Financial crisis newspaper headline
VoxEU Column

Why so many crises happen when we know why they happen and how to prevent them

Financial crises are not complicated, and many claim to know why they happen and how to prevent them. Why then do they happen with such alarming frequency? This column argues that a key reason is failures in regulation, and proposes that the financial authorities instead adopt a fundamental concept from finance – diversification. By doing so, we can build a more resilient system that not only better absorbs shocks but is also more efficient and less susceptible to crises than the current frameworks.

One of my favourite exam questions is: "Given our extensive knowledge about the causes of financial crises and the measures needed to prevent them, why do they happen so frequently?"

We have had a deep understanding of financial crises for over 200 years. A 19th-century central banker dealing with the severe crisis of 1866 would find few surprises in the more recent ones.

All crises share the same fundamental causes. Excessive leverage renders financial institutions vulnerable to even small shocks. Self-preservation in times of stress drives market participants to prefer the most liquid assets. System opacity, complexity, and asymmetric information make market participants mistrust one another. These three fundamental vulnerabilities have been behind almost every financial crisis in the past 260 years, ever since the first modern one in 1763 (Danielsson 2022).

If we know why crises happen, preventing them should be straightforward. But given their alarming frequency, it does not appear to be so. When looking at the various alternative explanations and ignoring the political ones, we find two different narratives: consensus and diversification. Let’s start with the consensus one.

The consensus narrative

We find the consensus narrative in financial stability and ‘lessons learned’ reports published by the financial authorities. These reports have become very common ever since the authorities re-started taking the financial system seriously after the crisis in 2008.

At the risk of oversimplification, the consensus narrative is as follows: some financial institutions bypass the spirit of regulations – deliberately or inadvertently – amassing large, illiquid, and risky positions that are increasingly vulnerable to stress. One of the best manifestations of this view is the Financial Stability Board's 2020 holistic review of the Covid March 2020 market turmoil.

The consensus narrative drives the recommended responses. Most parts of the system are safe because of regulations implemented after 2008. However, some undesirable activities have slipped through the cracks, necessitating tighter supervision, expanded regulatory coverage, and stronger capital and liquidity buffers.

The problem with the consensus narrative

I fear that the consensus narrative and its ever-increasing regulatory intensity will not protect us.

Finance is crucial and requires risk to deliver on its promises to society. Unfortunately, the private sector's incentives are not fully aligned with society, giving rise to crises that have cost Europe and the US trillions of dollars. To mitigate that very high cost, after-the-fact public bailouts of private risk are unavoidable, further misaligning incentives for private risk-taking. This problem is the rationale for before-the-event regulation.

This then begs the question: why not regulate finance heavily? Well, we already do, and it has become very hard to do even more, as it appears that we are getting close to the upper limits of regulatory intensity.

The reason it is so difficult to regulate finance is that the financial system is one of the most complicated of all human constructs. In effect, it is infinitely complex. And when a system is infinitely complex, there are infinite areas where excessive risk and misbehaviour can emerge. When market participants optimise, they are actively searching for overlooked areas in which to take risk, so it is almost axiomatic that crises happen where nobody is looking. How can we regulate something we have yet to see?

If crises are to be prevented, the architects of regulations have to foresee all the areas where vulnerability can emerge, and the supervisors must patrol all of them. That is not enough. They also need to identify all the latent links between the disparate areas of the system, channels that only emerge in times of stress. Meanwhile, the authorities have to contend with political forces that benefit from the pre-crisis bubble and do not want regulations that threaten the perceived benefits to society. Add to this the dismissive attitude of the monetary policy and supervision authorities to the macroprudential agenda.

The objectives of the consensus-founded macroprudential narrative are impossible to achieve, in part because effectiveness demands much higher resources – human capital, politics, data, compute – than those available to the authorities. Even worse, it requires far more resources than the private sector needs.

The consequence is a cat-and-mouse game where the mice have the advantage.

Of course, this is well understood, and there is a consensus solution: build tall buffers to protect financial institutions against bad outcomes. Unfortunately, that will not prevent crises. There are several reasons why.

The first is how consensus narrative regulations harmonise beliefs and action. The practical implementation of the regulations not only favours market concentration because of high fixed costs but also compels financial institutions to measure risk and respond to it in the same way. This makes them behave like a herd. While that is fine if we are regulating visible conduct, such as traffic police measuring speeding and issuing tickets, it is different with finance since the risk is latent (Danielsson 2024). The consequence is procyclical amplification of the financial cycle, increasing booms and deepening busts.

Regulations based on the consensus narrative also amplify the complacency channel of financial instability. If we believe that the authorities understand the system, have everything under control, and are confident that they will step in with bailouts if needed, it leads to overconfidence and excessive risk-taking, particularly in the parts of the system that the supervisors are not patrolling, which is most of it. The resulting complacency and short-termism are key factors in most crises, such as the one in 2008. Such Minsky-type responses make crises more rather than less likely, as empirically shown in Danielsson et al. (2018).

The consensus narrative further amplifies the political channel for instability. When the government takes increasing responsibility for financial activities, controlling risk and protecting us from the adverse consequences of that risk, it makes the state, rather than the private sector, responsible for finance. That, in turn, has two consequences. The first, as argued by Chwieroth and Walter (2019), is that bailouts become a middle-class good that cannot be politically forsworn. The more the state gets involved with finance, the higher the chance of bailouts, which makes crises more likely.

Furthermore, when the state regulates risk-taking and underwrites losses, how can one respond to political extremists who question why we have a private-sector financial system in the first place?

Meanwhile, the fiscal and monetary resources to fight crises have mostly been exhausted by fighting minor stress. We could marshal very significant fiscal and monetary resources in 2008. If the same event happened today, that would not be possible. Knowing this undermines the credibility of financial policy, making crises more frequent and severe.

In addition, the consensus approach to regulations neglects efficiency while stressing stability. The aim of regulations is not financial stability. It is to support prosperous and stable economic activity. Not recognising that makes regulations subject to increasingly vicious political attacks. If the cost of regulating increases faster than the economy grows, the authorities will be forced to change direction.

Finally, the consensus narrative has led to rapid concentration and ever larger too-big-to-fail financial institutions, exacerbating systemic risk and inducing the further raising of protective buffers.

A different way forward

There are three related reasons why a financial system composed of a large number of relatively small and diverse financial institutions is more stable and prosperous than one with few large and similar institutions.

First, when financial institutions differ from one another, excessive systemwide risk-taking is less likely because, at any given time, the actions of some institutions inflate bubbles while others do the opposite.

Second, when faced with shocks, relatively homogeneous institutions will respond similarly, buying and selling the same assets at the same time. This leads to disastrous selling spirals. In contrast, when they are diverse, some institutions will buy and others will sell, dissipating shocks. In other words, a system with relatively homogeneous institutions acts as a shock amplifier, whereas a more diversified system absorbs and dissipates shocks and, hence, is more stable.

Finally, a system with many small and diverse institutions will be more prosperous. It allows better tailoring of financial services to the needs of the economy while also requiring lower buffers against systemic risk. This means it offers the cheaper provision of financial services.

The benefit of diversification

The consensus approach to regulating reduces institutional diversity since it is a partial equilibrium approach subject to a fallacy of composition: if every individual part of the system is made safe, the system is safe. The problem is that the consensus approach makes financial institutions increasingly homogeneous, leading to herd behaviour that causes booms and busts. These financial institutions amplify shocks when stress occurs because they are compelled to seek safety in the same way.

It is better to borrow a basic principle from finance: diversification. Just as we should not put all our savings into one investment, a system composed of diverse institutions is more stable and prosperous.

This suggests learning from how competition authorities use their powers to increase competition. In practice, we can actively use the licensing regime to facilitate start-ups with diverse business models and tailor regulatory regimes to suit different types of institutions, including reducing the fixed cost of compliance.

Why does diversification not happen?

A competitive financial system simultaneously promotes and opposes diversification. Start-ups compete against incumbents by having better business models, diversifying the system. The profit motive drives incumbents towards the short-term selection of successful business models and, hence, homogeneity.

Since financial institutions are highly regulated, the authorities wield powerful tools that can either help or hinder the forces of positive diversification.

The incumbents' incentives are clear: keep entrants out. That means lobbying for regulations with high fixed costs and uniform licensing regimes and regulations, using arguments of ‘fair play’ and ‘level playing fields’.

The consensus narrative for financial regulations also pushes for homogeneity. There are three reasons for this:

  • first, regulatory capture, which is driven by the two-way traffic between banks and the authorities and political pressure;
  • second, the mistaken belief that there is no trade-off between stability and efficiency;
  • finally, the erroneous impression that a system composed of a few large players is easier to understand and to control than one that is more complex and diverse.


There are two acceptable answers to my favourite exam question: "Given our extensive knowledge about the causes of financial crises and the measures needed to prevent them, why do they happen so frequently?" A coherent answer based on the consensus approach to regulations will earn the student an A.

But there is a better answer. Use the authorities' powers to push for a more diversified financial system – one that absorbs shocks and increases efficiency – instead of the current set-up, which drives homogeneity, procyclicality and deadweight loss. We can easily achieve this by leveraging the licensing regime to actively permit start-ups that use innovative business models, and by tailoring regulations to the type of institution.

Such a diversified financial system will be more efficient, robust and stable – a win-win-win.


Chwieroth, J and A Walter (2019), “The wealth effect: The middle class and the changing politics of banking crises”,, 3 June.

Danielsson, J (2024), “When risk models hallucinate”,, 3 February.

Danielsson, J (2022), The Illusion of Control, Yale University Press.

Danielsson, J, M Valenzuela and I Zer (2018), “Low risk as a predictor of financial crises”,, 26 March.

FSB – Financial Stability Board (2020), “Holistic Review of the March Market Turmoil”, 17 November.