Why Was Asia Resilient? Lessons from the Past and for the Future
Author:
Mr. Phakawa Jeasakul
Search for other papers by Mr. Phakawa Jeasakul in
Current site
Google Scholar
Close
https://orcid.org/0000-0002-2071-7987
,
Cheng Hoon Lim
Search for other papers by Cheng Hoon Lim in
Current site
Google Scholar
Close
, and
Mr. Erik J. Lundback
Search for other papers by Mr. Erik J. Lundback in
Current site
Google Scholar
Close

Contributor Notes

Authors’ E-Mail Addresses: PJeasakul@imf.org; CLim@imf.org; ELundback@imf.org

Asia proved to be remarkably resilient in the face of the global financial crisis, but why was its output performance stronger than that of other regions? The paper shows that better initial conditions—in the form of lower external and financial vulnerabilities—contributed significantly to Asia’s resilience. Key pre-crisis factors included moderate credit expansion, reliance on deposit funding, enhanced bank asset quality, reduced external financing, and improved current accounts. These improvements reflected the lessons from the Asian financial crisis in the late 1990s, which helped reshape both public policies and private sector behavior. For example, several countries stepped up their use of macroprudential policies, well before they were recognized as an essential component of the financial stability toolkit. They also overhauled financial regulations and strengthened oversight of financial institutions, which helped reduce risk-taking by households and firms before the global financial crisis. Looking ahead, Asia is in the process of adjusting to more volatile external conditions and higher risk premiums. By drawing the right lessons from its pre-crisis experiences, Asia’s economies will be better equipped to address new risks associated with increased cross-border capital flows and greater integration with the rest of the world.

Abstract

Asia proved to be remarkably resilient in the face of the global financial crisis, but why was its output performance stronger than that of other regions? The paper shows that better initial conditions—in the form of lower external and financial vulnerabilities—contributed significantly to Asia’s resilience. Key pre-crisis factors included moderate credit expansion, reliance on deposit funding, enhanced bank asset quality, reduced external financing, and improved current accounts. These improvements reflected the lessons from the Asian financial crisis in the late 1990s, which helped reshape both public policies and private sector behavior. For example, several countries stepped up their use of macroprudential policies, well before they were recognized as an essential component of the financial stability toolkit. They also overhauled financial regulations and strengthened oversight of financial institutions, which helped reduce risk-taking by households and firms before the global financial crisis. Looking ahead, Asia is in the process of adjusting to more volatile external conditions and higher risk premiums. By drawing the right lessons from its pre-crisis experiences, Asia’s economies will be better equipped to address new risks associated with increased cross-border capital flows and greater integration with the rest of the world.

“Those who cannot remember the past are condemned to repeat it.”

George Santayana

Why Was Asia Resilient?: Lessons From the Past and For the Future

I. Introduction

The widespread default of U.S. subprime loans in 2007 erupted into a major financial crisis that spilled over to Asia and the rest of the world. With liquidity conditions rapidly deteriorating toward the end of October 2008 (Figure 1), banks and other financial institutions at the core of the global financial system pared back their positions overseas, forcing asset fire-sales and the withdrawal of credit lines around the world. Trade financing dried up and worldwide demand plunged. The economies of the major advanced countries ground to a halt in the fourth quarter of 2008 and Asia, along with the rest of the world, suffered a sharp decline in output in the ensuing global credit crunch (Figure 12).

Figure 1.
Figure 1.

Interbank Spreads, 2006-20111/

(Basis points)

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: Bloomberg; and IMF staff calculations.1/ Based on the difference between 3-month LIBOR and 3-month treasury bill rate.

The initial impact of the global financial crisis (GFC) recalled the trauma that engulfed Asia only a decade ago. The scale of capital outflows and the collapse in real activity in late 2008 were as large as that experienced during the height of the Asian financial crisis (AFC). The AFC culminated in a full blown financial crisis in Korea, Indonesia, Malaysia, the Philippines and Thailand, with sharp corporate and bank deleveraging, which led to significant output contractions, along with business failures, unemployment, and poverty.1 Yet, this time the outcome for Asia was different from ten years ago and from other similar economies. There was no full-blown financial crisis or sharp destructive external adjustments. Asia was relatively resilient and able to preserve systemic stability, while the euro area encountered its worst economic and financial crisis in history and other major advanced economies struggled to regain their footing. A number of economies such as Australia, China, and Indonesia continued growing throughout the GFC while the economies that saw an initial steep decline in output, such as Korea, Malaysia and Singapore, posted swift and robust recoveries.

Why was Asia more resilient and will it continue to be resilient? This paper provides an analysis of the factors underpinning Asia’s resilience during the GFC. It attempts to determine whether better initial conditions in Asia helped contribute to its resilience relative to the rest of the world during the GFC. The marginal impact of each factor is estimated using OLS regressions.2 An analysis of what went right offers important lessons for the future as the international community strives to complete the ambitious agenda on financial sector reform. Indeed, the need for Asia to reduce macro-financial vulnerabilities and to speed up regulatory reform have once again become prominent as investors pulled out of the region in reaction to news that the Fed will be ending quantitative easing.

II. The Global Financial Crisis

The GFC hit Asia hard in two ways:3

  • Capital fled the region (Figure 2). BIS-reporting banks’ cross-border claims on Asia declined by about 15 percent between the third quarter of 2008 and the first quarter of 2009. This was roughly twice the reduction experienced in other regions and surpassed the decline seen during the worst of the AFC (Appendix Table 2). The deleveraging was mostly carried out by European banks (both inside and outside the euro area), reducing their consolidated claims in virtually all Asian countries (Figure 13).4

  • Exports fell. Between September 2008 and February 2009, exports plummeted by 30 percent. This was comparable to the decline seen in other regions, and three times more severe than during the AFC. Industrial production for highly export dependent economies such as Hong Kong SAR, Malaysia, Singapore, and Thailand, was sharply lower. Even the larger economies that were not as export dependent, such as Australia, China, India and Indonesia, experienced a small decline.5

Figure 2.
Figure 2.

Asia: Financial and Trade Shocks during the Global Financial Crisis

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: BIS, Locational Banking Statistics; IMF, Direction of Trade Statistics, International Financial Statistics, and IMF staff calculations.

As a result, output shrank for Asia as a whole for two consecutive quarters (Figure 14). Real GDP in Asia, excluding China and India, fell by 11 percent in the fourth quarter of 2008 and a further 8 percent in the first quarter of 2009, on an annualized basis. The initial drop in output was more acute than that in other regions, including those countries that were at the core of the crisis.

The external shocks were also followed by significant exchange rate depreciation in several countries. In particular, countries running current account deficits and accumulating foreign liabilities in their banking systems, or with relatively open capital markets (notably, Australia, Korea, India, Indonesia, and New Zealand) were most affected by the initial shockwave and in many cases saw their nominal effective exchange rates depreciate significantly. The depreciations were nevertheless smaller and smoother than during the AFC, as were the current account adjustments (Figure 3). The relatively modest adjustment in the exchange rate and current account reflected external imbalances that were not as large as during the AFC.

Figure 3.
Figure 3.

Asia: Current Account Adjustments and Exchange Rate Movements

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: IMF, Information Notice System and World Economic Outlook databases; and IMF staff calculations.

To mitigate the effects of the GFC shock, Asia quickly adopted monetary and fiscal stimulus. Low inflation and government debt, credible monetary policies, and fiscal surpluses meant that many countries in Asia were able to mount a decisive and comprehensive countercyclical policy response. Aggressive monetary easing through cuts in policy rates or reserve requirements, large fiscal stimulus packages amounting to 3-5 percentage points on a cyclically adjusted basis (Table 1), and unprecedented actions by central banks to ensure that financial systems had adequate liquidity and support,6 all served to mitigate the effects of the crisis.

Table 1.

Regions: Change in Fiscal Balance, 2007-2009

(Percentage points)

article image
Sources: IMF, Fiscal Monitor October 2012; and IMF staff calculations.

In percent of GDP

In percent of potential GDP

Starting in March 2009, Asia’s exports and economy began to revive. Strong competitive positions, in some cases aided by exchange rate depreciation, and robust import demand within the region, in particular from China, helped propel Asian exports to above the pre-crisis level by the second quarter of 2010, three quarters ahead of the full recovery of global trade (Figure 15). By end 2010 and through 2011, Asia was driving the global recovery, the first time that Asia’s contribution to a global recovery had outstripped that of other regions. At the same time, unemployment rates were contained and only in New Zealand did rates increase significantly, by about 2½ percentage points.7

Capital inflows also quickly resumed. Within just 6 quarters, BIS-reporting banks’ cross-border claims on Asia rose from a trough in early 2009 to a new high in late 2010, much shorter than the decade it took to recover from the AFC. By end-2011, worldwide cross-border claims of international banks on Asia had reached another new high even though claims on other parts of the world were still 10 percent lower than the pre-crisis level. Korea was the only country with claims remaining about 20 percent below its pre-crisis level, but this reflected the reduced reliance of banks on foreign wholesale funding following macroprudential measures implemented to rein in foreign currency risk.

Asia’s financial sector remained stable through the crisis. There were no full blown banking or balance of payment crises, or very sharp current account adjustments, as there were during the AFC (Figure 4).8 Most countries did not see a significant deterioration in financial soundness indicators and Asian currencies strengthened relative to their pre-crisis levels in a short span of time. As the banking systems in Asia remained stable, private domestic credit generally held up and resident deposits continued to gradually expand in response to concerted efforts by banks to change their funding composition away from wholesale to retail. In countries such as Australia, Korea and New Zealand that relied more on wholesale funding, domestic credit growth dropped quite substantially, in part because of a significant decline in credit demand by households and businesses. In Japan, the stagnant credit growth was emblematic of a decade long structural trend.

Figure 4.
Figure 4.

Banking Crises and Credit Growth

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: IMF, International Financial Statistics; Laeven and Valencia (2012); and IMF staff calculations.

III. Factors Underpinning Asia’s Resilience

Macroeconomic policies in Asia were sound going into the GFC but this was also true for many of the countries at the epicenter of the crisis. Asia and many of the advanced economies shared common attributes: low inflation, fiscal surpluses or small deficits except for Greece, and public debt that was largely below 60 percent of GDP (Figure 16). These indicators were not exceptionally more favorable in Asia compared with the rest of the world. Thus, while credible and consistent macroeconomic policies were necessary to support a stable economy, they alone were not sufficient to explain cross country differences in resilience during the GFC.

Asia was more resilient than other regions because of relatively low financial and external vulnerabilities, the result of a decade of financial and structural reform following the AFC. The AFC had illustrated how financial imbalances in banks and corporations can become a threat to overall macroeconomic stability, thus raising the awareness of their close interdependence, and the need for financial sector reform to reduce key vulnerabilities (Box 1). As a result, policymakers in Asia adopted a more proactive and intensive approach to banking supervision to ensure that idiosyncratic risks were closely monitored and addressed, while making use of macroprudential instruments to respond to emerging systemic risks in the financial sector. Indeed, Asia was ahead of many other countries in deploying instruments, such as restrictions on loan-to-value, debt-to-income and credit growth, limits on currency and maturity mismatches, and adjustments in reserve requirements and risk weights to contain excessive financial imbalances.9 Equally important, the move to a more flexible exchange rate policy in the region also acted as an effective shock absorber (Appendix Table 3).

In particular, Asia’s financial sector was not highly leveraged or dependent on wholesale funding for expansion, and perceived to be well regulated.

  • Private sector credit growth in the five years preceding the GFC was modest compared to the advanced countries, and in some cases the expansion was slower than the rate of growth in GDP (Figure 17). In the countries most affected by the AFC, there was a long period of post-crisis deleveraging when private credit fell by some 40-50 percent of GDP.10

  • Most Asian countries had credit to deposit ratios of around or below 100 percent before the GFC in stark contrast to many advanced countries and emerging European economies.

  • The banking system was well capitalized with high quality (common equity tier-1) capital. Nonperforming loans were a small share of total loans and, given tight regulatory restrictions, exposure to subprime loans or structured credit products, such as collateralized debt obligations was minimal (Figure 5).

Figure 5.
Figure 5.

Subprime Market Losses and Writedowns

(US$ billions)

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: Bloomberg; and IMF staff calculations.
  • Banks’ predominantly held a net foreign asset position and were less vulnerable to external shocks during the GFC. They did not face a sustained debilitating funding squeeze, and were also able to easily absorb losses on their overseas securities portfolios, including Japanese banks, which had the highest leverage and exposure to Lehman Brothers (US$4.2 billion).11 Significant steps were taken to strengthen prudential regulation and supervision, particularly among the ASEAN-5 countries following the AFC (Appendix 2).

In addition, Asian economies were not highly exposed to short-term external liabilities, and rollover risk was low, thus enabling the falling exchange rate to absorb the initial shock.

  • Most countries were running current account surpluses, ranging from 2 percent of GDP to more than 25 percent, limiting the need for net foreign financing.12

  • In 2007, net external debt was about 25 percent of GDP or less except for Australia and New Zealand.

  • The ratio of short-term external debt to foreign reserves was below 100 percent in the nine Asian countries for which data is published; only Hong Kong SAR had a ratio above 100 percent, albeit it had a strong overall net foreign asset position.13

  • The stock of foreign reserves had been built up significantly after the AFC for countries with floating as well as managed or pegged arrangements, and speculative attacks were largely avoided.

Within Asia, the economies most affected by contagion from the GFC were precisely those with faster credit growth and larger current account deficits, funded by foreign currency borrowing. Australia, Korea, and New Zealand fell into this group (Figure 17). However, in these economies, the worst effects were avoided owing to effective policy responses, including central bank liquidity support. The exchange rates were allowed to depreciate vis-à-vis the U.S. dollar—by about 30 percent in Korea, which helped turn its current account from a deficit in the third quarter of 2008 to a surplus in early 2009, and by about 25 percent in Australia.14 Furthermore, Australia, Korea, and New Zealand established U.S. dollar swap lines with the Federal Reserve, which in the case of Korea was in addition to its already large pool of foreign reserves, to further bolster market confidence. New Zealand banks received funding support from their Australian parents, which were perceived to be sound. Eventually, markets participants were able to distinguish the strong macroeconomic fundamentals in these countries and their economies recovered quickly soon after.

Another important, albeit idiosyncratic, factor contributing to Asia’s resilience was its regional dynamism and China’s strong economic performance. In the years before the GFC, Asia was one of the most dynamic and fast growing regions in the world. As recovery took hold, individual economies helped each other sustain the growth momentum. In particular China’s robust growth supported demand for Asian countries’ export, including through higher commodity prices that benefitted countries such as Australia, Indonesia, and New Zealand. Moreover, in India and Indonesia, domestic demand held up fairly well and supported growth. In a scatter plot showing the relationship between real GDP growth and the GDP growth of trading partners during 2009-10, Asia gained from having partners that were strong, especially China, with growth 1.2 percentage points higher for each additional percentage point increase in the growth of its partners (Figure 6).

Figure 6.
Figure 6.

Real GDP Growth and Partners’ Economic Performance, 2009-2010

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: IMF, World Economic Outlook database; and IMF staff calculations.

IV. Estimating Asia’s Resilience During the GFC

A. How Resilient Was Asia?

By mid 2012, output in most Asian countries was significantly higher than their pre-crisis levels, a sharp contrast to some other parts of the world (Figure 7). In this section, we estimate the “magnitude” of Asia’s resiliency and the relative importance of financial and external strength, as noted in the section above, in underpinning the recovery from the GFC.

Figure 7.
Figure 7.

Real GDP, 2012H1

(Percentage points, relative to pre-crisis peak)

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: IMF, World Economic Outlook database; and IMF staff calculations.

To assess economic resilience, a measure of output performance is derived based on the depth and length of output decline to give a cumulative output loss/gain for each country:

  • Depth: The extent of the decline in real GDP from the pre-crisis peak to the trough (Figure 8). The pre-crisis peak is the highest output level that occurs between 2007Q3–and 2010Q4. The distance from peak to trough captures the immediate impact of the GFC shocks on output. In cases of no output contraction, the depth is set to zero.

  • Length: The time it takes for output to recover to the 2008Q3 level with a cut-off point at 2010Q4, to focus directly on the effects of the GFC. Thus, the longest recovery period is restricted to 9 quarters.

  • Loss: The cumulative loss of output relative to the 2008Q3 level. In cases where the economy expanded rapidly after an initial drop, or did not contract at all, there would be a positive cumulative output gain (which is a negative output loss).

  • Economic resilience minimizes the output loss for each country.

Figure 8.
Figure 8.

Measuring Output Performance

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Using the measure of output performance and accounting for the size of the trade and financial shocks, Asia is shown to be resilient in a sample of 82 countries.15 The result is obtained using a cross-country regression of output performance (y) on financial shock (fs) defined as the change in the cross border claims of international banks, and trade shock (ts) defined as the change in merchandise exports between 2008Q4 to 2009Q1,16 and a dummy variable (AD) to represent Asian countries:

y i = α + β i Fs i + β i ts i + A D i .

On average, the positive “Asia effect” meant that countries in the region had a shallower decline in output, a faster recovery, and a smaller cumulative output loss compared to other regions (Figure 9).17 Specifically, comparing Asia to the rest of the world, the depth of the output decline was smaller by 2.8 percentage points, the recovery to the 2008Q3 output levels was more than 3 quarters quicker, and the cumulative output loss was lower by 11 percent of annualized 2008Q3 GDP. This is in contrast to the results for Europe, where the trade and financial shocks caused significant output losses and some countries have yet to emerge from recession five years after the onset of the GFC.

Figure 9.
Figure 9.

Output Performance and Shocks during the Global Financial Crisis

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: Haver Analytics; IMF, International Financial Statistics and World Economic Outlook database; and IMF staff calculations.1/ Excluding Singapore due to its significant decline in mechandise exports (65 percent of GDP).2/ Excluding countries experienced extremely large change in cross-border claims of international banks.3/ Output performance is based on the period 2008Q3-2010Q4. Trade and financial shocks cover the period 2008Q4-2009Q1.

B. Testing Asia’s Financial and External Strength

The next step is to evaluate how much financial and external strength contributed to Asia’s resilience and explained cross-country differences in output performance during the GFC.18 To examine this relationship, a series of bivariate regressions were estimated:

y i = α + βx i + i ,

where yt is output performance in country i (i.e. depth, length or loss), and x¿ is an indicator capturing the degree of financial or external vulnerabilities in country i prior to the GFC (Appendix 3). Five indicators were used to capture financial vulnerabilities: (i) credit growth with respect to GDP; (ii) the change in the ratio of non-performing loans to total loans to measure asset quality; (iii) the ratio of private credit to total deposits as a proxy for the economy’s reliance on wholesale funding; (iv) the change in total capital to assets as a measure of banking system leverage; and (v) the net foreign assets of the banking system to measure its exposure to foreign funding risk.19 The indicators used to capture external vulnerabilities measure the country’s reliance on external financing in one form or another: (i) gross external debt; (ii) net external debt; (iii) the ratio of foreign reserves to short-term external debt; (iv) the current account deficit; (v) net non-direct investment inflows. Of course, financial and external vulnerabilities are also interconnected: for example, current account deficits generate external borrowing, which is typically denominated in foreign currency and intermediated through the banking system. Table 2 and Appendix Table 7 show the pre-GFC initial conditions of these vulnerability indicators. On this basis, Asia was in a much stronger position than Europe and Western Hemisphere, including Latin America.

Table 2.

Initial Conditions

article image
Source: IMF staff estimates.

Based on initial conditions of all countries in each region.

The empirical evidence suggests that countries with less prior financial and external vulnerabilities had more favorable output performance during the GFC.20

  • Most of the estimated coefficients are statistically significant and have the expected sign, including financial soundness indicators such as capital and non-performing loans. Countries with a slower pace of expansion in credit to GDP and less accumulation of external debt, particularly of the non-direct investment kind, also fare better in output performance (Table 3).21

  • On a weighted average basis, the drop in output in Asia was 3.8 and 1.4 percentage points lower compared to Europe and Western Hemisphere, respectively; the recovery period was shorter by about 5.4 and 5 quarters; and the cumulative output loss was lower by 21 and 16 percent of annualized 2008Q3 GDP.

  • With respect to financial vulnerabilities, credit growth appears as the most important factor in terms of explaining the difference in output performance across countries. The modest pre-GFC credit expansion in Asia can account for about 40 percent of the difference in the cumulative output loss vis-à-vis non-Asian economies (Figure 10 and Table 4), and 37 and 24 percent vis-à-vis Europe and Western Hemisphere, respectively.

  • With respect to external vulnerabilities, the level of external debt and foreign reserves matter the most, individually able to explain about 40-45 percent of the difference in the cumulative output loss between Asia and non-Asian economies.

  • The combined contribution of financial and external vulnerabilities to output performance can be estimated using a multivariate regression analysis. Based on preferred specifications,22 lower financial and external vulnerabilities in Asia account for about 60-84 percent of the difference in the cumulative output loss vis-à-vis Europe, and 49-65 percent vis-à-vis Western Hemisphere.

Figure 10.
Figure 10.

The Relative Importance of Financial and External Vulnerabilities in Explaining Differences in Output Performance between Asia and Other Regions

Share of Cumualtive Output Loss Explained by Initial Conditions 1/

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Source: IMF estiamtes.1/ For each region, the share of cumulative output loss (relative to Asia) explained by a particular initial condition is calculated as its estimated output loss (relative to Asia) divided by its actual output loss (relative to Asia). Then, the shares of individual regions are aggregated and weighted by regional GDP. The estimated output loss is based on the bivariate regression analysis.

The economic importance of these initial conditions is larger for emerging markets and developing economies.

Table 3.

Output Performance and Initial Conditions Based on Financial and External Vulnerabilities

Estimated Bivariate Regression Coeffcients

article image
Source: IMF staff estimates. Note: ***, ** and * indicate 1, 5 and 10 percent statistical significance, respectively.
Table 4.

Estimated Impact of Initial Conditions on Output Performance

article image
Source: IMF staff estimates.

Based on initial conditions of all countries in each region.

Multivariate regressions are based on change in credit to GDP, increase in bank NPLs to total loans, credit to deposit ratio, and one of the following external vulnerability indicators: gross external debt, net external debt, and foreign reserves t short-term external debt.

Based on all 82 countries included in the regression analysis.

However, had the initial conditions for external and financial vulnerabilities been similar to those leading up to the AFC, Asia would be have been severely tested. The region would have suffered an output loss that is broadly similar to Europe and Western Hemisphere during the GFC. In particular, the cumulative output loss for Asia relative to the 2008Q3 annualized output level would have been larger by 4 percentage points given the rapid increase in credit to GDP of 1.9 percentage points annually pre-AFC, or by 5 percentage points after further incorporating the effects of rising dependence on wholesale funding and smaller foreign reserve buffers (Table 5).

Table 5.

Estimated Impact on Output Performance if Asia were to have Weak Initial Conditions similar to those before the Asian Financial Crisis1/

article image
Source: IMF staff estimates.

AFC stands for Asian financial crisis; GFC stands for global financial crisis.

All Asia include all 14 major Asian countries included in the regression analysis. AFC countries include Indonesia, Korea, Malaysia, Philippines, and Thailand, all of which were severely hit by the AFC.

Multivariate regressions are based on change in credit to GDP, credit to deposit ratio, and one of the following external vulnerability indicators: gross external debt and foreign reserves to short-term external debt.

Consistent with the existing literature on banking crisis, higher financial and external vulnerabilities before the GFC increased the likelihood of a banking crisis. Financial soundness indicators turn out to be significant—low capital to total assets, increasing non-performing loans to total loans, and high corporate leverage all increase the likelihood of a banking crisis. Likewise, higher credit growth and reliance on wholesale funding of credit also increase the likelihood of a banking crisis. External vulnerability indicators have more mixed results, but more non-FDI capital inflows and higher external debt are both associated with a higher likelihood of a banking crisis. To interpret the estimated coefficients beyond the direction of the effect, the elasticity of the probability to a change in the explanatory variables at the mean was calculated. The results show that the probability of a banking crisis is most sensitive to the credit to deposit ratio and corporate leverage, where an increase of one percent increases the probability by about 2 percent. The calculated elasticities are also quite high for the bank capital to asset ratio and the level of foreign reserves to GDP (Box 2).

V. Will Asia Continue To Be Resilient?

Asia’s resilience was again tested in 2013 as news about the Fed’s decision to taper quantitative easing led to a broad sell-off of emerging market assets. With investors discriminating among fundamentals and policies, those economies with larger financial and external vulnerabilities (e.g. higher wholesale funding or higher current account deficits) or have seen an increase in vulnerabilities, have been affected the most. Within Asia, India and Indonesia fall in this group. In both countries, current accounts deficits had widened and an increasing share was being financed by debt-creating inflows. Corporate leverage, including via direct external borrowing by unhedged firms, had also risen quickly in India.23 In the three months from end-May to end-August 2013, the Indian rupee depreciated by 24 percent and Indonesian rupiah by 15 percent, compared with 3 percent for the rest of emerging and advanced Asia (Figure 11). In the process, Indonesia lost about 14 percent and India 6 percent of their foreign reserves. Market conditions have since calmed down, but risk premia in emerging markets have risen, leading to tighter funding conditions and higher volatility in almost all emerging market currencies.

Figure 11.
Figure 11.

Exchange Rates, 2013

(National Currency per US$; May 22, 2013=100)

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: Bloomberg; and IMF staff calculations.
Figure 12.
Figure 12.

Global Financial Market Conditions

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: BIS, Locational and Consolidated Banking Statistics; Bloomberg; IMF, World Economic Outlook database; and IMF staff calculations.1/ Based on banking systems’ foreign assets to total assets in Japan, Euro area, the United Kingdom, and the United States.
Figure 13.
Figure 13.
Figure 13.

Asia: Financial and Trade Shocks during the Global Financial Crisis

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: BIS, Locational and Consolidated Banking Statistics; IMF, Direction of Trade Statistics, International Financial Statistics, and World Economic Outlook database; and IMF staff calculations.1/ International banks’ withdrawal of funds from banks in Hong Kong and Sinapore amounted to 50 and 10 percent of GDP. New Zealand banks received inflows by about 15 percent of GDP, mainly from their parents.2/ Based on balance of payments data. Trade credit received by Asia reflecrs net liability inflows, and trade credit provided by G7 countries reflects net asset outflows.
Figure 14.
Figure 14.

Asia: Initial Impact of the Global Financial Crisis

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: IMF, Information Notice System, International Financial Statistics, and World Economic Outlook database; and IMF staff calculations.1/ Including Australia, Hong Kong SAR, Japan, Korea, New Zealand, Singapore, and Taiwan Province of China.2/ Including Indonesia, Malaysia, Philippines, Thailand, and Vietnam.
Figure 15.
Figure 15.

Asia: Recovery from the Global Financial Crisis

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: BIS, Locational Banking Statistics; IMF, Direction of Trade Statistics and World Economic Outlook database; and IMF staff calculations.1/ Excluding Hong Kong SAR and Singapore given their large exports relative to GDP. They also experienced a strong rebound of exports.2/ In 2011Q4, international banks’ cross-border claims on China and Thailand were 230 and 170 percent of their 2008Q3 level, respectively.
Figure 16.
Figure 16.
Figure 16.

Key Indicators of Macroeconomic Stability

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: IMF, World Economic Outlook; and IMF staff calculations.1/ Inflation was 11 and 12 percent in Russia and Turkey, respectively. Japan has a minimal negative inflation of 0.05 percent.2/ Government debt amouted to about 180 percent of GDP in Japan, 170 percent of GDP in Lebanon.3/ In Kuwait and Norway, the fiscal balance was in surplus of 39 and 18 percent of GDP, respectively.. In Lebanon, the fiscal deficit was 11 percent of GDP.
Figure 17.
Figure 17.
Figure 17.

Key Indicators of Financial and External Vulnerabilities

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: Haver Analytics; IMF, International Financial Statistics and World Economic Outlook database; and IMF staff calculations.

Asia’s underlying macroeconomic and external conditions have shifted since the GFC. In the post-GFC environment of low interest rates and abundant liquidity, Asia has been the recipient of substantial capital inflows from investors seeking higher rates of return. These inflows have led to the accumulation of foreign liabilities and rapid increases in domestic credit and asset prices, although it should be recognized that the acceleration in credit growth comes after a period of negative growth in the run up to the GFC (Table 6). On balance, financial vulnerabilities have not changed enough to make a significant difference to resilience, positive or negative. The deterioration in external factors, however, is more broad based and consistent with a weakening in resilience. Applying the estimated coefficients to the initial conditions pre-GFC and at end-2012, shows that Asia’s output resilience has weakened relative to the pre-GFC period. If history is a guide, these emerging risks, if unchecked, could lead to a significant build-up of risk in the economy.

Table 6.

Estimated Impact on Output Performance if Asia were to have Today Initial Conditions1/

article image
Source: IMF staff estimates.

GFC stands for global financial crisis.

Multivariate regressions are based on change in credit to GDP, increase in bank NPLs to total loans, credit to deposit ration and one of the following external vulnerability indicators: gross external debt, net external debt, and foreign reserves to short-term external debt.

VI. Conclusion

Asia was hit hard by the GFC, but remained resilient as the region went into the crisis from a position of strength. A key reason for this was the experience of the AFC, which was a critical factor in shaping public sector policies and private sector behavior. The AFC triggered wide ranging financial and structural reforms that led to stronger banks and corporations. The AFC also made abundantly clear that sound macroeconomic policies, such as containing fiscal deficits and inflation, were necessary but not sufficient to ensure economic and financial prosperity; it was crucial to also take financial and external imbalances into consideration when formulating policies. The major improvements made included moderating credit expansion and reducing leverage in the financial system to a level that is more consistent with economic fundamentals, enhancing asset quality in the banking system, maintaining a more sustainable current account balance, thus containing reliance on foreign funding, and accumulating adequate foreign reserves to cushion a sudden reversal of capital inflows. These improvements were facilitated by the active use of macroprudential policies well before they were recognized as an essential component of the financial stability toolkit, and the overhaul of the financial regulation and oversight framework that forced changes in risk-taking by households and firms. These are the areas where strengthening initial conditions help improve output resilience the most. In addition, Asia benefited by being part of a fast growing region, with momentum from trading partner growth helping Asia to recover quickly.

Going forward, learning from experience must be an ongoing process that can enable Asia to fortify itself against future risks. Thus, in much the same way as with the AFC, the GFC provides an opportunity for Asia to take stock and draw lessons on how it should address new challenges, particularly from increased cross-border flows and greater integration with the rest of the world. Asia was visibly shaken by the recent turbulence in financial markets and, while it bounced back, the episode served as a timely reminder that complacency should be avoided at all cost.

The AFC and Post-Crisis Reform

The AFC was triggered in July 1997 when Thailand depleted its foreign reserves in defense of the baht which came under severe speculative attack. It was not until 1999 when real GDP recovered to its pre-crisis levels. The five Asian countries most severely and directly affected by the AFC were Indonesia, Korea, Malaysia, the Philippines, and Thailand. While the exact conditions in each country varied, the crisis was largely the result of financial imbalances due to significant reliance on foreign funding and inadequate financial sector oversight (Appendix Tables 4 and 5).

  • Strong private capital inflows fueled a domestic demand boom, built up large unhedged foreign currency liabilities, and led to highly leveraged corporations. Private domestic credit, funded by capital inflows, increased substantially and fueled an unsustainable increase in real estate and stock prices. In most cases, the capital inflows were short-term and denominated in foreign currency, increasing the vulnerability to sudden stops.

  • Tightly managed exchange rates made countries vulnerable to speculative attacks on their currencies. There were widespread currency and maturity mismatches, mainly due to the false impression that the exchange rate risk was small or nonexistent. Foreign reserves were not large enough to fully support fixed exchange rates, with short-term external debt exceeding foreign reserves in Indonesia, Korea, the Philippines, and Thailand, and, concurrently, several countries ran substantial and persistent current account deficits.

  • The banking systems were highly vulnerable as a result of insufficient oversight following earlier liberalization. Many banks had inadequate risk management practices, while banking supervision was lax and financial legislation and regulation were lagging. Imprudent bank lending and loan portfolio risk management resulted in rapid credit expansion that eventually ended with substantial nonperforming loans.

A01ufig01

Foreign Reserves, 1996-2007

(Percent of GDP)

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: IMF, International Financial Statistics,; and IMF staff calculations.1/ Foreign reserves declined by 0.8 percent of GDP during 1997-2007.

The AFC experience prompted many countries in Asia, including those less directly affected by the AFC, to embark on an ambitious financial sector reform agenda. New laws and institutions were introduced to fill identified gaps in the regulatory and supervisory framework. Failed institutions were closed while the remaining viable banks were recapitalized and their legacy nonperforming loans removed and sold to restore profitability. Risk management policies, including rules on corporate governance and disclosure, were revamped with stiffer penalties for unsafe and unsound banking practices and expanded supervisory powers to intervene and conduct regular examinations. Among advanced economies, investment in modern market infrastructure became a priority to ensure the financial sector was able to cope with the demands of a rapidly growing region. In addition, many Asian countries have build up their stock of foreign reserves to provide a cushion against adverse external shocks.24

The private sector, including banks and corporations, strengthened their balance sheets, and credit growth was contained. In the short-term, rapid balance sheet restructuring was reflected in a sharp decline in banks’ credit to the private sector, particularly in Indonesia, Malaysia, and Thailand. Over time, financial institutions cleaned up their balance sheets, improved risk management, and became more prudent in their risk taking and lending. Likewise, corporations undertook substantial deleveraging, enhanced corporate governance, and became more conservative in undertaking investment, which eventually restored corporate profitability, along with better transparency and competitiveness.

A01ufig02

Nonperforming Loans, 2005-12

(Percent of total gross loans)

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: IMF, Financial Soundness Indicators database; and IMF staff calculations.
A01ufig03

Private Domestic Credit, 1997-2007

(Percent of GDP)

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

Sources: IMF, Corporate Vulnerability Utility based on WorldScope database, and International Monetary Statistics; WorldScope database; and IMF staff calculations.
A01ufig04

Corporate Sector Deleveraging, 1997-2007

(Percent; Debt to assets)

Citation: IMF Working Papers 2014, 038; 10.5089/9781475540284.001.A001

In late 2000s, the IMF’s financial sector assessment programs (FSAPs) gave Asia high marks for strengthening its supervisory and regulatory regime in line with Basel Core Principles. By the time of the GFC, Asia’s financial sector was in good health, and the supervisory and regulatory regimes for banking, insurance and securities were well developed and on par with international standards (Appendix 4).

The Effect of Initial Conditions on the Likelihood of a Banking Crisis

A probit analysis was used to assess how financial and economic conditions before the GFC affected the probability of a banking crisis. The dependent variable takes the value one 1 for countries which had a systemic or borderline systemic banking crisis during the GFC according to the database assembled by Laeven and Valencia (2012). Financial and external variables were then individually used as the explanatory variables to examine their effect on the probability of a banking crisis (Appendix 3). Overall, financial and macro-financial indicators have more impact on than external vulnerabilities on the probability of a banking crises.25

  • Financial soundness indicators (FSIs) mostly have the expected signs and are statistically significant. Low capital to total assets, increasing non-performing loans to total loans, and high corporate leverage all increases the likelihood of a banking crisis.

  • Macro-financial indicators have the expected signs and are mostly statistically significant. Lower credit growth, lower credit to deposit ratio / reliance on non-deposit funding for credit, and lower corporate leverage decrease the likelihood of a banking crisis.

  • External vulnerability indicators have mixed results. Smaller non-FDI capital inflows, lower gross external debt, and larger foreign reserves decrease the likelihood of a banking crisis, and have statistically significant coefficients. Other external vulnerability variables have insignificant results; notably the size of the current account deficit. However, no external variable comes out significant with the wrong sign.

  • Excluding outliers changes the results slightly. Net non-FDI capital inflows are no longer statistically significant, and (ii) net external debt becomes significant with the expected sign.

Banking Crisis and Intitial Conditions Based on Financial and External Vulnerabilities

Estimated Bivariate Probit Regression Coeffcients

article image
Note: ***, ** and * indicate 1, 5 and 10 percent statistical significance, respecti

Using probability implies that the interpretation of the coefficients is not straight forward. The direction of the effect from a change in an explanatory variable on the probability of an event can be seen from the sign of the coefficient, but the marginal probability effect depends of the value of the explanatory variable. To get some sense of the effect from a change in the explanatory variables on the probability of a banking crisis, the percentage change in the probability of a banking crisis in response to a percentage change in the explanatory variables at the mean was calculated (see Demirguc-Kunt and Detragiache, 1998). The highest elasticities were found for the credit to deposit ratio and corporate leverage, where an increase of one percent increases the probability by about 2 percent, but the calculated elasticities are also quite high for the bank capital to asset ratio and the level of foreign reserves to GDP.

Effect on Probability of Banking Crisis*

article image

Percentage change in the probability in response to a percentage change in explanatory variable

Appendix I. Tables and Figures

Appendix Table 1.

Asia: GDP and Size of Financial Sector, 2007

article image
Sources: BIS, Securities Statistics; Bloomberg; Haver Analytics; IMF, International Financial Statistics and World Economic Outlook database; and Imf staff calculations.
Appendix Table 2.

Asia: Effects on Output, Inflation and Unemployment

article image
Sources: IMF, World Economic Outlook database; and IMF staff calculations.
Appendix Table 3.

Asia: Stocks

article image
Sources: BIS, Locational Banking Statistics; Haver Analytics; IMF, Information Notice System, International Financial Statistics, and World Economic Outlook database; and IMF staff calculations.

Based on a weighted average of 12-month volatility of changes in the nominal effective exchange rate, foreign reserves and money-market interest rates.

Based on the decline in cross-border claims of BIS-reporting banks.

Appendix Table 4.

Asia: Policy Buffers

article image
Sources: IMF, International Financial Statistics, Information Notice System, and World Economic Outlook database; and IMF staff calculations.

Variance of monthly exchange rate movements over 12 months: July 1996 - June 1997 and September 2007 - August 2008.

Appendix Table 5.

Asia: Credit Expansion, Bank Funding, and Corporate Sector Strength

article image
Sources: IMF, Corporate Vulernability Utility based on WorldScope database, International Financial Statistics and World Economic Outlook; and IMF staff calculations.

For Vietnam, the AFC figure is 1993-1996.

Probability of default over the one-year ahead period based on the Black-Scholes-Merton model.

Appendix Table 6.

Asia: External Vulnerabilities

article image
Sources: IMF, Information Notice System and World Economic Outlook database; and IMF staff calculations.

For Korea, the AFC figure is 1998.

Real exchange rate appreciation over 26 months: April 1995 - June 1997 and June 2006 - August 2008.