Currency crises, speculative attacks and financial instability in a
global world: a Post Keynesian approach with reference to
*Brazilian currency crisis
Antonio J. ALVES, Jr.
Associate Professor of Economics at the Rural Federal University of Rio de Janeiro.
E-mail: antoniojbr@yahoo.com.br
Fernando FERRARI-FILHO
Visiting Research Fellow at the University of Oxford‟s Centre for Brazilian Studies and
Professor of Economics at the Federal University of Rio Grande do Sul.
E-mail: ferrari@ufrgs.br
Luiz Fernando de PAULA
Visiting Research Fellow at the University of Oxford‟s Centre for Brazilian Studies and
Associate Professor of Economics at the State University of Rio de Janeiro.
E-mail: lfpaula@alternex.com.br
Abstract
This paper develops a critical view of the conventional currency crisis models and presents a Post Keynesian view on financial instability and speculative attack. It also analyzes the 1998/1999 Brazilian currency crisis, according to the Post Keynesian approach.
Key words: Currency crisis models; Keynesian and Post Keynesian theories; Brazilian
currency crisis.
1. Introduction
In a recent article, Krugman (1997) reviews currency crisis models, dividing them in two types: “canonical” crisis models and second-generation crisis models. While the
former explains different experiences of speculative attacks, the latter seems appropriate for understanding the European monetary crisis in 1992/1993. Subsequently, recognizing the
* Paper published in Revista Venezolana de Análises de Coyntura, vol. X, n. 1, pp. 173-200, Jan.-June 2004.
2
failure of conventional theory in providing consistent answers for the East Asian crisis, Krugman (1998) develops a new approach on currency crisis in order to explain this specific crisis, based on the moral hazard/asset bubble view.
As stated by Krugman (Ibid.:6), conventional theory presumes that foreign exchange markets are efficient, that is to say they make the best use of available information, while, in the real world, foreign exchange market exhibits strong “anomalies”. Efficient market theory claims that economic agents analyze past and present market data, which means that price signals are presumed to provide enough information about forming rational expectations as a basis for making utility maximizing decisions. All relevant information about “economic fundamentals” regarding the future currently exists and is
available to the agents that are market participants and this information is embodied in the current market price signals. However, the author recognizes that foreign exchange markets
1can be inefficient.
Thus, currency crises can be generated either by self-fulfilling rational expectations or by irrational herding behavior involving bandwagon effects. But, even in models with self-fulfilling features it is only when fundamentals are sufficiently weak that a country is potentially vulnerable to speculative attack. In other words, currency crises are explained, even in last resort, by the inconsistency of economic fundamentals.
As is well known, efficient market theory has its foundation in the ergodic axiom, which means that the expected value of an objective probability can be always estimated from observed data that provides reliable information about the conditional probability function that will govern future outcomes. In this system, the decision maker believes that an immutable real objective probability distribution governs both current and future market outcomes. Therefore, market fundamentals are immutable in the sense that they cannot be changed by human actions; they also determine the conditional probabilities of future outcomes. According to the efficient market theory, short-run speculation can interfere with the efficient capital allocation function of financial markets and speculative volatility is explained by the existence of foolish “noise traders”. Otherwise, the observed secular trend
of financial markets is determined by immutable real sector fundamentals, which means that in the long-run irrational traders are made extinct by an efficient market (Davidson,
1 See, also, Stiglitz (1989) and Shleifer & Summers (1990) for a New Keynesian approach.
3
1998).
By contrast, Keynes and Post Keynesians reject the ergodic axiom of efficient market theory to explain the financial market behavior. In an uncertain world, in which fundamentals do not provide a reliable guide to the future, which is subject to sudden and violent changes, future market valuations are always subject to disappointments. Thus, speculation is not an “anomaly”, explained by the existence of foolish “noise traders”; on the contrary, it is a consequence of the operational way in which financial markets work in the real world. For Keynes and Post Keynesians, the outcome of speculation is ambiguous, because it can be disruptive with real consequences, devastating particular sectors as well as whole economies, once it can create speculative whirlpools; but at the same time it provides liquidity assets, an essential role of the financial markets.
This paper aims at analyzing the currency crisis models from a Post Keynesian
2perspective and explaining the 1998/1999 Brazilian currency crisis, according to the Post Keynesian theoretical framework. Besides this introduction, the paper is structured as follows: Section 2 describes the main currency crisis models as well as develops a critical view of the conventional theory. Section 3 presents a Post Keynesian view of financial markets and speculative activity and also develops a Post Keynesian perspective on financial globalization and speculative attack. Section 4 realizes an analysis on 1998/1999 Brazilian currency crisis. Finally, some concluding comments are presented in Section 5.
2. Currency crises models: the conventional theory
32.1. Speculative attack and currency crisis: the conventional theory
Currency crisis can be defined as “a sort of circular logic – in which investors flee a
currency because they expect it to be devalued, and much (though usually not all) of the pressure on the currency comes precisely because of this investor lack of confidence” (Krugman, 1997:1). On the other hand, a speculative attack on government‟s reserves “can
2 While we intend to contrast the differences between mainstream theory and Post Keynesian approach, Andrade & Silva (1998) explore the convergence between these different views on currency crises.
3 This section is based mainly on Krugman‟s ideas.
4
be viewed as a process by which investors change the composition of their portfolios, reducing the proportion of domestic currency and raising the proportion of foreign currency. This change in composition is then justified by a change in relative yields, for when the government is no longer able to defend the exchange rate [that is, a currency depreciation begins]” (Krugman, 1995:2).
Currency crisis and speculative attack are used almost as synonymous, but in reality a speculative attack on government‟s reserves may or may not result in a currency crisis. It depends on the ability or the will of the government to defend the domestic currency. In this context, a currency crisis happens when the government cannot (perhaps, it does not want) to support the exchange rate.
According to Krugman (1997), in general, conventional currency crisis theory can be divided in two types of models: “canonical” crisis models and “second-generation”
4crisis models. In a more recent work, Krugman (1998) added another view on currency crises for the understanding of the East Asian crisis – i.e., the “third-generation” model –,
which can be more properly referred as financial crisis models.
The “canonical” crises are the first generation crisis models (Krugman, 1979, and Flood & Garber, 1984) and they are derived from the models applied to commodity boards trying to stabilize commodity prices, known as “hotelling model”. The logic of the currency crisis in “canonical” models is the following: at the point in which speculators are supposed to wait until the reserves exhausted in the natural course of events, they would know that the price of foreign exchange rate, fixed up to now, will begin rising. In this situation, people would hold foreign currency instead of holding domestic currency, leading to a jump in the exchange rate, and by doing so advance the date of the exhaustion of reserves. When reserves fall to some critical level there would be an abrupt speculative attack that would quickly drive those reserves near to zero and, as a result, force an abandonment of the fixed exchange rate.
But what explains such crisis? According to this model, such crisis results of a fundamental inconsistency between domestic policies – typically the existence of money-
4 Besides these ones, it is important to say that there is what is known as contagious crisis; that is, a
phenomenon in which a currency crisis in one country seems to trigger crisis in other countries. A contagious
crisis can involve real linkages between countries (a currency crisis in country A can worsen the fundamentals of country B, or vice-versa) or not (as in the case of Mexico/Argentina, for instance), but the countries are perceived as a group with some common, even imperfectly, observed characteristics.
5
financial budget deficits – and the attempt to maintain a fixed exchange rate, once the
government is assumed to use a limited stock of reserves to peg its exchange rate. As this policy reveals to be unsustainable, the attempt of investors to anticipate the inevitable collapse would generate a speculative attack on the currency when reserves had fallen to some critical level. The main criticism on this model is that it represents government policy
5in a mechanical way, once the role of central bank in the model is passive.
The “second-generation” crisis models (Obstfeld, 1994) are more sophisticated than
“canonical” crisis models and the government policy in these models is less mechanical. In these models, government chooses to defend or not a pegged exchange rate by making a trade-off between short-run macroeconomic flexibility and long-term credibility.
The government must have a reason why it would like to abandon its fixed exchange rate or to defend it. Besides, the cost of defending a fixed exchange rate must itself increase when people expect that the exchange rate might be abandoned. In general, the main reason to allow currency depreciation in a country can be related to an increasing in unemployment due to downward rigid nominal wage rate, while a specific motive to fix the exchange rate can be related to the possibility of facilitating international trade and investment. According to these models, a fixed exchange rate will be costly to defend due to the fact that people, in the past, expected it would be depreciated at any time and/or because economic agents now expect it will be depreciated in the future. Thus, the logic of a crisis arises from the fact that defending a parity is more expensive (i.e., it requires higher
interest rates) if the market believes that defense will ultimately fail.
If a country‟s trade-off between the cost of maintaining the current parity and the costs of abandoning the fixed exchange rate is predictable, at some future date the country would be likely to devaluate its currency even in the absence of a speculative attack. In this case the speculators would try to get out of the currency ahead of that devaluation, but in doing so they would worsen the government‟s trade-off, probably leading to an earlier
devaluation. The end of the story can be a crisis that ends the fixed exchange rate regime before the fundamentals appear to make devaluation necessary.
Summing up, currency crisis may result from a conflict between domestic
5 For example, central bank does not make use of a variety of tools other than exchange market intervention to defend the exchange rate, as its ability to tighten domestic monetary policies.
6
objectives and the currency peg, which can make an eventual collapse of the currency peg inevitable. According to this approach, a speculative attack on a currency can also develop as a consequence of a predictable future deterioration in economic fundamentals, or purely through self-fulfilling prophecy, caused by a self-confirming pessimism, a case in which a country would suffer an “unnecessary” crisis. But even in the second-generation crisis, a
currency crisis is essentially the result of inconsistent policies with the long-run maintenance of a fixed exchange rate. In other words, it is only when fundamentals – such
as foreign exchange reserves, the government fiscal position and the political commitment of the government to exchange regime – are sufficiently weak that the country is potentially
vulnerable to speculative attack.
If a predictable crisis can happen before the fundamentals have reached the point where the exchange rate would have collapsed, then it can be provoked by a speculative attack not justified by current fundamentals. But, what prevents them? According to Krugman (1997), microeconomic frictions – such as transaction costs, the difficulty of
arranging credit lines, and so on – may prevent a subjectively low probability crisis from
ballooning into a full-fledged speculative attack.
Krugman (1998:1-2) now recognizes that “in order to make sense of what happened to [the 1997 East Asian crisis], it is necessary to adopt an approach quite different from that of traditional currency crisis theory”, as in the East Asia the “currency crises were only part of a broader financial crisis, which had very little to do with currencies or even monetary issues per se”. Thus, he develops a new approach on currency crisis – that is, “third-
6generation” crises models – in order to explain this crisis based on the moral hazard/asset bubble view.
According to the “third-generation” crisis models, currency crisis is viewed as an
integrated part of a general crisis of the economy, in which currency crises are pre-announced by financial crises. The logic of the analysis is that capital inflows increase the lending capacity of the banking system. So, the certainty of “bailout” of the financial
institutions by the monetary authorities explains “bad lending” practices used by the banks. Finally, a growing financial fragility path is followed leading the way to speculative crises,
6 The “third-generation” approach was also developed in Calvo & Mendoza‟s (1996) analysis of the Mexican
peso crisis.
7
once the increase in the money supply validates the bank run and, as a result, the economy begins to lose reserves.
In the case of the Asian crisis none of the fundamentals that drive “first generation” crisis models seem to have been observed in any of the afflicted Asian economies, and there did not seem to have any incentive to abandon the fixed exchange rate to pursue a more expansive monetary policy (as in the case of the 1992/1993 European monetary crisis). In other words, “Asian crisis is best seen not as a problem brought on by fiscal
deficits, as in „first-generation‟ models, nor as one brought on by macroeconomic
temptation, as in „second-generation‟ models, but as one brought on by financial excess and then financial collapse (...) The Asian story is really about a bubble in and subsequent collapse of asset values in general with the currency crisis more a symptom than a cause of this underlying (in both senses of the word) malady” (Ibid.:3).
In the East Asia, a boom-bust cycle created by financial excess preceded the currency crisis because the financial crisis was the real driver of the whole process. According to the moral hazard/asset bubble view, “the problem with financial intermediaries – institutions whose liabilities were perceived as having an implicit government guarantee, but were essentially unregulated and therefore subject to serve moral hazard problems [and] the excessive risky lending of these institutions created inflation – not of goods but of asset prices” (Ibid.:3), resulting in overpricing of assets.
However, the bubble of prices caused a deflation of assets and a deterioration in banking credits. Once a bank crisis has just been burst, the running against domestic currency was the natural consequence of the financial panic.
2.2. Some critiques on conventional view of speculative activity and currency crises
As stated before, according to the efficient market theory, agents with rational expectations make the best use of the available information, so that stock prices always reflect fundamental values. The social function of financial markets is to correctly allocate capital among enterprises in accordance with reliable information about future rates of returns determined by fundamentals.
In this sense, how can someone explain the speculative activity into this theory?
8
Speculation is the activity of buying (selling) and reselling (rebuying) assets in order to anticipate market value and making money by exploring “delays” in adjusting market prices to new economic fundamentals. Therefore, mainstream cannot explain why there is speculation without ad hoc assumptions. Speculators can only survive if there are (i)
informational problems, and (ii) waves of irrationality, which are attributes of “delay” markets. The problem in adopting ad hoc assumptions is that they provide excessive
freedom to the formulation of the models, generating, in this case, a contradiction inside the general equilibrium model framework.
Stiglitz (1989), for example, points out that short-term traders only include “noise
traders” – investors who believe that know more than the market and therefore do not have to acquire the correct information regarding future outcomes from the fundamentals. Of course, these phenomena can only occur in the short-run due to the fact that rational people conduct market towards long-run trend. Therefore, in spite of speculation, the economy will go towards the long-run equilibrium. Otherwise, is spite of short-term effects, speculation “affects how the pie is divided, but does not affect the size of the pie” (Ibid.:103). As it appears, the mainstream categorically supports that there is a kind of speculation neutrality
axiom, since, at least in the long-run, the size of the pie is determined by fundamentals. In other words, there are no long-run real effects if we assume (ad hoc) short-term speculation.
As conventional theory presumes that foreign exchange markets are efficient, according to the currency crisis models, speculative attack only occurs if there exists any sort of real “market fundamentals”, in general associated to a current or predictable future deterioration in economic fundamentals: an inconsistency or a conflict between domestic policies and exchange rate policy. These models also describe currency crises that are not driven by fundamentals, generated by self-fulfilling rational expectations or by irrational herding behavior – the case in which a wave of selling, whatever its initial case, could be magnified through sheer imitation or turn. The point is that they have difficulty in finding consistent explanations for currency crises that are not driven by fundamentals.
In an ergodic world, in which market fundamentals determine the conditional probabilities of future outcomes, speculative activity in foreign exchange markets is explained by the actions of foolish “noise traders”. Krugman (1997), for instance, utilizes
9
microeconomic fundamentals that make market inefficiencies to explain the existence of “herding” or the possibility of self-fulfilling crisis – as investors with access to private
information, creating asymmetric information in exchange rate market, or investment funds being managed by professional agents rather than directly by principals, that can result in investing money in crisis-prone countries.
Besides utilizing ad hoc microeconomic fundamentals for the explanation of
“irrational” crisis, conventional theory is always trying to find an ex post explanation for
each “new” currency crisis, as in the cases of European monetary crisis in 1992/1993, Mexican peso crisis in 1994/1995, East Asian crisis in 1997, Russian crisis in 1998, and Brazilian crisis in 1998/1999. For each new crisis, a new and in general more sophisticated model is developed, an evidence that speculative activity in foreign exchange markets is difficult to model as Krugman (Ibid.) recognizes.
3. A Post Keynesian approach on financial instability and speculative attack in an
7uncertain world
3.1. Financial markets and speculative activity in a nonergodic world
Keynes and Post Keynesians reject the classical ergodic axiom of efficient market theory to explain the financial market behavior. This is so since in an uncertain world future market valuations are always uncertain because the future is subject to sudden and
8violent changes and fundamentals do not provide a reliable guide to the future. In such a
world, speculation is not an anomaly but it results from the operational way in which financial markets work!
In different works, Keynes separated uncertainty from probable events, especially in relation to decisions involving the accumulation of wealth and the possession of liquidity. By uncertainty, he meant that “human decisions affecting the future, whether personal or political or economic, cannot depend on strict mathematical expectation, since the basis for making such calculations does not exist” (Keynes, 1964:162-3), which means “there is no
7 Sections 3.1 and 3.2 are an adaptation of Alves, Jr. et alli (1999/2000).
8 See, particularly, Keynes (1964, Chapter 12) and Davidson (1997, 1998).
10
scientific basis on which to form any calculable probability whatever. We simply do not know” (Keynes, 1973:114).
Therefore, Keynes rejected the belief that some observed economic phenomena are the outcome of any stochastic process, because for some occurrences, agents do not possess adequate information to construct useful future probabilities. The future is not calculable nor is the statistical reflection of the past, since, as Davidson (1994:89) points out, “the decision maker believes that during the lapse of calendar time between the moment of choice and the date(s) of payoff, unforeseeable changes can occur. In other words, the decision maker believes that reliable information regarding future prospects does not exist today”.
It is because uncertainty exists that future market valuations are neither predictable nor calculable by probability. Economic agents in financial markets have heterogeneous expectations, once one can never expect whatever data sets exist today to provide a reliable guide to future outcomes. In this sense, expectations that drive spot financial market are not rational, because the conventional valuation based on psychological forecasting of the market cannot be statistically reliable. Therefore, financial markets cannot be presumed to be efficient in the sense stated by efficient market theory (Davidson, 1998).
In the Post Keynesian view, the axiom of money neutrality does not work, because in a world under incalculable uncertainty, money – as the object that liquidates contractual
commitments denominated in the money account – can be held as a safety asset in moments
9of greater uncertainty by its characteristic of transporting purchasing power over time. So
liquidity preference can grow if entrepreneurs and speculators have contractual obligations and there is some degradation in the state of confidence. As the state of confidence is subjective, there is room for diversity of opinions about the future. And, if there is diversity of opinions and organized markets designed to give liquidity to assets, then there will be several opportunities for speculative activities to emerge.
According to Kaldor (1980), speculation is the act of purchasing an asset with the intention of reselling it latter, at higher prices, in the expectation of favorable changes taking place in the concerned markets. The role of the speculator is essential in these markets, because he/she can take the risks of acting against the market in anticipation of
9 This idea is clearly developed in Keynes (1964, Chapter 17) and Davidson (1994, Chapter 6).