The Asian financial crisis: Lessons learnt and unlearnt
In considering whether the lessons of the Asian crisis were really absorbed, Yilmaz Akyüz observes that while governments often draw lessons from financial crises to adopt measures to prevent their recurrence, such measures are often designed to address the root causes of the last crisis rather than the next one. More importantly, they can actually become new sources of instability and crisis.
The crisis revisited
GOVERNMENTS in both mature and emerging economies no doubt draw lessons from financial crises in order to adopt measures to prevent their recurrence. However, it is often the case that such measures are designed to address the root causes of the last crisis but not the next one. More importantly, they can actually become new sources of instability and crisis. This is indeed the case in emerging economies that experienced recurrent bouts of instability and crises in the second half of the 1990s and early 2000s, including several East Asian economies hit by a virulent crisis in 1997.
The Asian crisis was caused by a combination of misguided financial policies with overreaction of foreign lenders to temporary shortfalls in international liquidity rather than structural imbalances and excessive indebtedness. It was basically a liquidity crisis but it led to insolvencies because of misguided interventions, notably by the International Monetary Fund (IMF). Like crises almost everywhere else, it was preceded by a sharp increase in capital inflows, notably short-term lending by international commercial banks to both banks and firms in the region. Most such lending was directed to non-financial private firms, but in Korea, and to a lesser extent elsewhere, the financial sector was also an important recipient of funds.
An important reason for the surge in international lending to East Asia was the ‘yield famine’ in advanced economies due to low interest rates resulting from monetary policy response to economic slowdown in the early 1990s. Higher returns in high-growth, low-risk Asian economies with a record of relatively stable exchange rates made them attractive locations for international lenders. Moral hazard also played a role. The Mexican bailout encouraged imprudent lending and governments in East Asia looked ready to bail out private debtors.
An important part of capital inflows consisted of short-term arbitrage funds seeking to profit from interest rate differentials. Further, borrowing from cheaper foreign markets allowed local firms to reduce their financing costs. Firms were also driven by eroding competitiveness and reduced export earnings resulting from the entry of low-cost producers, particularly in Korea. They reacted by augmenting investment to increase productivity and market shares. In doing so, they also added to global excess supply in several manufacturing products exported from East Asia. As in Japan in the second half of the 1980s, the rapid expansion of production capacity was a key factor in the subsequent financial difficulties. However, not all international borrowers were engaged in export activity. There was a speculative surge in the property market supported by funds borrowed abroad, notably in Thailand. Similarly, some private firms in the region invested heavily in other non-traded activities, including infrastructure.
Both borrowers and lenders underestimated the exchange rate risk because of the history of stable exchange rates in the region. Exchange rate policies in the region were widely criticised for encouraging excessive borrowing abroad and giving one-way bets to speculators. However, the question of appropriate exchange rate regime under free capital mobility remains unresolved. No regime of exchange rates can guarantee stable rates. Evidence shows that currency crises can occur under flexible exchange rates as under fixed exchange rates. When capital inflows are strong, floating could lead to nominal appreciations, pushing up real exchange rates even further. It is probable that if currencies in East Asia had been allowed to float in the first half of the 1990s when inflows were in excess of what was needed for current-account financing, the result could have been nominal appreciations, pushing up the real exchange rate further and encouraging even more inflows in pursuit of capital gains from currency movements. On the other hand, greater flexibility in times of turmoil cannot prevent a freefall, as seen in East Asia in 1997, notably in Indonesia.
The main policy error relates to domestic financial deregulation and capital account liberalisation. The East Asian economies had been urged to follow Japan on a path of liberalisation, granting financial institutions more freedom in their borrowing and lending decisions, and introducing market-based monetary policy by loosening direct regulatory controls. In Korea the departure from the postwar practice in two key areas, control over external borrowing and state guidance of private investment, played an important role. Financial liberalisation went further in Southeast Asia. Thailand created the Bangkok International Banking Facility to intermediate foreign investment in the region. In reality, it served instead as a conduit for short-term foreign lending to the liberalised Thai banks and finance houses. Leveraged lending for property funded abroad was allowed to go unchecked, leading to a boom in property markets, making borrowers highly vulnerable to a downturn in property prices, a rise in interest rates or a depreciation of the baht.
Thus, in the build-up of external financial fragility, overinvestment in manufacturing, speculative investment in property and excessive short-term borrowing in foreign currencies played a crucial role. However, contrary to the contention of mainstream ideologues at the time, the main reason for these was not that there was too much government intervention and control, but too little.
The crisis broke out in Thailand when its reserves fell rapidly as net capital inflows fell short of the funds needed to meet the widening current account deficits which had reached 8% of GDP at the end of 1996, and the Bank of Thailand could no longer maintain the currency within the fluctuation band. Other economies in the region with better balance-of-payments fundamentals suffered primarily from contagion through the exchange rate. The decision to float the baht called into question the assumption of exchange rate stability upon which the existing regional division of labour had been built. As exchange rates came under pressure, markets soon became aware of the similarities in financial vulnerability and inadequacy of reserves, and governments were forced to float.
As the panic spread to the whole region, foreign speculators selling domestic currencies were joined by domestic financial and non-financial firms seeking to escape from the squeeze on their balance sheets caused by rising domestic cash needs to service foreign debt and falling cash flows to meet them. Although Korea had not experienced a speculative property bubble, it also suffered corporate bankruptcies. The Southeast Asian scenario was repeated in Korea as domestic debtors attempted to hedge or reduce their foreign exposure, causing a downward spiral in the currency market.
Lessons and policy responses
Recurrent currency, balance-of-payments and financial crises in emerging economies in the 1990s and early 2000s, including the 1997 Asian crisis, show that at times of surges in capital inflows, vulnerabilities can emerge in at least four areas: (i) currency and maturity mismatches in private balance sheets; (ii) domestic credit, asset and spending bubbles; (iii) unsustainable currency appreciations and external deficits; and (iv) reliance on IMF assistance and policy advice rather than self-insurance against sudden stops and reversals of capital flows.
In the new millennium governments in many emerging economies have taken measures to remove vulnerabilities in some of these areas, particularly as they faced a new surge in capital inflows, first thanks to the very same credit and spending bubbles that culminated in a severe crisis in the US and Europe in 2008 and then the ultra-easy monetary policies pursued in these economies in response to the crisis. However, they also liberalised further the capital account for non-residents and residents, leading to a deeper integration into the international financial system and creating new channels of transmission of external financial shocks without removing the traditional channels.
In some respects the boom in capital flows in the new millennium has been somewhat better managed in East Asia than the boom of the 1990s, and better than in most other emerging economies. One of the first steps taken was to move to more flexible exchange rate regimes. However, unlike other emerging economies which used monetary policy primarily for inflation targeting and left the currency to the whims of capital flows, most East Asian economies avoided significant currency appreciations despite strong surges in capital inflows. They have done this not only through interventions in foreign exchange markets, but also by using market disincentives for certain types of capital inflows, such as taxes on interest income and capital gains from foreign holdings of local securities, taxes on banks’ short positions, and higher reserve requirements for non-resident local currency deposits. Korea used such measures to such an extent that the won became one of the weakest currencies in the aftermath of the 2008 crisis when there was a strong surge in capital inflows. However, it should be kept in mind that while Thailand and Malaysia had moderate real appreciations in the run-up to the 1997 crisis, this was not the case in Korea and Indonesia.
Second, East Asian economies, like many others, made strong efforts to build self-insurance by accumulating large amounts of international reserves. Unlike most other emerging economies, in East Asia reserves did not just come from capital inflows. An important part has been generated by current account surpluses – that is, they are earned reserves rather than borrowed reserves. All countries hit by the 1997 crisis made significant progress in the management of their current accounts in the new millennium, running sizeable surpluses or moderate deficits. They also sought to strengthen regional cooperation in contingency financing by extending and multilateralising the Chiang Mai Initiative.
Third, in order to reduce vulnerability to external debt crises, East Asian economies, like several emerging economies, have sought to move from debt finance to equity finance on grounds that equity liabilities are less risky and more stable. Foreign direct investment regimes have been liberalised and overall limits and sectoral caps over direct and portfolio equity inflows have been relaxed or removed. As a result, non-resident holding of equities as a percentage of market capitalisation rose sharply, reaching 30-40% and even exceeding 50% in some, compared with 15% in the US. It has been in the order of 20% in Malaysia and Indonesia, 30% in Thailand and almost 50% in Korea.
While the Korean equity market is quite deep, coming in among the top 12 globally in capitalisation, many emerging economies lack a strong local investor base. Consequently, the entry and exit of even relatively small amounts of foreign investment can result in large price swings. Even in countries with little foreign presence, such as China, equity prices have thus become highly susceptible to changes in the global risk appetite because local investors now act with a global perspective.
Fourth, since currency mismatches in balance sheets played a central role in crises in emerging economies, governments have sought to reduce their exposure to exchange rate risk by opening local bond markets to non-residents and borrowing in local currencies. In East Asia the development of regional bond markets was also seen as a solution to the problems of currency and maturity mismatches, culminating in the Asian Bond Market Initiative in 2003. Governments in several emerging economies have effectively stopped issuing foreign-currency debt in international markets. A much higher proportion of public debt held by non-residents is now issued locally, denominated in local currencies and subject to local jurisdiction.
Domestically issued local-currency debt held by non-residents is not always included in external debt statistics even though according to the conventional definition based on the residency of holders, such debt is part of external debt. Because of this discrepancy, the external debt of emerging economies is often underestimated. For instance, when the Malaysian central bank started using a new definition of external debt in 2013, including all debt owed to non-residents irrespective of currency and place of issue, total external debt of Malaysia went up from 30.5% of GDP to over 60%.
Whether in local currency or dollars, foreign ownership of debt is a key indicator of external vulnerability. For instance, the US has always been uneasy about foreign holdings of its treasuries. Around one-third of US treasuries are held by non-residents. Sovereign debt in many emerging economies is now internationalised to a greater extent. In some emerging economies the share of non-residents in local government bond markets exceeds 50%. In Indonesia and Malaysia this proportion has varied between 30% and 40% in recent years. The proportion is much higher when internationally issued government debt is included. Furthermore, unlike US treasuries, this debt is not in the hands of foreign central banks and other official bodies, but mostly in the portfolios of fickle investors.
Opening local bond markets and borrowing from non-residents in local currency have no doubt allowed the sovereign to pass the currency risk to lenders. However, it has also led to a significant exposure to interest rate shocks and loss of autonomy in controlling domestic long-term rates and heightening their sensitivity to fluctuations in debt markets of major advanced economies. It has impaired the ability of local markets to act as a ‘spare tyre’ for local borrowers at times of interruptions to access to external financing. This could prove equally or even more damaging than currency exposure in the transition of central banks of major advanced economies from low-interest to high-interest regimes and normalisation of their balance sheets.
Fifth, most emerging economies have also shifted from cross-border borrowing to local borrowing from international banks by opening up their banking sector to them. There has been a sharp increase in the share of foreign banks in emerging economies in the new millennium even though the crisis in the US and Europe resulted in a certain degree of withdrawal of their banks from these economies. In Indonesia half of banks are foreign. Korea had no foreign banks in 1996, but their number increased rapidly in the new millennium. Local currency claims of international banks on residents of emerging economies rose from 15% of their total claims in the mid-1990s to 40% on the eve of the global crisis. Local lending by foreign banks in all currencies, including foreign currencies, is now greater than their cross-border lending. As seen during the eurozone crisis, foreign banks tend to act as a conduit of financial instability in advanced economies, transmitting credit crunches from home to host countries, rather than insulating domestic credit markets from international financial shocks.
Sixth, in East Asia banking regulations and supervision have improved, promoting more prudent lending and restricting currency and maturity mismatches in bank balance sheets. However, banks now play a much less prominent role in the intermediation of international capital flows than in the 1990s. International bond issues by corporations have grown much faster than cross-border bank lending directly or through local banks. More importantly, a very large part of capital inflows now go into local securities markets, bypassing the banking system.
Seventh, the opening of domestic asset and credit markets to non-residents has been accompanied by extensive liberalisation of the capital account for residents in East Asia and elsewhere. Since the global crisis, there has been a massive accumulation of debt in dollars by non-financial corporations, mainly through international bond issues. In major emerging economies, such issues have also been made through foreign subsidiaries. These are not always repatriated and registered as capital inflows and external debt, but they have a similar impact on corporate fragility. In East Asia, dollar debt accumulation is particularly notable in Indonesia and Korea. This means that the reduction in currency mismatches in balance sheets is largely limited to the sovereign while private corporations have been building up debt in low-interest reserve currencies very much in the same way as in the 1990s.
Eighth, most Asian emerging economies have also allowed and even encouraged corporations to invest abroad and become global players, occasionally by leveraging internationally. Limits on the acquisition of foreign securities, real estate assets and deposits by individuals and institutional investors have been raised or abolished in Malaysia, Korea and Thailand. During the surges in capital inflows, a main motive for outward liberalisation was to relieve upward pressures on currencies and avoid costly interventions in foreign exchange markets. In other words, liberalisation of resident outflows was used as a substitute for restrictions over non-resident inflows.
Finally, like many others, East Asian economies have not been able to prevent ultra-easy monetary policies in the US, Europe and Japan from producing domestic credit and asset market bubbles in the past 10 years. Increases in non-financial corporate debt in Korea and Malaysia are among the fastest, between 15 and 20 percentage points of GDP, including both external and domestic debt. At around 90% of GDP, Malaysia has the highest household debt in the developing world. In Korea the ratio of household debt to GDP is higher than the ratio in the US and the average of the OECD. Thailand has also seen a significant increase in household indebtedness since 2007, by some 25 percentage points of GDP.
Vulnerability to global financial shocks
Capital account regimes of emerging economies, including in East Asia, are much more liberal today both for residents and for non-residents than in the 1990s. Foreign presence in credit, equity and debt markets has reached unprecedented levels, strongly affecting their liquidity and valuation dynamics and making them highly susceptible to global financial conditions. In the same vein, residents of these economies have increasingly become active in international financial markets as borrowers and investors. As a result, all emerging economies have now become susceptible to global financial cycles and shocks irrespective of their balance-of-payments, external debt, net foreign assets and international reserves positions although these play an important role in the way such shocks could impinge on them.
Indeed, asset and currency markets of all emerging economies, including China and other East Asian economies with strong international reserves and investment positions, were hit on several occasions in the past 10 years, starting with the collapse of Lehman Brothers in September 2008. The Lehman impact was strong but shortlived because of the easy money policy introduced in response by the US. Subsequently these markets came under pressure again during the ‘taper tantrum’ in May 2013 when the US Federal Reserve revealed its intention to start reducing its bond purchases; in October 2014 due to growing fears over global growth and the impact of an eventual rise in US interest rates; and in late 2015 on the eve of the increase in policy rates in the US for the first time in seven years. These bouts of instability did not inflict severe damage because they were temporary, shortlived dislocations caused by shifts in market sentiments without any fundamental departure from the policy of easy money. But they give strong warnings of the kind of turmoil emerging economies could face in the event of a normalisation of monetary policy in the US, hikes in interest rates and contraction in global liquidity.
After the Asian crisis, external vulnerability came to be assessed in terms of adequacy of reserves to meet short-term external debt in foreign currencies, defined as debt with a remaining maturity of up to one year. While this is the most widely used indicator of external sustainability, empirical evidence does not always show a strong correlation between pressure on reserves and short-term external debt. Often, in countries suffering large reserve losses, sources other than short-term foreign currency debt played a greater role.
Vulnerability to liquidity and currency crises is not restricted to short-term foreign currency debt. Countries with extensive foreign participation in equity, bond and deposit markets could be highly vulnerable even in the absence of high levels of short-term foreign-currency debt. Currencies can come under stress if there is a significant foreign presence in domestic deposit and securities markets and the capital account is open for residents. A rapid and generalised exit could create significant turbulence with broader macroeconomic consequences, even though losses due to declines in asset prices and currencies fall on foreign investors and mitigate the drain of reserves. Financial turmoil could be aggravated if foreign exit is accompanied by resident capital flight. Indeed resident outflows rather than exit by foreign investors may well play a leading role in the drain of reserves and currency declines as seen in some previous episodes including in the $1 trillion decline in China’s reserves during 2015-16.
Such market pressures have emerged in Malaysia from mid-2014 onwards mainly due to political instability when foreign holders of domestic securities started to unload ringgit-denominated assets. Equity and currency markets fell sharply and foreign reserves declined from over $130 billion to $97 billion by June 2015. In October 2015, the ringgit came under strong pressure, hitting the lowest level since September 1998 when it was pegged to the dollar. Although it showed some recovery subsequently, at the end of 2016 it reached below the lows seen during the turmoil in January 1998 as investors continued to download domestic assets, reacting to measures restricting currency speculation as well as prospects of higher US interest rates.
In all four East Asian countries directly hit by the 1997 crisis, international reserves now meet short-term external dollar debt. But they do not always leave much room to accommodate a sizeable and sustained exit of foreign investors from domestic securities and deposit markets and capital flight by residents. This is particularly the case in Malaysia where the margin of reserves over short-term dollar debt appears to be quite small while foreign holdings in local debt and equity markets are sizeable.1 According to the latest figures by the country’s central bank, international reserves are RM425 billion while short-term external debt, including short-term loans obtained and bonds and notes issued abroad and non-resident holdings of ringgit-denominated short-term debt securities and deposits, is about RM413 billion. However, the latter does not include long-term local-currency debt held by non-residents, which, together with large equity holdings by them, constitutes an important source of drain on reserves in the event of market stress, as seen after 2014.
By contrast, Thailand’s foreign reserves position looks comfortable, exceeding its short-term dollar debt by a large margin (some $150 billion) and providing ample buffer against a rapid exit of foreign investors from its securities markets. In Indonesia reserves exceed short-term dollar debt also by a large margin ($80 billion), but foreign holdings in its local bond and equity markets are also substantial and the current account is in deficit. The country was included among the ‘Fragile Five’ in 2013 by Morgan Stanley economists for being too dependent on unreliable foreign investment to finance growth. In Korea too, the margin is large, over $250 billion, but foreign holdings of domestic securities are more than twice as much. Thus a rapid exit from the securities market can also put pressure on the won. Indeed when Korea was hit by fallouts from the US crisis in 2008, it lost some $60 billion in reserves and was given a swap line by the US Federal Reserve.
There has been no severe financial crisis in major emerging economies in the last decade and a half when global financial conditions have remained highly favourable thanks to policies of easy money in the US, Europe and Japan. This has created addiction to cheap funds, a massive accumulation of debt and a sharp increase in foreign presence in securities, credit and property markets of emerging economies. As a result, they have become highly vulnerable to a severe and sustained reversal of these conditions.
The self-insurance they have built up in international reserves may prove inadequate in the event of a sudden stop in capital inflows, massive exit of foreign investors and capital flight by residents. Nor can they count on South-South cooperation such as the Chiang Mai Initiative Multilateralisation (CMIM) of East Asian countries and the Contingent Reserve Arrangement (CRA) of BRICS. The CMIM is inadequate in size and flawed in design – some 1.5% of total GDP of the countries involved and access beyond 30% of quotas is tied to an IMF programme. The initiative has never been called upon; during the Lehman collapse, Korea and Singapore approached, instead, the US Federal Reserve, and Indonesia secured finance with a consortium led by the World Bank. The CRA does not look very much different from the CMIM. It is designed to complement rather than substitute the existing IMF facilities. Its size is even smaller and access beyond certain limits is also tied to the conclusion of an IMF programme.
That leaves two options in the event of a serious liquidity crisis – seek assistance from the IMF and central banks of reserve-currency countries, or engineer an unorthodox response, even going beyond what Malaysia did during the 1997 crisis, bailing in international creditors and investors by introducing, inter alia, exchange restrictions and temporary debt standstills, and using selective controls in trade and finance to safeguard economic activity and employment. The East Asian countries, like most emerging economies, appear to be determined not to go to the IMF again. But, serious obstacles may be encountered in implementing unilateral heterodox measures, including creditor litigation and sanctions by creditor countries.
Consequently, deepening integration into the inherently unstable international financial system without securing multilateral mechanisms for orderly and equitable resolution of external liquidity and debt crises could prove to be very costly.
Yilmaz Akyüz is chief economist at the South Centre and former director at the United Nations Conference on Trade and Development (UNCTAD) in Geneva. This article draws on various writings of the author including a recent book, Playing with Fire: Deepened Financial Integration and Changing Vulnerabilities of the Global South (Oxford University Press, 2017); The Management of Capital Flows and Financial Vulnerability in Asia (TWN Global Economy Series No. 17, 2009, Penang); and Causes and Sources of the Asian Financial Crisis (paper presented at the Host Country Event: Symposium on Economic and Financial Recovery in Asia, at the tenth session of the United Nations Conference on Trade and Development, Bangkok, 17 February 2000; printed in TWN Global Economy Series No. 1, 2000, Penang).
1 According to the latest figures given by the Malaysian central bank on 14 July 2017, short-term external debts of banks and non-banks add up to RM398 billion. At the current exchange rate this comes to more than $90 billion while reserves are $99 billion. Since much of this private short-term debt is in dollars (or in other reserve currencies), the margin of reserves over short-term external dollar debt can be estimated to be relatively small, possibly less than $20 billion.
*Third World Resurgence No. 321, May 2017, pp 8-12