The growing use of artificial intelligence (AI) poses difficult challenges for the financial authorities. AI allows private-sector firms to optimise against existing regulatory frameworks, helps those who want to damage the financial system, amplifies wrong-way risk and speeds up financial crises. It also gives the authorities new tools for executing their mandate. The authorities could become more effective stewards of the financial system by gaining expertise in AI, setting up AI-to-AI links, developing triggered facilities and incorporating AI into their monitoring frameworks.
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.
Financial institutions are rapidly embracing AI – but at what cost to financial stability? This column argues that AI introduces novel stability risks that the financial authorities may be unprepared for, raising the spectre of faster, more vicious financial crises. The authorities need to (1) establish internal AI expertise and AI systems, (2) make AI a core function of the financial stability divisions, (3) acquire AI systems that can interface directly with the AI engines of financial institutions, (4) set up automatically triggered liquidity facilities, and (5) outsource critical AI functions to third-party vendors.
Financial crises usually inflict the most damage when banks suddenly shift from pursuing profits to survival. This column argues that such drastic behavioural changes render statistical analyses based on normal times ineffective. That is why we cannot predict the likelihood of crises, or what banks will do during those crises. Since this behaviour arises from a natural desire for self-preservation, it cannot be regulated away.
Artificial intelligence can act to either stabilise the financial system or to increase the frequency and severity of financial crises. This second column in a two-part series argues that the way things turn out may depend on how the financial authorities choose to engage with AI. The authorities are at a considerable disadvantage because private-sector financial institutions have access to expertise
The rapid adoption of artificial intelligence is transforming the financial industry. This first of a two-column series argues that AI may either increase systemic financial risk or act to stabilise the system, depending on endogenous responses, strategic complementarities, the severity of events it faces, and the objectives it is given. Stress that might have taken days or weeks to unfold can now happen in minutes or hours. AI's ability to master complexity and respond rapidly to shocks means future crises will likely be more intense than those we have seen so far.
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.
Artificial intelligence will both be of considerable help to the financial authorities and bring new challenges. This column argues the authorities risk irrelevance if they are reluctant and slow in engaging with AI, and discusses how the authorities might want to approach AI, where it can help, and what to watch out for.
Risk model hallucination happens when models are forced to forecast the likelihood of extreme events in cases where they have not been trained with such extreme outcomes. This is surprisingly common in applications such as financial regulations, pension funds, reinsurance, as well as the containment of market upheaval and financial crises. This column argues that measuring systemic financial risk needs to acknowledge the changing behaviour of agents and institutions during periods of market stress. This approach is more challenging and difficult to formalise, but crucial for learning how to develop financial resilience.
As artificial intelligence makes inroads into the financial system, it exacerbates existing channels of instability and creates new ones. This column identifies several such channels, malicious and misinformed use, misalignment and the evasion of control, and finally risk monoculture and oligopolies. All arise when AI vulnerabilities interact with economic fragilities like strategic complementarities, problems of incentives, and incomplete contracts.
The use of artificial intelligence in the private sector is accelerating, and the financial authorities have no choice but to follow if they are to remain effective. Even when preferring prudence, their use of AI will probably grow by stealth. This column argues that although AI will bring considerable benefits, it also raises new challenges and can even destabilise the financial system.
Artificial intelligence is expected to be widely used by central banks as it brings considerable cost saving and efficiency benefits. However, as this column argues, it also raises difficult questions around which tasks can safely be outsourced to AI and what needs to stay in the hands of human decision makers. Senior decision makers will need to appreciate how AI advice differs from that produced by human specialists, and shape their human resource policies and organisational structure to allow for the most efficient use of AI without it threatening the mission of the organisation.
Crypto-promoters and financial authorities are split on the future of cryptocurrencies. Should crypto join the mainstream or remain in the wilderness? Should the authorities control and extinguish crypto, ignore it, or embrace it? This column argues by adopting the best ideas from crypto, it can be leveraged to improve the efficiency of the financial system and reduce rent extraction.
The financial regulators have recently taken an active interest in cryptocurrencies, more than a decade after their law enforcement counterparts did. So why are they doing this now, and what will the consequences be? Regulators feel compelled to respond due to political pressures and their actions may backfire.
The downfall of Silicon Valley Bank and Credit Suisse has exposed failures in how we regulate the financial system. This column argues that the problems we now see in the system have arisen because the financial authorities have been trying to do the impossible: maintain growth while keeping inflation under control and financial stability high. The best way forward would be to focus on shock absorption and moral hazard, not the current approach of buffers and risk measurements.
The collapse of Silicon Valley Bank shows that banks still pose risks. Are they systemic? While it is unlikely that the failure of SVB will lead to a crisis, it shows us that the financial system is much more fragile than the public had been led to believe.
Cryptocurrencies have enjoyed a remarkable run from obscurity to a trillion dollar valuation in just over a decade. This column argues, however, that they have now reached the beginning of the end as the factors fuelling their success have come to a standstill. While crypto currencies are not much of a public concern, they will leave a fine legacy by forcing much-needed reform on the financial system.
Global macroprudential and microprudential regulatory changes since 2008 have been designed to make a repeat of a systemic crisis much less likely. This column argues that one-size-fits-all regulation constrains market response to short-term fluctuations in measured risk. Without more diversity in regulation, this will increase the likelihood of systemic crises in future.
Economic researchers can choose between four general purpose programming languages – Julia, MATLAB, Python, and R. This column evaluates the four, comparing and contrasting the various language features and making recommendations in particular use cases.
Bitcoin is often sold as protection against adverse macroeconomic outcomes. This column argues that this depiction as a macro hedge does not stand up to scrutiny. Recent events — from Covid and inflation to war — provide a window into what sort of asset bitcoin actually is. It if were a true macro hedge, bitcoin would be positively correlated with macroeconomic uncertainty. Instead, it acts as a leveraged bet for speculators. And unlike gold, which has been trusted as a macro hedge for millennia, bitcoin requires access to electricity and the internet, precarious services in times of turmoil.
The cryptocurrency exchanges have only done what is legally required of them when sanctioning Russia for its invasion of Ukraine, unlike the mainstream financial institutions whose restrictions on the Russians generally exceeds what is required by law. This column argues that the implications for the future of cryptocurrencies will be considerable.
The Western countries have sanctioned Russia in a way not applied to any globally integrated major power in over a century, ever since 1914. This column argues that there are lessons to be learned from the 1914 systemic crisis and that high inflation and government debt will make it difficult to contain a crisis today if one emerges.
The relationship between financial risk and economic growth is complex. This column finds that perceptions of high risk unambiguously harm growth, while perceived low risk has an initial positive impact, which eventually turns negative. Global risk has a stronger effect on growth than local risk, via its impact on capital flows, investment, and debt-issuer quality, challenging monetary policy independence.
As the price of bitcoin continues to rise, this column argues that most of us would not want to live in a society where bitcoin succeeds. Fortunately, the internal contradictions and perverse consequences of cryptocurrencies success mean that they are destined for failure. Until then, it might make sense for speculators to ride the cryptocurrency bubble, so long as they get out in time.
While the direct economic consequences of Covid-19 have been significant, the impact on the financial markets has been more nuanced. This column uses a unique data set on the financial markets’ fears and perceptions of long-run financial risk to identify how Covid-19, and particularly Fed policy responses to Covid-19, affected global market fears. While some Fed interventions had little or no impact on market fear, the most powerful were the US dollar swap lines, which strongly reduced the perceived likelihood of global market losses decades into the future. The results suggest that the Fed's relative global role has been strengthened, possibly at the cost of increased moral hazard.
The most widely used programming languages for economic research are Julia, Matlab, Python and R. This column uses three criteria to compare the languages: the power of available libraries, the speed and possibilities when handling large datasets, and the speed and ease-of-use for a computationally intensive task. While R is still a good choice, Julia is the language the authors now tend to pick for new projects and generally recommend.
Many comparisons have been made between the coronavirus crisis and the global systemic crisis in 2008. This column argues that seen through the lens of exogenous and endogenous risk, these two crises are quite different. Coronavirus is unlikely to cause a global systemic crisis, and the policy response should be different.
Artificial intelligence, such as the Bank of England Bot, is set to take over an increasing number of central bank functions. This column argues that the increased use of AI in central banking will bring significant cost and efficiency benefits, but also raise important concerns that are so far unresolved.
Financial institutions are increasingly outsourcing information technology to the cloud, motivated by efficiency, security, and cost. This column argues that the consequence is likely to be short- and medium-term stability at the cost of the increased likelihood of catastrophic systemic events. Cloud providers are systemically important and should be regulated as such.
The type of risk we most care about is long-term, what happens over years or decades, but we tend to manage that risk over short periods. This column argues that the dissonance of risk is that we measure and manage what we don't care about and ignore what we do.
As central banks accumulate ever more job functions, their reputation risk increases. This column offers a cautionary tale from Iceland where, after the central bank was put in charge of capital controls, it was subject to severe attacks because of perceived mistakes in how the capital controls were enforced. The accumulation of powers erodes a central bank’s independence and subjects it to regulatory paralysis.
If private cryptocurrencies were to find widespread economic use the result would be increased financial instability, inequality, and social instability.
On top is the authority in charge of fiscal policy, followed by those running monetary, microprudential, and finally macroprudential policies. This ranking can cause conflicts in terms of policy effectiveness and legitimacy.
Julia, MATLAB, Python and R are among the most commonly used numerical programming languages by economic researchers. In this post, Jon Danielsson and Jia Rong Fan compare and contrast these four, reaching a very subjective conclusion as to which is best and which is worst.
Reliable indicators of future financial crises are important for policymakers and practitioners. While most indicators consider an observation of high volatility as a warning signal, this column argues that such an alarm comes too late, arriving only once a crisis is already under way. A better warning is provided by low volatility, which is a reliable indication of an increased likelihood of a future crisis.
Cryptocurrencies are supposedly a new and superior form of money and investments – the way of the future. The author of this column, however, does not see the point of cryptocurrencies, finding them no better than existing fiat money or good investments.
Artificial intelligence is increasingly used to tackle all sorts of problems facing people and societies. This column considers the potential benefits and risks of employing AI in financial markets. While it may well revolutionise risk management and financial supervision, it also threatens to destabilise markets and increase systemic risk.
Brexit is likely to cause considerable disruption for financial markets. Some worry that it may also increase systemic risk. This column revisits the debate and argues that an increase in systemic risk is unlikely. While legal ‘plumbing’ and institutional and regulatory equivalence are of concern, systemic risk is more likely to fall due to increased financial fragmentation and caution by market participants in the face of uncertainty.
Discretionary macroprudential policies aim to be countercyclical by adjusting risk-taking across the financial cycle. This column argues that the opposite effect may happen in certain cases. Depending on how regulators measure risk and how they react, the eventual outcome may well be procyclical, with serious unintended consequences.
Political risk is a major cause of systemic financial risk. This column argues that both the integrity and the legitimacy of macroprudential policy, or ‘macropru’, depends on political risk being included with other risk factors. Yet it is usually excluded from macropru, and that could be a fatal flaw.
Investor demand for bonds is very high. This column argues that this is surprising because under almost any likely inflation scenario, including central banks merely hitting their target inflation rates, bondholders suffer large losses. The beneficiaries are sovereign and corporate borrowers; the losers are pension funds, insurance companies and some foreign exchange reserve funds. Meanwhile, the systemic risk from a bond crisis is increasing.
Brexit creates new opportunities and new risks for the British and EU financial markets. Both could benefit, but a more likely outcome is a fall in the quality of financial regulations, more inefficiency, more protectionism, and more systemic risk.
The threat to the financial system posed by cyber risk is often claimed to be systemic. This column argues against this, pointing out that almost all cyber risk is microprudential. For a cyber attack to lead to a systemic crisis, it would need to be timed impeccably to coincide with other non-cyber events that undermine confidence in the financial system and the authorities. The only actors with enough resources to affect such an event are large sovereign states, and they could likely create the required uncertainty through simpler, financial means.
Macroprudential policy has become increasingly popular in the aftermath of the Global Crisis, but it remains controversial. This column argues that vigorous disagreement is both inevitable and healthy, reflecting differing fundamental views of how the financial system really works. By embracing the divergence of views instead of seeing it as problematic, macroprudential policymaking will be made easier and more effective. åÊ
There has always been conflict between macro- and microeconomic regulation. Microeconomic policy reigns supreme during good times, and macro during bad. This column explains that while the macro and micro objectives have always been present in regulatory design, their relative importance has varied according to the changing requirements of economic, financial and political cycles. The conflict between the two seems set to deepen and so, regardless of which wins, policymakers must not undermine the central bank's execution of monetary policy.
The Greek and the Icelandic crisis have much in common, not the least the heavy pressure from foreign countries and the hectoring from their public officials. In Iceland and in Greece this was counterproductive, hardening the opposition to any settlement. The will to reform needs to come from within, and the sooner the Troika realizes this, the easier it will be to deal with the Greek situation.
The long-running Greek crisis and China's recent stock market crash are the latest threats to the stability of the global financial system. But as this column explains, systemic risk is an inevitable part of any market-based economy. While we won't eliminate systemic risk entirely, the agenda for researchers and policymakers should be to create a more resilient financial system that is less prone to disastrous crises and that still delivers benefits for the economy and for society.
Some financial authorities have proposed designating asset managers as systemically important financial institutions (SIFIs). This column argues that this would be premature and probably ill conceived. The motivation for such a step comes from an inappropriate application of macroprudential thought from banking, rather than the underlying externalities that might cause asset managers to contribute to systemic risk. Further, policy authorities are silent on the question of what SIFI designation should mean in practice, despite the inherent link between identification and remedy.
Iceland has just announced it is getting rid of its capital controls. This column argues that the government's plan is a credible, efficient and fair plan to lift the costly and misguided controls.
Regulators and financial institutions increasingly depend on statistical risk forecasting. This column argues that most risk modelling approaches are highly inaccurate and confidence intervals should be provided along with point estimates. Two major approaches, value-at-risk and expected shortfall are compared, and while the former is found to be superior in practice, it is also easier to be manipulated by forecasters.åÊ
This column introduces a new Vox eBook collecting some of the best Vox columns on financial regulations, starting with the fundamentals of financial regulations, moving on to bank capital and the Basel regulations, and finishing with the wider considerations of the regulatory agenda and the political dimension. Collecting columns from over the past six years, this eBook maps the evolution of leading thought on banking regulation.
The proposed EU capital markets union aims to revitalise Europe's economy by creating efficient funding channels between providers of loanable funds and firms best placed to use them. This column argues that a successful union would deliver investment, innovation, and growth, but it depends on overcoming difficult regulatory challenges. A successful union would also change the nature of systemic risk in Europe.
The Swiss central bank last week abandoned its euro exchange rate ceiling. This column argues that the fallout from the decision demonstrates the inherent weaknesses of the regulator-approved standard risk models used in financial institutions. These models under-forecast risk before the announcement and over-forecast risk after the announcement, getting it wrong in all states of the world.
Risk forecasting is central to financial regulations, risk management, and macroprudential policy. This column raises concerns about the reliance on risk forecasting, since risk forecast models have high levels of model risk - especially when the models are needed the most, during crises. Policymakers should be wary of relying solely on such models. Formal model-risk analysis should be a part of the regulatory design process.
Basel III is coming into focus. The fundamental logic of the regulatory changes seems sensible, but the devil is in the detail - empirical implementation. This column discusses a detailed quantitative study, incorporating analytical calculations, Monte Carlo simulations and results from observed data. It concludes that the Basel Committee has taken three and a half steps backwards and half a step forward. If implemented, the framework is likely to lead to less robust risk forecasts than current methodologies.
Europe is set to finally approve new insurance regulation, Solvency II. This column argues that the final text should respect three fundamental principles to ensure solvency.
Central banks frequently lead the macroprudential policy implementation. The hope is that their credibility in conquering inflation might rub off on macroprudential policy. This column argues the opposite. The fuzziness of the macroprudential agenda and the interplay of political pressures may undermine monetary policy.
Icelandic voters recently ejected its post-Crisis government - a government that successfully avoided economic collapse when the odds were stacked against it. The new government comprises the same parties that were originally responsible for the Crisis. What's going on? This column argues that this switch is, in fact, logical given the outgoing government's mishandling of the economy and their deference towards foreign creditors.
Cyprus has imposed temporary capital controls. This column sheds light on how temporary and how damaging they are likely to be, based on Iceland's experience. The longer controls exist, the harder they are to abolish. Icelandic capital controls, which have been Û÷temporary' for half a decade, deeply damage the economy by discouraging investment. We can only hope the authorities that created the chaos in the first place realise that temporary really needs to mean temporary.
Is the fact that different banks have different risk models problematic? Contrary to the Basel Committee and the European Banking Authority, this column argues that heterogeneity is a good thing. It leads to countercyclicality, and thereby reduces instances of procyclical price movements. Both the Basel Committee and the European Banking Authority have indicated that they are troubled by heterogeneity and are seeking to rectify the problem. Their conclusion is plainly wrong.
The current EZ crisis is not Europe's first sovereign-debt crisis. This column shows parallels can be drawn from an all-but-forgotten episode, i.e. the 1990 Faroese crisis. Just like Greece, the Faroes got into difficulty because of excess borrowing facilitated by a currency union with an AAA-rated partner undeterred as the sovereign debt spiralled upwards. In the Faroese case, the crisis was eventually resolved when political necessities outweighed the cost of the bailout.
October 2011 saw the latest draft of Solvency II, the European Union's code for regulation of the insurance industry. This column argues that the latest proposals need to be drafted again, urgently.
October 2011 saw the latest draft of Solvency II, the European Union's code for regulation of the insurance industry. This column argues that the latest proposals need to be drafted again, urgently.
The IMF has emerged from the global crisis bigger and more powerful. But this column argues that the capital controls it required Iceland to adopt in 2008 are not of the soft and cuddly modern type that slow hot money flows. Instead they are akin to the draconian controls common in the 1950s. They violate the civil rights of Icelanders and significantly hamper economic growth.
Much of macroeconomic policymaking is trial and error. This column discusses calamitous error on the part of Iceland's policymakers, in the hope that others can at least try something else.
A debate is raging on capital adequacy requirements for banks. The UK wants to be allowed to be on ÛÏtop upÛ the agreed levels, i.e. to impose stricter capital standards than the EU minimum. This column argues the UK is right, and that the German and French opposition might be motivated by weaknesses in their banking systems.
Financial risk models have been widely criticised for both theoretical and practical failures, especially during the recent financial crisis. In the second of two columns, the authors outline why the shortcomings of risk models matter before making suggestions for how the financial industry and supervisors should use models in practice.
Risk models are at the heart of the financial sector's self-monitoring as well as supervision by regulators. This column, the first of two, addresses the question of how risk models are misused in practice by practitioners and supervisors alike. This misuse causes risk management to fail when it is most needed.
Icelanders have voted against providing a government guarantee for claims made by the UK and the Dutch governments against Iceland's deposit insurance fund. This column argues that the heated debates surrounding the referendum may provide a glimpse into the challenges that lie ahead for European policymakers as they attempt to allocate losses suffered by banks between the taxpayers of different countries.
A delicate regulatory question is under consideration on the capital (reserve) requirements at the heart of Solvency II (the insurance industry equivalent of Basel III), which is scheduled to come into effect by 2013. This decision will have implications for both regulation of insurers and for macroprudential stability. The six authors of this article were invited to discuss the issues and concluded that more public scrutiny over this important question is urgently needed.
In crises, insurance companies' asset values may fall significantly without a corresponding drop in their liabilities. European insurers have argued that their liabilities should be discounted by a higher rate during crises, lest regulations force them to raise more capital at exactly the wrong time. This column argues that that would be the wrong approach to the problem.
Financial models are widely blamed for underestimating and thus mispricing risk prior to the crisis. This column analyses how the models failed and questions their prominent use in the post-crisis reform process. It argues that over-relying on market data and statistical forecasting models has the potential to further destabilise the financial system and increase systemic risk.
Icelanders may well reject the terms of the financial deal with Britain and the Dutch in a March referendum. This column introduces a new CEPR Policy Insight arguing that responsibility for Icesave losses falls jointly on Iceland, Britain and the Netherlands. Regardless of the vote, the three governments should come to a more reasonable agreement that enables Iceland to pay its obligations without tipping the economy into the abyss.
The Icelandic parliament has decided to apply for EU membership. This column warns that domestic opposition and outstanding disputes with EU member countries on Icesave may derail the agreement.
Bank bonuses have been blamed for contributing to the crisis, and regulators and politicians are now demanding changes in compensation arrangements. Most of these calls are based on a misconception of the nature of financial risk, an inflated view of the efficacy of risk models, and an incorrect view of the incentive issues facing financial institutions. This column proposes reforms that would discipline senior managers by exposing them to the dangers of junior managers' risk taking.
Many are calling for significant new financial regulations. This column says that if the regulate everything that moves crowd has its way, we will repeat past mistakes and impose significant costs on the economy, to little or no benefit. The next crisis is years away, we have time to do bank regulation right.
By incorporating endogenous risk into a standard asset-pricing model, this column shows how banks' capacity to bear risk seemingly evaporates in the face of market turmoil, pushing the financial system further into a tailspin. It suggests that risk-sensitive prudential regulation, in the spirit of Basel II, makes systemic financial crises sharper, larger, and more costly.
Much of today's financial regulation assumes that risk can be accurately measured so that financial engineers, like civil engineers, can design safe products with sophisticated maths informed by historical estimates. But, as the crisis has shown, the laws of finance react to financial engineers' creations, rendering risk calculations invalid. Regulators should rely on simpler methods.
Some view Iceland's crisis as holding lessons for any country with an outsized financial sector, e.g. the UK. This column disagrees, arguing that Iceland's downfall is explained by the way its unique history, inappropriate policy responses, and weaknesses in EU banking regulations created a perfect storm, unlikely to happen elsewhere.
Iceland's banking system is ruined. GDP is down 65% in euro terms. Many companies face bankruptcy; others think of moving abroad. A third of the population is considering emigration. The British and Dutch governments demand compensation, amounting to over 100% of Icelandic GDP, for their citizens who held high-interest deposits in local branches of Icelandic banks. Europe's leaders urgently need to take step to prevent similar things from happening to small nations with big banking sectors.
Complex financial models and intricate assets structures meant extraordinary profits before the crisis. Markets for structured products became overly inflated as even the banks did not have a clear view of the state of their investments. Given complexity's role in today's mess, future regulation should focus on variables that are easy to measure and hard to manipulate (e.g. leverage ratios).
In response to financial turmoil, supervisors are demanding more risk calculations. But model-driven mispricing produced the crisis, and risk models don't perform during crisis conditions. The belief that a really complicated statistical model must be right is merely foolish sophistication.
Models and risk | Financial Regulation, Systemic Risk, Stability and AI
Blogs and appendices on artificial intelligence, financial crises, systemic risk, financial risk, models, regulations, financial policy, cryptocurrencies and related topics