Navigation – Plan du site
Time in Economics, part II

Financial Regulation and the Speed of Financial Risks

La règlementation financière et la vitesse des risques financiers
Ralph Chami, Thomas Cosimano, Connel Fullenkamp et Celine Rochon
p. 161-190


Cet article montre que la vitesse des risques financiers, plutôt que la vitesse des régulateurs, est la cause principale des crises financières. Notre argument contredit l’affirmation couramment avancée dans la littérature selon laquelle les crises financières surviennent car les régulateurs ne peuvent suivre le rythme rapide des changements dans les marchés financiers. Nous séparons les risques financiers en deux catégories — risques lents et risques rapides — et affirmons que la majorité des crises sont causées par des risques lents que les régulateurs auraient pu atténuer. La crise financière mondiale a néanmoins montré que la combinaison de risques rapides et lents conduit également à l’instabilité financière. Cet article offre un argument nouveau pour limiter la prise de risques, incluant la séparation des institutions financières suivant le type de risques qu’elles gèrent.

Haut de page

Texte intégral

  • 1 Recent examples of this sentiment include McMillan (2015), Dewatripont, Rochet, and Tirole (2010), (...)

1Time plays a central role in a common narrative purporting to explain why financial regulation fails to prevent crises. The essence of the story is that regulators move slowly, while the financial industry moves quickly. This leads to the common complaint that financial regulators are always (at least) one step behind the industry. The sentiment is repeated so frequently in the literature on financial regulation that it is taken to be one of the primary causes of financial crises1.

2Lagging behind the industry is dangerous, according to this narrative, because the financial sector is in a state of ever-accelerating innovation that has potentially dangerous consequences for the financial system and society. Banks and other financial institutions constantly create new products that quickly take hold in the markets. Although only a few of the new products are inherently dangerous, nearly all of them have the potential to destabilize both individual institutions and the entire financial system through overuse or abuse. The more time it takes regulators to understand innovations and respond to them, the more likely it is that one of these new products will cause a financial crisis. Thus, according to this narrative, the fact that regulators lag behind the industry is a major failing of regulation—a virtual recipe for disaster.

3Like all pieces of conventional wisdom, this story does contain important truths. But it also oversimplifies the situation, possibly overstating its actual importance and obscuring subtleties that could produce useful guidance for making and enforcing policy. Thus, in this paper, we undertake a more careful analysis of slow-moving regulators and fast-moving financial markets. First, we argue that although financial markets are continuously innovating, the risks they produce do not move at the same speed. Some financial risks are fast-moving, but there are also slow-moving financial risks. Fast-moving markets and risks can be thought of as evolving in continuous time, while slow-moving regulators and slow-moving risks operate and evolve in discrete time. The essential difference between fast-moving and slow-moving risks is that fast-moving risks produce large losses virtually instantaneously, while slow-moving risks cause losses that build up over several periods.

4After categorizing the major financial risks as fast-moving or slow-moving, we then examine how large a role the difference in speed between regulators and private actors has actually played in causing financial crises. We argue that this difference in speed only began to play a primary, causal role during the runup to the global financial crisis of 2007-2009. The reason for this change also involves a time or speed mismatch that had not been present in previous crises: the mixing of slow-moving financial risks with fast-moving risks. In addition to discussing how the particular mixing of fast-moving and slow-moving risks contributed to the global financial crisis, we use a recent theoretical paper by Chami et al. (2017) to argue that in general, mixing fast-moving and slow-moving risks in the same financial institution poses a danger to financial stability.

5Given that the mixing of fast-moving risks and slow-moving risks is the key issue we identify, we then explore its implications for financial regulation. We ask whether there exists a regulatory strategy that can be successful in permitting financial institutions to take on exposures to both fast-moving and slow-moving risks without threatening financial stability. After examining several alternatives, we argue that the only feasible strategy for dealing with continuous financial risks is to limit the mixing of fast-moving and slow-moving risks in financial institutions. This may include the segregation of financial institutions according to the type of risk they are allowed to take on.

6This paper makes several contributions to the literature on risk and financial regulation. First, we introduce a novel way to distinguish between different types of risk that proves to be a useful lens through which to view risk for policymakers. Finding useful distinctions between different categories of risks or dimensions of risk has given rise to many important insights in economics and finance. For example, the distinction between systematic and idiosyncratic risk is one of the building blocks of asset pricing theory. In addition, Merton (2005) proposes a breakdown between value-adding and non-value-adding risks that can be used as a general guide for hedging. Similarly, using our speed criterion to distinguish between risks naturally divides financial risks into two distinct groups that need to be managed by regulators in very different ways. Understanding the speed of a financial risk gives good insight into how it can be managed by regulators concerned with financial stability.

7The second contribution of this paper is to apply our distinction between fast-moving and slow-moving risks to examples from the historical record, in order to argue that the vast majority of large financial crises that have taken place over the past several decades have been caused by slow-moving risks, not slow-moving regulators (or fast-moving risks). Ideally, an econometric analysis would also provide such evidence, and it is hoped that this paper will motivate future research in this direction. We believe the analysis of historical episodes that we present is quite suggestive, and it will help dispel the misconception that financial innovation causes financial crisis, as well as the misconceptions either that financial regulation is doomed to be ineffective, or that financial regulators must somehow accelerate their activities in order to regulate effectively. Our analysis improves society’s understanding of what financial regulation can and cannot accomplish, so that society can focus on matching the appropriate kind of financial regulation to fast-moving and slow-moving risks, and on fixing the problems of regulation that can actually help reduce the incidence and severity of financial crises.

8Finally, our paper also yields a novel argument for limiting the risk exposures that banks should take on (as well as limiting the risk exposures that nonbank financial institutions should take on). Recently, several jurisdictions such as the U.S., U.K., and the E.U. have imposed limits on the types of risks that banks are allowed to take. The arguments justifying these actions have been based on fairness, on the goal of returning to the intended purposes of deposit insurance, and on claims that allowing banks to engage in activities such as proprietary trading is a threat to financial stability. Our decomposition of financial risks into fast-moving and slow-moving risks helps show how and why restricting banks to activities centered on deposit-taking and lending may benefit society. We explain why the mixing of fast-moving and slow-moving risks significantly elevates the possibility of financial distress and panic, which in turn implies that limiting the extent to which financial institutions mix exposures to these risks could improve financial stability.

9In the following section of this essay, we introduce and explain the distinction between fast-moving and slow-moving financial risks, and we categorize each of the main financial risks as fast-moving and slow-moving. The next section discusses the recent history of financial crises and attributes the causes of these crises to fast-moving and slow-moving risks. As a part of this section, we discuss how the mixing of fast- and slow-moving risks within financial institutions was a key departure from the past that helps explain the origins of the global financial crisis of 2008-09. Because the different risks were mixed in a particular way, we then argue in the following section that any mixing of slow-moving and fast-moving risks in the same financial institution is a threat to the institution’s viability and in turn to financial stability. After that, we move to a discussion of whether it is possible to regulate fast-moving risks in a way that would permit the safe mixing of fast-moving and slow-moving risks within financial institutions. We conclude by discussing the implications of our argument that such a regulatory system has not yet been designed.

1. Fast-Moving and Slow-Moving Risks

10Financial institutions continually adjust their balance sheets and off-balance-sheet commitments by entering into contracts and trading financial instruments. Regulators, on the other hand, attempt to monitor financial institutions closely, but may not be able to track the institutions’ actions perfectly. In addition, regulators generally only take actions once in a while—that is, they act at discrete intervals. This difference between the private sector’s continuous actions and the regulators’ discrete supervisory actions thus appears to be the reason why regulators are always one step behind the financial institutions. The institutions’ ability to continuously change their exposures also appears to be a primary cause of financial instability. If financial institutions take on risks more quickly than regulators can detect or restrict the risk-taking, then this does raise the possibility that the private sector’s speed of action is the main factor that causes financial crises. For example, a bank can rapidly increase the risk of its loan portfolio in the time between examinations, dramatically increasing its chances of entering into financial distress.

11But as we argue below, the speed with which financial institutions can adjust their exposures is not necessarily the ultimate cause of financial instability. Rather, we argue that the speed with which a particular exposure can inflict significant damage on an institution presents the greater challenge to regulating the financial system. The main types of financial risk vary greatly in terms of this speed, and this leads to significant differences in how each of these risks can be successfully regulated. Thus, we first need to examine the major types of financial risk and the speed with which each risk inflicts damage.

  • 2 See, for example, Hull (2015).
  • 3 Because operational risk involves losses resulting from people, processes and systems, it can be bo (...)
  • 4 Technically, credit risk is a type of counterparty risk, but we are following a common convention u (...)

12Discussions of financial risk management2 usually focus on five main risks faced by banks and other financial services firms: market risk, interest rate risk, credit risk, liquidity risk, and operational risk.3 Four of these risks are distinct, while interest rate risk is, strictly speaking, a type of market risk. But because the methods of measuring and managing interest rate risk are quite different from those used to measure market risk, it is usually considered separately from market risk. In our analysis below, we will also argue that interest rate risk behaves differently than other forms of market risk, such as losses from price movements of equities or exchange rates. In addition, in our analysis we will follow a common practice of including counterparty risk—the losses that occur when a counterparty to a financial contract fails to perform a promised action—as a type of credit risk.4 Systemic risk is also a risk faced by financial institutions, but the way in which systemic risk causes losses to an individual financial institution is through one or more of the other four financial risks listed above. For example, systemic risk can cause an otherwise healthy bank to suffer losses because of cascading defaults of other financial institutions that have borrowed from the bank.

  • 5 Risks may interact with each other and one risk can be transformed into another. For example, in th (...)

13Each of the losses faced by financial institutions has two main characteristics that are important to our analysis. The first characteristic is the particular way (or ways) that each risk can inflict losses on a financial institution. The second is the speed with which each risk inflicts losses. Although most economists are in agreement about the first characteristic of each risk,5 the profession does not appear to appreciate the importance of the second characteristic, speed. As we argue below, understanding the differences in the speed with which the various risks inflict damage on a financial institution is essential to making both business and policy decisions about how to manage these risks.

14In our analysis, we distinguish between fast-moving and slow-moving risks. Like the economic terms short run and long run, the terms fast-moving and slow-moving risk are not defined according to absolute measures of time, but rather in a relative sense. We define slow-moving risks as risks whose loss events generally emerge or intensify slowly—that is, over multiple (discrete) periods. The slow development of these risks means that they often give reliable advanced signals that significant loss events from these risks are likely to occur. The distinguishing feature of slow-moving or discrete risks is that these loss events unfold slowly enough such that human agents can take significant actions to limit the damage from these risks once the loss event starts to occur.

15Of the financial risks listed above, credit risk is a slow-moving risk. This is partly because many financial contracts call for multiple payments over time, so that a late or missed payment today is a strong signal of future default. Even when a credit contract calls for a single repayment, such as a zero-coupon bond, potential credit losses can be foreseen. For example, firm characteristics such as sales and earnings (and financial ratios based on these and other accounting measures) signal increased default risk several periods in advance of actual defaults. This gives lenders sufficient time to take steps to limit or even possibly prevent credit losses, such as working with borrowers to address problems in their firms. Derivatives based on credit risk, such as credit default swaps, are also best characterized as exhibiting slow-moving risks, since their prices will mostly reflect changes in the underlying credit risk of the reference entity.

16Fast-moving risks, on the other hand, cause loss events that emerge or intensify very quickly, and in some cases virtually instantaneously. Hence, they can be thought of as occurring in continuous time. Significant losses either emerge very quickly, or they come as surprises—that is, they do not give reliable advanced signals of their imminent occurrence. In either case, human agents do not have sufficient time to limit the damage from these risks once the loss event starts to occur, which is the defining feature of a fast-moving risk.

  • 6 See United States Securities Exchange Commission and Commodity Futures Trading Commission (2010).

17Market risk is a fast-moving risk. Thanks to advances in electronic communications technology, asset trading can take place more or less continually, even when traditional exchanges and trading venues are closed. Another consequence of the application of technology to trading is that multiple trades can take place within each second. But it is the combination of fast trading with unpredictability of price movements that makes market risk a fast-moving risk, in terms of the definition introduced above. The level of volatility in many markets also severely limits the ability to extract reliable signals that large price changes are about to occur. Thus, large changes in market prices can occur at any time and seemingly without warning. For example, in the U.S.’ so-called Flash Crash in 2010, many individual share prices fell to fractions of their previous values within seconds, without any prior indication of market turbulence.6 Not only can market risk be realized as a single, devastating loss that occurs without warning, but it can also be realized as a series of smaller losses that occurs so quickly that the firm lacks sufficient time to prevent or mitigate the damage.

  • 7 In addition, interest rates on private debt securities include a credit spread component that refle (...)

18Because interest rate risk is a type of market risk, one might infer that it is also a fast-moving risk. Although changes in market interest rates can occur quickly, they are heavily influenced by current and expected future monetary policy. In most countries, monetary policy actions are taken only at discrete intervals, such as at meetings of the Federal Reserve’s Open Market Committee, which take place roughly eight times per year. Therefore, because the policy actions that strongly influence interest rate risk move slowly—that is, they unfold over multiple discrete periods—interest rate risk is more accurately characterized as a slow-moving risk than a fast-moving risk.7 Interest-rate derivatives are also best characterized as exposure to slow-moving risk, for the same reason.

  • 8 See Cohan (2010).

19Liquidity risk, on the other hand, is also a fast-moving risk. Liquidity risk generally manifests as a sudden decrease in the ability to sell a financial instrument for the full value implied by its customary pricing method (market liquidity) or in the willingness of lenders to make short-term loans (funding liquidity). For example, an instrument that has many willing buyers (at the midpoint of the bid-ask spread) one moment can have none the next. The classic funding liquidity risk for banks is a sudden perception on the part of depositors of insufficient liquidity at the bank, which then becomes a self-fulfilling prophecy within hours or even minutes. Although there may be some advanced warning signs of potential liquidity problems in the market for an instrument, or on a bank’s balance sheet, liquidity can remain adequate for long periods until a tipping point is reached and there is a sudden loss or flight of liquidity. Once this tipping point is reached, moreover, there seems to be little that can be done to address the loss of liquidity, at least in the near term. For example, the investment bank Bear Stearns found that its frequent and public claims to have billions of dollars of cash on hand had no effect on its counterparties, once market sentiment about the company’s liquidity had turned negative.8

20The distinction between fast-moving and slow-moving risks is important because of what it implies for the ability of managers and regulators to predict whether loss events caused by each source of risk are likely to occur in the near future, and to mitigate losses once an event begins. Slow-moving risks generally offer advanced warnings that the likelihood of loss events is rising, which provides an opportunity for regulators to compel managers to take steps to prevent loss events from occurring. In addition, once a loss event begins, it unfolds slowly enough that regulators can compel managers to increase loss-reserves or take other steps to limit the damage. In other words, slow-moving regulators can generally be effective at mitigating slow-moving risks. As long as the intervals between their examinations, or between their finding a problem and taking action, are not excessively long, regulators will generally have sufficient time to detect and significantly mitigate slow-moving financial risks taken by the institutions they oversee.

21But the definition of fast-moving risks implies that they cannot be successfully mitigated in the same ways that slow-moving risks can. Fast-moving risks are unpredictable, and they intensify so quickly that they inflict damage before regulators can require managers to take effective countermeasures. The regulators’ limited ability to mitigate these risks implies that they need to set different goals with respect to dealing with fast-moving risks, and they also need to follow different regulatory strategies to manage these risks. We will discuss these strategies in detail in a later section.

22For much of the past century, fast-moving and slow-moving risks were largely segregated into different financial institutions, with banks specializing in slow-moving risks and other financial institutions taking on fast-moving risks, possibly in addition to slow-moving risks. During the era before financial deregulation, when interest rates were commonly regulated by the state, banks primarily took on credit risk as their main value-creating risk. After interest rates were deregulated, banks also added interest rate risk to their list of value-creating risks, but as we argued above, this is also a slow-moving risk. Other financial institutions such as broker-dealers (investment banks) and finance companies focused on fast-moving risks such as market risk and liquidity risk.

  • 9 It is also important to note that even in countries that lacked formal deposit insurance, an implic (...)

23It is possible to object to the idea that banks took on only slow-moving risks for two main reasons. First, because banks engage in maturity transformation, it is true that they have also faced significant funding liquidity risk in the past. This risk has largely been removed from banks through two types of government intervention, however. First, lenders of last resort who stand ready to address banks’ liquidity shortfalls have been implemented in most economies, usually in the form of central banks. Second, and perhaps more effective on a day-to-day basis, has been the introduction of government-administered deposit insurance in most banking systems. Deposit insurance removes depositors’ incentives to engage in bank runs, and in fact has removed many depositors’ concern with bank liquidity altogether.9

  • 10 See for example Rochet (2008).

24The second way in which banks could be thought to take on fast-moving risks is through fast accumulation of large quantities of risky loans. In particular, it is always possible for banks to go on lending binges between supervisory examinations and in fact banks may have the incentive to do so when their financial condition is already deteriorating. This strategy, which is called a gamble for resurrection, is well known in the banking literature.10 Certainly, although some gambles for resurrection have resulted in the collapse of the bank making the gamble, in general this strategy takes time to execute and it still takes time for the loans to default. Thus, while this strategy generally increases the losses once a bank defaults or is seized by regulators, it does not create fast moving risk per se.

25The segregation of risks in different institutions broke down as a result of the cycle of disintermediation, innovation, and deregulation that began in the 1960s. Disintermediation led to innovation, both in terms of new products and new practices. In the face of fast-paced and large-scale innovations, regulators often had little choice but to accept the innovations, particularly when the innovations moved certain financial activities beyond their jurisdiction. Both in order to prevent further innovation of this nature, as well as because of changes in regulatory philosophy in favor of free competition, regulators began to deregulate financial markets more quickly and intentionally. But this also led to further disintermediation, which placed additional revenue pressure on banks and also gave other financial institutions new opportunities to expand.

26Innovation and deregulation enabled financial institutions of all kinds to mix fast-moving and slow-moving risks in new ways. Banks sought new income streams to replace decreased lending revenues as large corporations decreased their reliance on banks and instead funded an increasing share of their debt in the money market, by issuing commercial paper. When deregulation allowed it, many money-center banks turned to market-making and trading as new sources of revenue. In addition, banks of all sizes increasingly took advantage of new types of securities for funding their activities, such as medium-term notes.

27Similarly, the creation of new instruments made it increasingly attractive for nonbank companies to mix slow-moving and fast-moving risks as they used the new instruments to exploit the difference between capital-market yields and short-term funding costs. For example, the creation of the money-market mutual fund established a new path through which household savings could reach companies in search of funding. Money-market funds were the natural purchasers of commercial paper issued by finance companies, who in turn used this funding to expand their lending portfolios. Later, creation of asset-backed securities enabled structured investment vehicles (SIVs) to issue asset-backed commercial paper and invest the proceeds in tranches of collateralized debt obligations (CDOs).

28Mixing risks also became attractive to all financial institutions because some slow-moving risks became market traded, through the creation of new derivatives and structured products. Although these products are frequently used for hedging purposes, slow-moving but market-traded risks can also be converted to fast-moving risks through the application of leverage. Because leverage multiplies the payoff to a risk exposure, it can transform the typically small payoff from a change in a slow-moving risk into a large payoff to the holder of a leveraged exposure to the risk. For example, a small price change on an interest-rate future or credit default swap can become a source of large profits if the position is highly leveraged. But the use of leverage also makes it possible for a sudden but small adverse price change to impose large losses that impair a firm’s capital. This situation corresponds to the definition of fast-moving risk given above.

29To the extent that banks specialized in managing slow-moving risks in the past, and continue to do so today despite taking on some fast-moving risks as well, this implies that banking crises tend to be slow-moving crises. In other words, they develop gradually over several discrete periods and give some advanced warning of increased future losses, like the underlying risks that cause them. This in turn implies that regulators should have enough time to intervene to mitigate banking crises. In other words, this suggests that the story of the one-step-behind regulator is not an accurate description of banking crises. This does not imply, of course, that banking crises will necessarily be infrequent or mild, only that they are not well described by the one-step-behind story. In the next section, we discuss the historical experience of financial crises and examine whether this analysis is valid.

2. Financial Crises: Fast-Moving or Slow-Moving Risks?

30We now use the definitions of slow-moving and fast-moving risks set out in the previous section to examine the characteristics of recent financial crises. The narrative of the one-step-behind regulator suggests that new financial products get out of control and cause crises before regulators are either aware of the danger or have sufficient time to extend appropriate regulation to the product. This makes for a compelling story, but is it an accurate representation of most financial crises?

31A careful consideration of the financial crises that have taken place during the era of financial innovation—which began in the 1960s—suggests that it is not. In other words, financial crises seem more often to have been caused by slow-moving risks than by fast-moving risks. Of course, fast-moving risks doubtlessly play an important role in many crises along with slow-moving risks. But when we look for the main, underlying causes of historical crises, we find that slow-moving risks are more prevalent than fast-moving risks.

  • 11 See Caprio (1997) and Laeven and Valencia (2012).
  • 12 See Montiel (2014) for detailed summaries of these crises.

32The overwhelming majority of financial crises during the past half-century have been banking crises. Banking crises in the United States, the Nordic countries, and Japan that took place in the 1980s and 1990s have received the most attention internationally, but nearly every country with a banking system experienced at least one banking crisis during the 1970-2000 period.11 In a typical banking crisis, banks overexpand their loan portfolios and after some time has passed, suffer excessive rates of default, which often trigger bank failures and recessions. Many major international financial crises of the past half century, such as the Chilean crisis of 1982, the East Asian crisis of the late 1990s, the Turkish crisis of 2001, and the Icelandic crisis of 2008 are fundamentally banking crises.12

33Banking crises are slow-moving crises, in the sense that it takes several years to incubate one. Stefan Ingves, the head of Sweden’s central bank and chair of the Basel Committee on Banking Supervision, observed based on his own country’s experience that it takes three years to ruin a bank—that is, to go through the entire cycle from rapid loan expansion to the insolvency of the bank.

34Thus, the characteristics of a banking crisis violate two essential premises of the one-step-behind narrative. First, the instruments that cause these crises are not new products but rather one of the oldest financial products, the bank loan. Second, the risks of excessive lending manifest themselves so slowly that even a regulator who is one step behind should have sufficient time to perceive the rising risk and then act in ways to prevent or at least mitigate a developing banking crisis.

35These crises did take place in large part because of regulatory lags. But as the literature has shown, these lags were created by intentional temporizing and forbearance on the part of regulators. Choosing a course of inaction is fundamentally different from lacking sufficient time to understand or deal with the danger. Thus, in these crises, giving the regulators more time would not have solved the problem, but only allowed the problem to grow or intensify. The experience of Japan is instructive in this regard. This country lost at least one and possibly multiple decades of economic growth due to the fact that regulators were allowed to act slowly, cautiously, and gradually in response to the country’s banking crisis.

  • 13 See Montiel (2014) for detailed discussions of these crises.
  • 14 See Montiel (2014).

36Another major type of financial crisis that has occurred during the era of financial innovation is the sovereign debt crisis. Examples include the Mexican “Tequila” crisis of 1994, the Brazilian crisis of 1999,13 as well as the European sovereign debt crisis that began in 2010. Like banking crises, sovereign debt crises involve instruments that have existed for hundreds of years. In addition, these are also slow-moving crises. Debt-to-GDP levels in countries like Greece and Italy, for example, were known to be excessively high for years before the crisis emerged. And finally, sovereign debt crises almost inevitably involve banks, who purchase large quantities of this debt because of its attractive yields. In fact, it is usually the banks’ holdings of sovereign debt that turn a possible sovereign default into a crisis situation. The Russian default of 1998, for example, damaged domestic Russian banks that held large quantities of government securities and caused problems in the financial markets through its knock-on effects.14

  • 15 See Lowenstein (2000) and Jacque (2010). LTCM both took on direct exposure to market risks and conv (...)

37The 1998 Russian default is a valuable example because it shows what financial regulators can accomplish, even when they are one step behind the financial industry. The Russian default’s main international impact was that it triggered a flight to quality, which in turn caused the Long Term Capital Management (LTCM) hedge fund to implode due to fast-moving risks.15 The impending failure of this institution became a systemic danger because of the large positions in sovereign debt (and other financial instruments) that LTCM had accumulated, through leverage. Had LTCM been forced to liquidate its positions quickly, this could have caused a financial panic.

38But because the leverage used by LTCM originated predominantly in banks, banking regulators were able to respond by coordinating a private bailout of LTCM by a consortium of banks that had loaned the fund billions of dollars. This coordination, moreover, took place over a weekend. The banking regulators did not anticipate the changes in market prices that caused LTCM’s positions to lose most of their value, but they did anticipate what LTCM’s losses would do to the fund’s ability to repay its loans and the consequences for the financial health of the banks under their purview of an LTCM default. The risk of default on the banks’ loans to LTCM was a slow-moving risk whose effects the regulators understood and could mitigate in a timely way.

39The other instructive aspect of the LTCM case is that even though the cause of the problem was fast-moving risks related to trading, the banks were sufficiently insulated from these fast-moving risks because their exposure to these risks was indirect, through the loans they extended to LTCM. That is, the hedge fund LTCM effectively transformed fast-moving market risk into slow-moving credit risk, so that the banks were not directly exposed to fast-moving market risks. Banks, and their regulators, did not have to deal with fast-moving risks.

40The Russian default and ensuing LTCM crisis demonstrates how the segregation of risks according to speed can insulate the financial system from fast-moving risks. Although market risk caused by the Russian default led to the collapse of a prominent hedge fund, it did not cause widespread financial panic. Again, this was largely due to regulators who were fast enough to anticipate the consequences for banks’ credit risk exposures to the fund and then coordinate an effective private bank response to the likely default of LTCM. In this case, the segregation of risks provided enough of a time lag to enable regulatory mitigation of the risk before it precipitated a full-blown crisis.

41Another example of the value of risk segregation is the Black Monday stock market crash of 1987. In this case, market risk was amplified by financial innovation. Synthetic put options sold as “portfolio insurance” set up a feedback loop in stock and stock index futures that caused an extremely rapid fall in stock prices, amounting to an over 20 percent loss in one day. This fall in prices put market makers and other broker-dealers into potential financial distress. But these losses did not affect banks, due to the segregation of risks between banks and nonbank intermediaries. In fact, an announcement by the Fed late on Black Monday that the Fed stood ready to provide needed liquidity to the broker-dealer system—intermediated through the banks—was instrumental in calming the markets and averting wider (or lasting) damage from the stock market crash. The Black Monday crash was one of the most clear examples of financial innovation leading to problems, but it failed to precipitate a financial crisis, at least in part because of the segregation of risks in the system.

42It was only when this risk segregation broke down that the one-step-behind narrative began to play out in reality. This happened as banks began to take on direct exposures to fast-moving risks in the early 2000s. In some cases, banks took on exposure to fast-moving market risk by engaging in proprietary trading. But it was actually liquidity risk—not from deposits, but market-based, wholesale funding sources such as commercial paper and repurchase agreements (repo)—that ended up doing significant damage to banks. During the early 2000s, banks increasingly turned to capital markets and away from deposits for their funding. For example, the British bank Northern Rock turned to capital markets to fund its massive expansion into mortgage lending during this period. Capital markets can provide huge amounts of wholesale funding, and can do so quickly, but this funding can leave just as quickly, which is a source of liquidity risk that moves faster than classic deposit runs. The sudden departure of capital-market-sourced liquidity is actually what placed banks in jeopardy faster than regulators could react.

  • 16 See Bruni and Llewellyn (2009).

43The failure of Northern Rock in 2007 was the first major example.16 When the capital markets began to fear that this institution had overexpanded its mortgage lending, the bank lost access to the capital market funding it had become dependent on. The bank could not replace this wholesale funding quickly, which caused it to go into distress. But regulators were unprepared for the sudden loss of funding and did nothing to ameliorate the situation. Indeed, Mervyn King, the head of the Bank of England, responded to this development by drawing a line in the sand, declaring that there would be no bailouts. Four months after this statement, however, the bank had been nationalized. Regulators who were a step behind the industry found themselves forced to do the one thing they explicitly said they would not do, because the risks moved faster than they were prepared to deal with.

  • 17 See Covitz, Liang, and Suarez (2009).

44The Structured Investment Vehicles (SIVs) crisis of late 2007 followed the same template.17 Large, internationally active banks such as Citibank, HSBC, and ABN-Amro had created huge special-purpose vehicles called Structured Investment Vehicles that purchased high-yielding securities—mostly subprime CDO tranches—and funded these purchases by issuing asset-backed commercial paper (ABCP). As in the case of Northern Rock, banks chose to expose themselves directly to capital-market-based liquidity risk rather than fund this lending via deposits. In the case of the SIVs, the risk moved especially quickly because the exposure was to commercial paper, which has very short maturities. When capital market participants refused to roll over the maturing ABCP issued by the SIVs, this set off a crisis within the companies. As in the case of Northern Rock, regulators were generally unprepared to respond to this fast-moving risk. Their response was to propose a joint SIV backed by the government, a proposal that was not only unattractive but late in coming. By the time this proposal was made, the banks had already been forced to adopt their worst-case contingency plans of funding the SIV assets themselves, necessitating the consolidation of the SIV assets on the balance sheets of the banks. Taking these “toxic” assets onto the banks’ balance sheets was a devastating blow to the firms’ financial health that nearly proved fatal, and probably stoked the rising sense of impending trouble in the financial system.

  • 18 See Acharya et al (2011).

45These examples are suggestive, but they raise two important questions that will help us refine our understanding of the one-step-behind narrative. First, how does the mixing of fast-moving and slow-moving risk lead to crisis? And second, does the one-step-behind narrative apply to the regulation of all financial institutions, or exclusively to the regulation of commercial banks? In other words, there are many types of nonbank financial institutions that regularly expose themselves to fast-moving risks, and yet the only financial crisis of the past fifty years which seems to have emanated from these institutions was the global financial crisis of 2007-09. And even in this case, banks (and quasi-banks such as the U.S.’s so-called Government Sponsored Enterprises, Fannie Mae and Freddie Mac) arguably play the leading role.18 In the following section, we discuss how the mixing of fast-moving and slow-moving risks can send financial institutions into distress. In the subsequent section, we use the second question to motivate a broad examination of the options that regulators have for dealing with fast-moving risks.

3. How Mixing Risks Can Lead to Crisis

  • 19 In previous sections, we also considered the use of leverage to convert slow-moving risk into fast- (...)

46In this section of the paper, we consider the mechanics of how the mixing of fast-moving and slow-moving risks can lead to institutional failure and possible systemic crises. We do this in order to support our claim that the mixing of risks better explains the causes of the recent global financial crisis than the one-step-behind narrative. We consider the two most common cases of mixing slow-moving and fast-moving risks: mixing these risks in the asset portfolio; and using liabilities with fast-moving risks to fund assets with slow-moving risks.19

3.1. Combining Exposures to Fast- and Slow-Moving Risks in the Asset Portfolio

47By the early 2000s, many large, international commercial banks were actively conducting proprietary trading, which is trading for the benefit of the bank, rather than trading on behalf of its customers. So-called “FICC” desks, which stands for fixed income, currencies and commodities, generated significant profits for banks through market making and, increasingly, through proprietary trading. The potential danger of adding proprietary trading to the asset portfolio, which is discussed in detail below, is that fast-moving market risks may generate losses that are so large they reduce the capital of the institution significantly, threatening its solvency.

  • 20 See for example IMF (2013).

48Proprietary trading did not appear to push any banks into distress, even during the global financial crisis, but this does not necessarily imply that it could not happen. And given the regulatory responses to the financial crisis, it appears that regulators believe that this is not only possible but likely. These responses include strict limits on proprietary trading by bank holding companies, in the form of the so-called Volcker Rule in the U.S., as well as segregation of lending from trading activities in the form of so-called ringfencing rules in the U.K.20 And across jurisdictions, regulators have used supervisory tools such as aggressive monitoring and moral suasion to raise the cost of engaging in proprietary trading and thus encourage banks to downsize their FICC desks or drop these activities altogether. But these actions have been driven more by regulators’ beliefs than by experience.

49Because we lack real-world cases of banks that went into distress because of market risk, we have to rely on the next-best evidence in order to envision how this could occur, which exists in at least one theoretical model of bank behavior. Recently, Chami et al. (2017) have developed a model in which a bank holding company engages in both trading and traditional bank lending. Their framework enables them to examine the impact of fast-moving market risks on the institution’s business decisions as well as its solvency. Their paper describes how the mixing of fast-moving market risks and slow-moving credit risks in the asset portfolio of a financial corporation can lead to distress. Although the argument is presented in terms of a bank holding company, the basic story can be applied to any financial institution that mixes fast-moving and slow-moving risks in their asset portfolio.

50In the Chami et al. (2017) model, a bank holding company combines exposures to fast-moving and slow-moving risks in the following way. The holding company owns both a bank, which specializes in taking on exposure to slow-moving credit risk, and an investment fund that specializes in taking on exposure to fast-moving market risk. The managers of the bank holding company must decide how to allocate resources, including capital, to each of its businesses. Thus, the two businesses are in competition with each other for scarce firm resources. If the managers of the bank holding company are more optimistic about the future profitability of taking on exposure to market risk, then they will allocate more capital and other resources to the investment fund business and fewer to the bank.

  • 21 Bebchuk and Spamann (2010) discuss the incentives of bank holding company managers to gamble with i (...)

51In the bank holding company structure, fast-moving risks can endanger the institution in two ways. First, the market risks taken on by the investment fund may lead to losses that are so large, and occur so quickly, that they bankrupt the entire bank holding company, including the banking division. In this case, regulators may well be placed in their worst-case scenario: having to deploy deposit insurance funds to resolve a bank that was ruined by gambling. Even though in this model the banking and the investment businesses are separate, one could argue that the bank holding company is still gambling, albeit indirectly, with insured deposits. Second, the managers may allocate an insufficient amount of capital and other resources to the banking division but refuse to reduce the scale or credit risk exposures of this division, increasing the likelihood of financial distress originating in the banking business. If the managers of the investment company fully deploy their capital and other resources, then they may not have any excess capital to enable the holding company to absorb unanticipated losses to the banking business. Again, this basically amounts to gambling with insured deposits on the part of the bank holding company’s managers.21

  • 22 See United States Senate Permanent Subcommittee on Investigations (2013).
  • 23 See Jacque (2010).
  • 24 See Fortado (2015).

52The Chami et al (2017) model suggests that mixing fast-moving and slow-moving risks in the same financial institution by placing both risks on the asset side of the balance sheet can be dangerous. Although this mixing of risks has not resulted in a financial crisis per se, significant loss episodes in bank holding companies or universal banks related to market risk have occurred. One such example is the “London Whale” losses that JP Morgan Chase sustained in 2013,22 while other trading losses occurred at Société Générale in 200823 and UBS in 2011.24 These losses, which amounted to billions of dollars, euros, and pounds sterling respectively, could potentially have become large enough to be an existential threat to these firms, had they gone undiscovered for longer, perhaps only a few more months in the case of the London Whale.

3.2. Funding Slow-Moving Risk Assets with Fast-Moving Risk Liabilities

53The other main way that both banks and other financial institutions mix fast-moving and slow-moving risks is by raising funds through issuing liabilities exposed to fast-moving risks and using the proceeds to invest in assets exposed to slow-moving risks. In particular, the fast-moving risk on the liability side of the balance sheet tends to be liquidity risk (including both market and funding liquidity risk), while the slow-moving risk on the asset side of the balance sheet tends to be credit risk.

54It is possible that funding liquidity simply dries up for reasons beyond the institution’s control. For example, this appears to be the case in many of the 19th century American financial panics, in which liquid reserves tended to pool in large city-based banks and could not be deployed quickly to banks in smaller towns and rural areas, causing bank runs on the small banks. This could (and in many cases did) lead to solvency problems when banks responded to the lack of liquidity by calling in their short-term loans.

55In the modern financial system, however, what seems to occur most frequently is that concerns over the slow-moving risks in the company’s asset portfolio drive a loss in funding liquidity that in turn pushes the institution into distress. For banks, problems (or rumors) regarding non-performing loans may prompt depositors to withdraw their funds, leading to a liquidity crisis. This was evident in the “silent run” on Continental Illinois—a large lender to oil production and exploration companies that failed when oil prices collapsed in the mid-1980s—that led to its bailout and the emergence of “too big to fail”. One of the early events of the global financial crisis involved investment funds operated by Bear Stearns and BNP Paribas. Investors in these funds grew concerned over the quality of the CDO tranches these funds invested in, and the subsequent run forced the companies to suspend withdrawals or bail out their funds by purchasing the questionable assets. Even in the case of Northern Rock, the funding liquidity problem was caused by a sudden lack of confidence in securitized assets. Northern Rock’s securitized assets may have been perfectly sound, but investors lost faith in all securitized assets, and refused to purchase them just at the time when Northern Rock needed to renew its funding.

  • 25 See Adrian and Shin (2008).

56Another possible way that liquidity risk causes financial distress is through market liquidity risk related to collateral. If a financial institution takes on debt or other commitments for which it needs to post collateral, it is vulnerable to collateral calls requiring the institution to post additional collateral that often must be paid in cash. If the firm needs to raise this cash collateral quickly, it may have to resort to selling some of the securities it owns, which may be illiquid, leading to losses. In the extreme case, collateral calls may escalate and the firm may have to liquidate its assets at successively lower prices, triggering solvency concerns as the firm’s asset values fall. During the financial crisis, several firms underwent a “collateral death spiral” in which the collateral they posted for short-term borrowing via repo itself fell in value, leading the firms to sell off more of the same assets to raise cash, thus initiating a negative feedback loop.25

57As in the case of the loss of funding liquidity discussed above, however, the driving force behind the problem generally appears to be slow-moving credit risk. The collateral call may be initiated because the lender revises the borrower’s creditworthiness downward, for example. Or, as in the case of the collateral death spirals during the financial crisis, the collateral itself loses value because of a fall in its credit quality, triggering the initial collateral call.

58As this discussion shows, there are multiple ways that liquidity risk in liabilities can interact with credit risk in assets to produce financial distress. But in general, it appears that the mixing of fast-moving risks on the liability side of the institution with slow-moving credit risks on the asset side is an especially volatile combination. Shocks to credit risk, such as downward revisions in a firm’s or an asset’s credit quality, tend to set off the liquidity problems that ultimately cause the firm to go into distress. Mervyn King (2016, 254) argues that this is what triggered the global financial crisis: “And in the recent crisis, concern about solvency was the main driver of the liquidity problems facing banks. When creditors started to worry that bank equity was insufficient to absorb potential losses, they decided that it was better to get out while the going was good”.

59The above discussion supports our claim that the one-step-behind narrative is not the best explanation for the recent financial crisis. Interestingly, slow-moving risks once again appear to be the ultimate cause of the crisis, but with additional problems resulting from the mixing of fast-moving risk liabilities and slow-moving risk assets. In retrospect, this mixture made the financial system less robust to shocks, because they would quickly trigger liquidity problems. In addition, the mixing of risks made the damage from the crisis worse, both by enabling a greater expansion of the asset portfolio on the way up, and inflicting more severe losses on the way down. Thus, it appears that mixing fast-moving and slow-moving risks amplifies both the likelihood and the severity of financial crisis. As in previous crises, the slow regulatory response to evidence of increasing credit risk, such as decreased lending standards, plays a more important role than the pace of financial innovation. Nonetheless, the mixing of fast-moving and slow-moving risks is an innovation that should be addressed, if possible, through regulation. We turn to the challenges of doing so next.

4. Strategies for Regulating Fast-Moving Risks

60We have been arguing that combining fast-moving risks and slow-moving risks in the same institution is the main way that fast-moving risks present a threat to financial stability. But perhaps it is possible to regulate financial institutions in a way that would allow them to take on exposures to both risks without posing a systemic threat. Therefore, in this section, we consider how to regulate fast-moving risks in order to ensure that they do not threaten financial stability. We take lessons both from the ways that these risks are actually managed, as well as suggestions from the literature regarding how they could be managed.

61An observation that helps frame this section is that nonbank financial institutions, which routinely take on direct exposures to fast-moving risks, have not caused many financial crises. The key question raised by this observation is whether this stylized fact is better explained by effective risk management at the institutional level, or by regulatory strategies that allow for institutional failure without incurring systemic risk.

62Let us consider the first explanation. How do the firms that take on exposures to fast-moving risks mitigate them? The first step in mitigating these risks is to measure them, and it turns out that even the act of measuring fast-moving risks has severe limitations. The standard model for measuring fast-moving market risk, for example, is Value at Risk (VaR), which is prone to many weaknesses such as procyclicality and data limitations. In fact, Rebonato (2007) argues convincingly that the most critical questions that VaR attempts to address are intrinsically impossible to answer, given our limited data. For some types of fast-moving risk, such as liquidity risk, we lack adequate models for measuring this risk.

  • 26 See Adrian and Shin (2008).

63Despite the problems of measuring fast-moving risks, many strategies for mitigating these risks do exist, such as posting collateral, purchasing insurance, or hedging. But these strategies also have significant weaknesses that may prevent them from adequately mitigating the risks they are designed to manage. To begin with, the instruments used to implement all of these strategies are all subject to fast-moving risks themselves, which can limit their ability to reduce risks or even make them amplify risks. For example, a well-known problem faced by most financial instruments is that the correlations of their returns converge to one during times of stress. In other words, prices all tend to move in the same direction in a crisis. During the global financial crisis of 2008-2009, for instance, collateral posted to absorb market risk lost a significant share of its value during times of market stress, placing further financial pressure on firms to find additional collateral.26

64A further problem with these strategies is that the instruments themselves may not be well suited to mitigating each fast-moving risk. In other cases, even when the instrument is well suited to mitigating the risk, its use may be quite costly. Options and other insurance arrangements, for example, are usually expensive during normal times and their prices rise during times of stress to unaffordable levels. In the case of liquidity risk, the only good hedging instrument for this risk is cash—and holding large amounts of cash incurs a very high opportunity cost for businesses, especially financial intermediaries.

65Because of these problems with measuring and mitigating continuous risks, relying on the private sector does not appear to be sufficient to prevent fast-moving risks from causing distress in financial institutions that mix exposures to fast-moving and slow-moving risks. Earlier in the paper, however, we discussed how the presence of deposit insurance enables banks to mix fast-moving funding liquidity risk and slow-moving credit risk on their balance sheets. Would it be feasible and desirable to provide deposit or other insurance to financial intermediaries that take on exposures to both fast-moving and slow-moving risks?

  • 27 To the extent that the banks also constitute the bulk of the payment system, this is another justif (...)

66In general, the answer to this question appears to be negative. Insurance of any kind brings with it moral hazard problems, and even in the case of deposit insurance, these moral hazard problems have occasionally been severe. The benefit of protecting individual depositors—households, which generally have limited means and are financially unsophisticated—has provided a compelling reason in the past to tolerate these moral hazard problems.27 In addition, because funding liquidity crises at banks are generally triggered by doubts regarding the performance of the banks’ loan portfolios, the liquidity risk event is actually derived from credit risk, which is a slow-moving risk. This aspect provides the possibility that regulatory intervention can mitigate credit risks before they trigger a liquidity risk event. This in turn lessens the moral hazard problem and reduces the effective risk of a liquidity event that would trigger the deposit insurance. Thus, regular examinations aimed at ensuring bank safety and soundness have become part of the price of extending deposit insurance to banks.

67It is doubtful that the moral hazard problem caused by extending insurance to other fast-moving risks could be effectively mitigated in the same way that deposit insurance mitigates liquidity risk in banks. If a bank’s liquidity risk is derived from market risk, then government supervision may not be able to mitigate the market risk sufficiently. For example, if a bank uses insured deposit funds to finance purchases of shares or other market-traded assets, these investments can lose their value without warning, triggering a liquidity crisis in the bank that could not have been foreseen or prevented by supervisors. Although VaR or other methods could be used ex ante to try to limit the bank’s exposure to continuous risks and thus ensure some measure of safety and soundness in the context of fast-moving risks, we have discussed above how these methods of measuring fast-moving risks have significant weaknesses that limit their effectiveness.

68In the case of market liquidity risk, Mehrling (2011), Mehrling et al. (2014) and King (2016) present interesting proposals for public liquidity insurance. Mehrling argues that the Fed already successfully became the “dealer of last resort” during the financial crisis, and that it and other central banks should continue in this role permanently. As dealer of last resort, a central bank would stand ready to buy securities from market participants at prices that are sufficiently below the current market price to discourage all but those desperate for liquidity from selling them to the central bank. King’s proposal, which is similar, calls upon central banks to act as a “pawnbroker for all seasons” that similarly stands ready to buy collateral. In King’s proposal, however, financial institutions would be able to precommit to sell the central bank specific collateral in case they run into liquidity problems, and the central banks would precommit to pay specified prices for the contingently pledged collateral. King uses this arrangement as a new basis for regulating commercial banks: they would only be allowed to raise as much deposit funding as the value that the central bank promises to pay for the collateral the bank has pledged to it.

69Formalizing the provision of market liquidity into a standing central bank facility could potentially enhance the robustness of the financial markets. But in their proposals, both Mehrling and King abstract from credit risk. As we have argued above, a main source of the liquidity problems associated with financial crises is a perceived fall either in the overall creditworthiness of financial institutions, or in the quality of the individual assets they hold. If credit risk is taken into account, the attractiveness of these proposals becomes much less clear.

70If market liquidity falls because of an exogenous shock or an accident, then these proposed liquidity facilities would probably be effective. For example, it can be argued that the loss of liquidity in the mortgage-backed securities market during the financial crisis was an accident caused by faulty disclosure of the contents of collateralized debt obligations. In short, all mortgage-backed securities were shunned because it was too difficult to determine whether any particular security contained exposure to bad mortgage loans. Therefore, the Fed was able to supply liquidity by purchasing some types of illiquid but high quality mortgage-backed securities (basically, those guaranteed by Fannie Mae and Freddie Mac) that the market was temporarily undervaluing. But this example suggests that accidental market liquidity problems are the result of misperception or mispricing on a grand scale. In reality, how often do accidents like these occur?

71On the other hand, if the source of market liquidity problems is an actual decline in the credit quality of securities used for collateral, a dealer of last resort may not actually be able to provide liquidity by purchasing these securities, since central banks want to avoid exposure to credit risk. And if the dealer of last resort does purchase securities that have declined in credit quality, it runs the risk of stoking the same moral hazard problem that affects deposit insurance. This is a significant concern, because historical experience indicates that market liquidity problems tend to originate in the deterioration of the credit quality of the securities. To the extent that market liquidity problems do originate in credit risk or other slow-moving risks, this suggests that a safety-and-soundness approach could mitigate the problem. But this would in turn require that safety and soundness regulation and supervision be imposed on all issuers of securities, which is quite impractical.

72The work of Chami et al. (2017) suggests another approach to safety and soundness regulation in the context of fast-moving risks. In their paper, they show that it is theoretically possible for the supervisors of bank holding companies to induce these companies to choose the socially optimal levels of risk exposure for the investment and lending businesses owned by the holding company. They can do this by limiting the institution’s leverage, and by intervening in its governance to limit the risk appetite of management. In order to do this, however, the supervisors must be able to estimate the relevant probability distributions of the risk exposures and also measure the investment manager’s coefficient of risk aversion. In addition, bank supervisors must either possess the power to persuade bank managers to adjust their risk appetites or be given the effective ability to replace bank managers at their discretion. All these tasks are challenging and it is unclear whether supervisors can perform them well enough in the real world to achieve the optimal outcome. In all likelihood, some will be able to do this while many others will not.

73Thus, it does not appear that safety and soundness regulation and supervision can be used in the context of fast-moving risks in order to limit the liabilities of the insurance system ex ante. Moreover, the inability of regulators to mitigate the damage from fast-moving risks, once a loss event occurs, severely exacerbates the moral hazard problem. In other words, if the cause of a loss is a fast-moving risk, regulators will not have time to make the bank managers change their behavior in order to limit the loss, as they do in the case of discrete risks. Instead, regulators will be forced to make payouts from the insurance fund if a loss event occurs. The increased certainty of insurance fund payouts in cases of loss will incentivize firms to take on larger exposures to these risks ex ante.

74Therefore, the successful use of deposit insurance to mitigate funding liquidity risk in banks arises from a specific set of conditions that do not hold for exposures to other types of fast-moving risk. In particular, the fact that the fast-moving liquidity risk faced by banks and other financial institutions generally derives from slow-moving credit risk creates the possibility that effective supervision can sufficiently limit the moral hazard problem in this insurance contract, so that it becomes worthwhile from a societal perspective to provide the insurance. In the case of insuring against funding liquidity risks derived from other fast-moving risks, supervision will be ineffective at limiting the moral hazard problem and in fact the increased likelihood of payout in cases of loss will worsen the moral hazard problem. In the case of insuring against market liquidity risk, the insurance mechanism is problematic and supervision of all securities-issuing institutions is impractical. Therefore, providing insurance to financial institutions against fast-moving risks is not a good strategy for allowing institutions to take on exposures to these risks.

75Since private-sector risk-mitigation programs cannot sufficiently mitigate fast-moving risks, and insuring these risks is a counterproductive strategy, this leaves only one main way to prevent financial crises caused by the fast-moving risks taken on by financial institutions: limit the damage to the system caused by distress in an individual institution. This appears to be the main way in which regulation has actually developed. In particular, regulators of nonbank financial institutions, such as the Securities Exchange Commission (SEC) in the U.S., pursued a regulatory strategy of allowing failures and limiting the damage from the failure of an institution rather than preventing failures. For example, one of the SEC’s strictest policies is the enforcement of segregation of customer funds from a broker-dealer’s proprietary funds. Thus, in the case of the failure of a broker-dealer due to losses on a proprietary trading portfolio, customers’ funds will be unaffected. This largely removes the incentives for runs that could damage the broker-dealers and also destabilize trading markets.

76In addition, the strategy of allowing firms to fail creates an incentive for nonbank financial institutions to design and use mechanisms to protect themselves. These mechanisms further limit the damage from one institution’s failure. For example, the practice of demanding collateral for trades and other transactions limits the losses when a counterparty defaults. Similarly, institutions developed a system of institution-by-institution exposure and credit limits they imposed on their counterparties, as well as position limits on individual traders. With the advent of measurement tools such as Value at Risk (VaR), moreover, these position limits were imposed in terms of risk as well as in monetary terms. Even though we argued above that these private-sector tools are insufficient for preventing failures, they can still achieve the regulators’ goal of limiting contagion risk from one institution’s failure.

  • 28 See Jacque (2010) and United States Financial Crisis Inquiry Commmission (2011).

77The strategy of limiting systemic damage from one institution’s failure still poses challenges to regulators. In particular, problems in monitoring and limiting inter-institutional exposures to fast-moving risks were exposed during the global financial crisis of 2008-09 by the emergence of institutions that were “too connected to fail” due to the large number of counterparties and total exposures on their trading books. For example, AIG Financial Products’ immense exposure to credit default swaps that it had made with virtually every major financial institution did not come to light until these exposures threatened to bankrupt the entire company.28 In response to this problem, however, regulators have mandated that as many contracts as possible be moved onto central counterparties such as clearing houses, in order to both monitor exposures to fast-moving risks as well as to limit the damages to the system from the possible distress of one institution. This example shows that regulators are improving the ways in which they limit the damage from one institution’s distress, and are still using this as their main approach to managing fast-moving risks taken on by financial firms.


  • 29 See Fullenkamp and Rochon (2016) for a regulation proposal that emphasizes prompt, corrective actio (...)

78Regulators will always be one step behind the markets. But as we have argued in this paper, this is not a primary cause of financial crisis. Instead, we argued that the speed with which various financial risks inflict losses is a more important concern when thinking about how to design or reform financial regulation. We characterized different financial risks as fast-moving and slow-moving, and showed that slow-moving risks, especially credit risk, have been responsible for most of the significant financial crises that have taken place during the past several decades. The historical experience of crises suggests that the political economy of regulation, which enables regulators to temporize and exercise forbearance rather than take action against slow-moving risks, bears primary responsibility for the occurrence and severity of financial crises. Redesigning regulations in a way that ensures prompt, corrective action on the part of regulators to prevent slow-moving risks from inflicting losses, or to mitigate the damage from slow-moving risks that have already become loss events, should therefore be a focus of efforts to improve financial regulation.29

79In addition, we argued that financial innovation per se does not lead to financial crisis. One innovation that does threaten financial stability, however, is the mixing of fast-moving and slow-moving risks within financial institutions. The mixing of risks in a particular way—taking on fast-moving liquidity risk on the funding or liability side of banks—was one of the main causes of the global financial crisis. We further argued that mixing fast-moving and slow-moving risks in any way increases the likelihood of institutional failure and hence of financial contagion and crisis.

80We also considered whether there exist mechanisms that can enable financial institutions to engage in mixing these two risks without endangering financial stability. But we showed that neither private risk management techniques nor (government-provided) insurance is an effective or desirable means of accomplishing this goal. Indeed, the main strategy that regulators now use to limit the systemic danger posed by fast-moving risks is simply to limit the amount of damage, including contagion, that will result when fast-moving risks produce loss events.

  • 30 See, for example, Bliss and Kaufman (2011) and Jackson and Skeel (2012), as well as Basel Committee (...)

81It is conceivable, however, that an effective set of macroprudential regulations, when paired with a bankruptcy regime for financial institutions, could allow financial institutions to take on exposures to both fast-moving and slow-moving risks and still provide a high level of protection against financial crisis. An improved bankruptcy regime30 would result in the timely and fast resolution of financial institutions that succumb to financial risks of any kind, therefore limiting the potential damage and contagion risk resulting from this failure. In addition, an effective set of macroprudential regulations would further limit the ability of one institution’s distress to spread to other institutions. Unfortunately, however, neither member of this pair of policies is currently in place now, nor is likely to be completely implemented in the near future.

82At least for the meantime, then, placing strict limits regarding the amount of risk-mixing that may take place in an institution, with the option of returning to a policy of segregating fast-moving and slow-moving risks in separate institutions, appears to be the best way to prevent fast-moving risks from causing financial crises. Although this prescription is not new, having been implemented in various forms and to varying degrees in the U.S., U.K., and E.U., this paper has presented a novel argument to justify this separation. The strategies for dealing with fast-moving and slow-moving financial risks, both from a private risk management perspective as well as a regulatory perspective, are fundamentally incompatible. The remaining alternative is to limit the systemic damage that the failure of an institution would cause, which is facilitated by limiting the mixing of risks or the outright segregation of fast-moving risks and slow-moving risks into separate financial institutions.

83As the results of Chami et al. (2017) suggest, the limits on mixing risks depend mostly on the capacity of supervisors to adequately measure an institution’s exposures and discipline its managers. If supervisory capacity or authority are low, or the probability distributions of the risks prove excessively difficult to estimate, then segregation of slow-moving risks in commercial banks and fast-moving risks in broker-dealers will be the best policy.

  • 31 See Mehrling et al (2014) for a definition of shadow banks.
  • 32 See, for example, IMF (2014).

84Unfortunately, however, the financial world does not fit neatly into the two categories of bank and broker-dealer. In particular, institutions now known as shadow banks31 have always been part of the financial markets, and in recent years have exhibited tremendous rates of growth in many countries32. Finding a sensible way to regulate shadow banks and prevent them from causing financial instability, while preserving the services they currently provide, is a challenge that must be addressed in the near future.

85The analysis of this paper suggests that shadow banks must each be evaluated on the degree to which they are mixing fast-moving and slow-moving risks. In particular, some shadow banks may have found techniques to minimize or mitigate the fast-moving risks they take on. For example, a shadow bank that closely matches the maturity of its borrowing to the maturity of its lending may sufficiently minimize the funding liquidity risk it takes on so that the mixing of risks may not pose a hazard to financial stability. On the other hand, shadow banks that are taking on significant exposures to fast-moving risks such as liquidity risk do pose potential threats to financial stability and should be required to alter their exposures, and perhaps to cease operations altogether.

86As the example of shadow banking demonstrates, analyzing financial risks in terms of their speed improves our understanding of financial crises and yields practical prescriptions for financial regulation. This way of categorizing and thinking about financial risks should therefore become a standard tool for financial risk management.

The authors thank Mark Levonian, Stijn Claessens, and the anonymous referees for helpful comments and suggestions. All remaining errors are the authors’ own.

Haut de page


Acharya, Vidal, Matthew Richardson, Stijn Van Nieuwerburgh and Lawrence J. White. 2011. Guaranteed to Fail: Fannie Mae, Freddie Mac, and the Debacle of Mortgage Finance. Princeton, NJ: Princeton University Press.

Adrian, Tobias and Hyung Song Shin. 2008. Liquidity and Financial Contagion. Banque de France Financial Stability Review, No. 11(February): 1-7.

Basel Committee on Banking Supervision. 2010. Proposal to Ensure the Loss-Absorbency of Regulatory Capital at the Point of Insolvency, August. Basel: Bank for International Settlements.

Bebchuk, Lucian A. and Holger Spamann. 2010. Regulating Bankers’ Pay. Georgetown Law Journal, 98(2): 247-287.

Bliss, Robert R. and George G. Kaufman. 2011. Resolving Insolvent Large Complex Financial Institutions: A Better Way. Banking Law Journal, 128(4): 339-363.

Bruni, Franco and David T. Llewellyn. 2009. The Failure of Northern Rock: A Multidimensional Case Study. Vienna: SUERF, The European Money and Finance Forum.

Caprio, Gerard. 1997. Safe and Sound Banking in Developing Countries: We’re Not in Kansas Anymore. World Bank Policy Research Working Paper 1739. Washington, DC: The World Bank Group.

Chami, Ralph, Thomas Cosimano, Jun Ma and Celine Rochon. 2017. What’s Different About Bank Holding Companies. IMF Working Paper 17/26. Washington, DC: International Monetary Fund.

Cohan, William D. 2009. House of Cards: A Tale of Hubris and Wretched Excess on Wall Street. New York: Doubleday.

Covitz, Daniel, Nellie Liang and Gustavo Suarez, 2009. The Evolution of a Financial Crisis: Panic in the Asset-Backed Commercial Paper Market. Federal Reserve Board Finance and Economics Discussion Series Working Paper 2009-36. Washington, DC: Federal Reserve Board of Governors.

Dewatripont, Mathias, Jean-Charles Rochet and Jean Tirole. 2010. Balancing the Banks: Global Lessons from the Financial Crisis. Princeton, NJ: Princeton University Press.

Eichengreen, Barry, 2003. Strengthening the International Financial Architecture: Where Do We Stand? In Dilip Das (ed.), An International Finance Reader. New York: Routledge, 65-86.

Fortado, Lindsay. 2015. A Rogue Trader’s Tale. Financial Times, 24 October: 19-20.

Fullenkamp, Connel and Celine Rochon. 2016. Reconsidering Bank Capital Regulation: A New Combination of Rules, Regulators, and Market Discipline. Journal of Economic Policy Reform.

Hull, John. 2015. Risk Management and Financial Institutions, 4th edition. Hoboken, NJ: John Wiley & Sons Inc.

International Monetary Fund (IMF). 2013. Creating a Safer Financial System: Will the Volcker, Vickers and Liikanen Structural Measures Help? IMF Staff Discussion Note 13/4. Washington, DC: International Monetary Fund.

International Monetary Fund (IMF). 2014. Risk Taking, Liquidity and Shadow Banking: Curbing Excess While Promoting Growth. IMF Global Financial Stability Report, October. Washington, DC: International Monetary Fund.

Jackson, Thomas H. and David A. Skeel Jr. 2012. Dynamic Resolution of Large Financial Institutions. Harvard Business Law Review, 2(2): 435-460.

Jacque, Laurent L. 2010. Global Derivative Debacles: From Theory to Malpractice. Hackensack, NJ: World Scientific Publishing Co.

Kane, Edward J. 2003. What Economic Principles Should Policymakers in Other Countries Have Learned from the S&L Mess? Business Economics, 38(1): 21-30.

King, Mervyn. 2016. The End of Alchemy. New York: Norton.

Laeven, Luc and Fabian Valencia. 2012. Systemic Banking Crises Database: An Update. IMF Working Paper WP/12/163. Washington, DC: International Monetary Fund.

Lowenstein, Roger. 2000. When Genius Failed: The Rise and Fall of Long Term Capital Management. New York: Random House.

McMillan, Jonathan. 2015. Banking in the Digital Age: The Failure of Financial Regulation. The Guardian, 20 January.

Mehrling, Perry. 2011. The New Lombard Street. Princeton, NJ: Princeton University Press.

Mehrling, Perry, Zoltan Pozsar, James Sweeney and Daniel H. Neilson. 2014. Bagehot Was a Shadow Banker: Shadow Banking, Central Banking, and the Future of Global Finance. In Stijn Claessens, Douglas Evanoff, George Kaufman, and Luc Laeven (eds), Shadow Banking Within and Across National Borders. Hackensack, NJ: World Scientific Publishing, 81-98.

Merton, Robert C. 2005. You Have More Capital Than You Think. Harvard Business Review, 83(11): 84-94.

Montiel, Peter J. 2014. Ten Crises. New York: Routledge.

Rebonato, Ricardo. 2007. Plight of the Fortune Tellers: Why We Need to Manage Financial Risk Differently. Princeton, NJ: Princeton University Press.

Rochet, Jean-Charles. 2008. Why Are There So Many Banking Crises? The Politics and Policy of Bank Regulation. Princeton, NJ: Princeton University Press.

Steil, Benn and Robert E. Litan. 2006. Financial Statecraft: The Role of Financial Markets in American Foreign Policy. New Haven: Yale University Press.

Tsingou, Eleni. 2010. Reactions to the Global Credit Crisis: Analyzing a Policy Community Under Stress. In Eric Helleiner, Stefano Pagliari and Hubert Zimmerman (eds), Global Finance in Crisis: The Politics of International Regulatory Change. New York: Routledge, 21-36.

United States Financial Crisis Inquiry Commission. 2011. Final Report of the National Commission on the Causes of the Financial and Economic Crisis in the United States. Washington, DC: Government Printing Office.

United States Senate Permanent Subcommittee on Investigations, 2013. J.P. Morgan Chase Whale Trades: A Case Study of Derivatives Risks and Abuses. Washington, DC: U.S. Government Printing Office.

United States Securities Exchange Commission and Commodity Futures Trading Commission. 2010. Findings Regarding the Market Events of May 6, 2010. Washington, DC: U.S. Government Printing Office.

Haut de page


1 Recent examples of this sentiment include McMillan (2015), Dewatripont, Rochet, and Tirole (2010), Tsingou (2010), Steil and Litan (2006) and Eichengreen (2003).

2 See, for example, Hull (2015).

3 Because operational risk involves losses resulting from people, processes and systems, it can be both a slow-moving and a fast-moving risk. In addition, it does not originate from financial activity per se, so we do not consider it in this paper.

4 Technically, credit risk is a type of counterparty risk, but we are following a common convention used in the markets.

5 Risks may interact with each other and one risk can be transformed into another. For example, in the global financial crisis of 2007-2009, liquidity risk surprisingly caused large losses at financial institutions when market liquidity problems led to so-called “collateral death spirals”. See Adrian and Shin (2008).

6 See United States Securities Exchange Commission and Commodity Futures Trading Commission (2010).

7 In addition, interest rates on private debt securities include a credit spread component that reflects the credit risk of the borrower, and hence is also a slow-moving component of interest rates.

8 See Cohan (2010).

9 It is also important to note that even in countries that lacked formal deposit insurance, an implicit guarantee on deposits was often in place, which was also effective in removing liquidity risk.

10 See for example Rochet (2008).

11 See Caprio (1997) and Laeven and Valencia (2012).

12 See Montiel (2014) for detailed summaries of these crises.

13 See Montiel (2014) for detailed discussions of these crises.

14 See Montiel (2014).

15 See Lowenstein (2000) and Jacque (2010). LTCM both took on direct exposure to market risks and converted slow-moving risks to fast-moving risks through the application of significant leverage.

16 See Bruni and Llewellyn (2009).

17 See Covitz, Liang, and Suarez (2009).

18 See Acharya et al (2011).

19 In previous sections, we also considered the use of leverage to convert slow-moving risk into fast-moving risk.

20 See for example IMF (2013).

21 Bebchuk and Spamann (2010) discuss the incentives of bank holding company managers to gamble with insured deposits.

22 See United States Senate Permanent Subcommittee on Investigations (2013).

23 See Jacque (2010).

24 See Fortado (2015).

25 See Adrian and Shin (2008).

26 See Adrian and Shin (2008).

27 To the extent that the banks also constitute the bulk of the payment system, this is another justification for desposit insurance.

28 See Jacque (2010) and United States Financial Crisis Inquiry Commmission (2011).

29 See Fullenkamp and Rochon (2016) for a regulation proposal that emphasizes prompt, corrective action.

30 See, for example, Bliss and Kaufman (2011) and Jackson and Skeel (2012), as well as Basel Committee on Banking Supervision (2010).

31 See Mehrling et al (2014) for a definition of shadow banks.

32 See, for example, IMF (2014).

Haut de page

Pour citer cet article

Référence papier

Ralph Chami, Thomas Cosimano, Connel Fullenkamp et Celine Rochon, « Financial Regulation and the Speed of Financial Risks », Œconomia, 7-2 | 2017, 161-190.

Référence électronique

Ralph Chami, Thomas Cosimano, Connel Fullenkamp et Celine Rochon, « Financial Regulation and the Speed of Financial Risks », Œconomia [En ligne], 7-2 | 2017, mis en ligne le 01 juin 2017, consulté le 18 octobre 2017. URL : ; DOI : 10.4000/oeconomia.2669

Haut de page


Ralph Chami

International Monetary Fund, Washington, DC 20431 USA.

Thomas Cosimano

University of Notre Dame, Notre Dame IN 46556 USA.

Connel Fullenkamp

Corresponding Author. Department of Economics, Duke University, Durham NC 27708 USA.

Celine Rochon

International Monetary Fund, Washington, DC 20431 USA.

Haut de page

Droits d’auteur

Licence Creative Commons
Les contenus d’Œconomia sont mis à disposition selon les termes de la Licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page
  • Logo Association Œconomia
  • Logo CNRS
  • Les cahiers de