On Financial Crises

Arjen Boin
Leiden University
University Crisis Research Center, Department of Public
Administration, PO Box 9555, 2300 RB Leiden, The Netherlands

British Journal of Management
Volume 15 Issue 2 Page 191 - June 2004
doi:10.1111/j.1467-8551.2004.00414.x

Financial Crises (And What to Do About Them), Eichengreen, B. (2002). Oxford University Press, pp. 194, ISBN 0-19-925743-4

Capital Flows and Crises, Eichengreen, B. (2003). The MIT Press, pp. 377, ISBN 0-262-05067-6

Coping with Crisis: International Financial Institutions in the Interwar PeriodKasuya, M. (ed.), (2003). Oxford University Press, pp. xv+235, ISBN 0-19-925931-3

Why Stock Markets Crash: Critical Events in Complex Financial Systems, Sornette, D. (2003). Princeton University Press, pp. xx+421, ISBN 0-691-09630-9

In the wake of September 11th, the crisis field has gained a great deal of relevance in both academic and practitioner circles. Suddenly, policymakers and managers have become interested in crisis research findings, funding is forthcoming, and academics of many a feather are flocking to the scene. Crisis used to be a 'sexy' topic, but it is now red hot.

The crisis field is marked by ill-defined boundaries. It is made up of specialized academics drawn from many disciplines (i.e. disaster sociology, public administration, political science, international relations and management). They tend to define crisis in terms of some basic threat to the core values of a system, necessitating urgent response under conditions of severe uncertainty. It is this catch-all character of the crisis definition that allows for communication between these academics and makes for what I here refer to as a 'generic' crisis field.

So what interesting research findings has this field yielded? This is best discussed and evaluated on the basis of two crucial questions, which, incidentally, signal the societal relevance of this research field. The first question asks why a social system a firm, a town, a nation or a global network experiences a crisis. The second question asks why some systems manage to minimize the crisis impact where others suffer severe damages.

A general consensus is emerging in the crisis field with regard to these questions, and can be summarized in a handful of principles. The first principle, which can be considered the bottom line of this research consensus, holds that crises will always be with us. We may learn from previous out-of-the-box events and develop tailor-made coping repertoires only to discover that the nature of crisis is continuously changing. The implications are sobering: crisis prevention is a good idea, but it will never make us safe from new crises. Increased airport safety may be great, but it will not protect us from future terrorist attacks.

The second principle is deduced logically from the first. If crisis prevention is essentially impossible, organizational and societal resilience must be the proper way to prepare for and deal with crises. The idea of resilience, perhaps explained best by the late Aaron Wildavsky (1988) in his classic Searching for Safety, directs our energy toward the design of organizational structures that facilitate flexible and resourceful answers to unknown future problems. This translates into a formidable challenge. Whereas modern organizations are typically geared toward routine production effective and efficient this principle of crisis management demands inherent redundancy. However, this is not something stockholders, stakeholders or voters tend to reward.

The third principle draws our attention to the unintended effects of crisis management efforts. The crisis management capacity of policymakers and managers is, as the research makes abundantly clear, limited at best. Decision-making under conditions of crisis is marred by very difficult but pervasive pathologies, which tend to fuel rather than dampen the crisis. Guided by good intentions, crisis managers often discover that their efforts produce spiralling circles of distrust and long-term aftershocks that keep the crisis alive beyond expectation. Ambitious crisis management is discouraged for these reasons.

This evolving consensus on the crucial principles of crisis management is informed by detailed studies of many different actual crises such as natural disasters, riots, terrorist acts and factory explosions. One category remains conspicuously absent in most of these case-study banks: the category of financial crises. This 'oversight' is remarkable, to say the least. The financial crises of the 1930s caused untold hardship and worldwide depression. More recently, in 19971998, financial crises spread from Asia to Latin America, then to Russia, and finally to New York with the near failure of Long-Term Capital Management. In 2000, the stock market took a dive as the Internet Bubble finally gave way. These financial crises ate away at household savings and pension funds, directly attacking livelihood and welfare. Yet, the field of economic crisis studies remains largely untapped by researchers who consider themselves at home in the 'generic' crisis field.

The books under review here represent a small sample of what appears to be a blossoming field. The authors of these books clearly do not view themselves as crisis researchers and, quite unsurprisingly, do not show any awareness of the crisis studies and findings summarized above. Nevertheless, their work is of great relevance to self-described crisis researchers as it shows how design errors, non-linear interactions and tight couplings in our financial architecture produce 'normal' crises that cause untold damage. Moreover, it documents the effects of the many mechanisms now in place to manage financial crises.

But there is added value to these books. The collective findings drawn from these works question the emerging consensus described above. The authors believe that financial crises can be understood and suggest that prevention, therefore, may become a real possibility. They expect little from crisis management efforts.

In Financial Crises, Barry Eichengreen presents a wonderfully lucid introduction to the subject. He neatly summarizes the state of the art in financial crisis management and spells out his vision on the causes and management of these crises. His argument is deceptively simple, as the deeper-delving Capital Flows and Crises reveals. For the non-economists solely interested in the research findings, Eichengreen's primer will suffice. For those who wish to explore his underlying theoretical work, the collection of essays in the second book will provide good reading. In this review, we stick with the primer. Didier Sornette's Why Stock Markets Crash poses much more of a challenge, but the payback for the dedicated reader is handsome. Larded with intimidating mathematical exercises not quite optional, I am afraid and spiced with illuminating examples, this text will bring new insights to most readers. Both authors arrive at similar viewpoints, as we will see.

Crises must be viewed as 'an unavoidable concomitant of the operation of financial markets', Eichengreen writes in the introduction of Financial Crises (p. 4). Eichengreen and Sornette both view crises as 'normal' outcomes of systemic vulnerabilities. If too much money is invested in something (be it stocks or some country's economy), a correction will follow one way or another, sooner or later. If this correction turns into a major disruption, the system has assumed crisis proportions.

Eichengreen is particularly concerned with crises developing in low-income countries 'that are as yet scarcely on the international financial community's radar screen' (p. 6). These countries must borrow to finance their development, but the heavy loan load typically proves too much of a strain on their fragile economies. This may result in debt crises, currency crises, banking crises or 'the most debilitating strain of the disease' twin crises. All of them have the potential to cascade through the global financial system, invading 'developed' economies with the pent-up energy of a tsunami.

Eichengreen begins by prescribing the obvious medicine: developing countries must build stronger financial and political systems. Few will argue against the proposition that healthy fundamentals 'stronger economic, financial and political institutions' will help prevent crises. In addition, Eichengreen discusses a set of 'important steps that have been taken to reduce the frequency and severity of crises' (p. 7). In a discussion of prevention-oriented organizations and mechanisms (Chapter 2), Eichengreen shows how much is actually understood about the nature of financial crises. The key to successful crisis prevention is twofold: allowing markets to discipline investors and designing standards that bear on institution building in emerging markets.

Market discipline is the first line of defence against financial excesses and imbalances. Market participants will not lend to governments with unsustainable policies if they suffer the consequences. Lenders therefore need accurate information about the countries to which they lend. Alas, as information is asymmetric, moral hazard and adverse selection will continue to undermine the operation of the market.

Some progress has been made with the numerous transparency-related initiatives under way. But more information, Eichengreen helpfully informs us, is not always beneficial. In fact, additional information may sometimes hurt more than it helps. When relatively little is known about some country's financial system, as was the case in Brazil in 1999, the sudden release of accurate information on (in Brazil's case) foreign-exchange reserve flows may have a destabilizing effect as investors become aware of how serious the situation really is. The question is how emerging markets can reach the stage where transparency has the desired effect without passing through stages of sudden shock among foreign investors.

The international community has been working hard on establishing a second line of defence, by promulgating up to 70 codes and standards, which would allow for prudential supervision and regulation. Eichengreen has little patience with these efforts, because 'regulation does not work: the regulators are always one step behind' (p. 16). The very proliferation of standards undermines the credibility and effectiveness of the international effort. More importantly, Eichengreen argues that these standards are simply not suited for developing countries. International standards will not work in the absence of a complementary body of domestic law, which usually does not exist. The imposition of these standards may provoke a crisis in the process of institution building, which, in the long run, is even less productive.

These crisis prevention efforts will not be perfect and crises will therefore continue to happen. Crisis management, therefore, remains important. The International Monetary Fund (IMF), our global financial crisis manager, thus comes into play. In theory, what the IMF must do is very simple: 'only lend to countries where there exists a strong disposition toward reform and where political countries are strong enough for the government to commit to its implementation' (p. 63). Too many political forces operate on the IMF decision-making process, however, which leads to failed interventions that, in the end, are paid for by the citizens of the poor countries. Eichengreen is not alone in thinking that disappointingly little has been achieved in resolving financial crises.

If financial crisis management fails, as Eichengreen asserts, it has been that way since the inter-war depression years. This we learn from the edited collection of conference papers entitled Coping with Crisis. The chapter authors describe, in rather sweeping terms, how the various financial institutions in the five great nations of the interwar period reacted to the pressures exerted by the Great Depression. Unfortunately, the case studies are primarily descriptive, raising questions rather than answering them. The most important finding is that the financial institutions in and across the countries under study varied greatly in their coping efforts. But we do not learn why some institutions managed to stick to their traditional way of banking whereas others adopted fundamental reforms, successful or not. In the absence of some encompassing (crisis) theory, these essays mainly satisfy a historical interest. They provide additional cases to the great crises bank, without generating new perspectives or insights.

Sornette certainly offers a fresh perspective. In fact, his book Why Stock Markets Crash summarizes years of thinking about the causes of 'ruptures' in complex systems. In this line of thinking, popularized by members of the Santa Fe Institute such as Stuart Kauffman (1995), 'extreme events' spontaneously emerge as a natural result of the repeated interactions among the constituents of a given system. 'The outstanding scientific question' in Sornette's mind is how crises evolve from a series of routine interactions 'on the smallest and increasingly larger scales' (p. 19). The answer to this question may shift our understanding of crises, not only in the realm of stock markets and international financial systems but in other complex systems as well. Lest we get ahead of ourselves, let us first see how Sornette explains stock-market crashes.

The true cause of a stock market crash is systemic instability. The market enters an unstable phase, which is marked by an accelerating ascent of the market price. This phase is commonly known as 'the bubble'. Speculative bubbles flow from positive feedbacks such as imitative behavior and 'herding' between investors. All investors understand this phenomenon (or will learn the hard way).

Sornette then applies the insights from complex systems theory and focuses on the critical point the tipping point at which price acceleration suddenly reverses into a steep decline. Much of Sornette's book aims to show the reader that the process leading up to this critical point has a 'robust signature', which opens up the possibility of foreseeing these critical points. Sornette then revisits all well-known crashes to demonstrate that these 'precursory patterns' have been documented in essentially all of them.

What does a precursory pattern or 'fingerprint' look like? Keep an eye out for 'an overall super-exponential power-law acceleration in the price decorated by log-periodic precursors' (p. 24). This is, believe it or not, less complicated than it seems. The exponential power law describes the rapid acceleration toward the critical point as a steeply inclining line (we are still in the realm of basic math). The real news comes with the oscillations that 'decorate' this steeply inclining line 'with frequencies accelerating as the critical time is approached' (p. 172). It is this 'complimentary signature of impending criticality' that is most fascinating, as it harbours predictions about the emergent critical point at which a small trigger can burst the bubble.

Clearly, this predictive power is not available to just anyone. A minimum requirement is a thorough understanding of mathematics, which, if properly applied, may help to 'read' the signs of impending catastrophe. Sornette does not let the reader in on all his secrets, but he does seem to prove that the model works both in retrospective explanations of past crashes across the world as well as in a predictive sense (Sornette reports to have actually profited financially from his model).

Sornette presents us with powerful material. He shows how normal behaviour of seemingly rational actors can produce critical events. Moreover, the road toward the critical point is 'decorated' by subtle warnings, which can be read by the trained observer (Sornette does not tell us what will happen if everybody could read these signs). Understanding complex systems in terms of self-organization has clear implications for crisis management as well. The debate, as Eichengreen summarizes it, is whether markets can stabilize by themselves or require a little help. Sornette explains how the theory of self-organizing complex systems prescribes a let-it-burn approach, as complex non-linear systems tend to find the optimal solution (p. 316). Eichengreen is likely to sympathize with this prescription.

The most intriguing question, however, asks whether this theory fashionable as it has become in many domains of science can be applied to other impending critical events. Sornette certainly thinks this is possible. Even though he cautions against 'jumping into the prediction game' (p. 321), he easily succumbs to the 'this-theory-explains-everything' temptation as he enthusiastically points out similarities in evolutionary discourse (p. 216). He does make clear, and convincingly so, that 'regions of decreasing uncertainty' exist in chaotic and apparently random systems (p. 324). If we can learn how to find and map these regions, an entirely new perspective on crisis causation and possible prevention may unfold.

Barry Turner (1978) and Charles Perrow (1984) arguably the best writers on organizations and crisis have already laid the foundations for this perspective, as they focused on the understanding of escalation and feedback loops that prevent organizations from seeing disasters coming and render useless their efforts to deal with these events. Both conceptualized catastrophe as the outcome of escalation, a process driven by endogenous forces. Turner pointed to the rational design of modern organizations, which helps augment small errors into big disasters. The obsession with efficient production blinds the organization from seeing the nearing disaster. Perrow treated critical events as the 'normal' outcome of highly complex and tightly coupled organizational structures, which allow seemingly meaningless glitches to cascade into life-threatening ruptures. All this resonates in the most powerful work on collective behaviour, which identifies rapid escalation of standard behaviour as the key to understanding riots.

Eichengreen describes financial crises in terms of declining trust spirals. Consider, for instance, the question of a proper exchange-rate regime. In Eichengreen's view, soft pegs are the problem and require a continuous reconfirmation of commitment. Any doubt with regard to the commitment to uphold the peg invites massive speculation, which, in turn, tends to solicit 'unconditional support' on the part of monetary authorities. In the end, market forces override this unconditional support, more or less destroying the credibility of these authorities. The inevitable recession is then fed by the loss of investor confidence in these same authorities.

Sornette, in short, is on to something. More precisely, Sornette has demonstrated that crisis researchers can make use of seemingly esoteric complex system theory. Some work has been done in applying these insights to patterns of crisis mitigation (Comfort, 1999). Sornette makes a case that this theoretical perspective could be explored fruitfully to jump-start the stalled search for crisis causes. Taken together, these books on financial crises provide an eye-opener to a crisis field that in itself sometimes resembles a self-organizing system. As crisis researchers increasingly invest in consultancy pay-offs at the expense of theoretical research, the field may be approaching its own critical point. These works remind crisis researchers that much work remains to be done. To paraphrase Sornette's final sentence in his book: Only when we understand the origin of crisis can we be prepared and prepare others for subtle but significant precursors!

Arjen Boin References Go to:

Comfort, L. (1999). Shared Risk: Complex Systems in Seismic Response. Pergamon, New York.

Kauffman, S. (1995). At Home in the Universe: The Search for Laws of Self-Organization and Complexity. Oxford University Press, New York.

Perrow, C. (1984). Normal Accidents: Living With High-Risk technology. Basic Books, New York.

Turner, B. A. (1978). Man-Made Disasters. Wykeham Publications, London.

Wildavsky, A. (1988). Searching for Safety. Transaction Books, New Brunswick.