论文部分内容阅读
The main goal of this paper is to trace the long record of financial crises from the perspective of an emerging economy. Two questions are addressed. First, what explains the incidence and severity of financial crises in an emerging market economy? And, second, what is the role of learning; how does the country learn from its past experience in financial crises to improve institutions and develop better techniques so as to successfully manage successive crisis events? To the best of our knowledge, this is the first attempt to provide a crisis event taxonomy looking at a systematic categorisation of the crises episodes that the country experienced over its 200-year life span, from its independence and the foundation of the Modern Greek state in 1829 to the recent 2008 crisis. To answer the above questions, I first present evidence on financial crises in Greece over a long time span of two centuries. Greece is chosen as our working template since it is a country with a rich history in financial crises. In particular, we try to identify different varieties of crises events, providing thus a crisis chronology. Moreover, we present some stylised facts on the incidence, the frequency and the severity of crises events. And second, we discuss the key determinants of the crises episodes, closely related to country specific factors, such as credit expansion, fiscal imbalances and the limited reserve coverage of the monetary base.
JEL Classification: E5, N2
Keywords: financial crises, emerging economies, sudden stops, country specific factors.
Acknowledgments: I would like to acknowledge helpful comments by a referee of this Journal. This paper was presented at XXX Conference of the Portuguese Economic and Social History Association on Economic and Social Crises, Lisbon, 19-20 November 2010. I am grateful to Nuno Valério, Rita Martins de Sousa, Amélia Branco and Rui Pedro Esteves for their useful comments. An extended version entitled financial crises and financial market regulation: the historical record of an emerger was presented at the 5th SEEMHN Annual Conference on Monetary policy during economic crises: A comparative and historical perspective, hosted by the Central Bank of the Republic of Turkey in Istanbul on 15-16 April 2010. I would like to thank the conference participants and especially George Chouliarakis, Kalina Dimitrova, evket Pamuk, Milan Sojic and Ali Cokun Tuncer for their comments and suggestions. An earlier version of the first part of this paper was presented at the workshop on Financial market regulation in the wake of financial crises: The historical experience, organized by the Banca d’Italia on 16-17 April 2009 in Rome. I would like to acknowledge helpful comments by the workshop’s participants and especially Alfredo Gigliobianco and Claire Giordano. Finally, I am most grateful to Luis Catao for kindly providing Stone’s data on foreign capital inflows to Greece. The views expressed here are those of the author and do not necessarily reflect those of the Bank of Greece. Any errors remain my responsibility.‘We have been here before…This time is (not) different. It almost never is’ (Reinhart & Rogoff, 2009).
Introduction
In the present era of globalization, financial crises are a recurring phenomenon in the history of emerging market economies. The literature is rich in evidence, both narrative and empirical, on external shock-driven sudden stops of capital inflows, current account reversals, currency drops, a rise in sovereign risk spreads, output losses, and severe financial turbulence (Calvo, Izquierdo, & Mejía, 2004; Calvo, Izquierdo & Talvi, 2005, 2006; Calvo & Talvi, 2005; Krugman, 2009). The pattern has been common to all emerging economies during recent years. A rapid and sudden cut-off in capital inflows is often accompanied by currency, debt, and banking crises, while the impact of the crisis on the real economy and its incidence differ markedly across countries, depending on the special features of the affected country’s economic structure. Prominent examples of this analysis are the 1998 Russian default lending crisis, the 1998 Brazil crisis, and the 2001 crises in Argentina and Turkey.
More recently, Reinhart and Rogoff (2009, 2010), covering a large number of mature and emerging countries over eight centuries, have provided a comprehensive look at the varieties of financial crises. They find a strong link between sovereign debt crises and banking crises across rich and poor countries alike. Researchers’ interest has now focused on the comparison of the emergers’ crisis experience across the two periods of globalization: the present era from 1980 to present, and the first era of ‘golden’ globalization from 1870 to 1914. They have concluded that the crisis pattern for the 1880s and the 1890s emergers was strikingly similar to the experience of today’s emerging market economies (Reinhart, 2010; Bordo, 2008; Bordo et al., 2001; Eichengreen & Lindert, 1989). This means that many driving forces that were present during the recent emerging market crises were also at work in the emergers’ crises a century ago.
In this paper, our main task is to trace the history of financial crises from the perspective of an emerging economy. Two questions are addressed. First, what explains the incidence and severity of financial crises in an emerging market economy? And second, what is the role of learning? In other words, how does the country learn from its past experience in financial crises to improve institutions and develop better techniques so as to successfully manage successive crisis events? Learning concerns both institutional learning and how to follow policies consistent with the aims of these institutions.
The case-study of my analysis is Greece, a country with a rich history in financial crises. Pre-WWII Greece was a typical example of a South-Eastern European (SEE) ‘emerging market economy’,1 as it was on the ‘periphery’ of the international monetary system.2 Most of the historical literature so far has concerned the experience of the advanced countries of Western Europe. Interest in studying the behaviour and response of emerging countries has appeared only recently. However, historical research has largely focused on the peripheral areas of Latin America and Asia and only partly on the emerging economies of SEE. Indeed, the history of financial crises in that particular region of the European periphery—which faced a more turbulent financial, economic, and political environment—is still unexplored or only partly studied. To the best of our knowledge, this paper is the first attempt to provide a crisis event taxonomy looking at a systematic categorisation of the crises episodes that the country experienced over its 200-year life span. Well till now, in all empirical studies Greece has been systematically neglected or included only occasionally and sporadically in their country samples and thus the Greek crisis experience is more or less terra incognita. Hence, tracing the past experience of an SEE emerger, like Greece, enriches our knowledge to answer the following questions. First, why do crisis events hit more frequently and more severely countries with backward and unsound economic and political institutions? Second, how does the absence of ‘country’ and ‘currency trust’ exaggerate the effects of a financial crisis on the real economy? And third, how do the emergers learn to prevent and contain crisis episodes?
To answer the above questions, I will first present evidence on financial crises in Greece over a long time span. In particular, I will try to identify different varieties of crisis events, providing thus a crisis chronology. Moreover, I will present a number of facts about the incidence, frequency, and severity of crisis events. To this end, I use the standard definition of financial crises. And, second, I will discuss the key determinants of crisis events, closely related to country-specific factors, such as monetary expansion, fiscal imbalances, and the limited reserve coverage of domestic money. I will conclude by presenting important lessons derived from the country’s historical experience. The main lesson of historical experience is that crisis events in Greece were largely credit driven. A comparison of the key characteristics and the sources of the two historical episodes of financial instability (‘then’ and ‘now’) reveals first that excess credit was the main crisis determinant in both instances, and, second, the absence of ‘country’ and ‘currency trust were the key aggravating factors of all crisis episodes. Spendthrift governments proved to be unable to tax sufficiently to cover their expenditures and to undertake a budget reform, i.e. to manufacture a politically difficult reform to eliminate fiscal excess.
Currency crises, banking crises and debt crises: defining the terms
Before revisiting the long record on financial crisis episodes from the perspective of an emerging market economy, it might be useful to set the theoretical framework by defining the terms used. Following Eichengreen and Portes (1987), the typology of financial crises can be thought as including three types of crises explained by different sets of variables: currency crises, banking crises, and debt crises (see also Goldstein, 2007; Bordo, 2006, 2008; Reinhart & Rogoff, 2009, Part I)3.
Currency crises are viewed as being caused either by weak economic fundamentals or certain government policy actions, self-fulfilling expectations of market participants, and possibilities of multiple equilibria, or they are viewed as a run on a currency or a financial panic (see, inter alia, Krugman, 1979, 1997, 1999; Obstfeld, 1996). Attention has been drawn to the importance of balance sheet effects for the sustainability of a currency target. Measures of real exchange rate overvaluation, external imbalances, foreign exchange reserves and export growth are explanatory variables for currency crises (Catao, 2006; Catao & Solomou, 2005). For identifying the period of a currency crisis event, in the empirical work that follows, we have based upon the following definition: a currency crisis is defined as a speculative attack on the exchange rate, resulting in a sharp devaluation or depreciation within a given year in the case of a flexible rate regime, or in a large outflow of the country’s foreign reserves, accompanied by a forced abandonment of an exchange rate commitment in the case of a fixed rate regime.
In the case of banking crises, attention should be paid to the special features of banks, such as maturity and currency transformation and asymmetric information. These features make banks vulnerable to runs and collapses following adverse shocks of either domestic or external origin. Theory and practice provide us with four typical sources of bank failures: imprudent credit policies, imprudent investment policies, poor asset-liability management, and bank panics.
Bank runs have been modelled as asymmetric information problems between depositors and banks, random manifestations of mass hysteria, or self-fulfilling expectations of depositors (Kindlerberger, 1978; Diamond & Dybrig, 1983). Hence, banking crises encompass both liquidity and solvency crises. A bank is illiquid when it is not able to meet short-term expected and unexpected obligations when they fall due. A bank is insolvent when its liabilities are greater than the value of its assets. As long as profits are sufficient to cover loan loss provisions, the level of bank capital and its capital adequacy ratio remain unchanged. However, when profits fall short, the amount of the bank’s capital declines4. To sum up, real economy disturbances may adversely affect the banks’ loan portfolio (credit risk); bubbles in equity and bond markets often collapse and destroy the value of the banks’ securities portfolio (market risk); and exchange rate disturbances induce interest rate changes which change the position and the slope of the yield curve(exchange rate and interest rate risk).
In my analysis, I identify banking crises as periods of severe difficulty, either because of a severe liquidity shortfall, and/or a reduction of the banks’ capital position. In twin crises, both currency and banking crises occur together. Severe currency volatility results in a sharp rise in the risk premium and the interest rates, thus making banks and corporations face increasing lending costs. Furthermore, currency mismatches bring to the surface serious weaknesses in the banks’ balance sheets5.
A debt crisis is a situation in which the country either has sizeable arrears of principal and/or interest on its obligations to its lenders, or in which there is debt restructuring or debt repudiation with commercial creditors(Detregiache & Schilimbergo, 2001; Manasse, Roubini & Schimmerlfenning, 2003). In other words, a debt crisis occurs when investors conclude that the debt ratio has become unsustainable. Thus, debt crises might be better explained by economic factors that measure the weight of domestic and foreign debt, the amount of debt service obligations, or the ability of a country to generate sufficient resources to meet these obligations. In emerging market economies, currency crises are closely linked to the probability of sovereign default, in the sense that a currency crisis increases the probability of fiscal distress. It will activate currency mismatches that result in a ‘balance sheet crisis’ both for the government and the private sector.
In general, a sovereign country’s negative decision on its debt obligations may have two forms: repudiation and rescheduling. ‘Debt repudiation is an outright cancellation of all current and future debt obligations by a borrower […] [while] debt rescheduling is the change of the contractual terms of a loan such as its maturity and interest payments’ (Saunders & Cornett, 2003, p398). Rescheduling has been the most common form of sovereign risk events in the period after World War II, whereas a large proportion of debt problems were met with repudiations before World War II (Eichengreen & Portes, 1987). This difference in behaviour is explained by the fact that before the war most international debt was in the form of foreign bonds, while after the war debt was in bank loans. However, in the present era of globalization, bond financing is again the main instrument of sovereign borrowing. This is clearly a similarity to the earlier era of ‘golden’globalization6. Finally, a third generation crisis or a ‘balance sheet crisis’ is defined as a twin crisis accompanied by a debt crisis.
Apparently, to analyse financial crises we need a ‘minimal structure’ around which historical observations can be placed (Eichengreen & Portes, 1987). This structure is thought to be made up of four elements. First, it consists of a chain in a financial crisis. The linkages run from exchange rate disturbances to debt defaults and bank failures, and eventually to the real macro-economy. Second, the propagation of the disturbance and the probability of its turning into a generalized crisis are influenced by the specific institutional character of the country’s financial system. How well do the banks manage risk? Are they supervised and regulated by an independent authority or not? Third, the government should be actively involved in creating a regulatory framework that insulates the banking system and the real economy from a shock. And fourth, there is a need to link monetary stability with financial stability. The exchange rate and monetary arrangements influence both the volatility of the exchange rates and the interest rates. In the context of the so called ‘balance sheet approach’to financial crises (Goldstein, 2007; Borio, 2004; Beim & Calomiris, 2001; Allen et al., 2002), a financial crisis can be considered a chain. It draws attention to the fact that a crisis can emerge from weaknesses in banks’balance sheets, vulnerabilities in corporate and household balance sheets, and problems in the government’s balance sheet (see Chart 1).7
Chart 1
Interpreting financial crises: the ‘balance sheet approach’
Source. Sheikh and Heijmans 2004.
The pattern of financial crises: the Greek story line
The pattern of financial crises in Greece is shown in Figures 1 and 2. Figure 1 shows the frequency of crises measured as percent in total, where we divide the number of crisis events that occurred in each sub-period. The data set is separated into five different sub-periods: the years prior to 1880, the gold standard period (1880-1913), the interwar period (1919-1939), the Bretton Woods period (1945-1974), and the most recent period (1975-2008). As can be seen, in the period before World War I the predominant form of the crisis were currency crises, followed by banking crises, although debt crises were the second most frequent type of the crisis before 1880. Before 1913, currency crises accounted for more than half of all crisis events. On the contrary, banking crises predominated in the interwar years; they account for more than two-thirds of the total events. The pattern of crises did not change notably after World War II. Currency collapses remained the most frequent crisis event, followed by banking crises, while no debt crisis event occurred during that period8,9.There is no evidence of a twin crisis, while a third-generation crisis occurred only once, in 1932.10
Figure 2 shows the frequency of crises measured as percent per year; the years in crisis are divided by the total number of years in each data sub-sample. The figure suggests that, with the exception of the interwar period, crises appear to be more frequent after World War II. Surprisingly, the crisis frequency of 36.7 percent during the Bretton Woods period is much greater when compared to 17.7 percent in the period after 1975, while it is cut only in half compared to the most unstable period of the interwar era (71.4 percent). Another interesting point to be stressed here is that the frequency of crises was much lower in the early era of globalization before 1913, when compared to the post-1975 era. Finally, as expected, all types of crisis show the greatest frequency during the interwar period, confirming that the interwar crisis event still remains the ‘mother’ of all subsequent historical crisis episodes.
It may be interesting to compare the Greek crisis experience in the two eras of ‘finance capitalism’, that is, pre-1913 and post-1975. Then, improvements in technologies of communication and transfer—for example, the transatlantic telegraph cable in the late 1860s, which was the only way to transmit information across oceans—reduced the cost of money and goods transformation, thereby contributing to greater economic integration. Both periods are very similar to each other in many respects, but very distinct too. Clear similarities are the use of bond financing as an instrument of sovereign borrowing and the high degree of world capital and goods market integration. However, a stark difference is the post-1975 strong increase in financial leverage, caused by the decoupling of money and credit aggregates. This means that after 1975 banks funded growth mainly through non-monetary liabilities, and thus their role in credit creation via bank loans is of great importance (Schularick & Taylor, 2009)11.
A closer look at the 1890s crisis episode and the current crisis event reveals that financial instability in Greece has been largely credit-driven. As Reinhart and Rogoff (2008) have argued, periods of financial distress have been associated with economic downturns and typically follow waves of domestic and international credit expansion. Figures 3 and 4 tell us the story. Figure 3 plots the ratio of bank private credit to GDP over the pre-World War I period; this ratio shows the importance of the main function of the banks, namely channelling savings to investors. The data refer to bank gross loans to firms and households gathered by the balance sheets of eight domestic banks with a long presence on the Greek money market. Based on a simple inspection of the data, two periods of soaring credit can be easily detected. The first one took place from the mid-1870s to the early 1890s, and the second one from 1898 until the outbreak of World War I.
Systematic attempts to industrialize the country can be dated to the last quarter of the nineteenth century. The economic policy of the governments placed emphasis on the development of the private sector and the introduction of new technology to the production process, with parallel attempts to free the economy from stifling state control. It was at this time that the first efforts were made to create infrastructure, which helped to modernize the country as well as to restructure the economy. During that period, agricultural production increased, industry was introduced, trade developed, credit grew rapidly, and the number of functioning banks increased. New banking products appeared, such as mortgage-backed loans, mortgage-backed credit lines, and long-term bank bonds. Rich Greek emigrants were attracted by the high deposit rates and kept their money within Greek banks12. Moreover, the country’s modernization and the implementation of many large public works attracted foreign private investment capital. Foreign and Greek businessmen heavily invested in residential and commercial properties, as well as in portfolio investments13. A lending boom soon occurred14, which made land and resource prices soar.
Figure 1. Crisis frequency (percent in total). Notes. A currency crisis is identified as a year when there was either a forced abandonment of an exchange rate commitment, an official depreciation, or an abnormally large devaluation of the value of the exchange rate in a given year. A debt crisis is identified as a year when government debt repudiation or restructuring occurred. A banking crisis is identified as a year when either a wide-spread deposit bank run occurred, banks went bankrupt, or when they asked for bail-out when the function of the lender of last resort was activated to confront a wholesale or a deposit bank run. Isolated events are not taken into account, but only the year when the event took place, excluding the years of oncoming or ongoing crises. See also Table A in the appendix.
Figure 2. Crisis frequency (percent per year). Note: See Figure 1.
Figure 3. Credit soaring in the first era of globalization, 1846-1914 (percent to GDP). Note. bank gross loans to firms and households (private bank credit). Computations are based on yearly averages, using a five-year rolling time span. They remain unchanged when we use a longer (ten-year) rolling window. Using a rolling time span helps to minimize the resulting loss of information, which is the principle drawback of averaging. End-of-year data. Source. See appendix.
Economic growth in the 1880s was based on cheap and easy foreign lending. After the country’s foreign debt compromise in 1878-1879, Greek government bonds were again traded on the London stock market, and the governments started to heavily borrow from abroad to cover excess spending, both consumption and investment. During that period, the country’s creditworthiness was rebuilt, since the governments considered the gold standard as a ‘good housekeeping seal of approval’ along the lines considered by Bordo and Rockoff(1996), in order to improve the country’s access to the foreign capital markets and attract cheap lending(Lazaretou, 2005). Ultimately, the government accumulated an enormous external debt burden with gold clauses that triggered its ability to repay it, and thus the currency was pressed by a heavy speculation attack. In 1895, the spot exchange rate almost doubled.
With the successful rescheduling of foreign debt in 1898, a second period of soaring credit was put in motion. This time the credit increase was strong and rapid: the ratio of bank loans to GDP grew by a factor of four, reaching a peak in 1910-1911, when the country credibly joined gold after a long, painful process of effective fiscal adjustment and monetary restraint. The key driver was the rapid output growth that the country generally experienced in the years prior to World War I. Export trade increased mainly after 1905, and the growth of domestic production accelerated. The shipping industry marked considerable progress, expanding its activities to the transit trade of third countries. Greater economic activity brought with it an increase in the number of commercial banks. The country’s international credit standing was rebuilt, resulting in foreign capital inflow.
In the current era of financial capitalism, soaring private and public credit has also been the key driving force. As seen in Figure 4, the ratio of private bank loans to GDP soared in the few years prior to the current crisis. Public debt also soared. While in the 1990s the mean rate of private credit stood at very low levels (36.6 percent), in the first half of the 2000s it quickly soared to 54.4 percent. Thereafter, it increased even more rapidly, approximating 84 percent in 2008 and 81 percent in 2009, becoming twice as high as in 2000. Similarly, general government debt as a percent to GDP increased from 107.4 in 2007 to 144.9 in 2010. Lending became very cheap and easy, just after the country’s entrance into the euro area, soon resulting in a‘liquidity overhang’ that triggered an asset price bubble and a consumption and investment boom. However, in the wake of the international financial crisis in 2008, a bust soon followed, resurfacing the country’s structural inefficiencies—such as excess public indebtedness, a sharp enlargement of both the current account deficit and fiscal deficit, and a heavy loss in the country’s international credit standing (Bank of Greece, 2010).
What went wrong? The economic and financial environment
With the help of some simple descriptive statistics and using the Greek case as a working template, in this section I will highlight the key driving forces of emerging market crises. My interest is focused on the pre-1913 golden era of globalization. This era was characterized globally as a period of a low frequency of crises events. During that period, crises were phenomena observable in emerging rather than advanced countries.15 This was because emerging market economies lacked a sound institutional framework and had not learned to follow successful policies within that framework. This framework includes sound fiscal and monetary institutions, as well as regulation and supervision authorities. This is what Caballero et al. (2004) have named ‘country’ and‘currency trust’.16 In particular, sound fiscal institutions mean the existence of an efficient tax system, the avoidance of excessive public debt exposure, and the credible commitment to balance the budget. Sound monetary institutions include the credible adherence to the classical gold standard, by holding sufficiently large gold reserves to minimize a currency mismatch between hard currency liabilities and domestic currency revenue.
Figure 4. Credit soaring in the current era of globalization, 1950-2009 (percent to GDP). Note: total bank credit, i.e. loans to firms and households. Yearly averages are calculated by using a five-year rolling window. End-of-year data. Source: See appendix.
The advanced countries of that time had developed a sound and solid institutional framework that allowed them to manage crisis episodes successfully. They had developed financial markets, which had largely overcome the problem of asymmetric information, and they had efficiently channelled savings to productive investment. They enacted balanced fiscal policies and non-accommodative monetary policy targeting the fixed rate. They operated within the framework of globalization with free trade, free capital, and labour movements. They had successfully established important institutions such as a legal system to protect property rights and constitutional democracy, which greatly supported financial growth (Bordo & Rousseau, 2006).
By contrast, emerging countries of that time were in the process of developing fiscal and monetary institutions. They faced a more turbulent financial environment and experienced long periods of political instability. This in turn meant that they were prone to the occurrence of crises and that they had not learned to deal with them. What were the ‘risk factors’ for them? In the era of laissez faire laissez passé, excess saving in the advanced world flowed towards the then poor peripheral countries17. The massive capital inflows boosted asset prices in the emerging countries, particularly of land and unexploited physical resources, and fed a credit and growth boom. Inflows were accompanied by high budget deficits and government debt accumulation. In conjunction with an accommodative monetary policy, inflows fuelled domestic inflation and fed speculative attacks on the currency. The real burden of debt (both private and public) soared, asset prices failed, and lending went bust. Ultimately, the government and the banks faced a balance sheet crisis.
Historical accounts
As we saw in the previous section, the Greek record of financial crises verifies the story outlined above. What was the economic and financial landscape that made an ‘emerger’ like Greece more vulnerable to crises18? First, pre-World War II Greece was a poor agricultural economy19; it financed its growth process through heavy foreign lending. Second, capital inflows were attracted by higher returns on land and other resources and led to lending booms. Booms were accompanied by fiscal expansion, financed by money creation and government debt accumulation. However, the interplay between fiscal imbalances and monetary disturbances resulted in frequent convertibility crises. The country responded to the speculative attacks on the currency by adopting, if only occasionally, the gold standard as ‘a good housekeeping seal of approval’. This enabled the country to continue attracting cheap foreign capital.
Third, the country was prone to a debt crisis when the domestic economy collapsed as a consequence of a lending bust. This was because the country’s fiscal and monetary institutions were extremely fragile. The governmental financial system was weak due to the inherent structural inefficiencies and the frequent external spending shocks, such as wartime emergencies. In particular, the tax system, based largely on indirect taxes and custom duties, was procyclical. Thus, the governments were unable to raise sufficient tax revenues to service the debt. Further, monetary policy was loose and accommodated fiscal policy20.
Fourth, the country suffered from ‘original sin’, as it has been put forth by Eichengreen and Hausmann(1999) and Eichengreen et al. (2003). That is, it was not able to borrow abroad or even at home in terms of its own currency; debt issue required gold or exchange rate clauses. This in turn meant that in the case of a nominal currency devaluation, the debt burden soared, increasing the likelihood of the government’s default on its outstanding debt, with certain repercussions adversely affecting the quality of the balance sheet both of the banks and the private sector. As a result, risk premia rose.
Fifth, the country had the experience of ‘sudden stops’. Sudden stops were a common feature in the economic history of the emergers (Catao, 2006, Bordo et al., 2010). Every time the economic and financial circumstances in the advanced lending countries changed and thereby led to a cut-off of cheap capital inflows to the emerging economies, the latter soon faced a balance of payments crisis and a debt crisis. This was exactly the case of a sudden stop event that Greece experienced in the early 1890s and again in the early 1930s.
Sixth, the country was financially less developed; financial depth measured as the ratio of bank assets to GDP and/or broad money to GDP exhibited a positive long-run trend, even though it was relatively weak and excessively volatile21. Throughout the whole sample period from 1846 to 1939, banking intermediation as measured by both indices stood at very low levels: it was on average 51 and 37.7 percent, respectively.
Empirical accounts
The data strongly support the historical accounts. Figure 5 shows the pattern of gross capital inflows to Greece based on Stone’s portfolio calls on London data from 1870 to 1913 (Stone, 1999). Due to lack of data, portfolio calls on London can be considered a proxy for foreign capital inflows to the country, since London was then the most important lending centre for emerging economies. As is evident, the time series seems to be stationary22, although it was extremely volatile. However, an upward trend was present from 1879 to 1890. Gross foreign capital inflows pointed to a peak in 1898, 1903, and again in 1910. Following the same working definition as in Catao (2006, p7), I define a sudden stop as a drop from peak to trough of no less than two standard deviations of the deviations of the respective series, from a linear trend and/or any drop that exceeds 3 percent of GDP over a period shorter than four years. Based on this definition, the figure clearly shows a significant rise in the second half of the 1880s and a sudden drop in the 1890s. This pattern is in accordance with the average pattern of capital inflows for a large panel of emerging countries during that period23. Precisely, three episodes of sudden stops can be detected and are denoted in the shaded areas: 1881-1883, 1890-1897, and 1911-191324. The second episode was associated with the 1893 debt repudiation.
Figure 5. Gross portfolio calls on London, 1870-1913 (million of pounds). Note. Sudden stops are in the shaded areas. Source. the data are from Stone (1999).
Figure 6 shows the pattern of sovereign bond spread over the period between 1870 and 1914. The bond spread is measured as the difference between the Greek bond yield and the British consol yield25. Although it shows a declining long-term trend reflecting a greater world-wide financial market integration, it sharply increased above its trend in the years of a sudden stop, especially in the years when a sudden stop was followed by a debt crisis, as was the case in the 1890s.
Figure 7 depicts the time series behavior of the nominal and the real exchange rate of the drachma against the British pound. For the larger part of that time, Greece did not adhere continuously, but only partially, to the classical gold standard. As shown, the country experienced high depreciating rates during sudden stop episodes. The most notable example is the episode of 1890-1897. The drop in capital inflows was accompanied by a sizeable depreciation in nominal terms between 1892 and 1895. The nominal value of the currency rate peaked in 1895, two years after the 1893 debt crisis. Depreciation was also sizeable in 1886, followed by the sudden stop event of 1881-1883. Following Catao (2006, p11), I define a currency crash as nominal exchange rate depreciation greater than at least one standard deviation of the annual percentage change of the spot exchange rate of the drachma against the British pound over the period 1885 to 1914. I also assume that this depreciation was not fully reversed within a three-year window. Based on this working assumption, I date four such events: 1886, 1892, 1893, and 1894.
Figure 6. Greek sovereign bond spread, 1870-1914 (percentage points). Note. Greek gold bond yield minus British consol yield; percent per annum. Source. See appendix.
Figure 8 depicts the evolution over time of the real GDP per capita. In the same figure, the incidence of crisis events is also indicated and the sudden stop episodes are shaded. One can notice that sudden stops and crisis events were associated with notable drops in the level of the real GDP per capita26.
Figure 9 shows output losses associated with each sudden stop event. Following Adelet and Eichengreen(2005) and Bordo (2006), the output loss is measured as the change between the average growth rate three years before the crisis and the average growth rate three years after the crisis. A three-year growth average is considered to serve as a proxy for the trend growth. As seen, the sudden stop episode of 1890-1897 was associated with a considerable output loss.
Country specific factors
So far, the inspection of the Greek historical data has indicated that drops in foreign capital inflows were associated with currency crises and debt crises. These fluctuations in capital flows were largely determined by external shocks, such as a change in the central bank rate of the lending countries. For example, the 1890-1897 crisis in Greece can be considered the outcome of an array of adverse shocks in the international economic and financial environment, which worried foreign creditors who, until the late 1880s, were generously supplying cheap loans to emerging economies without any guarantee and at low interest rates27. Specifically, the lending boom of the 1880s ended suddenly in 1890-1891, when the advanced lending countries stopped lending as economic conditions in advanced Europe gradually improved and investment opportunities reappeared. The core countries’ central banks responded by increasing their discount rates, which caused a massive slowdown in investment abroad.
Figure 7. The nominal and real exchange rate of the drachma against the British pound, 1885-1914. Notes. Cc = dates of the currency crises. Sudden stops are in shaded areas. Spot rates, yearly averages. LHS = real (1887-1914), RHS = nominal. The real exchange rate has been calculated as the ratio of British wholesale prices (Sauerbeck index) to prices for basic foodstuffs in Greece, using the bilateral nominal exchange rate as the conversion ratio. The food price index (1866-1877 = 100) has been constructed as a simple geometric average of the relative prices of five traded food products. Since data on quantities consumed are not available, a Laspeyres index cannot be calculated. Moreover, as Fisher (1927) and Mitchell (1938) have pointed out, the simple geometric average has the advantage of smoothing the time series of prices with regard to extreme values. This is very important in the case of food products, because their prices exhibit high volatility. Pre-1914 official data for a price index do not exist. The data for the Sauerbeck index have been taken from the Journal of Royal Statistical Society and Grilli and Kaminsky (1991). Source. See appendix.
Figure 8. Real GDP per capita, 1870-1914. Note: In 1990 international US dollars. Cc = dates of currency crises events; Bc = dates of a banking crisis; Dc = dates of a debt crisis; sudden stops are in the shaded areas. Source. See appendix.
Figure 9. Output loss during sudden stops. Note. Change in real output growth rates, percentage points. Output loss has been calculated as the change in the average growth rates three years before and after the sudden stop event.
However, external shocks affected emerging countries differently. Country-specific factors closely related to fiscal, monetary and financial behaviour were the key drivers of this effect. Figure 10 shows the trend of the fiscal and monetary policy indicators over time. One can observe that fiscal behaviour as proxied by the ratio of total government spending to total tax revenues was generally lax, noticeably cyclical without however exhibiting a long-run trend. Sudden stop episodes and currency crashes were preceded by periods of high expenditure, which were accompanied by rapid monetary expansion as shown by the ratio of M3 to GDP. The ratio doubled between 1870 and 1880 and tripled between 1885 and 1890. As a result, a concomitant sharp drop in the international reserve coverage of the domestic currency occurred, implying that the country’s monetary authorities were no longer able to adhere to the specie convertibility rule. Further, by regressing the cyclical component of real government expenditure on the cyclical component of real GDP28, bivariate regression estimates verify the well-known conclusion that in countries that experience currency crises fiscal policy is procyclical29. A procyclical response of fiscal policy to the business cycle means that during cyclical upswings fiscal policy exacerbates imbalances, which in turn causes nominal exchange rate depreciation and a drop in the reserve coverage of the monetary base.
Figure 10. Fiscal and monetary policy stance indicators, 1870-1914. Note. percent per annum. The ratio of spending to revenues is measured on the right hand axis. Source. See appendix.
Figure 11 plots the M3 multiplier—that is, the ratio of broad money (M3) to monetary base (M0). It corroborates the view that monetary expansion played a significant role in aggravating currency crises during sudden stop events. As with the ratio of broad money to GDP, domestic bank credit as proxied by the M3 multiplier is also cyclical throughout the period under study. From the turn of the century, however, it exhibited a strong upward trend, as the result of the country’s rapid financial development process that started in those years.
Figure 11. M3 multiplier, 1870-1914. Note. M3 to M0 ratio. End-of-year data.
To recapitulate, in the pre-1913 period Greece experienced sudden stops in foreign capital inflows, accompanied by currency and debt crises. Country-specific factors were the key driving forces of this effect. Indeed, as the data show, all crisis events were preceded by periods of fiscal laxity, rapid monetary expansion, and limited reserve coverage of domestic money. Moreover, probit estimates strongly support the above conclusion. The empirical estimates conclude that crisis predictor coefficients on m3/y, reserves/m0, and spending/reserves are shown to be statistically significant in spite of the small number of crisis events. 30
Concluding remarks
Several important lessons can be drawn from Greece’s long record of crisis experience. First, a comparison of the key characteristics and the sources of the two historical episodes of financial instability(‘then’ and ‘now’) reveal that excess credit was the main crisis determinant in both instances. Second, country-specific features such as fiscal laxity, credit expansion, and the limited reserve coverage of the domestic currency were the key aggravating factors of the earlier crisis. And third, data and historical accounts verify that the government’s response to a crisis event was weak, slow, and delayed. This means that the government officials learned only with serious difficulty and after a long delay how to manage successive crises and develop better techniques.
Recently, several scholars have raised the following question:31 Can strict regulation produce innovations that might in turn produce future crises? This is not the general rule. Certainly, there are many others risk factors. For example, riskiness might arise from global macroeconomic imbalances combined with inherent financial inefficiencies and weak institutional determinants that put to work destabilizing incentives.
A crucial issue is thus addressed: How can we strengthen the deep institutional determinants of the country’s financial development (such as the rule of law, property rights, corporate governance, and constitutional democracy) to avoid a new crisis? Therefore, an interesting topic for further research might be to identify the predicted probabilities of a crisis event. By enlarging the country sample to include all SEE countries, probit regression results might show if and to what extent country-specific features would be the most important predictor of crashes.
To recall Mark Twain, ‘…history does not repeat itself, but it does rhyme’.32 In other words, history has shown that, even though financial crises were accompanied by considerable output loss, they were always followed by a period of growth that lasted many years. However, as historical accounts demonstrate, crisis events will always occur, and it is very difficult to avoid repetition. On the other hand, policy-makers and governments often have a short historical memory. This means that their past attempts to strengthen macro-prudential supervision either failed or were proven to be ineffective in preventing similar situations in the near future.
References
Adalet, M., & Eichengreen, B. (2005). Current Account Reversals: Always a Problem. NBER Working Paper 11634.
Allen, M., Rosenberg, C., Keller, C., & Roubini. N. (2002). A Balance Sheet Approach to Financial Crises, IMF Working Paper 02/210.
Bank of Greece. (2009). Monetary Time Series of South-eastern Europe from the 1870s to 1914. SEEMHN and Bank of Greece Working Paper no 94, February.
Beim, D., & Calomiris,C. (2001). Emerging Financial Markets. London, McGraw Hill.
Bernanke, B. S., & M. Getler and S. Gilchrit. (1999). The Financial Accelerator in a Quantitative Business Cycle Framework. In J. B. Tayloer, & M. Woodford (Eds.), Handbook of Macroeconomics (Vol.1, pp.1341-1393). Elsevier.
Bordo, M. (2006). Sudden Stops, Financial Crises and Original Sin in Emerging Markets: Déjà vu. NBER Working Paper no 12393.
Bordo, M. D. (2008). Growing up to Financial Stability. SEEMHN Bank of Greece Working Paper no 85.
Bordo, M. D, & Meissner C. M. (2006a). The Role of Foreign Currency Debt in Financial Crises: 1880-1913 versus 1972-1997. Journal of Banking and Finance, 30, 3299-3329.
Bordo, M. D., & Meissner, C. M. (2006b). Financial Crises, 1880-1913: the Role of Foreign Currency Debt. In S. Edwards, G. Esquivel, & G. Marquez (eds.), Growth, Protection and Crises: Latin America from an Historical Perspective, Chicago, University of Chicago Press.
Bordo, M. D., & Rousseau, P.B. (2006). Legal-Political Factors and the Historical Evolution of the Finance-Growth Link. European Review of Economic History, 10, 421-444.
Bordo, M. D, & Rockoff, H. (1996). The Gold Standard as a Good Housekeeping Seal of Approval .Journal of Economic History, 56, 389-428.
Bordo, M. D., Cavallo, A. F., & Meissner, C. M. (2010). Sudden Stops: Determinants and Output Effects in the First Era of Globalisation 1880-1913. Journal of Development Economics, 91, 227-241.
Bordo, M. D., Eichengreen, B., Klingebiel, D., & Martinez-Peria, M.S. (2001). Is the Crisis Problem Growing more Severe?. Economic Policy, 32, 51-75.
Caballero, R., Cowan, K., & Kearns. J (2004), Fear of Sudden Stops: Lessons from Australia and Chile. NBER Working Paper 10519.
Calvo, G., Izquierdo, A., & Mejía, L. F. (2004). On the Empirics of Sudden Stops: the Relevance of Balance Sheet Effects. NBER Working Paper 10520.
Calvo, G., Izquierdo, A., & Talvi, E. (2005). Sudden Stops, the Real Exchange Rate and Fiscal Sustainability: Argentina’s Lessons. In V. Alexander, J. Melitz, & G. V. Fustenberg. (Eds.), Monetary Unions and Hard Pegs: Effects on Trade, Financial Development and Stability (pp.151-181). N.Y, Oxford University Press.
Calvo, G., Izquierdo, A., & Talvi, E. (2006). Phoenix Miracles in Emerging Markets: Recovery Without Credit from Systemic Financial Crises. NBER Working Paper 12101, March.
Calvo, G., Talvi, E. (2005). Sudden Stops, Financial Factors and Economic Collapses in Latin America: Lessons from Argentina and Chile. NBER Working Paper 11153.
Casu, B., Giradrone, C. & Molyneaux, P. (2006). Introduction to Banking. FT Prentice Hall.
Catao, L. (2006), ‘Sudden Stops and Currency Drops: A Historical Look’, IMF Working Paper 06/133.
Catao, L., & Solomou, S. N. (2005). Effective Exchange Rates and the Classical Gold Standard Adjustment. American Economic Review, 95, 1259-1275.
Della Paolera, G., & Taylor, A. (2001), Straining at the Anchor: The Argentina Currency Board and the Search for Macroeconomic Stability, 1880-1935. Chicago ,Chicago University Press.
Detragiache, E. , & Spilimbergo, A. (2001). Crises and Liquidity: Evidence and Interpretation. IMF Working Paper no WP01/2.
Diamond, D. W., & Dybvig, P. M. (1983).Bank Runs, Deposit Insurance and Liquidity.Journal of Political Economy, 91, 401-419.
Eichengreen, B. (1992). Golden Fetters, New York, Oxford University Press.
Eichengreen, B., & Hausmann, R. (1999). Exchange Rates and Financial Fragility. In Federal Reserve Bank of Kansas City (Ed.), New Challenges for Monetary Policy (pp.329-368).
Eichengreen, B., Hausmann, R., & Panizza, U. (2003).Currency Mismatches, Debt Intolerance and Original Sin: Why they are not the same and why it matters. NBER Working Paper 10036.
Eichengreen B., & Portes, R. (1987). The Anatomy of Financial Crises. In R. Portes, & A. Swoboda (Eds.), Threats to International Financial Stability (pp.10-66). CEPR and Cambridge University Press.
Fisher, I. (1927). The Making of Index Numbers. A Study of Their Varieties, Tests and Reliability. N. York Economic Classics 1967.
Friedman, M., & Schwartz, A. J. (1963). A Monetary History of the United States: 1867-1960. Princeton, N.J.: Princeton University Press.
Giordano, C. (2009). Prudential Regulation and Supervision Instruments and Aims: A general Framework. In A. Gigliobianco, & G. Toniolo (Eds.), Financial Market Regulation in the Wake of Financial Crises: the Historical Experience (pp. 243-251). Banca d’Italia, Workshops and Conferences.
Gigliobianco, A., Giordano, C., & Toniolo, G. (2009), Innovation and regulation in the wake of financial crises in Italy(1880s-1930s). In A. Gigliobianco, & G. Toniolo (Eds), Financial Market Regulation in the Wake of Financial Crises: The Historical Experience (pp.45-73). Rome: Banca d’Italia.
Goldstein, M. (2007). Emerging Market Financial Crises: Lessons and Prospects. Speech delivered at the 25th Anniversary Membership Meeting of the Institute of International Finance, 20 October.
Grilli, V. (1990). Managing Exchange Rate Crises: Evidence from the 1890s. Journal of International Money and Finance, 9, 258-275.
Grilli, V., & Kaminsky, G. (1991). Nominal Exchange Rate Regimes and the Real Exchange Rate. Journal of Monetary Economics, 27, 191-212.
Kindleberger, C. (1978), Manias, Panics and Crashes: A History of Financial Crises, New York, Basic Books.
Kostelenos, G., Vasiliou, D., Kounaris, E., Petmezas, S., & Sfakianakis, M. (2007), Sources of the Economic History of Modern Greece: Quantitative Data and Statistical Series: Greek Domestic Product, 1830-1939, Centre for Planning and Economic Research and the National Bank of Greece.
Krugman, P. (1979). A Model of Balance of Payments Crises.Journal of Money. Credit and Banking, 11, 311-325.
Krugman, P. (1997).Are Currency Crises Self Fulfilling?. NBER Macroeconomic Annals. Cambridge MIT Press, 345-378.
Krugman, P. (1999).Balance Sheets, the Transfer Problem and Financial Crises. In P. Isard, A. Razin, & A. K. Rose (Eds.), International Finance and Financial Crises: Essays in Honor of Robert P. Flood Jr. Kluwer Academic Publishers.
Krugman, P. (2009).The Return of Depression Economics and the Crisis of 2008. Allen Lane, The Penguin Press.
Lazaretou, S. (2010). Money Supply and Greek Historical Monetary Statistics: Definition, Construction, Sources and Data. Economies et Sociétés, 41, 47-67.
Lazaretou, S. (2008). Banking and Central Banking in pre-WWII Greece: money and currency developments. SEEMHN Bank of Greece Working Paper 86.
Lazaretou, S. (2005). The Drachma, Foreign Creditors and the International Monetary System: tales of a currency during the 19th and the early 20th centuries. Explorations in Economic History, 42, 202-236.
Lazaretou, S. (1995). Government Spending, Monetary Policies and Exchange Rate Regime Switches: the drachma in the gold standard period. Explorations in Economic History, 35, 28-50.
Manasse, P., Roubini, N., & Schimmelpfenning, A. (2003), ‘Predicting Sovereign Debt Crises’, IMF Working Paper WP03/221.
Mitchell, W. C. (1938). The Making and Using of Index Numbers. New York, Economic Classics, 1965.
Obstfeld, M. (1996). Models of Currency Crises with Self Fulfilling Features. European Economic Review.40, 1037-1047.
Obstfeld, M., & Taylor, A. (2003). Sovereign Risk, Credibility and the Gold Standard: 1870-1913 versus 1925-1931. Economic Journal, 113, 1-35.
Reinhart, C. M. (2010).This Time Is Different Chart Book: Country Histories on Debt, Default and Financial Crises’ NBER Working Paper no 15815.
Reinhart, C. M. and K. Rogoff (2009), This Time is Different: Eight Centuries of Financial Folly. Princeton: Princeton University Press.
Reinhart, C. M., & Rogoff, K. S. (2008). Is the 2007 US Subprime Crisis So Different: An International Historical Comparison?. American Economic Review, 98, 339-344.
Reis, J. (1992).The Gold Standard in Portugal, 1854-1891. Mimeo Universidale Nova de Lisbon.
Saunders, A., & Cornett, M. (2003). Financial Institutions Management: A Risk Management Approach (4th edition). McGraw Hill.
Schularick, M., & Taylor, A. (2009). Credit Booms Gone Bust: Monetary Policy, Leverage Cycles and Financial Crises 1870-2008. NBER Working Paper 15512.
Sheikh, K., & Heijmans, P. (2004). New Developments in Country Risk: the New Anatomy of Crises, Rising Public Debt Burdens and Basel II. In S. Wilkin (Ed.). Country and Political Risk: Practical Insights for Global Finance. Risk Books, 3, 37-56.
Stone, I. (1999). The Global Export of Capital from Great Britain 1860-1914. London: Macmillan.
Twain, M. (1940). Eruption: Hitherto Unpublished Paper about Men and Events, edited and introduced by Bernard Devoto (1st edition). New York, First Printing Harper and Brothers.
Van Agtmael, A. (2006). The Emerging Market Century. Free Press.
Appendix
Table A
Dates of Crises Events
Block I
Notes: 1 the data series starts as late as 1885. Debt crises= we include only the year of debt repudiation or debt restructuring. Negotiations years are excluded. Currency crises= *definition 1: we refer to the year of a heavy speculation attack on the domestic money causing abnormal fluctuations of the exchange rate or the year of a nominal currency depreciation or devaluation;**definition 2: following Catao’s working assumption (2006) we define a currency crisis (i.e. currency drop) as nominal exchange rate depreciation greater than at least one standard deviation of the annual percentage change of the spot exchange rate of the drachma against sterling over the sample period. We also assume that this depreciation is not fully reversed within a three year window. Banking crises= we include the years of a wide spread bank panic and/or the years of massive bank failures. Isolated events are not taken into account. ***Following Catao’s (2006) working assumption, we define sudden stop as a drop from peak to trough of no less than two standard deviations of the deviations of the respective series from a linear trend and/or any drop that exceeds 3 per cent of GDP over a period shorter than four years.
Note. 3 gold hoarding; 4 speculation attacks resulting in an abnormal increase in the interest rates to defend currency; 5 an isolated event of a bank failure due to a corporate governance failure (fraud risk); it is not therefore taken into account; 6 bail out, rescue package.
Data Appendix
For the period from 1846 to 1914, bank loans to firms and households refer to total gross loans to the private sector. The data on lending are taken directly from the Balance Sheets of the eight large commercial banks of Greece. For the period from 1950 to 2009, total bank credit includes loans to firms and households; the data are from the Bank of Greece. The data for the level of the Gross Domestic Product (1833-1938) in current prices and in constant 1914 prices, both aggregate and per capita, are from Kostelenos et al. (2007). GDP deflator is a Paasche type index of the prices of 10 products from the primary and secondary sectors and covered over 23 percent of the total value of GDP. GDP has been computed ‘based on estimates made directly using the production (value added) method, the most notable exception being the analysis of the tertiary sector, where a combination of the income method and of an indirect approach has been used’ (Kostelenos et al. 2007, p251). For the period after World War II, the data for the GDP in current prices are from the National Statistical Service of Greece. Total bank assets for the same sample of banks are also taken directly from their balance sheets. Broad money (M3) and monetary base (M0) are from Lazaretou (2010). Demand deposits, vault cash, and short-term lending (advances and discounts to traders) refer to all commercial banks for the period from 1894 to 1932 and are taken from the banks’ balance sheets. The equity to assets ratio has been computed for all commercial banks for the period between 1894 and 1932. Return on equity has been assessed for six large banks for the period between 1910 and 1932. The financial statements of those banks are my primary data source. The data for the nominal exchange rate of the drachma against the British pound are taken from Bank of Greece (2009). Total reserves, foreign exchange and metallic, are from Bank of Greece (2009). The data for total government spending and total tax revenues are from the Greek Government Budgets, Annual Reports, and concerned realized entries.
JEL Classification: E5, N2
Keywords: financial crises, emerging economies, sudden stops, country specific factors.
Acknowledgments: I would like to acknowledge helpful comments by a referee of this Journal. This paper was presented at XXX Conference of the Portuguese Economic and Social History Association on Economic and Social Crises, Lisbon, 19-20 November 2010. I am grateful to Nuno Valério, Rita Martins de Sousa, Amélia Branco and Rui Pedro Esteves for their useful comments. An extended version entitled financial crises and financial market regulation: the historical record of an emerger was presented at the 5th SEEMHN Annual Conference on Monetary policy during economic crises: A comparative and historical perspective, hosted by the Central Bank of the Republic of Turkey in Istanbul on 15-16 April 2010. I would like to thank the conference participants and especially George Chouliarakis, Kalina Dimitrova, evket Pamuk, Milan Sojic and Ali Cokun Tuncer for their comments and suggestions. An earlier version of the first part of this paper was presented at the workshop on Financial market regulation in the wake of financial crises: The historical experience, organized by the Banca d’Italia on 16-17 April 2009 in Rome. I would like to acknowledge helpful comments by the workshop’s participants and especially Alfredo Gigliobianco and Claire Giordano. Finally, I am most grateful to Luis Catao for kindly providing Stone’s data on foreign capital inflows to Greece. The views expressed here are those of the author and do not necessarily reflect those of the Bank of Greece. Any errors remain my responsibility.‘We have been here before…This time is (not) different. It almost never is’ (Reinhart & Rogoff, 2009).
Introduction
In the present era of globalization, financial crises are a recurring phenomenon in the history of emerging market economies. The literature is rich in evidence, both narrative and empirical, on external shock-driven sudden stops of capital inflows, current account reversals, currency drops, a rise in sovereign risk spreads, output losses, and severe financial turbulence (Calvo, Izquierdo, & Mejía, 2004; Calvo, Izquierdo & Talvi, 2005, 2006; Calvo & Talvi, 2005; Krugman, 2009). The pattern has been common to all emerging economies during recent years. A rapid and sudden cut-off in capital inflows is often accompanied by currency, debt, and banking crises, while the impact of the crisis on the real economy and its incidence differ markedly across countries, depending on the special features of the affected country’s economic structure. Prominent examples of this analysis are the 1998 Russian default lending crisis, the 1998 Brazil crisis, and the 2001 crises in Argentina and Turkey.
More recently, Reinhart and Rogoff (2009, 2010), covering a large number of mature and emerging countries over eight centuries, have provided a comprehensive look at the varieties of financial crises. They find a strong link between sovereign debt crises and banking crises across rich and poor countries alike. Researchers’ interest has now focused on the comparison of the emergers’ crisis experience across the two periods of globalization: the present era from 1980 to present, and the first era of ‘golden’ globalization from 1870 to 1914. They have concluded that the crisis pattern for the 1880s and the 1890s emergers was strikingly similar to the experience of today’s emerging market economies (Reinhart, 2010; Bordo, 2008; Bordo et al., 2001; Eichengreen & Lindert, 1989). This means that many driving forces that were present during the recent emerging market crises were also at work in the emergers’ crises a century ago.
In this paper, our main task is to trace the history of financial crises from the perspective of an emerging economy. Two questions are addressed. First, what explains the incidence and severity of financial crises in an emerging market economy? And second, what is the role of learning? In other words, how does the country learn from its past experience in financial crises to improve institutions and develop better techniques so as to successfully manage successive crisis events? Learning concerns both institutional learning and how to follow policies consistent with the aims of these institutions.
The case-study of my analysis is Greece, a country with a rich history in financial crises. Pre-WWII Greece was a typical example of a South-Eastern European (SEE) ‘emerging market economy’,1 as it was on the ‘periphery’ of the international monetary system.2 Most of the historical literature so far has concerned the experience of the advanced countries of Western Europe. Interest in studying the behaviour and response of emerging countries has appeared only recently. However, historical research has largely focused on the peripheral areas of Latin America and Asia and only partly on the emerging economies of SEE. Indeed, the history of financial crises in that particular region of the European periphery—which faced a more turbulent financial, economic, and political environment—is still unexplored or only partly studied. To the best of our knowledge, this paper is the first attempt to provide a crisis event taxonomy looking at a systematic categorisation of the crises episodes that the country experienced over its 200-year life span. Well till now, in all empirical studies Greece has been systematically neglected or included only occasionally and sporadically in their country samples and thus the Greek crisis experience is more or less terra incognita. Hence, tracing the past experience of an SEE emerger, like Greece, enriches our knowledge to answer the following questions. First, why do crisis events hit more frequently and more severely countries with backward and unsound economic and political institutions? Second, how does the absence of ‘country’ and ‘currency trust’ exaggerate the effects of a financial crisis on the real economy? And third, how do the emergers learn to prevent and contain crisis episodes?
To answer the above questions, I will first present evidence on financial crises in Greece over a long time span. In particular, I will try to identify different varieties of crisis events, providing thus a crisis chronology. Moreover, I will present a number of facts about the incidence, frequency, and severity of crisis events. To this end, I use the standard definition of financial crises. And, second, I will discuss the key determinants of crisis events, closely related to country-specific factors, such as monetary expansion, fiscal imbalances, and the limited reserve coverage of domestic money. I will conclude by presenting important lessons derived from the country’s historical experience. The main lesson of historical experience is that crisis events in Greece were largely credit driven. A comparison of the key characteristics and the sources of the two historical episodes of financial instability (‘then’ and ‘now’) reveals first that excess credit was the main crisis determinant in both instances, and, second, the absence of ‘country’ and ‘currency trust were the key aggravating factors of all crisis episodes. Spendthrift governments proved to be unable to tax sufficiently to cover their expenditures and to undertake a budget reform, i.e. to manufacture a politically difficult reform to eliminate fiscal excess.
Currency crises, banking crises and debt crises: defining the terms
Before revisiting the long record on financial crisis episodes from the perspective of an emerging market economy, it might be useful to set the theoretical framework by defining the terms used. Following Eichengreen and Portes (1987), the typology of financial crises can be thought as including three types of crises explained by different sets of variables: currency crises, banking crises, and debt crises (see also Goldstein, 2007; Bordo, 2006, 2008; Reinhart & Rogoff, 2009, Part I)3.
Currency crises are viewed as being caused either by weak economic fundamentals or certain government policy actions, self-fulfilling expectations of market participants, and possibilities of multiple equilibria, or they are viewed as a run on a currency or a financial panic (see, inter alia, Krugman, 1979, 1997, 1999; Obstfeld, 1996). Attention has been drawn to the importance of balance sheet effects for the sustainability of a currency target. Measures of real exchange rate overvaluation, external imbalances, foreign exchange reserves and export growth are explanatory variables for currency crises (Catao, 2006; Catao & Solomou, 2005). For identifying the period of a currency crisis event, in the empirical work that follows, we have based upon the following definition: a currency crisis is defined as a speculative attack on the exchange rate, resulting in a sharp devaluation or depreciation within a given year in the case of a flexible rate regime, or in a large outflow of the country’s foreign reserves, accompanied by a forced abandonment of an exchange rate commitment in the case of a fixed rate regime.
In the case of banking crises, attention should be paid to the special features of banks, such as maturity and currency transformation and asymmetric information. These features make banks vulnerable to runs and collapses following adverse shocks of either domestic or external origin. Theory and practice provide us with four typical sources of bank failures: imprudent credit policies, imprudent investment policies, poor asset-liability management, and bank panics.
Bank runs have been modelled as asymmetric information problems between depositors and banks, random manifestations of mass hysteria, or self-fulfilling expectations of depositors (Kindlerberger, 1978; Diamond & Dybrig, 1983). Hence, banking crises encompass both liquidity and solvency crises. A bank is illiquid when it is not able to meet short-term expected and unexpected obligations when they fall due. A bank is insolvent when its liabilities are greater than the value of its assets. As long as profits are sufficient to cover loan loss provisions, the level of bank capital and its capital adequacy ratio remain unchanged. However, when profits fall short, the amount of the bank’s capital declines4. To sum up, real economy disturbances may adversely affect the banks’ loan portfolio (credit risk); bubbles in equity and bond markets often collapse and destroy the value of the banks’ securities portfolio (market risk); and exchange rate disturbances induce interest rate changes which change the position and the slope of the yield curve(exchange rate and interest rate risk).
In my analysis, I identify banking crises as periods of severe difficulty, either because of a severe liquidity shortfall, and/or a reduction of the banks’ capital position. In twin crises, both currency and banking crises occur together. Severe currency volatility results in a sharp rise in the risk premium and the interest rates, thus making banks and corporations face increasing lending costs. Furthermore, currency mismatches bring to the surface serious weaknesses in the banks’ balance sheets5.
A debt crisis is a situation in which the country either has sizeable arrears of principal and/or interest on its obligations to its lenders, or in which there is debt restructuring or debt repudiation with commercial creditors(Detregiache & Schilimbergo, 2001; Manasse, Roubini & Schimmerlfenning, 2003). In other words, a debt crisis occurs when investors conclude that the debt ratio has become unsustainable. Thus, debt crises might be better explained by economic factors that measure the weight of domestic and foreign debt, the amount of debt service obligations, or the ability of a country to generate sufficient resources to meet these obligations. In emerging market economies, currency crises are closely linked to the probability of sovereign default, in the sense that a currency crisis increases the probability of fiscal distress. It will activate currency mismatches that result in a ‘balance sheet crisis’ both for the government and the private sector.
In general, a sovereign country’s negative decision on its debt obligations may have two forms: repudiation and rescheduling. ‘Debt repudiation is an outright cancellation of all current and future debt obligations by a borrower […] [while] debt rescheduling is the change of the contractual terms of a loan such as its maturity and interest payments’ (Saunders & Cornett, 2003, p398). Rescheduling has been the most common form of sovereign risk events in the period after World War II, whereas a large proportion of debt problems were met with repudiations before World War II (Eichengreen & Portes, 1987). This difference in behaviour is explained by the fact that before the war most international debt was in the form of foreign bonds, while after the war debt was in bank loans. However, in the present era of globalization, bond financing is again the main instrument of sovereign borrowing. This is clearly a similarity to the earlier era of ‘golden’globalization6. Finally, a third generation crisis or a ‘balance sheet crisis’ is defined as a twin crisis accompanied by a debt crisis.
Apparently, to analyse financial crises we need a ‘minimal structure’ around which historical observations can be placed (Eichengreen & Portes, 1987). This structure is thought to be made up of four elements. First, it consists of a chain in a financial crisis. The linkages run from exchange rate disturbances to debt defaults and bank failures, and eventually to the real macro-economy. Second, the propagation of the disturbance and the probability of its turning into a generalized crisis are influenced by the specific institutional character of the country’s financial system. How well do the banks manage risk? Are they supervised and regulated by an independent authority or not? Third, the government should be actively involved in creating a regulatory framework that insulates the banking system and the real economy from a shock. And fourth, there is a need to link monetary stability with financial stability. The exchange rate and monetary arrangements influence both the volatility of the exchange rates and the interest rates. In the context of the so called ‘balance sheet approach’to financial crises (Goldstein, 2007; Borio, 2004; Beim & Calomiris, 2001; Allen et al., 2002), a financial crisis can be considered a chain. It draws attention to the fact that a crisis can emerge from weaknesses in banks’balance sheets, vulnerabilities in corporate and household balance sheets, and problems in the government’s balance sheet (see Chart 1).7
Chart 1
Interpreting financial crises: the ‘balance sheet approach’
Source. Sheikh and Heijmans 2004.
The pattern of financial crises: the Greek story line
The pattern of financial crises in Greece is shown in Figures 1 and 2. Figure 1 shows the frequency of crises measured as percent in total, where we divide the number of crisis events that occurred in each sub-period. The data set is separated into five different sub-periods: the years prior to 1880, the gold standard period (1880-1913), the interwar period (1919-1939), the Bretton Woods period (1945-1974), and the most recent period (1975-2008). As can be seen, in the period before World War I the predominant form of the crisis were currency crises, followed by banking crises, although debt crises were the second most frequent type of the crisis before 1880. Before 1913, currency crises accounted for more than half of all crisis events. On the contrary, banking crises predominated in the interwar years; they account for more than two-thirds of the total events. The pattern of crises did not change notably after World War II. Currency collapses remained the most frequent crisis event, followed by banking crises, while no debt crisis event occurred during that period8,9.There is no evidence of a twin crisis, while a third-generation crisis occurred only once, in 1932.10
Figure 2 shows the frequency of crises measured as percent per year; the years in crisis are divided by the total number of years in each data sub-sample. The figure suggests that, with the exception of the interwar period, crises appear to be more frequent after World War II. Surprisingly, the crisis frequency of 36.7 percent during the Bretton Woods period is much greater when compared to 17.7 percent in the period after 1975, while it is cut only in half compared to the most unstable period of the interwar era (71.4 percent). Another interesting point to be stressed here is that the frequency of crises was much lower in the early era of globalization before 1913, when compared to the post-1975 era. Finally, as expected, all types of crisis show the greatest frequency during the interwar period, confirming that the interwar crisis event still remains the ‘mother’ of all subsequent historical crisis episodes.
It may be interesting to compare the Greek crisis experience in the two eras of ‘finance capitalism’, that is, pre-1913 and post-1975. Then, improvements in technologies of communication and transfer—for example, the transatlantic telegraph cable in the late 1860s, which was the only way to transmit information across oceans—reduced the cost of money and goods transformation, thereby contributing to greater economic integration. Both periods are very similar to each other in many respects, but very distinct too. Clear similarities are the use of bond financing as an instrument of sovereign borrowing and the high degree of world capital and goods market integration. However, a stark difference is the post-1975 strong increase in financial leverage, caused by the decoupling of money and credit aggregates. This means that after 1975 banks funded growth mainly through non-monetary liabilities, and thus their role in credit creation via bank loans is of great importance (Schularick & Taylor, 2009)11.
A closer look at the 1890s crisis episode and the current crisis event reveals that financial instability in Greece has been largely credit-driven. As Reinhart and Rogoff (2008) have argued, periods of financial distress have been associated with economic downturns and typically follow waves of domestic and international credit expansion. Figures 3 and 4 tell us the story. Figure 3 plots the ratio of bank private credit to GDP over the pre-World War I period; this ratio shows the importance of the main function of the banks, namely channelling savings to investors. The data refer to bank gross loans to firms and households gathered by the balance sheets of eight domestic banks with a long presence on the Greek money market. Based on a simple inspection of the data, two periods of soaring credit can be easily detected. The first one took place from the mid-1870s to the early 1890s, and the second one from 1898 until the outbreak of World War I.
Systematic attempts to industrialize the country can be dated to the last quarter of the nineteenth century. The economic policy of the governments placed emphasis on the development of the private sector and the introduction of new technology to the production process, with parallel attempts to free the economy from stifling state control. It was at this time that the first efforts were made to create infrastructure, which helped to modernize the country as well as to restructure the economy. During that period, agricultural production increased, industry was introduced, trade developed, credit grew rapidly, and the number of functioning banks increased. New banking products appeared, such as mortgage-backed loans, mortgage-backed credit lines, and long-term bank bonds. Rich Greek emigrants were attracted by the high deposit rates and kept their money within Greek banks12. Moreover, the country’s modernization and the implementation of many large public works attracted foreign private investment capital. Foreign and Greek businessmen heavily invested in residential and commercial properties, as well as in portfolio investments13. A lending boom soon occurred14, which made land and resource prices soar.
Figure 1. Crisis frequency (percent in total). Notes. A currency crisis is identified as a year when there was either a forced abandonment of an exchange rate commitment, an official depreciation, or an abnormally large devaluation of the value of the exchange rate in a given year. A debt crisis is identified as a year when government debt repudiation or restructuring occurred. A banking crisis is identified as a year when either a wide-spread deposit bank run occurred, banks went bankrupt, or when they asked for bail-out when the function of the lender of last resort was activated to confront a wholesale or a deposit bank run. Isolated events are not taken into account, but only the year when the event took place, excluding the years of oncoming or ongoing crises. See also Table A in the appendix.
Figure 2. Crisis frequency (percent per year). Note: See Figure 1.
Figure 3. Credit soaring in the first era of globalization, 1846-1914 (percent to GDP). Note. bank gross loans to firms and households (private bank credit). Computations are based on yearly averages, using a five-year rolling time span. They remain unchanged when we use a longer (ten-year) rolling window. Using a rolling time span helps to minimize the resulting loss of information, which is the principle drawback of averaging. End-of-year data. Source. See appendix.
Economic growth in the 1880s was based on cheap and easy foreign lending. After the country’s foreign debt compromise in 1878-1879, Greek government bonds were again traded on the London stock market, and the governments started to heavily borrow from abroad to cover excess spending, both consumption and investment. During that period, the country’s creditworthiness was rebuilt, since the governments considered the gold standard as a ‘good housekeeping seal of approval’ along the lines considered by Bordo and Rockoff(1996), in order to improve the country’s access to the foreign capital markets and attract cheap lending(Lazaretou, 2005). Ultimately, the government accumulated an enormous external debt burden with gold clauses that triggered its ability to repay it, and thus the currency was pressed by a heavy speculation attack. In 1895, the spot exchange rate almost doubled.
With the successful rescheduling of foreign debt in 1898, a second period of soaring credit was put in motion. This time the credit increase was strong and rapid: the ratio of bank loans to GDP grew by a factor of four, reaching a peak in 1910-1911, when the country credibly joined gold after a long, painful process of effective fiscal adjustment and monetary restraint. The key driver was the rapid output growth that the country generally experienced in the years prior to World War I. Export trade increased mainly after 1905, and the growth of domestic production accelerated. The shipping industry marked considerable progress, expanding its activities to the transit trade of third countries. Greater economic activity brought with it an increase in the number of commercial banks. The country’s international credit standing was rebuilt, resulting in foreign capital inflow.
In the current era of financial capitalism, soaring private and public credit has also been the key driving force. As seen in Figure 4, the ratio of private bank loans to GDP soared in the few years prior to the current crisis. Public debt also soared. While in the 1990s the mean rate of private credit stood at very low levels (36.6 percent), in the first half of the 2000s it quickly soared to 54.4 percent. Thereafter, it increased even more rapidly, approximating 84 percent in 2008 and 81 percent in 2009, becoming twice as high as in 2000. Similarly, general government debt as a percent to GDP increased from 107.4 in 2007 to 144.9 in 2010. Lending became very cheap and easy, just after the country’s entrance into the euro area, soon resulting in a‘liquidity overhang’ that triggered an asset price bubble and a consumption and investment boom. However, in the wake of the international financial crisis in 2008, a bust soon followed, resurfacing the country’s structural inefficiencies—such as excess public indebtedness, a sharp enlargement of both the current account deficit and fiscal deficit, and a heavy loss in the country’s international credit standing (Bank of Greece, 2010).
What went wrong? The economic and financial environment
With the help of some simple descriptive statistics and using the Greek case as a working template, in this section I will highlight the key driving forces of emerging market crises. My interest is focused on the pre-1913 golden era of globalization. This era was characterized globally as a period of a low frequency of crises events. During that period, crises were phenomena observable in emerging rather than advanced countries.15 This was because emerging market economies lacked a sound institutional framework and had not learned to follow successful policies within that framework. This framework includes sound fiscal and monetary institutions, as well as regulation and supervision authorities. This is what Caballero et al. (2004) have named ‘country’ and‘currency trust’.16 In particular, sound fiscal institutions mean the existence of an efficient tax system, the avoidance of excessive public debt exposure, and the credible commitment to balance the budget. Sound monetary institutions include the credible adherence to the classical gold standard, by holding sufficiently large gold reserves to minimize a currency mismatch between hard currency liabilities and domestic currency revenue.
Figure 4. Credit soaring in the current era of globalization, 1950-2009 (percent to GDP). Note: total bank credit, i.e. loans to firms and households. Yearly averages are calculated by using a five-year rolling window. End-of-year data. Source: See appendix.
The advanced countries of that time had developed a sound and solid institutional framework that allowed them to manage crisis episodes successfully. They had developed financial markets, which had largely overcome the problem of asymmetric information, and they had efficiently channelled savings to productive investment. They enacted balanced fiscal policies and non-accommodative monetary policy targeting the fixed rate. They operated within the framework of globalization with free trade, free capital, and labour movements. They had successfully established important institutions such as a legal system to protect property rights and constitutional democracy, which greatly supported financial growth (Bordo & Rousseau, 2006).
By contrast, emerging countries of that time were in the process of developing fiscal and monetary institutions. They faced a more turbulent financial environment and experienced long periods of political instability. This in turn meant that they were prone to the occurrence of crises and that they had not learned to deal with them. What were the ‘risk factors’ for them? In the era of laissez faire laissez passé, excess saving in the advanced world flowed towards the then poor peripheral countries17. The massive capital inflows boosted asset prices in the emerging countries, particularly of land and unexploited physical resources, and fed a credit and growth boom. Inflows were accompanied by high budget deficits and government debt accumulation. In conjunction with an accommodative monetary policy, inflows fuelled domestic inflation and fed speculative attacks on the currency. The real burden of debt (both private and public) soared, asset prices failed, and lending went bust. Ultimately, the government and the banks faced a balance sheet crisis.
Historical accounts
As we saw in the previous section, the Greek record of financial crises verifies the story outlined above. What was the economic and financial landscape that made an ‘emerger’ like Greece more vulnerable to crises18? First, pre-World War II Greece was a poor agricultural economy19; it financed its growth process through heavy foreign lending. Second, capital inflows were attracted by higher returns on land and other resources and led to lending booms. Booms were accompanied by fiscal expansion, financed by money creation and government debt accumulation. However, the interplay between fiscal imbalances and monetary disturbances resulted in frequent convertibility crises. The country responded to the speculative attacks on the currency by adopting, if only occasionally, the gold standard as ‘a good housekeeping seal of approval’. This enabled the country to continue attracting cheap foreign capital.
Third, the country was prone to a debt crisis when the domestic economy collapsed as a consequence of a lending bust. This was because the country’s fiscal and monetary institutions were extremely fragile. The governmental financial system was weak due to the inherent structural inefficiencies and the frequent external spending shocks, such as wartime emergencies. In particular, the tax system, based largely on indirect taxes and custom duties, was procyclical. Thus, the governments were unable to raise sufficient tax revenues to service the debt. Further, monetary policy was loose and accommodated fiscal policy20.
Fourth, the country suffered from ‘original sin’, as it has been put forth by Eichengreen and Hausmann(1999) and Eichengreen et al. (2003). That is, it was not able to borrow abroad or even at home in terms of its own currency; debt issue required gold or exchange rate clauses. This in turn meant that in the case of a nominal currency devaluation, the debt burden soared, increasing the likelihood of the government’s default on its outstanding debt, with certain repercussions adversely affecting the quality of the balance sheet both of the banks and the private sector. As a result, risk premia rose.
Fifth, the country had the experience of ‘sudden stops’. Sudden stops were a common feature in the economic history of the emergers (Catao, 2006, Bordo et al., 2010). Every time the economic and financial circumstances in the advanced lending countries changed and thereby led to a cut-off of cheap capital inflows to the emerging economies, the latter soon faced a balance of payments crisis and a debt crisis. This was exactly the case of a sudden stop event that Greece experienced in the early 1890s and again in the early 1930s.
Sixth, the country was financially less developed; financial depth measured as the ratio of bank assets to GDP and/or broad money to GDP exhibited a positive long-run trend, even though it was relatively weak and excessively volatile21. Throughout the whole sample period from 1846 to 1939, banking intermediation as measured by both indices stood at very low levels: it was on average 51 and 37.7 percent, respectively.
Empirical accounts
The data strongly support the historical accounts. Figure 5 shows the pattern of gross capital inflows to Greece based on Stone’s portfolio calls on London data from 1870 to 1913 (Stone, 1999). Due to lack of data, portfolio calls on London can be considered a proxy for foreign capital inflows to the country, since London was then the most important lending centre for emerging economies. As is evident, the time series seems to be stationary22, although it was extremely volatile. However, an upward trend was present from 1879 to 1890. Gross foreign capital inflows pointed to a peak in 1898, 1903, and again in 1910. Following the same working definition as in Catao (2006, p7), I define a sudden stop as a drop from peak to trough of no less than two standard deviations of the deviations of the respective series, from a linear trend and/or any drop that exceeds 3 percent of GDP over a period shorter than four years. Based on this definition, the figure clearly shows a significant rise in the second half of the 1880s and a sudden drop in the 1890s. This pattern is in accordance with the average pattern of capital inflows for a large panel of emerging countries during that period23. Precisely, three episodes of sudden stops can be detected and are denoted in the shaded areas: 1881-1883, 1890-1897, and 1911-191324. The second episode was associated with the 1893 debt repudiation.
Figure 5. Gross portfolio calls on London, 1870-1913 (million of pounds). Note. Sudden stops are in the shaded areas. Source. the data are from Stone (1999).
Figure 6 shows the pattern of sovereign bond spread over the period between 1870 and 1914. The bond spread is measured as the difference between the Greek bond yield and the British consol yield25. Although it shows a declining long-term trend reflecting a greater world-wide financial market integration, it sharply increased above its trend in the years of a sudden stop, especially in the years when a sudden stop was followed by a debt crisis, as was the case in the 1890s.
Figure 7 depicts the time series behavior of the nominal and the real exchange rate of the drachma against the British pound. For the larger part of that time, Greece did not adhere continuously, but only partially, to the classical gold standard. As shown, the country experienced high depreciating rates during sudden stop episodes. The most notable example is the episode of 1890-1897. The drop in capital inflows was accompanied by a sizeable depreciation in nominal terms between 1892 and 1895. The nominal value of the currency rate peaked in 1895, two years after the 1893 debt crisis. Depreciation was also sizeable in 1886, followed by the sudden stop event of 1881-1883. Following Catao (2006, p11), I define a currency crash as nominal exchange rate depreciation greater than at least one standard deviation of the annual percentage change of the spot exchange rate of the drachma against the British pound over the period 1885 to 1914. I also assume that this depreciation was not fully reversed within a three-year window. Based on this working assumption, I date four such events: 1886, 1892, 1893, and 1894.
Figure 6. Greek sovereign bond spread, 1870-1914 (percentage points). Note. Greek gold bond yield minus British consol yield; percent per annum. Source. See appendix.
Figure 8 depicts the evolution over time of the real GDP per capita. In the same figure, the incidence of crisis events is also indicated and the sudden stop episodes are shaded. One can notice that sudden stops and crisis events were associated with notable drops in the level of the real GDP per capita26.
Figure 9 shows output losses associated with each sudden stop event. Following Adelet and Eichengreen(2005) and Bordo (2006), the output loss is measured as the change between the average growth rate three years before the crisis and the average growth rate three years after the crisis. A three-year growth average is considered to serve as a proxy for the trend growth. As seen, the sudden stop episode of 1890-1897 was associated with a considerable output loss.
Country specific factors
So far, the inspection of the Greek historical data has indicated that drops in foreign capital inflows were associated with currency crises and debt crises. These fluctuations in capital flows were largely determined by external shocks, such as a change in the central bank rate of the lending countries. For example, the 1890-1897 crisis in Greece can be considered the outcome of an array of adverse shocks in the international economic and financial environment, which worried foreign creditors who, until the late 1880s, were generously supplying cheap loans to emerging economies without any guarantee and at low interest rates27. Specifically, the lending boom of the 1880s ended suddenly in 1890-1891, when the advanced lending countries stopped lending as economic conditions in advanced Europe gradually improved and investment opportunities reappeared. The core countries’ central banks responded by increasing their discount rates, which caused a massive slowdown in investment abroad.
Figure 7. The nominal and real exchange rate of the drachma against the British pound, 1885-1914. Notes. Cc = dates of the currency crises. Sudden stops are in shaded areas. Spot rates, yearly averages. LHS = real (1887-1914), RHS = nominal. The real exchange rate has been calculated as the ratio of British wholesale prices (Sauerbeck index) to prices for basic foodstuffs in Greece, using the bilateral nominal exchange rate as the conversion ratio. The food price index (1866-1877 = 100) has been constructed as a simple geometric average of the relative prices of five traded food products. Since data on quantities consumed are not available, a Laspeyres index cannot be calculated. Moreover, as Fisher (1927) and Mitchell (1938) have pointed out, the simple geometric average has the advantage of smoothing the time series of prices with regard to extreme values. This is very important in the case of food products, because their prices exhibit high volatility. Pre-1914 official data for a price index do not exist. The data for the Sauerbeck index have been taken from the Journal of Royal Statistical Society and Grilli and Kaminsky (1991). Source. See appendix.
Figure 8. Real GDP per capita, 1870-1914. Note: In 1990 international US dollars. Cc = dates of currency crises events; Bc = dates of a banking crisis; Dc = dates of a debt crisis; sudden stops are in the shaded areas. Source. See appendix.
Figure 9. Output loss during sudden stops. Note. Change in real output growth rates, percentage points. Output loss has been calculated as the change in the average growth rates three years before and after the sudden stop event.
However, external shocks affected emerging countries differently. Country-specific factors closely related to fiscal, monetary and financial behaviour were the key drivers of this effect. Figure 10 shows the trend of the fiscal and monetary policy indicators over time. One can observe that fiscal behaviour as proxied by the ratio of total government spending to total tax revenues was generally lax, noticeably cyclical without however exhibiting a long-run trend. Sudden stop episodes and currency crashes were preceded by periods of high expenditure, which were accompanied by rapid monetary expansion as shown by the ratio of M3 to GDP. The ratio doubled between 1870 and 1880 and tripled between 1885 and 1890. As a result, a concomitant sharp drop in the international reserve coverage of the domestic currency occurred, implying that the country’s monetary authorities were no longer able to adhere to the specie convertibility rule. Further, by regressing the cyclical component of real government expenditure on the cyclical component of real GDP28, bivariate regression estimates verify the well-known conclusion that in countries that experience currency crises fiscal policy is procyclical29. A procyclical response of fiscal policy to the business cycle means that during cyclical upswings fiscal policy exacerbates imbalances, which in turn causes nominal exchange rate depreciation and a drop in the reserve coverage of the monetary base.
Figure 10. Fiscal and monetary policy stance indicators, 1870-1914. Note. percent per annum. The ratio of spending to revenues is measured on the right hand axis. Source. See appendix.
Figure 11 plots the M3 multiplier—that is, the ratio of broad money (M3) to monetary base (M0). It corroborates the view that monetary expansion played a significant role in aggravating currency crises during sudden stop events. As with the ratio of broad money to GDP, domestic bank credit as proxied by the M3 multiplier is also cyclical throughout the period under study. From the turn of the century, however, it exhibited a strong upward trend, as the result of the country’s rapid financial development process that started in those years.
Figure 11. M3 multiplier, 1870-1914. Note. M3 to M0 ratio. End-of-year data.
To recapitulate, in the pre-1913 period Greece experienced sudden stops in foreign capital inflows, accompanied by currency and debt crises. Country-specific factors were the key driving forces of this effect. Indeed, as the data show, all crisis events were preceded by periods of fiscal laxity, rapid monetary expansion, and limited reserve coverage of domestic money. Moreover, probit estimates strongly support the above conclusion. The empirical estimates conclude that crisis predictor coefficients on m3/y, reserves/m0, and spending/reserves are shown to be statistically significant in spite of the small number of crisis events. 30
Concluding remarks
Several important lessons can be drawn from Greece’s long record of crisis experience. First, a comparison of the key characteristics and the sources of the two historical episodes of financial instability(‘then’ and ‘now’) reveal that excess credit was the main crisis determinant in both instances. Second, country-specific features such as fiscal laxity, credit expansion, and the limited reserve coverage of the domestic currency were the key aggravating factors of the earlier crisis. And third, data and historical accounts verify that the government’s response to a crisis event was weak, slow, and delayed. This means that the government officials learned only with serious difficulty and after a long delay how to manage successive crises and develop better techniques.
Recently, several scholars have raised the following question:31 Can strict regulation produce innovations that might in turn produce future crises? This is not the general rule. Certainly, there are many others risk factors. For example, riskiness might arise from global macroeconomic imbalances combined with inherent financial inefficiencies and weak institutional determinants that put to work destabilizing incentives.
A crucial issue is thus addressed: How can we strengthen the deep institutional determinants of the country’s financial development (such as the rule of law, property rights, corporate governance, and constitutional democracy) to avoid a new crisis? Therefore, an interesting topic for further research might be to identify the predicted probabilities of a crisis event. By enlarging the country sample to include all SEE countries, probit regression results might show if and to what extent country-specific features would be the most important predictor of crashes.
To recall Mark Twain, ‘…history does not repeat itself, but it does rhyme’.32 In other words, history has shown that, even though financial crises were accompanied by considerable output loss, they were always followed by a period of growth that lasted many years. However, as historical accounts demonstrate, crisis events will always occur, and it is very difficult to avoid repetition. On the other hand, policy-makers and governments often have a short historical memory. This means that their past attempts to strengthen macro-prudential supervision either failed or were proven to be ineffective in preventing similar situations in the near future.
References
Adalet, M., & Eichengreen, B. (2005). Current Account Reversals: Always a Problem. NBER Working Paper 11634.
Allen, M., Rosenberg, C., Keller, C., & Roubini. N. (2002). A Balance Sheet Approach to Financial Crises, IMF Working Paper 02/210.
Bank of Greece. (2009). Monetary Time Series of South-eastern Europe from the 1870s to 1914. SEEMHN and Bank of Greece Working Paper no 94, February.
Beim, D., & Calomiris,C. (2001). Emerging Financial Markets. London, McGraw Hill.
Bernanke, B. S., & M. Getler and S. Gilchrit. (1999). The Financial Accelerator in a Quantitative Business Cycle Framework. In J. B. Tayloer, & M. Woodford (Eds.), Handbook of Macroeconomics (Vol.1, pp.1341-1393). Elsevier.
Bordo, M. (2006). Sudden Stops, Financial Crises and Original Sin in Emerging Markets: Déjà vu. NBER Working Paper no 12393.
Bordo, M. D. (2008). Growing up to Financial Stability. SEEMHN Bank of Greece Working Paper no 85.
Bordo, M. D, & Meissner C. M. (2006a). The Role of Foreign Currency Debt in Financial Crises: 1880-1913 versus 1972-1997. Journal of Banking and Finance, 30, 3299-3329.
Bordo, M. D., & Meissner, C. M. (2006b). Financial Crises, 1880-1913: the Role of Foreign Currency Debt. In S. Edwards, G. Esquivel, & G. Marquez (eds.), Growth, Protection and Crises: Latin America from an Historical Perspective, Chicago, University of Chicago Press.
Bordo, M. D., & Rousseau, P.B. (2006). Legal-Political Factors and the Historical Evolution of the Finance-Growth Link. European Review of Economic History, 10, 421-444.
Bordo, M. D, & Rockoff, H. (1996). The Gold Standard as a Good Housekeeping Seal of Approval .Journal of Economic History, 56, 389-428.
Bordo, M. D., Cavallo, A. F., & Meissner, C. M. (2010). Sudden Stops: Determinants and Output Effects in the First Era of Globalisation 1880-1913. Journal of Development Economics, 91, 227-241.
Bordo, M. D., Eichengreen, B., Klingebiel, D., & Martinez-Peria, M.S. (2001). Is the Crisis Problem Growing more Severe?. Economic Policy, 32, 51-75.
Caballero, R., Cowan, K., & Kearns. J (2004), Fear of Sudden Stops: Lessons from Australia and Chile. NBER Working Paper 10519.
Calvo, G., Izquierdo, A., & Mejía, L. F. (2004). On the Empirics of Sudden Stops: the Relevance of Balance Sheet Effects. NBER Working Paper 10520.
Calvo, G., Izquierdo, A., & Talvi, E. (2005). Sudden Stops, the Real Exchange Rate and Fiscal Sustainability: Argentina’s Lessons. In V. Alexander, J. Melitz, & G. V. Fustenberg. (Eds.), Monetary Unions and Hard Pegs: Effects on Trade, Financial Development and Stability (pp.151-181). N.Y, Oxford University Press.
Calvo, G., Izquierdo, A., & Talvi, E. (2006). Phoenix Miracles in Emerging Markets: Recovery Without Credit from Systemic Financial Crises. NBER Working Paper 12101, March.
Calvo, G., Talvi, E. (2005). Sudden Stops, Financial Factors and Economic Collapses in Latin America: Lessons from Argentina and Chile. NBER Working Paper 11153.
Casu, B., Giradrone, C. & Molyneaux, P. (2006). Introduction to Banking. FT Prentice Hall.
Catao, L. (2006), ‘Sudden Stops and Currency Drops: A Historical Look’, IMF Working Paper 06/133.
Catao, L., & Solomou, S. N. (2005). Effective Exchange Rates and the Classical Gold Standard Adjustment. American Economic Review, 95, 1259-1275.
Della Paolera, G., & Taylor, A. (2001), Straining at the Anchor: The Argentina Currency Board and the Search for Macroeconomic Stability, 1880-1935. Chicago ,Chicago University Press.
Detragiache, E. , & Spilimbergo, A. (2001). Crises and Liquidity: Evidence and Interpretation. IMF Working Paper no WP01/2.
Diamond, D. W., & Dybvig, P. M. (1983).Bank Runs, Deposit Insurance and Liquidity.Journal of Political Economy, 91, 401-419.
Eichengreen, B. (1992). Golden Fetters, New York, Oxford University Press.
Eichengreen, B., & Hausmann, R. (1999). Exchange Rates and Financial Fragility. In Federal Reserve Bank of Kansas City (Ed.), New Challenges for Monetary Policy (pp.329-368).
Eichengreen, B., Hausmann, R., & Panizza, U. (2003).Currency Mismatches, Debt Intolerance and Original Sin: Why they are not the same and why it matters. NBER Working Paper 10036.
Eichengreen B., & Portes, R. (1987). The Anatomy of Financial Crises. In R. Portes, & A. Swoboda (Eds.), Threats to International Financial Stability (pp.10-66). CEPR and Cambridge University Press.
Fisher, I. (1927). The Making of Index Numbers. A Study of Their Varieties, Tests and Reliability. N. York Economic Classics 1967.
Friedman, M., & Schwartz, A. J. (1963). A Monetary History of the United States: 1867-1960. Princeton, N.J.: Princeton University Press.
Giordano, C. (2009). Prudential Regulation and Supervision Instruments and Aims: A general Framework. In A. Gigliobianco, & G. Toniolo (Eds.), Financial Market Regulation in the Wake of Financial Crises: the Historical Experience (pp. 243-251). Banca d’Italia, Workshops and Conferences.
Gigliobianco, A., Giordano, C., & Toniolo, G. (2009), Innovation and regulation in the wake of financial crises in Italy(1880s-1930s). In A. Gigliobianco, & G. Toniolo (Eds), Financial Market Regulation in the Wake of Financial Crises: The Historical Experience (pp.45-73). Rome: Banca d’Italia.
Goldstein, M. (2007). Emerging Market Financial Crises: Lessons and Prospects. Speech delivered at the 25th Anniversary Membership Meeting of the Institute of International Finance, 20 October.
Grilli, V. (1990). Managing Exchange Rate Crises: Evidence from the 1890s. Journal of International Money and Finance, 9, 258-275.
Grilli, V., & Kaminsky, G. (1991). Nominal Exchange Rate Regimes and the Real Exchange Rate. Journal of Monetary Economics, 27, 191-212.
Kindleberger, C. (1978), Manias, Panics and Crashes: A History of Financial Crises, New York, Basic Books.
Kostelenos, G., Vasiliou, D., Kounaris, E., Petmezas, S., & Sfakianakis, M. (2007), Sources of the Economic History of Modern Greece: Quantitative Data and Statistical Series: Greek Domestic Product, 1830-1939, Centre for Planning and Economic Research and the National Bank of Greece.
Krugman, P. (1979). A Model of Balance of Payments Crises.Journal of Money. Credit and Banking, 11, 311-325.
Krugman, P. (1997).Are Currency Crises Self Fulfilling?. NBER Macroeconomic Annals. Cambridge MIT Press, 345-378.
Krugman, P. (1999).Balance Sheets, the Transfer Problem and Financial Crises. In P. Isard, A. Razin, & A. K. Rose (Eds.), International Finance and Financial Crises: Essays in Honor of Robert P. Flood Jr. Kluwer Academic Publishers.
Krugman, P. (2009).The Return of Depression Economics and the Crisis of 2008. Allen Lane, The Penguin Press.
Lazaretou, S. (2010). Money Supply and Greek Historical Monetary Statistics: Definition, Construction, Sources and Data. Economies et Sociétés, 41, 47-67.
Lazaretou, S. (2008). Banking and Central Banking in pre-WWII Greece: money and currency developments. SEEMHN Bank of Greece Working Paper 86.
Lazaretou, S. (2005). The Drachma, Foreign Creditors and the International Monetary System: tales of a currency during the 19th and the early 20th centuries. Explorations in Economic History, 42, 202-236.
Lazaretou, S. (1995). Government Spending, Monetary Policies and Exchange Rate Regime Switches: the drachma in the gold standard period. Explorations in Economic History, 35, 28-50.
Manasse, P., Roubini, N., & Schimmelpfenning, A. (2003), ‘Predicting Sovereign Debt Crises’, IMF Working Paper WP03/221.
Mitchell, W. C. (1938). The Making and Using of Index Numbers. New York, Economic Classics, 1965.
Obstfeld, M. (1996). Models of Currency Crises with Self Fulfilling Features. European Economic Review.40, 1037-1047.
Obstfeld, M., & Taylor, A. (2003). Sovereign Risk, Credibility and the Gold Standard: 1870-1913 versus 1925-1931. Economic Journal, 113, 1-35.
Reinhart, C. M. (2010).This Time Is Different Chart Book: Country Histories on Debt, Default and Financial Crises’ NBER Working Paper no 15815.
Reinhart, C. M. and K. Rogoff (2009), This Time is Different: Eight Centuries of Financial Folly. Princeton: Princeton University Press.
Reinhart, C. M., & Rogoff, K. S. (2008). Is the 2007 US Subprime Crisis So Different: An International Historical Comparison?. American Economic Review, 98, 339-344.
Reis, J. (1992).The Gold Standard in Portugal, 1854-1891. Mimeo Universidale Nova de Lisbon.
Saunders, A., & Cornett, M. (2003). Financial Institutions Management: A Risk Management Approach (4th edition). McGraw Hill.
Schularick, M., & Taylor, A. (2009). Credit Booms Gone Bust: Monetary Policy, Leverage Cycles and Financial Crises 1870-2008. NBER Working Paper 15512.
Sheikh, K., & Heijmans, P. (2004). New Developments in Country Risk: the New Anatomy of Crises, Rising Public Debt Burdens and Basel II. In S. Wilkin (Ed.). Country and Political Risk: Practical Insights for Global Finance. Risk Books, 3, 37-56.
Stone, I. (1999). The Global Export of Capital from Great Britain 1860-1914. London: Macmillan.
Twain, M. (1940). Eruption: Hitherto Unpublished Paper about Men and Events, edited and introduced by Bernard Devoto (1st edition). New York, First Printing Harper and Brothers.
Van Agtmael, A. (2006). The Emerging Market Century. Free Press.
Appendix
Table A
Dates of Crises Events
Block I
Notes: 1 the data series starts as late as 1885. Debt crises= we include only the year of debt repudiation or debt restructuring. Negotiations years are excluded. Currency crises= *definition 1: we refer to the year of a heavy speculation attack on the domestic money causing abnormal fluctuations of the exchange rate or the year of a nominal currency depreciation or devaluation;**definition 2: following Catao’s working assumption (2006) we define a currency crisis (i.e. currency drop) as nominal exchange rate depreciation greater than at least one standard deviation of the annual percentage change of the spot exchange rate of the drachma against sterling over the sample period. We also assume that this depreciation is not fully reversed within a three year window. Banking crises= we include the years of a wide spread bank panic and/or the years of massive bank failures. Isolated events are not taken into account. ***Following Catao’s (2006) working assumption, we define sudden stop as a drop from peak to trough of no less than two standard deviations of the deviations of the respective series from a linear trend and/or any drop that exceeds 3 per cent of GDP over a period shorter than four years.
Note. 3 gold hoarding; 4 speculation attacks resulting in an abnormal increase in the interest rates to defend currency; 5 an isolated event of a bank failure due to a corporate governance failure (fraud risk); it is not therefore taken into account; 6 bail out, rescue package.
Data Appendix
For the period from 1846 to 1914, bank loans to firms and households refer to total gross loans to the private sector. The data on lending are taken directly from the Balance Sheets of the eight large commercial banks of Greece. For the period from 1950 to 2009, total bank credit includes loans to firms and households; the data are from the Bank of Greece. The data for the level of the Gross Domestic Product (1833-1938) in current prices and in constant 1914 prices, both aggregate and per capita, are from Kostelenos et al. (2007). GDP deflator is a Paasche type index of the prices of 10 products from the primary and secondary sectors and covered over 23 percent of the total value of GDP. GDP has been computed ‘based on estimates made directly using the production (value added) method, the most notable exception being the analysis of the tertiary sector, where a combination of the income method and of an indirect approach has been used’ (Kostelenos et al. 2007, p251). For the period after World War II, the data for the GDP in current prices are from the National Statistical Service of Greece. Total bank assets for the same sample of banks are also taken directly from their balance sheets. Broad money (M3) and monetary base (M0) are from Lazaretou (2010). Demand deposits, vault cash, and short-term lending (advances and discounts to traders) refer to all commercial banks for the period from 1894 to 1932 and are taken from the banks’ balance sheets. The equity to assets ratio has been computed for all commercial banks for the period between 1894 and 1932. Return on equity has been assessed for six large banks for the period between 1910 and 1932. The financial statements of those banks are my primary data source. The data for the nominal exchange rate of the drachma against the British pound are taken from Bank of Greece (2009). Total reserves, foreign exchange and metallic, are from Bank of Greece (2009). The data for total government spending and total tax revenues are from the Greek Government Budgets, Annual Reports, and concerned realized entries.