of financial institutions, while shifting financial intermediation away from banks.

All these functions lie within the macroprudential mandate of the financial authorities and are vital for long-term prosperity. Perhaps paradoxically, economic growth, deregulation and institutional diversification are also the best ways to meet the objectives of the macroprudential authorities.

References

Barnichon, R, C Matthes and A Ziegenbein (2022), “Are the effects of financial market disruptions big or small?” Review of Economics and Statistics 104(3): 557–70. https://doi.org/10.1162/resta00972

Danielsson, J (2025), “Artificial Intelligence and Stability”, VoxEU.org, 6 February. https://cepr.org/voxeu/columns/artificial-intelligence-and-stability

Hemingway, E (1926), The Sun Also Rises. New York: Scribner.

Norvig, P and S Russell (2021), Artificial intelligence: A modern approach. London: Pearson.

Roberts, R (2013), Saving the City: The Great Financial Crisis of 1914. Oxford: Oxford University Press.

About the authors

Jón Danielsson

Jón Daníelsson is a SUERF Fellow, the Director of Systemic Risk Centre and Professor of Finance at the London School of Economics. Since receiving his PhD, Jón’s work has focused on how economic policy can lead to prosperity or disaster. He is an authority on both the technical aspects of risk forecasting and the optimal policies that governments and regulators should pursue in this area. Jón has written three highly regarded books: The Illusion of Control (Yale University Press, 2022), which was included on the Financial Times “Best books of 2022” list; Financial Risk Forecasting (Wiley, 2011); and Global Financial Systems: Stability and Risk (Pearson, 2013). He has also contributed numerous academic papers on systemic risk, artificial intelligence, financial risk forecasting, financial regulation and related topics.

Robert Macrae

Robert Macrae is a Research Associate at LSE’s Systemic Risk Centre. Robert has spent the last 30 years applying an engineering approach to problems in value investment and risk. Robert studied electrical engineering at Imperial College, London and moved into finance in 1990, eventually managing several value-based equity hedge funds. He is now a research associate at the LSE Financial Markets Group with areas of interest spanning value, hedge funds, risk control, tail risk, financial crises, financial regulation and machine learning.

More on these topics

Tags:
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.