External vulnerability, balance sheet effects, and the institutional framework — Lessons from the Asian crisis

International Review of Economics and Finance 21 (2012) 16–28

Contents lists available at ScienceDirect

International Review of Economics and Finance j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i r e f

External vulnerability, balance sheet effects, and the institutional framework — Lessons from the Asian crisis Christian Mulder a, Roberto Perrelli a, Manuel Duarte Rocha a,b,c,⁎ a b c

International Monetary Fund, 700 19th Street, N.W., Washington, D.C. 20431, USA CEF.UP, Faculdade de Economia, Universidade do Porto, Rua Dr. Roberto Frias, 4200-464 Porto, Portugal Faculdade de Economia, Universidade do Porto, Rua Dr. Roberto Frias, 4200-464 Porto, Portugal

a r t i c l e

i n f o

Article history: Received 3 September 2010 Received in revised form 5 April 2011 Accepted 5 April 2011 Available online 12 April 2011 JEL classification: F31 F47 G15

a b s t r a c t This study tests the balance sheet approach of “third-generation” explanations of external crises in emerging markets, looking in particular at the 1997–98 Asian crisis. Using unique datasets, we find that corporate sector balance sheets, macroeconomic balance sheets, and the legal environment have a significant impact on the likelihood of external crises, and some of which also on the depth of crises. These indicators supplement, rather than substitute, the traditional macroeconomic variables. Predictions point to potentially large improvements in the predictive power of models that include these indicators. The results highlight the importance of sound financial structures and institutional framework, alongside prudent macro policies, in limiting external vulnerability. © 2011 Elsevier Inc. All rights reserved.

Keywords: Currency crises Institutions Corporate indicators Early warning systems Emerging markets

1. Introduction The devastating and unexpected impact of the Asian crisis of 1997–98 on the economies of the affected countries has often been attributed to weaknesses in corporate sector balance sheets. This was not a feature that existing crisis models could explain. Early models of currency crises emphasized inconsistent macroeconomic policies leading to an erosion of reserves and eventual attack on the exchange rate. A second generation of models emphasized a combination of weak fundamentals and insufficient political stamina to fight costly currency crises, which provide an invitation to speculative attacks (see Flood & Marion, 1999 for an overview). The countries at the core of the Asian crisis did not suffer from the traditional macro imbalances and weak fundamentals: inflation was low and fiscal balances about neutral. Banking and corporate weaknesses, however, were widespread. This inspired a “third generation” of external crisis models. Focusing on a balance sheet approach, this literature points to weaknesses or mismatches in the balance sheets of individual sectors of the economy and the country's aggregate balance sheet1 – such as currency mismatches, maturity mismatches (illiquidity), excessive indebtedness (leverage), and low profitability. Some of these models focus on the corporate sector's side – viz. Krugman (1999), Caballero and Krishnamurthy (2001, 2003), Aghion, Bacchetta, ⁎ Corresponding author at: Present address: Faculdade de Economia da Universidade do Porto, Rua Dr. Roberto Frias, 4200-464 Porto, Portugal. Tel.: + 351 225571100; fax: + 351 225505050. E-mail address: [email protected] (M.D. Rocha). 1 See, for example, Allen, Rosenberg, Keler, Setser, and Roubini (2002). 1059-0560/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.iref.2011.04.002

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

17

and Banerjee (2001, 2004), Bris and Koskinen (2002), and Schneider and Tornell (2004). They center on the existence of incomplete financial markets, which cause an over-reliance by firms on debt and foreign financing. Domestic financing is assumed to depend on limited domestic collateral, whose value collapses when the external financing constraint is tightened, for example due to a drop in investors' confidence. As a result, loans are called and companies sharply curtail productive new investment or close, contributing to a sharp fall in demand and further declines in the value of collaterals. These explanations for the crises have testable implications. The worse the corporate balance sheets, the more vulnerable countries are to external crises. Some studies have tried to explain the differences in corporate financial structure, notably Claessens, Djankov, and Xu (2000) and Claessens, Djankov, and Nenova (2000), by including the legal and tax regimes. Stylized facts (e.g., Stone, 2000) indeed suggest that crises with a corporate element lead to sharp falls in investment and output. However, thus far little systematic empirical research has been undertaken to examine the impact of corporate balance sheets on the incidence and depth of crises. In particular, on how well the balance sheet approach helps to explain the likelihood and depth of the Asian crisis. This paper seeks to fill this void by using corporate sector indicators, derived from individual corporations' balance sheets, to test whether these affect the likelihood and depth of external crises in emerging market economies during the 1990s, with special reference to the Asian financial crisis—which inspired the third-generation models. The importance of studying these episodes is critical since the imbalances in international trade and reserves accumulation that originated in the aftermath of the Asian crisis have been linked to the global financial crisis of 2008–09 (see Obstfeld, 2010). This study also seeks to find evidence of the role of the corporate sector through the banking sector. Using data on corporate sector balance sheets allows us to study indirectly the impact that banks' balance sheets have on external crises. We do this by testing whether a large exposure of the banking sector to the corporate sector, in combination with weak corporate indicators, enhances the vulnerability to crises. In addition, we empirically investigate the effects of macroeconomic balance sheet indicators, as well as indicators of legal regimes. The study of macroeconomic balance sheet indicators allows capturing important details that are not available at the corporate level.2 We examine the influence of macro indicators and indicators of the legal regime because the quality of lending decisions arguably plays an important role in the incidence and depth of crises and the quality of such decisions is, in turn, affected by the overall institutional framework. The potential for government bailouts is, for example, often considered a prime indicator of the quality of the decisions. We use macro institutional indicators that may be indicative of the likelihood that the government will allow the private sector to service its debts without the imposition of exchange restrictions, on the reason that the potential imposition of such restrictions may lead to uncertainty and an early withdrawal of capital. Finally, we test the importance of the implementation of corporate governance standards, which the international community has highlighted in the wake of these crises, by considering indicators of the legal regime, including creditor and shareholder rights. Such variables may affect how soundly the private sector conducts its business, and their exposure to sudden large-scale withdrawal of external finance, and more generally may be indicative of the government's desire to let the private sector be responsible for its own business. To test the validity of these indicators, we examine their explanatory power in different models, so as to have a clear test of their influence over and above existing explanations. The first is a probit model of the likelihood of external crises over the coming 24 months, which outperforms acclaimed models in the literature.3 The second is an estimation of the depth of crises, which lends itself to evaluate policies that limit the impact of systemic emerging market crises on individual countries. The paper is organized as follows. In Section 2 we discuss the theory behind the indicators and their selection. In Section 3 we present the estimation results for the crisis probability, followed by those for the crisis depth in Section 4. Section 5 concludes. 2. The candidates for crisis indicators 2.1. Corporate balance sheet indicators Good data on corporate sector balance sheets are hard to come by. They are not part of usual statistical data collection sets. To overcome this problem, for this study we have tapped a large private database, Worldscope. This database contains data on corporations that publish annual reports for a range of countries, including the most important emerging market economies from 1991 onwards. We consider the balance sheet information for the 1991–99 period in a sample of 19 countries.4 From this database we selected the non-financial corporations and computed median observations for each year. Using the median rather than the average limits the risk of data pollution by large outliers caused by misclassification, or the presence of near defunct corporations. The Worldscope database contains a vast list of corporate variables. From this list we select a core set that are common in the business literature, and which are investigated for example by Claessens, Djankov and Xu (2000) in their study of corporate financial structure. They can be classified in four categories: variables that reflect (1) the degree of financial leverage (debt over equity), (2) the maturity structure of debt financing, (3) the availability of liquidity, and (4) the profitability and cash flow of a company.

2 In particular, there are usually no data on breakdowns by residency or currency available at a micro level, as the reporting of corporations in their annual reports does not focus on breakdowns that are relevant for balance of payments analysis. 3 See Frankel and Saravelos (2010) for a recent review of the literature on early warning indicators and application to the 2008–09 global financial crisis. 4 Our sample comprises the following 19 countries: Argentina, Brazil, Chile, Colombia, Egypt, India, Indonesia, Israel, Jordan, Korea, Malaysia, Mexico, Pakistan, Peru, Philippines, South Africa, Thailand, Turkey, and Venezuela. This sample was selected on the basis of data availability – i.e. for which it was possible to obtain balance sheets data from Worldscope – being close to the sample of 23 countries used by Berg and Pattillo (1999).

18

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

The impact of these variables on the likelihood and depth of a crisis is broadly as follows. The more a corporation's financing is leveraged, the more likely it is that a shock to its assets value will severely erode its balance sheet and result in the call of loans or the postponement of profitable investment. The maturity structure, as measured for example by a high ratio of short-term debt to total debt will, ceteris paribus, exacerbate cash flow problems. The availability of assets can offset liquidity risks, and these are captured by liquidity ratios, such as the quick and current ratios. These provide an indication of the extent to which an otherwise solvent corporation could run into problems because of lack of liquid assets to meet obligations. Financial institutions are the most leveraged of the private sector. To the extent they are exposed to corporations that are themselves highly geared or vulnerable to liquidity problems, this can exacerbate crises. To estimate the impact of the corporate sector through the banking sector we use composite variables: we multiply key financial structure variables of the corporate sector by the size of lending to the corporate and other private sector borrowers. This allows us to study indirectly the impact on external crises through banks' balance sheets.5 2.2. Macro balance sheet and institutional indicators Many have argued that foreign currency borrowing by companies that do not have commensurate foreign currency receipts (such as real estate firms in Thailand) exposes these companies to large exchange rate risks. In the absence of universal micro data for corporations on the extent of foreign currency financing and revenue, we construct variables based on macroeconomic statistics. For this purpose we tapped BIS data, a universal database with a corporate breakdown of external debt.6 Using these data we construct a ratio of corporate foreign debt (to banks) over exports (as a proxy for revenue), with a higher ratio expected to make countries and corporations more vulnerable to crises. We use these BIS data also to calculate the maturity structure of overall debt (to test if short maturity exacerbates cash flow problems), and augment the corporate debt ratio with bank debt to test if bank debt exacerbates the exposure. Finally, we calculate the share of the public sector in external borrowing, to test if a lower share of government debt reduces its incentive to default — it could run against a larger private sector interest to maintain access to financing. 2.3. Legal indicators The debate on currency crises has highlighted the importance of creditor and shareholder rights and legal regimes supportive of contract enforcement for the financial structure of enterprises. Strong creditor and debtor rights may not only improve the corporate financial structure, but also lend increased confidence to creditors in the country's ability to resolve potential financial crises.7 However, the information on the nature of the legal regimes that affects the financial structure of corporations has only recently been systematically collected. Thanks to the seminal work of La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998) such variables are now more readily available. Of the many variables collected by these authors, we consider five indicators representative of corporate governance and legal environment: (1) creditor's rights; (2) shareholder's rights; (3) the ability to enforce contracts; (4) accounting standards, and (5) the origin of the legal regime. In general these variables impact the operational climate of corporations, and can be used here to test whether they have amplified the effects of weak financial structures during crises. All variables – except the origin of the legal regime – are composite indexes of specific characteristics of the legal regime. These composite legal indicators reflect a series of underlying legal and other institutional features that affect the interest of creditors and shareholders, whether contracts are enforceable, and whether published annual reports met certain accounting standards. In general, the higher the indexes the better is the corporate sector operational environment. However, a key drawback of the data is that no time series are available, but only individual observations, which reduces their discriminatory power. 3. The impact of the indicators on the likelihood of crises To test the implications and relevance of these variables and assertions on the incidence of currency crises, we use a probit model that extends the path-breaking work of Berg and Pattillo (1999, BP). BP extended the seminal works of Frankel and Rose (1996) and Kaminsky, Lizondo, and Reinhart (1998, KLR), by estimating a probit model for the occurrence of currency crises where the variables enter the model as percentiles that depend on the range of the independent variable observed in each country.8 The model uses a dummy for the existence of a currency crisis in the next 24 months as the dependent variable. This dummy assumes a value 1 (one) when a crisis occurs in this period, and 0 (zero) otherwise.9 BP's best results (Berg, Borensztein, Milesi-Ferretti, & 5 Ideally, we would have tested for the balance in the foreign currency cash flow of corporations and the quality and maturity mismatch in the banks' foreign books, but such data are not widely available. 6 We use BIS consolidated statistics as they provide breakdown of debt by sectors and maturity. 7 The legal environment has also been found to influence the performance of stock markets across countries (Chiou, Lee, & Lee, 2010). 8 BP transform the data into percentiles instead of using thresholds and thereby limit the need to impose arbitrary cut-off points. The use of percentiles enables the distinction between usual and unusual values for each series in each country — this is especially relevant when working with such a distinct sample of countries. 9 A crisis is defined as occurring when the weighted average of monthly depreciations in the exchange rate and monthly declines in reserves exceeds by more than 3 standard deviations the country mean — the weights are calculated in order to equalize the variance of the two components of the index (Berg & Pattillo, 1999).

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

19

Pattillo, 1999) are quite parsimonious: just five variables explain the probability of a crisis: short-term debt-to-reserves ratio (STD/ R), real effective exchange rate (REER), current account-to-GDP ratio (CA/GDP), change in reserves (ΔR), and change in exports growth (ΔXGR). The impact of our proposed indicators can be tested simply by adding them to the parsimonious core equation, as in Eq. (1) below,   PrðCC =1jt = T Þr = pðXit Þr′ βi + p Yjt ′ δj + pðZk Þr′ λk + εtr r

ð1Þ

where Pr(CC = 1|t = T) is the time-t probability of a currency crisis in the next 24 months, and r refers to each emerging market economy in the sample. The binomial dependent variable we use is plotted in Fig. A1. The independent variables are all transformed into percentiles by the function p(.), and allocated in three distinct groups: Xi is the group of macroeconomic variables used in the core equation; Yj is the group of corporate sector indicators, including composite, macroeconomic balance sheet and macro institutional indicators (Tables A1 and A2), and Zk is the group of time invariant indicators of the legal regime (Table A3). The group Yj is collectively referred to as balance sheet indicators.10,11 Eq. (1) is estimated using a pooled probit regression, the same methodology used by BP. 3.1. Empirical results Thanks to the provision of the data by BP we could reproduce the results of the original paper, with minor differences because of data updates. Noteworthy is that two of the five BP variables are not significant when the model is estimated for the shorter period and with a smaller sample of 19 countries, as the estimation results in Table 1 show. The fit, the Akaike information criterion, and other selection criteria are nevertheless slightly better. The two variables that become insignificant are the export growth and reserves change. These two variables are those that are the least prone to policy interpretation as they pick up negative or positive trends. The three core variables that remain highly significant in explaining the likelihood of crises are the short-term debt to reserves ratio, the external current account deficit and real effective appreciation. Interestingly, they are the same as those that were found to explain the depth of crises (see Bussière & Mulder, 1999 and the discussion below). The indicators we propose, when included in the baseline estimation, turn out to be very powerful contributors to predicting the probability of a crisis and supplement rather than substitute the core macroeconomic variables — these remain highly significant and the size of their contribution remains very similar to the original estimation. The indicators we propose add quite substantially to the explanatory power of the equation, with the McFadden R-squared increasing from 0.19 to 0.30–0.37 for the final results. This is prima facie evidence of the importance of these variables for the promulgation of crises. The following variables are highly significant in the joint estimation: corporate financial structure—financial leverage (using the book value of common equity) and short-term debt over working capital; macroeconomic balance sheet—the overall indebtedness of the corporate and banking sectors in relation to exports; macro institutional indicator—the ratio of public debt to total debt; and legal/institutional indicators—shareholder rights. The strong significance of the macroeconomic balance sheet effect, the financial leverage, and the liquidity ratio support the balance sheet explanations of external crises. Another interesting result is that the profitability and interest coverage of corporations did not significantly impact the results. This suggests that corporate balance sheet factors are more important for crises than the corporate flow variables and lends further support to the theories that focus on the balance sheet as the channel through which the corporate sector affects external crises. Interestingly also, the significance of the debt to export ratio suggests that currency mismatches play an important role in increasing the likelihood of a crisis. High debt-to-export ratios often indicate such currency mismatches which cause balance sheet effects in case of exchange rate movements. Interesting is as well that the institutional or bail-out indicators are significant: a high share of public debt increases the crisis probability, suggesting that the quality of public external debt is poorer or, given the significance of the private debt to export ratio, that it enhances the overall currency mismatch. The solid significance of shareholder rights suggests that there is an important role for minority shareholders in promoting sound corporate governance beyond promoting a sound financial structure. However, it should be borne in mind that this variable might be closely related to other sound institutional policies, which together with such shareholder rights may be responsible for the overall favorable result. 3.2. Predictive quality In this section we evaluate how well our models perform. First, we start by looking at the particular case of the Asian crisis 1997. Table 2 presents the estimated crisis probabilities for the various countries as of June 1997, right before the Asian crisis. We compare these crisis probabilities, and respective countries' rank, given by various models of early warning systems (EWS) with an actual crisis index that captures the depth of the Asian crisis. 10 Corporate balance sheet data are available annually. We transform the annual data into a monthly frequency by replicating the year-end median value along the following months until the next year-end observation becomes available. This represents the information available to markets at the time of the crisis forecast. 11 Descriptive statistics for all variables are presented in Table A4 in the Appendix A. The variability of the explanatory variables is reflected in the standard deviation values and the differences between the 10th and the 90th percentiles, during both tranquil and crisis periods. We can also see that there is a broad tendency for the explanatory variables to display higher values for the mean and the percentiles during crisis periods relative to tranquil periods.

20

Table 1 Probit estimations of external crisis probability Sign

a/

.

EWS reduced sample

Testing the proposed indicators Corporate

(1)

a/

+ + + + +

0.011 0.010 0.019 0.001 –0.002

(2) (8.1) (5.7) (10.7) (0.8) (− 1.0)

Legal

(3) 0.012 0.009 0.014

(7.5) (4.6) (7.8)

All

(4) 0.008 0.013 0.016

(6.1) (7.7) (9.5)

Selection

(5) 0.014 0.011 0.018

(8.5) (5.4) (9.6)

0.011 0.008 0.015

+ + + – – –

0.010 − 0.002 0.003 − 0.001 0.001 − 0.005

(3.2) (− 1.6) (0.8) (− 0.7) (0.4) (− 2.4)

0.0004 − 0.008 0.008 − 0.007 0.007 − 0.006

+ + –

− 0.006 0.011 0.011

(− 1.8) (3.2) (6.2)

0.003 0.006 0.016

+ + + +

0.007 0.013 0.005 –0.002

(4.4) (8.7) (3.3) (− 1.8)

– – – – − 2.83 1870 0.79 0.81 0.79 − 0.39 0.19

(− 14.6)

− 3.64 1675 0.78 0.82 0.79 − 0.38 0.27

(− 13.7)

− 3.83 1864 0.77 0.80 0.78 − 0.38 0.23

(− 15.0)

0.004 0.019 − 0.017 –0.002 − 0.113 − 0.243 0.140 0.028 − 4.51 1487 0.74 0.76 0.75 − 0.36 0.27

(− 2.8) (− 7.1) (1.9) (6.6) (− 13.0)

− 0.275 − 0.337 0.246 0.039 − 6.20 1348 0.64 0.73 0.67 − 0.31 0.40

With legal

Without legal

(6)

(7)

(4.9) (3.7) (7.3)

0.013 0.007 0.014

(7.5) (3.8) (8.0)

0.011 0.009 0.014

(6.5) (5.1) (8.2)

(0.1) (− 4.2) (1.4) (− 2.6) (2.6) (− 2.1)

0.006

(3.9)

0.006

(4.1)

0.012

(3.6)

0.014

(8.0)

0.005

(1.6)

0.004 0.018

(2.8) (9.3)

0.009 0.020

(5.6) (11.4)

(0.6) (1.2) (3.7) (1.6) (8.3) (− 3.1) (− 1.2) (− 4.9) (− 8.2) (3.0) (6.8) (− 10.0)

− 0.381

− 3.91 1688 0.66 0.69 0.67 − 0.32 0.37

(− 12.5)

(− 15.0)

− 4.99 1730 0.72 0.75 0.73 − 0.36 0.30

(− 19.4)

Heteroskedasticity-consistent (QML Huber/White) z-statistics in brackets. Type I error probability of 1% corresponds to a z-statistic of about 2.57, 5% to 1.96, and 10% to 1.64. The column Sign indicates the expected sign.

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

Macro benchmark CA/GDP REER STD/R XGR RGR Corporate FlevBV STD/LTD STD/WC CurrentR IntCover Margin Composite FlevBV ⁎ Bnk STD/WC ⁎ Bnk QuickR ⁎ Bnk Macro-balance/institutional PubD/TotD BkCorpD/X Bnk StD/TotD Legal CredRight ShareRight ContrEnfor AccStan Constant No. of observations Akaike info criterion Schwarz criterion Hannan–Quinn criterion Avg. log likelihood McFadden R-squared

Macro B/I

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

21

Table 2 24 Months crisis probabilities as of June 1997 and the depth of the Asian crisis. Country

a/

Indonesia Korea Thailand Malaysia Philippines Colombia Turkey Brazil India Pakistan South Africa Chile Mexico Jordan Peru Venezuela Argentina Crisis ranks: estimated vs. actual Linear regression (R-squared) Pearson correlation coefficient

KLR

b/

BP

c/

EWS reduced sample

Balance sheet indicators only

Balance sheet & legal indicators

Actual

Prob.

Rank

Prob.

Rank

Prob.

Rank

Prob.

Rank

Prob.

Rank

Crisis index

11% 25% 12% 17% 41% 17% 17% 37% 11% 15% 22% 11% 14% 14% 15% 14% 15%

15 3 14 6 1 7 5 2 17 8 4 15 11 13 9 11 10

26% 26% 38% 38% 22% 36% 14% 25% 14% 28% 23% 18% 6% 21% 27% 9% 15%

6 6 1 1 10 3 14 8 14 4 9 12 17 11 5 16 13

49% 46% 78% 65% 29% 60% 18% 20% 6% 62% 8% 16% 8% 9% 28% 4% 18%

5 6 1 2 7 4 10 9 16 3 14 12 14 13 8 17 10

76% 54% 81% 83% 40% 30% 6% 35% 14% 53% 18% 22% 2% 0% 17% 0% 5%

3 4 2 1 6 8 13 7 12 5 10 9 15 16 11 16 14

94% 70% 89% 69% 35% 24% 21% 28% 7% 26% 5% 5% 7% 0% 10% 1% 7%

1 3 2 4 5 8 9 6 11 7 14 14 11 17 10 16 11

94 66 48 45 33 15 14 11 9 8 6 4 1 0 −1 −4 −8

0.05 0.22

0.34 0.58

0.45 0.67

0.69 0.83

d/

Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

0.73 0.86

a/

Egypt and Israel were excluded due to lack of data. The probabilities of Kaminsky, Lizondo, and Reinhart original specification are as reported in Berg and Pattillo (1999, Table 5), divided by 100. The ranks above reported correspond to countries' classification within the reduced sample considered here. c/ The probabilities are reported as in Berg and Pattillo (1999), linear model (Table 5). The ranks above reported correspond to countries' classification within the reduced sample considered here. d/ As reported in Bussière and Mulder (1999). This crisis index covers the core 5 month crisis period. b/

We can first observe that, in general, the two models that include our proposed indicators generate more pronounced crisis probabilities (i.e. both higher and lower probabilities) than the models without these variables. We have also calculated the contribution of our proposed indicators to these crisis probabilities and found that, for the Asian countries, the balance sheet indicators contributed 18% of probability (plus 3.3% by the legal indicators)12 while the standard EWS macroeconomic variables contribute 19.9%, suggesting that our proposed indicators are as important as the latter variables in their contribution to explaining the 1997 crisis in the Asian countries. In Table 2 we can also see that there is a high correlation between the actual ranking position across countries and the crisis ranks estimated by our models.13 For example, the Pearson correlation coefficient between our two models and the actual crisis index is 83% and 86%, while in the case of the KLR and the BP models it is 22% and 58%, respectively (the R-squared is also much smaller than in our models). Moreover, our models are successful in attributing a higher crisis probability to the cluster of Asian countries and in identifying which countries are in the top-four group (in the actual crisis index). In fact, our most complete model matches the positions of Indonesia (1st), Malaysia (4th), and Philippines (5th). The predictive power increases considerably when the indicators we propose are added, but does this also mean improved predictive power? The usual focus in evaluating these models is whether they predict crises that occur and avoid predicting crises that do not occur (false alarms). In terms of global performance, a good model should balance these two components, and present a reasonable measure of efficiency. We suggest a simple measure of efficiency based on the difference between these components: the signal-to-noise balance. According to this measure, the higher the pre-crisis correctly called or the lower the number of false alarms, the higher that difference would be and the better the model. In Table 3 we present a number of statistics that have been used (Berg & Pattillo, 1999) in evaluating these models, as well as the efficiency measure. As we can see, at both 25% and 50% cut-off probabilities, both our models present a higher percentage of correctly called pre-crisis periods and a lower percentage of false alarms than the KLR and BP models (only inferior to BP at the 50% cut-off level). In fact, our models present a much higher signal-to-noise balance than all other models at both cut-off probabilities, indicating a higher efficiency in balancing the correct calls with false alarms. The BP and KLR models did not perform too well when comparing the percentage of crises called correctly with the percentage of false alarms, with the latter actually exceeding the former by a fairly wide margin especially for KLR. The choice of cut-off levels to prevent currency crises is generally arbitrary, and the relative performance of any model can vary substantially among the different choices for the cut-off probabilities. Instead of taking just one or two cut-off points (as in Berg & 12 13

Contribution compared to average observation of variables (contributions assume that other variables are at actual observed level, and cannot be added up). There is also high correlation between the actual crisis index and the crisis probabilities.

22

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

Table 3 Accuracy of crisis probabilities for the next 24 months for various EWS models. KLR

a/

BP

b/

EWS reduced sample

Balance sheet indicators only

Balance sheet & legal indicators

Cut-off probability of 50 percent A Total correctly called observations c/ B Pre-crisis periods correctly called d/ C Tranquil periods correctly called e/ D False alarms f/ E Signal-to-noise balance (B-D)

82 9 98 44 − 35

84 7 100 11 −4

85 32 98 23 9

84 46 94 34 12

84 54 92 35 18

Cut-off probability of 25 percent A Total correctly called observations B Pre-crisis periods correctly called C Tranquil periods correctly called D False alarms E Signal-to-noise balance (B-D)

77 41 85 63 − 22

78 48 84 63 − 15

78 61 82 56 5

78 75 79 52 23

82 78 83 45 34

a/

Kaminsky, Lizondo, and Reinhart original specification as reported in Berg and Pattillo (1999, Table 2). As reported in Berg and Pattillo (1999), linear model (Table 2). This model uses reserves over broad money rather than reserves over short-term debt used in the benchmark model reported in Berg et al. (1999). c/ As percentage of the total number of observations. There are two situations in which the observation is correctly called: (i) when the crisis probability exceeds the cut-off point, and a crisis happens in the next 24 months, or (ii) when the crisis probability does not exceed the cut-off point, and no crisis happens in the next 24 months. d/ As percentage of the total number of actual crisis. A crisis is correctly called when situation (i) happens. e/ As percentage of the total number of tranquil periods. A tranquil period is correctly called when the situation (ii) happens. f/ As percentage of the total number of alarms. A false alarm occurs when the estimated crisis probability exceeds the cut-off probability (i.e., alarm) but no crisis ensues within 24 months. b/

Pattillo, 1999), this evaluation of the predictive power is best conducted by examining the performance of the models through the entire range of possible cut-off points. We therefore calculate the performance of our proposed models with balance sheet indicators only and with balance sheet and legal indicators, for a range of cut-off points, and compare with the EWS reduced sample model.14 The results plotted in Fig. 1 show that the superior performance of our proposed model with balance sheet and legal indicators over the benchmark EWS is confirmed at every cut-off level. This is also the case for our proposed model with balance sheet indicators only (except at 55% cut-off level). Arguably, the signal-to-noise balance can help policymakers choose the best model and focus on the range of cut-off points that are most useful. With this in mind, we present the efficiency measures for different possible goals of the policymaker in Table 4. In this exercise, the cut-off points for each model are chosen so that the objective of the strategy is attained. This constrained optimization allows the policymaker to maximize the correct calls given a limit on false alarms or minimize false alarms for a given target on correct calls. The results show that, whatever the strategy – zero tolerance for false alarms, at most 15% of false alarms, at least 75% of pre-crisis periods correctly called and 90% of pre-crisis periods correctly called – the models including balance sheet indicators (with and without legal indicators) outperform the reduced sample benchmark EWS model in terms of the signal-tonoise balance efficiency measure. 4. The impact of the indicators on the depth of crises The results so far also suggest that the models have power in explaining crisis depth. Crisis probabilities ahead of the Asian crisis show a very high correlation (both absolute probabilities as well as ranks) with crisis depth as measured by the crisis index (Table 2). Sachs, Tornell, and Velasco (1996) explored an empirically very tractable methodology for analyzing crises that focused on the depth rather than the likelihood of crises.15 Their work only covered the period of the Mexico crisis, but was extended by Tornell (1999) to the Asian crisis, and by Bussière and Mulder (1999) to the Russian crisis episodes. The latter executed a comparison between the variables used by Sachs, Tornell, and Velasco (1996) and the variables in the BP model, and found a core set of the latter variables to far outperform in out of sample predictions. This core set of variables – the real appreciation over the previous 4 years, the current account balance over GDP, and the ratio of short-term debt to reserves – provided a very good fit and proved hard to improve upon. As noted above, these variables are the same as those that are significant in BP's benchmark equation for the shorter and smaller sample. We also apply this approach by regressing, using pooled OLS, the crisis index for different countries and crises episodes (Fig. A2) on the vulnerability indicators we analyze.16 The estimation results, presented in Table 5, support the conclusion that the 14 Due to the lack of data, we could not evaluate the performance of KLR and BP original models in cut-offs different from the 25% and 50% ones published in their works. For this purpose only these models were chosen as they are the only ones we have estimated ourselves. 15 The depth of a crisis is measured as the weighted loss in reserves and the exchange rate, where the weights are equal to the precision, so that the depreciation gets (near) full weight under a fixed exchange rate regime and the loss in reserves a high weight under a floating regime with limited intervention. 16 Our sample is comprised by the following 17 countries: Argentina, Brazil, Chile, Colombia, India, Indonesia, Jordan, Korea, Malaysia, Mexico, Pakistan, Peru, Philippines, South Africa, Thailand, Turkey, and Venezuela. These are the countries for which it was possible to obtain balance sheets data from Worldscope, out of the 20 originally used by Sachs, Tornell, and Velasco (1996).

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

23

40 Balance Sheet & Legal Indicators

35

Signal-to-noise balance (%)

30 25 20

Balance Sheet Indicators only

15 10 5

EWS Reduced Sample

0 -5 -10 1% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 99%

Cut-off level Fig. 1. Efficiency measure for the various models.

indicators we propose in the present work are also strong contributors to explaining the depth of crises. The adjusted R-squared increases from about 0.3 to 0.4, and the Akaike information criterion also improves. As in the estimation of the probability of the crisis, our proposed indicators are supplementary to the macroeconomic variables. Financial leverage and the ratio of short-term debt over working capital are similarly the core corporate sector indicators. The estimations suggest that the banking channel, as measured by banking sector credit to the private sector (bnk) in conjunction with corporate indicators, is especially important in explaining the depth of the crises and more so than for the probability of crisis. Short-term debt over working capital is especially significant when multiplied by the ratio of private credit to GDP. The size of the coefficient implies very significant exacerbating or mitigating impacts of low company liquidity. Another liquidity variable, the quick ratio, is also significant when pre-multiplied by the ratio of private credit to GDP. Indicators of legal rights are not sufficiently significant to pass normal tests, and this is a fortiori the case with macroeconomic balance sheet and institutional indicators. Interestingly, a number of standard variables that feature in corporate analysis – notably the profit margin and interest coverage ratio – were not significant in either the likelihood or depth of crisis estimations. This again suggests that balance sheet variables are more important than the flow variables. Also surprising is that the quick ratio often showed the wrong sign while being fairly significant. This may be due to measurement problems for inventories, the pricing of which is a tricky issue and leaves much room for reporting differences, especially during downturns.

Table 4 Accuracy of crisis probabilities: policy calibration of cut-off levels

Zero false alarms Cut-off probability Percentage crisis periods called correctly Signal-to-noise balance 15 percent false alarms Cut-off probability Percentage crisis periods called correctly Signal-to-noise balance 75 percent crisis periods called correctly Cut-off probability Percentage of false alarms Signal-to-noise balance 90 percent crisis periods called correctly Cut-off probability Percentage of false alarms Signal-to-noise balance a/

See Table 3 footnotes for more details.

a/

.

EWS reduced sample

Balance sheet indicators only

Balance sheet & legal indicators

71 6 6

80 7 7

80 19 19

61 15 0

64 29 14

73 24 10

18 65 10

25 52 23

27 43 32

9 75 15

13 61 29

14 54 35

24

Table 5 OLS estimations of crisis depth 1994, 1997, 1998

a/

.

BM Reduced sample

(1) − 0.33 1.43 0.23

(− 3.6) (2.6) (2.7)

Macro B/I

Legal

All

(2)

(3)

(4)

(5)

− 0.36 2.17 0.23

(− 3.4) (2.3) (2.4)

12.67 − 4.46 1.33 24.95 − 1.56 37.02 0.02 0.56 –0.18

− 0.36 1.78 0.26

− 0.54 2.01 0.27

(− 2.3) (0.7) (1.2)

(1.3) (− 0.8) (0.2) (1.2) (− 0.8) (0.6)

− 0.27 − 8.12 − 5.12 21.03 − 4.66 78.75

(− 0.02) (− 0.5) (− 0.3) (0.6) (− 1.2) (0.6)

(0.4) (2.9) (− 1.5)

0.57 − 0.54 1.52

(0.02) (− 0.1) (0.3)

6.62 0.46 − 1.57 4.74

(0.2) (0.1) (− 0.4) (0.1)

5.59 − 0.53 − 1.69 0.69 − 75.35 39 0.722 0.413 1.569 8.636 9.532 2.339 0.038

(1.3) (− 0.1) (− 0.3) (1.1) (− 0.7)

− 2.98 − 1.84 − 0.01 − 16.89

− 16.47 51 0.349 0.307 1.775 8.746 8.897 8.381 0.0001

(− 2.9)

− 53.97 49 0.539 0.386 1.643 8.804 9.306 3.512 0.002

(− 1.4)

− 5.23 51 0.365 0.262 1.860 8.876 9.179 3.536 0.004

(− 2.8) (2.6) (2.3)

− 0.49 2.25 0.21

Selection

(− 3.9) (1.8) (2.6)

(− 0.1) (− 0.8) (− 0.1) (− 0.7)

(− 0.3)

2.02 − 3.83 0.45 0.42 − 37.76 39 0.541 0.437 1.706 8.472 8.814 5.212 0.0005

(0.7) (− 1.4) (0.1) (1.9) (− 1.4)

With legal

Without legal

(6)

(7)

− 0.31 2.20 0.17

(− 2.5) (2.5) (2.1)

− 0.32 1.69 0.20

(− 2.8) (2.4) (2.3)

8.49

(1.6)

8.41

(1.6)

12.89

(1.2)

12.51

(1.2)

0.50 –0.13

(3.6) (− 2.0)

0.62 –0.17

(7.9) (− 2.7)

− 2.29

(− 1.0)

− 14.75 49 0.504 0.405 1.476 8.714 9.062 5.083 0.0002

(− 1.6)

− 22.45 49 0.495 0.408 1.493 8.692 9.001 5.734 0.0001

(− 2.6)

a/ Newey–West heteroskedastic and autocorrelation consistent (HAC) t-statistics in brackets (lag truncation = 3). Type I error probability of 1% corresponds to a t-statistic of about 2.57, 5% to 1.96, and 10% to 1.64. Egypt and Israel were excluded due to lack of data.

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

Macro benchmark REER CA/GDP STD/R Corporate FlevBV STD/LTD STD/WC CurrentR IntCover Margin Composite FlevBV ⁎ Bnk STD/WC ⁎ Bnk QuickR ⁎ Bnk Macro-Balance/Institutional PubD/TotD BkCorpD/X Bnk StD/TotD Legal CredRight ShareRight ContrEnfor AccStan Constant No. of observations R-squared Adjusted R-squared Durbin-Watson stat. Akaike info criterion Schwarz criterion F-statistic Prob(F-statistic)

Testing the proposed indicators Corporate

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

25

5. Conclusions Our empirical results lend support for balance sheet explanations of external crises—the third generation models—in particular to the case of the Asian crisis. The results suggest that the financial structure of the corporate sector and the legal regimes of countries, along with key macroeconomic variables found in other studies, play a particularly important role in predicting the probability of crises, and some of these can also help explain the depth of crises. There is a fair degree of similarity between the variables that help predict the probability and those that explain the depth of crises. Corporate balance sheet variables such as leveraged financing and a high ratio of short-term debt to working capital are key significant indicators of external vulnerability. Their impact, especially on the depth of the crisis, depends on the total size of credit by the banking sector to the economy. This suggests that corporate weaknesses are transmitted through the banking system, and that having a financially weak corporate sector is especially costly if it is financed through the banking system. The estimates of the crisis probability highlight more the role of macroeconomic balance sheet and institutional indicators. The ratio of bank and corporate debt to exports is especially significant, and its presence strengthens the impact of key corporate indicators, suggesting that crises are more likely if banks and corporations are more exposed to foreign financing in relation to exports, likely on account of currency mismatches and the balance sheet effects caused by currency movements. Moreover, the likelihood of a crisis is higher for countries whose public sector share in external borrowing is larger, although it has no impact on the depth of crises. This could mean that a small share of public borrowing reflects a stronger private constituency in favor of continued debt-service and integration in world markets or, in conjunction with the private debt ratio, points to the overall currency mismatches.17 The fact that these variables are significant in explaining the probability of crises, rather than the depth of crises, may be due to the fact that the former sample extends over a longer period over which more structural changes in these ratios have taken place. Of the various legal indicators, shareholder rights have by far the greatest impact on the probability of crises. In these estimations, several legal/institutional indicators – notably the indicator of accounting standards and contract enforcement – have incorrect signs. This may be due to the fact that time series data are not available for these series. These results also have important policy implications. They suggest that both macroeconomic and microeconomic policies affect external vulnerability: the stronger the microeconomic policies, the less are macroeconomic policies a constraint. The impact of the microeconomic policies is such that they are broadly at par with the economic fundamentals in explaining the probability of crises.18 The depth of crises during periods of systemic emerging market crises, while significantly influenced by the microeconomic policies, is still dominated by the core economic variables — notably the ratio of short-term debt to reserves. The existence of a tradeoff is in line with the approach advocated in IMF (2000) to closely examine private sector risk management in assessing reserve adequacy, and consistent with the fact that the advanced industrial countries manage well with limited reserves and sizable external short-term debt exposure.

Acknowledgments We thank two anonymous referees for their comments and suggestions. Many thanks are due in particular to Andy Berg, Catherine Pattillo, and Eduardo Borensztein for sharing their data set and for discussion of the results, Stijn Claessens for advice on the corporate sector and legal indicators, Ydahlia Metzgen and Scott Brown for their comments and support of this project, and participants at various seminars at the International Monetary Fund (IMF), the University of Oxford, and the University of Illinois. CEF.UP (Center for Economics and Finance at the University of Porto) is a research center supported by Fundação para a Ciência e a Tecnologia, which is financed by European Union and Portuguese funds. The views expressed in this article are those of the authors and do not necessarily represent those of the IMF or IMF policy. All errors are ours.

17

It could also be a proxy for political regimes that adhere to more market oriented policies. Our findings revealing the importance of balance sheets and the institutional framework for the likelihood of currency crises are, therefore, also relevant for the recent literature analyzing the role of capital controls for currency stability (e.g., Glick & Hutchison, 2011). 18

26

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

Appendix A Table A1 Corporate sector balance sheet and composite indicators. Nature

Name

Source

Financial leverage Financial leverage Debt maturity structure Debt structure Liquidity Liquidity Liquidity Profitability

FlevMV FlevBV STD/LTD STD/WC CurrentR QuickR WC/TA IntCover

Worldscope Worldscope Worldscope Worldscope Worldscope Worldscope Worldscope Worldscope

Description

Profitability

Margin

Worldscope database

Composite variables Leverage Leverage Liquidity Liquidity

FlevMV ⁎ Bnk FlevBV ⁎ Bnk STD/WC ⁎ Bnk QuickR ⁎ Bnk

database database database database database database database database

Total debt/market value of common equity Total debt/book value of common equity Short-term debt/long-term debt Short-term debt/working capital Current assets/current liabilities (current ratio) Current assets net of inventory/current liabilities (quick ratio) Net working capital/total assets Operational cash-flow (i.e. before interest and taxes)/interest payments (interest coverage) Net income before preferred dividends/net sales or revenues

Balance sheet indicators are multiplied by Bnk, i.e. domestic credit to private sector and public enterprises/GDP

Table A2 Macroeconomic-balance sheet and macro institutional indicators. Function

Name

Source

Description

Private sector dominance Repayment capacity Repayment capacity Repayment capacity Repayment capacity Private credit importance

PubD/TotD CorpD/X BkCorpD/X TotD/X StD/TotD Bnk

BIS BIS, IFS BIS, IFS BIS, IFS BIS IFS

Public debt to foreign banks/total debt to foreign banks Corporate debt to foreign banks/exports Bank and corporate debt to foreign banks/exports Total debt to foreign banks/exports Short-term debt over total debt to foreign banks Domestic credit to the private sector and public enterprises/GDP

Table A3 Legal indicators. Variable

Name

Source

Description

Legal origin Creditor rights

LegOrig CredRight

La Porta et al. (1998) La Porta et al. (1998)

Shareholder rights

ShareRight

La Porta et al. (1998)

Enforcement

ContrEnfor

La Porta et al. (1998)

Accounting standards

AccStan

La Porta et al. (1998)

Dummy for countries with common law (0) or civil/roman law (1). Index of creditors rights, from 0 (low) to 4 (high). Reflects administration of property by debtor during bankruptcy, automatic stay, restrictions on filing for debt reorganization, ranking of secured creditors in distribution of bankruptcy proceeds. Index of shareholders rights, from 0 (low) to 6 (high). Reflects aspects such as proportional representation, shares needed to call shareholder meeting, anti-director rights, voting by mail. Index of contract enforcement level. Index from 0 (low) to 10 (high) corresponding to the average of five main variables of similar scale: the efficiency of the judicial system, the rule of law, the inverted level of corruption, the risk of expropriation, and the risk of contract repudiation. Index of accounting standards based on inclusion of 90 items in annual reports for 1990, such as balance sheet detail.

Table A4 Descriptive statistics. Variable

Tranquil periods No. obs.

Crisis indicator Macro benchmark CA/GDP REER STD/R XGR RGR

Mean

Crisis periods Std. dev.

Perc.10%

Perc.90%

No. obs.

Mean

Std. dev.

Perc.10%

Perc.90%

1540

0

0

0

0

369

1

0

1

1

1540 1520 1539 1539 1540

48 51 31 55 48

28 29 24 24 26

8 8 1 21 12

86 89 63 88 82

369 369 369 351 369

67 63 52 59 58

21 29 29 21 25

32 18 14 29 23

91 92 88 87 88

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

27

Table A4 (continued) Variable

Tranquil periods No. obs.

Mean

Std. dev.

1362 1391 1379 1391 1391 1386 1391 1379 1391

41 42 47 41 43 40 41 44 45

30 29 29 28 30 29 30 29 29

1345 1369 1354 1369

46 46 44 45

1539 1536 1536 1536 1539 1518 1540 1385 1540 1540 1269

Perc.10%

Perc.90%

No. obs.

Mean

Std. dev.

Perc.10%

Perc.90%

0 0 0 0 0 0 0 0 0

89 82 89 78 88 88 88 89 88

357 357 357 357 357 357 357 357 357

50 50 43 58 44 52 49 48 44

22 24 25 27 22 23 22 26 24

22 11 11 22 11 22 22 11 11

78 88 78 88 78 78 78 78 78

31 30 28 28

6 8 7 7

90 89 85 86

357 357 357 357

55 57 66 61

21 24 26 28

26 26 24 18

83 89 95 95

49 45 44 44 46 46

29 29 29 29 30 29

5 5 5 5 5 8

89 89 86 88 89 88

369 369 369 369 369 369

41 63 66 66 50 57

29 25 24 23 26 26

5 23 26 32 17 17

83 92 94 92 83 89

1 2 3 6 54

0 2 1 1 13

0 0 1 4 38

1 4 5 7 70

369 322 369 369 339

1 2 2 6 56

0 2 1 1 10

0 0 1 4 40

1 4 4 8 65

ARG

BRA

CHI

COL

EGY

INA

IND

ISR

JOR

KOR

MAL

MEX

PAK

PER

PHI

SAF

THA

TUR

VEN

0

.5

1

0

.5

1

0

.5

1

Corporate FlevMV FlevBV STD/LTD STD/WC CurrentR QuickR WC/TA IntCover Margin Composite FlevMV ⁎ Bnk FlevBV ⁎ Bnk STD/WC ⁎ Bnk QuickR ⁎ Bnk Macro-balance/institutional PubD/TotD CorpD/X BkCorpD/X TotD/X StD/TotD Bnk Legal LegOrig CredRight ShareRight ContrEnfor AccStan

Crisis periods

50

100

0

.5

1

0

0

50

100 0

50

100 0

50

100 0

50

100

Sample month Graphs by ID 1/

List of countries: Argentina, Brazil, Chile, Colombia, Egypt, India, Indonesia, Israel, Jordan, Korea, Malaysia, Mexico, Pakistan, Peru, Philippines, South Africa, Thailand, Turkey, and Ve nezuela. Fig. A1. Dependent variable for probit estimations

1/

.

28

C. Mulder et al. / International Review of Economics and Finance 21 (2012) 16–28

100

80

60

40

20

0

-20 1994

1997

1998

1/ The

crisis index is defined as the weighted loss in reserves and the exchange rate, as reported in Bussière and Mulder (1999). Fig. A2. Crisis index for depth of different crises episodes

1/

.

References Aghion, P., Bacchetta, P., & Banerjee, A. (2001). Currency crises and monetary policy in an economy with credit constraints. European Economic Review, 45, 1121–1150. Aghion, P., Bacchetta, P., & Banerjee, A. (2004). A corporate balance-sheet approach to currency crises. Journal of Economic Theory, 119, 6–30. Allen, M., Rosenberg, C., Keler, C., Setser, B., & Roubini, N. (2002). A balance sheet approach to financial crisis. IMF Working Paper, WP/02/210, Washington, D.C: International Monetary Fund. Berg, A., Borensztein, E., Milesi-Ferretti, G. -M., & Pattillo, C. (1999). Anticipating balance of payments crises — The role of early warning systems. IMF Occasional Paper, 186, Washington, D.C: International Monetary Fund. Berg, A., & Pattillo, C. (1999). Predicting currency crises: The indicators approach and an alternative. Journal of International Money and Finance, 18, 561–586. Bris, A., & Koskinen, Y. (2002). Corporate leverage and currency crises. Journal of Financial Economics, 63, 275–310. Bussière, M., & Mulder, C. (1999). External vulnerability in emerging market economies: How high liquidity can offset weak fundamentals and the effects of contagion. IMF Working Paper, WP/99/88, Washington, D.C: International Monetary Fund. Caballero, R., & Krishnamurthy, A. (2001). International and domestic collateral constraints in a model of emerging market crises. Journal of Monetary Economics, 48, 513–548. Caballero, R., & Krishnamurthy, A. (2003). Excessive dollar debt: Financial development and underinsurance. The Journal of Finance, 58, 867–893. Chiou, W. -J. P., Lee, A. C., & Lee, C. -F. (2010). Stock return, risk, and legal environment around the world. International Review of Economics and Finance, 19, 95–105. Claessens, S., Djankov, S., & Nenova, T. (2000). Corporate risk around the world. The World Bank, Policy Research Working Paper Series, 2271, Washington, D.C: The World Bank. Claessens, S., Djankov, S., & Xu, L. C. (2000). East Asian Corporations, Before and During the Recent Financial Crisis. Mimeo, Washington, D.C.: The World Bank. Flood, R., & Marion, N. (1999). Perspectives on the recent currency crisis literature. International Journal of Finance and Economics, 4, 1–26. Frankel, J. A., & Rose, A. (1996). Currency crises in emerging markets: An empirical treatment. Journal of International Economics, 41, 351–366. Frankel, J. A., & Saravelos, G. (2010). Are leading indicators of financial crisis useful for assessing country vulnerability? Evidence from the 2008–09 global crisis. NBER Working Paper, 16047, Cambridge, MA: National Bureau of Economic Research June. Glick, R., & Hutchison, M. (2011). The illusive quest: Do international capital controls contribute to currency stability? International Review of Economics and Finance, 20, 59–70. IMF (2000). Debt- and reserve-related indicators of external vulnerability. IMF Staff Paper for Board Seminar. Washington, D.C: International Monetary Fund. Kaminsky, G., Lizondo, S., & Reinhart, C. (1998). Leading indicators of currency crises. IMF Staff Papers, 45, Washington, D.C: International Monetary Fund. 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 (pp. 31–44). Washington, D.C: International Monetary Fund. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. W. (1998). Law and finance. Journal of Political Economy, 106, 1113–1155. Obstfeld, M. (2010). The immoderate world economy. Journal of International Money and Finance, 29, 603–614. Sachs, J., Tornell, A., & Velasco, A. (1996). Financial crisis in emerging markets: The lessons from 1995. Brooking Papers on Economic Activity: I (pp. 147–215). Brookings Institution. Schneider, M., & Tornell, A. (2004). Balance sheet effects, bailout guarantees and financial crises. Review of Economic Studies, 71, 883–913. Stone, M. (2000). The corporate sector dynamics of systemic financial crises. IMF Working Paper, WP/00/114, Washington, D.C: International Monetary Fund. Tornell, A. (1999). Common fundamentals in the Tequila and Asian crises. NBER Working Paper, 7139, Cambridge, MA: National Bureau of Economic Research.