Asset Bubbles and Minsky Crises in East Asia: A Spatialized Minsky Approach Gary A.

Dymski*
Associate Professor of Economics University of California, Riverside Riverside CA 92521-0427 USA dymski@mail.ucr.edu

April 1999
ABSTRACT Some economists have suggested that the economic crises in East Asia were preceded by asset bubbles because of some nations inadequate bank supervision and myopic (or even criminal) bank lending behavior. Others have developed an alternative approach based on Hyman Minskys financial instability hypothesis: in this view, these nations have experienced boom-bust financial cycles, with asset-price booms generating Minsky crises and eventually a series of crashes. Minskys policy prescriptions for financial cycles big banks and big governmentthen present remedies for these nations ills. This paper also applies Minskys ideas to the Asian financial crisis, but modifies them to to incorporate cross-border financial flows. Ironically, this modification is based on earlier work done by Minsky himself in the mid-1960s. This extension of Minskys framework shows first that asset bubbles can arise whenever cross-border inflows of financial claims on capital exceeds the growth rate of real asset production. Since crossborder imbalances occur frequently, asset bubbles and/or asset-price deflations are everpresent tendencies. Second, shifting cross-border flows can bring about Minsky crises even without Minsky cycles. Finally, this paper shows that Minskys big bank and big government prescription for recovery is more difficult to apply when cross-border constraints matter. Japans ability to apply these remedies in the current Asian crisis is complicated in ways that do not arise in the U.S. case highlighted in Minskys work.
*

The research reported here was supported by the Pacific Rim Research Program of the University of California and by the Center for Global Partnership. The author has received helpful comments on earlier versions of this manuscript from Jim Crotty, Tokutaro Shibata, Fernando Carvalho, Fernando Costa, and Luiz Fernando de Paula, from participants in seminars at Seoul National University, Fukyong National University, Ehime University, the University of Tokyo, UNICAMP, and the Federal University of Rio de Janeiro, and from all the participants in the March 1999 Housing Finance Futures conference held at the University of California, Riverside. All remaining errors are the authors responsibility.

The most significant economic event of the era since World War II is something that has not happened: there has not been a deep and long-lasting depression. -- Hyman Minsky, Can It Happen Again? A Reprise (1982) All things I pick up Are moving, awash upon The beach at low tide -- Chiyojo (trans. from the Japanese by Buchanan, 1973, page 69) 1. Introduction In one way, the Asian financial crisis has had a paradoxical effect on the economics profession. It has been a boon, giving rise to many new questions and some new answers. For a profession which increasingly uses rhetorical puzzles as launching points for theoretical inquiry, the Asian crisis has been a godsend: why did Southeast Asian equities markets with different fundamentals all lose value? Why did the Korean won collapse? Why was Brazil dragged in? However, the Asian crisis has jarred economists confidence that they understand the issues that policy-makers and economic agents now face in the real world. For one thing, economists efforts to explain this crisis have demonstrated their models ambiguity as much as their relevance.1 Further, this crisis challenges behavioral assumptions that many economists regard as self-evident. For example, the interrelated ideas that asset prices respond to fundamentals and that financial investors are rational have been profoundly damaged. The World Bank and IMF, two institutions whose economists have largely celebrated the centrality of fundamentals in economic outcomes, have put out a series of working papers on (expectationally arational, fundamentals-defying) contagion effects in Asian currency and asset markets. Finally,

For example, Krugman (1998) explains the Asian financial crisis as the result of an asset bubble resulting from perverse credit-market outcomes; when these nations domestic banks underwrote risky projects, they ignored the possibility of failure, and thus drove asset prices out of sight. The culprits in these price spirals, and hence in the crisis, are thus banks, bank regulatory systems, and East Asian governments provisions against bank failure. The asset bubbles burst when low returns generated bank losses that governments were unable or unwilling to absorb, setting off contagion effects ranging from depositor flight to currency crisis. Before the current crisis, however, incentive-based models with asymmetric information have been used to explain why East Asias banking structures have supported (not undermined) the East Asian miracle. For example, Stiglitz and Uy wrote, financial sector interventions in East Asia incorporated design features that improved the chances of success and reduced opportunities for abuse; interventions that did not work out were dropped unhesitatingly; and policies were adapted to reflect changing economic conditions (Stiglitz and Uy, 1996: 249). 1

the continually evolving character of this crisis has revealed many economic insights to be valid only for one moment: in effect, they are written on wet sand at low tide. The persistence of economic effects that werent supposed to happen has generated new interest in alternative approaches to financial phenomena; and the ideas of Hyman P. Minsky about financial crises, long overlooked by mainstream financial economists, have in this reconsideration come to the fore. Akyuz (1998) and Kregel (1998), among others, have demonstrated the significance of Minskian financial instability in the Asian financial crisis. This paper too uses Minskys ideas to explain the role of asset bubbles in creating, transmitting, and resolving financial crises in Japan and Korea. At the same time, this paper suggests some alterations to Minskys core ideas about the emergence and resolution of asset bubbles and boom-bust financial cycles. Specifically, exploring the role of financial instability in the Japanese and Korean crises requires that Minskys core framework be spatializedthat is, it must take explicit account of spatial economic borders and foreign-exchange constraints. Minskys core financial boom-bust model is aspatial--it implicitly assumes that the financial cycle plays itself out in a closed, homogeneous economy.2 Once we have allowed for trade imbalances and factor flows across borders, it is easy to see that cross-border imbalances in capital movements can be an independent contributing source of financial fragility, aside from the financial fragility that arises due to the economys cyclical momentum. Ironically, Minsky himself suggested the building blocks of a spatially-aware approach in work predating most of his writing on financial instability per se; in effect this paper unites two strands of thought that Minsky left unconnected. We go on to argue that financial fragility and instability in an economy depends not just on the business cycle factors emphasized by Minsky, but also on the real/financial-sector tensions inherent in flows of capital across its borders. Whether asset bubbles and tendencies toward financial crisis emerge depends in part on any economys institutional mechanisms for channeling inflows of capital into investment outlets. The paper concludes by considering whether Minskys policy prescriptions can be applied in the case of the spatialized financial and economic crisis of East Asia. 2. The Minskian approach to financial fragility and asset bubbles The twin entry points for any model building on the work of Hyman Minsky are Keynesian uncertainty and real time. Keynesian uncertainty refers to the key methodological insight in Keynes General Theory (1936): the future is uncertain because events that unfold in real time do not obey pre-given probability distributions. For example, there is no way to know objectively whether asset prices are rising because the fundamentals are improving, or instead because the fundamentals are being evaluated differently. In a world with uncertainty, people cannot be fully rational because parametric information is not out there to be had. Real time, in turn, means that agents seeking to accumulate wealth must purchase and hold partially or wholly illiquid assets for a series of short periods, during which market conditions may change substantially. Because of these two characteristics of the lived environment, humans will fluctuate between being willing to make irreversible commitments, and instead seeking liquid assets that maximize flexibility and minimize exposure to risk. Uncertainty also has implications
2

The ideas developed here are indebted to the insights of Gray and Gray (1994). 2

for how agents form expectations and make decisions: for one thing, they cannot formulate expectations of asset values that are time-invariant; further, they will have some degree of disbelief in their own estimates, as well as some doubt that these estimates will guide market price formation. An immediate implication of this dual entry point is that it becomes impossible to know when a given market is subject to a bubble. For even the perceived line separating sound from unsound investments is inconstant and shifting through time; and there is no objective grounds for deciding. As Grant observes, Every loan, even if fully secured, is a kind of speculation. (5, 1992) When decisions are made under uncertainty, the passage of time always threatens to undermine well-laid plans. We make our way through this world under a condition of fundamental uncertainty that denies us knowledge about those things we most need to know. These dimensions of an uncertain economic environment immediately suggest the importance of institutional structures. As Crotty (1993) has emphasized, people construct rules of thumb to guide their behavior; when widely adopted, these conventions can create conditional stability for periods of time by channeling behaviors in a way that bounds the consequences of uncertainty. These conventions provide people with assurance that they are minimizing their exposure to mistakes and permit them to make longer-term commitments (such as purchases of land and fixed assets). Crotty goes on to observe that while conventions will bound uncertainty, they do not reduce it to probabilistic risk; indeed, the very conventions that generate conditional stability for a time eventually have behavioral consequences that undermine that apparent stability and bring us face to face again with the consequences of uncertainty.3 Minskys financial instability hypothesis. But economic events are not only molded by behavioral conventions; they unfold within aggregate structural settings that both give shape to economic dynamics and set limits on feasible outcomes. Many theorists have followed Keynes and Kalecki in observing that the relation between investment and saving can give rise to business cycles. Further, Keynes observation that economic agents will discount liquidity in periods of economic growth and rising asset values, but run to liquidity when these magnitudes turn down, suggests a financial source of cyclical fluctuation. This was exposited definitively by Hyman Minsky (1975). He showed that just as expectations vary with the state of the business cycle, balance-sheet relationships too evolve systematically during the cycle. Initially, balance sheets are robust because assets are conservatively priced and debt commitments modest; but during the course of an expansion phase, asset prices rise and debt burdens grow until finally liability commitments outpace asset returns and a downturn is induced. An economy becomes more financially fragile as an expansion proceeds, with the consequence that a period of financial instability is eventually reached: asset values fall, and a debt-deflation cycle may be unleashed. In Minskys model, the evolution of balance sheets over the cycle would be accompanied by a cyclical pattern in the relationship between the stock-market and production-cost prices of capital assets, which he terms PK and PI respectively. Minsky argues that as a boom period proceeds, an asset bubble emerges: specifically, the asset bubble arises when the ratio (PK/PI) exceeds one for an extended period of time. Figure 1 shows a stylized picture of a Minsky crisis, An example of a self-undermining rule of thumb is the notion that real estate in a given market area is at once the highest-return and safest asset one can hold. 3
3

emphasizing the key role of asset bubbles in cyclical fluctuations. The rapid pace of output growth eventually exhausts industrial capacity and forces firms to take on debt to expand production. The combination of euphoric expectations and competitive pressure drives up debt/income ratios and asset prices simultaneously; leverage is rewarded. When the collapse comes, it comes fast. Note that two financial characteristics define the difference between prosperity and crash: the collapse of the (PK/PI) ratio and the collapse of asset prices relative to liabilities. Income flows too are low relative to debt obligations. Defining L as liabilities, the crash period can be characterized as one in which PK < PI = L. In the debt-deflation state, the inability to service L leads to defaults, forcing the sale of assets on markets with few buyers, sinking PK further. Note that a contraction in real prices (a fall in PI ) will only worsen the burden of debt. We denote this crash state herein as a Minsky crisis. Minskys writings on financial dynamics suggest that a Minsky cycle (as described in Figure 1) precedes a Minsky crisis; implicitly, a Minsky crisis can be generated in Minskys work by a prior Minsky cycle. Responding to Minsky crises: the big bank and big government. Minsky crises have devastating effects on advanced economies with sophisticated and interlinked financial markets. Consequently, recurring outbreaks of these crises lead to several systematic adaptations and institutional changes. Minsky (1975, 1986) described these adaptations in his 1970s and 1980s writing on financial instability; Kregel (1998) has suggested the colorful terms big bank and big government for the two central developments. Big bank means the intervention of a central bank as a lender of last resort, providing liquidity to support banks ability to provide credit to businesses that would otherwise default on their loans and cease to function. To intervene in this way, the central bank must provide as much liquidity as is required to meet financially fragile cash-flow needs; and of equal importance, market participants must perceive the big bank as being willing to play this role. The second adaptation is the rise of big government. Minsky shows how countercyclical government spending by a public sector that constitutes a significant share of aggregate demand can check the tendency toward debt-deflation that emerges in the crisis. If liquidity provision by banks and the central bank in the face of a crisis is to stabilize cash flows, someone must spend it that is, exchange liquidity for goods; and if business firms have little capacity to spend, and households are too scared to spend, then government can spend. This blocks the panic that otherwise emerges, and creates a floor on prices that previously would have been driven (as in the Great Depression) substantially below their current levels. The importance of big government in economic dynamics was so fundamental that Minsky divided the performance of the US economy into two periods: a small government era from the end of the Civil War to the Depression; and a big government era dating from World War II. It is indeed true that government spending as a share of GDP rose from about 5% prior to World War II to 20% or more thereafter. Pollin and Dymski (1994) show that, as Minsky suggests, many key aggregate variables in the US economy demonstrate very different cyclical behavior in the two eras. In particular, cyclical downturns in the small government era were typically accompanied by falling prices, debt deflation, and high bankruptcy rates among banks and non-financial firms alike. Because of the deflationary trend of prices, the real cost of capital rose precipitously in the turndown. 4

However, in the big government era, cyclical downturns had different effects. Because of the intervention of the big bank and big government, price levels did not fall, nor did debt deflation occur. The average real cost of capital changed little in recessionary periods during the 1947-79 period. Non-financial business failure rates were vastly reduced, and bank failure rates almost nil. This depiction of the big-government era makes it seem that the problems associated with macroeconomic and financial instability should recede into insignificance in this era. However, Pollin and Dymski went on to argue that the macroeconomic effects of the big-government/bigbank era were somewhat ambiguous, and that this eras macroeconomic benefits weakened substantially after 1980. For one thing, the reduction of business and bank failure rates was a mixed blessing: failures have the effect of cleansing balance sheets and opening the field for new capital investments; and weak businesses and banks that dont fail will continue to do business. For another thing, the price of avoiding the debt-deflations that killed off so many institutions in the small-government era was a persistent upward bias in the level of prices. This persistent inflationary pressure had deleterious effects on US competitiveness, and it was among the factors undercutting political support for the large-scale countercyclical fiscal policies and for welfare state policies. Pollin and Dymski also observed that tension began to emerge in the 1980s between the Federal Reserves lender-of-last-resort role and its monetary-policy role. When pushed, the Fed function as lender of last resort overrode its macroeconomic-policy commitments. In effect, the policy steps required to maintain financial stability became increasingly divorced from the policy steps required to maintain stable macroeconomic growth. This uncoupling was due in large part to the stresses placed on US banks and savings and loans by credit-market crises which could be traced to the effects of deregulation and to the globalization of US financial relations. We concluded that these changes had the cumulative effect of reducing the effectiveness of the tools available to government to respond effectively to episodes of macroeconomic and financial instability. The combined big-bank/big-government response to episodes of turbulence was becoming instead a big-bank-only response which focused on a narrower set of trigger events. Big-government stabilization became less feasible for several reasons: taxpayer revolts, reduced welfare benefits, and shifts in political ideology. 3. Spatializing Minskys model These warning signs about the effectiveness of big-bank/big-government policies in the face of macroeconomic and financial instability will be explored further in the discussion of the East Asian crisis which follows. We first must suggest some modifications in Minskys boombust model, to permit it to be better able to encompass events observed recently in Korea, Japan, and other countries affected in the current global financial crisis. Two modifications are necessary. First, Minskys core model does not include the possibly significant difficulties associated with current-account/reserve relationships. Minskys mature model took a closed-economy approach because the U.S. was, for him, the paradigm case of a large economy with mature financial markets. But in recent experience, financial crises are especially likely in immature (Malaysia) or semi-mature (Korea); and current-account/reserve relationships are crucial in these open economies. In turn, the currency-crisis model ignores the impact of balance sheet and debt-income imbalances on target country asset prices. The second blind spot is complementary: because of the imbalances that may arise in cross-border financial 5

flows, a Minsky crisis may arise even if not preceded by a Minsky cycle. So the two modifications suggested here both imply that there are more paths to Minsky crisis than Minskys own characteristic interpretation anticipates; and these depend crucially on the particularities of historical time and place that underly financial flows. Here we focus first on some structural factors associated with the emergence of asset bubbles; then we turn to factors associated with Minsky crises. The key structural feature of asset bubbles is that they are inherently spatial as well as temporal phenomena. Specifically, we want to understand economies as bordered spaces. The idea is very general: any contiguous enclosed spatial area can be treated analytically as a bordered economy. The term is intentionally plastic: a country is a bordered economy, but so is a city within a country; a neighborhood within a city; a street within a neighborhood; and even a home along a street. Every bordered economy has a current account balance which may be unstable due to the cross-border migration of factors of production and wealth. And even when wealth and labor dont flow across one border of an economy (such as the national boundary), these factors do flow across intra-national boundaries. It is highly unlikely that the incremental resources required to fuel a given regions economic growth can all be found within its borders. Interestingly, this idea originated with Minskys own (1965) analysis of California high growth rate relative to the rest of the U.S.4 The spatial segmentation that characterizes the division of income flows and asset and liability stocks gives rise to various structural complications for bordered economies, especially when flows across spatial borders are unbalanced. In neoclassical trade and finance theory, crossborder relationships are very simple--either a nation has an export surplus, and chooses between building up foreign exchange reserves or buying foreign assets; or it has an export deficit and must allow foreigners to acquire its assets. The country in question is viewed as passively adjusting to these flow equilibria. If movements of assets and goods are not impeded, regions with high consumption propensities will sell off their assets to support their spending habits; and vice versa. The countries in this simple story are just placeholders in a global consumption/saving-portfolio balance problem. The assets are out there and the right prices for them can be found as long as exchange rates and domestic asset prices are free to adjust. But there is more to these processes. As Minsky always emphasized, asset accumulation requires liability growth. Further, assets and liabilities must balance on the flow and stock balance sheets of nations, as well as those of regions within nations. The nominal value of the capital assets emitted in a given region or nation may notindeed is unlikely to--correspond precisely to these assets production cost. This leads to the central point. There is no reason to expect nominal and real assets to correspond over time: hence asset bubbles and/or asset collapses are omnipresent in bordered economies. Bubbles thus continually affect economic outcomes; their impact is a matter of degree and perception, not kind. This alternative approach to bubbles rests both on the idea of bordered economies and on the idea that cross-border wealth flows are not real flows, but nominal flows--abstract wealth claims chasing real assets. Wealth itself is an inherently abstract concept, disembedded from the particular assets in which it is stored at any point in time. Capital mobility involves Minsky pointed out that one of the truisms of economics is that the current account balance balances (1965). On this approach to regional growth, also see Thirlwall (1980). 6
4

movement in the geographic ownership pattern of wealth claims, not a shift in the spatial locus of real capital itself. This applies whether the capital movement in question consists of a reserve shift due to household migration, of foreign direct investment, or of portfolio investment. Several factors explain why asset bubbles arise in some bordered economies but not others. The first is the pattern of regional economic growth. We consider two scenarios vis--vis the growth rates of a bordered economy and of the region surrounding it: the bordered economy can grow faster than the surrounding region; or it can grow at the same pace. The faster-growth case here encompasses two cases: first, that the surrounding region is growing, but the bordered economy is growing faster; second, that the surrounding region is declining, in which case the bordered economy qualifies as a high-growth area even if it is declining less quickly. These two cases of uneven economic growth can give rise to asset bubbles in the bordered economy. Table 1 sets out the possible relationships between asset bubbles and high-growth economies. The factors emphasized in other models--asset bubbles due to asymmetricinformation problems and to unregulated banking marketsare independent of the unevengrowth case, as shown in cell (1,2). Interest in Table 1 centers on the next three possibilities. Note first that uneven growth could arise without asset bubbles in the faster-growing area, as long as its growth rate reflected superior productive capacity. An area with more skilled labor and more capital can generate more product and income than less well-endowed neighbors, all else equal; Japan in the 1970s is a case in point. Independent of productive capacity, however, an asset bubble may form in a bordered region due to its receiving wealth inflows at a higher rate than the surrounding region (of course, these inflows could originate within this surrounding region). This upward pressure on asset prices (relative to the real cost of producing assets) is evident, for example, in the virtually global rural-to-urban migration of the past four decades. Intra-national or international immigration, like that from foreign countries and other U.S. states into California can also generate asset-price pressure of this type. We can define a boom economy as a high-growth economy whose rapid growth is attributable in part or whole to significant, sustained inflows of labor and wealth. To understand the applicability of this framework to the current financial crisis, it is important to see that the capital-inflow scenario can also work in reverse. Suppose, for example, that all economies in a given region experience capital flight, including the failure of overseas lenders to renew short-term credits. Suppose further than one bordered area within this region has less capital flight than the region as a whole. Unless this bordered economy has some other advantages in productivity or capacity, its asset prices may be viewed as too highbubble prices-in the perverse sense that they have declined less than have other capital-asset prices in the region. And even when a bordered economy has avoided declines as large as those afflicting the remainder of the region, capital flight again can turn its asset prices into bubble prices. So the asset-bubble process works two ways, on the way up and on the way down. As Table 2 illustrates, the cross-border wealth flows of interest here are typically both intra-national and inter-national. The vertical dimension of Table 2 shows that wealth (and/or population) can move into a boom region from a hinterland or rustbelt region within a nation. This is equivalent, as Minsky (1965) points out, to a movement of funds across national borders under a strict gold standard. Every region is, of course, part of a nation that itself has a capital-account balance. If exports and imports balance, no wealth moves; and if exports are 7

greater than imports, then either foreign exchange will build up or foreign assets will be purchased. In the extreme cases constituted by 1980s California and 1980s and 1990s Seoul, wealth enters the boom economy across both intra-national and inter-national borders. At the other extreme is a region systematically losing wealth to more rapidly-growing regions within a nation that is a net buyer of overseas assetsthe case of rural Japan. Even when an economy has negative foreign savings, intra-national wealth shifts can create boom and hinterland economies within the nationthe case of Tokyo. The other ambiguous case combines a national trade deficit with a regional capital (and possibly population) outflow--the case of the U.S. Midwestern rustbelt in the 1980s. There are important feedbacks between the pace of capital in- and outflows and the relative rate of growth of any economic region. Indeed, Keynes notion of a currency union can be understood here as precisely an effort to resolve imbalances by forcing adjustments on the part of countries with trade surpluses (capital outflows). Apart from wealth flows and economic growth, other spatial factors can help generate or block asset bubbles. Two such factors are singled out in Table 3. Table 3 focuses attention on factors affecting the rate of transformation of cross-border wealth flows into real capital assets. The cases of foreign direct investment and portfolio investment are differentiated. With foreign direct investment, financial inflows lead directly to the purchase or construction of real assets. Examples are phenomena as overseas firms construction of new factories and Midwestern families purchases of Los Angeles homes. Portfolio investment is more problematic because the claim is only on financial assets per se. One key factor identified here is the strength of the bordered economys intermediation capacity: the lending capacity of its banks, the strength of borrower-lender relationships, and so on. A second key is pertinent to international capital flows: the presence or absence of controls or conditions for capital inflows and outflows. Inward capital controls and a strong intermediation sector can be used to improve the financial-real capital compression rate. Inflows, even when they take the form of portfolio investments, can be longer-term in nature, generating increases in the real-asset base and multiplier benefits for other companies. Conversely, with weak banks and no controls over capital movements, the amount of financial-real compression is likely to be small, all else equal. Our discussion has emphasized capital flows across borders; but movements of people across borders adds another dimension to boom-economy dynamics. While the interrelations between wealth and population movements deserve a lengthy discussion, here we make two points. First, a boom (capital-absorbing) economy like California grows seemingly without effort most of the time because it enjoys a steady stream of migrants who bring both their labor power and their wealth along with them. So Californias productive capacity expands even as more nominal wealth pours in. Second, a boom economy that enjoys inward migration and hence labor-absorption is more capable of sustained growth than a boom economy with a stationary population. Both factors of production are expanding in the former case, but only capital--or actually, wealth claims on capital--in the latter case. A boom economy without labor in-migration must continually shift to a higher capital/labor ratio to avoid asset-bubbles or unused resources. This of course is the case of Japan in the 1960s, as it shifts its textile manufactures to Korea; it is the case of Japan in the 1970s, as it automates its factories; it is the case of Japan in the 1980s, as it builds up productive capacity and office space too much, and suffers the collapse of a bubble.

Spatializing Minskian financial fragility. In his mature work on financial instability, Minsky argued that assets may often be overvalued in the late stages of a sustained period of cyclical expansion. Minsky asserts that the more powerful the boom, the greater the potential for an asset bubble; and the more rapid the growth of a bubble, the more certain are participants that the sky will not fall, the more likely a crash. So success breeds success, and this breeds fragility and eventually the reversal of the growth-generating conditions. At the core of this financial instability hypothesis is irreversibility. Agents in an overheated economy are carrying forward asset positions acquired in the past, which tie them to balance sheet commitments in the present period. They must meet commitments in real time on those asset positions--and these units viability depends on meeting these commitments in a timely manner. The downturn is driven by both the asset and the liability side. The spatial conception developed here adds to Minskys notions about financial fragility in two ways. First, it suggests a new source of financially-fragile asset bubbles, independent of Minskys cyclical conception. Combining uncertainty with a structural approach to bordered economies, recognizing that most economies have wealth and/or labor inflows or outflows, and acknowledging that asset-augmentation capacity varies widely, leads to the conclusion that asset bubbles arise everywhere. So boom economies are commonplace, not rare; and boom economies are very likely to become bubble economies; and boom economies are therefore likely to be financially fragile. Minsky himself recognized this in a 1965 volume predating his 1975 volume on Keynes and the financial instability hypothesis. Minsky wrote: Californias growth .. requires that capital be imported. The markedly greater growth rate of California proportional to the rest of the country has swamped, and will continue to swamp, the export surplus that California would enjoy if the two grew at the same rate. The reserve base of the states banks would tend to be dissipated unless it was supplemented by reserves acquired by way of capital imports. To import capital, California must either generate liabilities of the kind accepted throughout the country, or the migrants must carry with them sufficient capital to maintain the reserve base of the regions banks. (Minsky, 1965: 99) Because of uncertainty, the notion of an excessively high or low state of expectations cannot be well-defined, for there is no stable set of fundamentals to serve as a reference point. And even if one were to point to a set of firm characteristics as fundamentals, we have argued that border balances also matter. Indeed, there can be an asset bubble due to cross-border pressures even in the absence of expectational error--that is, as a purely structural supply and demand phenomenon. So real/nominal imbalances occur all the time; they become quantitatively important when enough wealth seeks out assets in a particular place with inflexible labor supplies that this places labor/capital wealth/asset relationships are disrupted. Then Minskys point about the liability side comes into play--the liability structures that geniuses of leverage exploited in the high-growth days become nooses for aggressive-growth firms when low growth hits. The second amendment to Minskys financial instability framework involves the recognition that border constraints create broader possibilities for asset-price reversals and Minsky crisis. Table 4 summarizes the relationships developed here. The middle row of Table 4 presents the view of the Minsky cycle and crisis emphasized by Minsky himself. There are, however, two more rows. The top row illustrates the case of an economy that is in surplus vis-9

vis the rest of the world.5 Such an economy is unlikely to have an asset bubble unless its asset prices are pulled up by an overseas asset bubble. This could occur if this economys wealthowners are consistently buying financial and other assets in overseas markets; if prices there are high, domestic securities prices may rise to attract wealth-owners savings. The boom economy case occupies the bottom row. As noted, a boom economy can encourage the development of a domestic asset bubble independent of the Minsky cycle. This bubble can then be punctured, setting off a Minsky crisis, if wealth inflows slow gradually or are suddenly reversed in a capitalflight episode. Implications for big-government, big-bank policies. Spatializing the Minsky framework emphasizes complications that arise because of open-economy relations, and unexplored by Minsky. He liked to say that the sky did not fall after different episodes of financial disorder on Wall Street, such as the 1987 Crash. Preventing the sky from falling was a Federal Reserve playing a lender-of-last-resort role with the worlds reserve currency, the dollar. Such interventions promised to stabilize the unstable economy, in Minskys phrase, at the cost of occasional inflationary pressure. Section 2 has already discussed how Minskys suggested interventions for countering financial disorder became less effective in the 1980s, with big government receding in importance and big bank interventions alone available to block financial disorders. Some additional problematic side-effects of big-bank/big-government interventions must be considered due to the special features of bordered economies. In particular, interventions by the Federal Reserve as lender-of-last-resort have an effect on exchange-market equilibria. Increasing liquidity without limit also increases the relative supply of dollars. What happens depends on whether a fixed or a flexible exchange-rate regime exists. If the former, then more dollars will end up offshore in foreign nations reserve holdings, and the trade balance will worsen. This creates the possibility of devaluation and makes foreign governments more reluctant to hold the dollars. If the latter, then the dollar will fall in value. This will, in turn, improve the US trade balance. Insofar as lender-of-last-resort interventions began in the fixed-exchange rate, these interventions would have been among the factors gradually undermining the Bretton Woods system. A second significant change in Minskys suggested interventions also arises because of the existence of multiple national currencies. This second change follows not from currency supply-demand balances, but from the fact that some currencies are more acceptable than others as international monies. The ability of any central bank to play a lender-of-last-resort role is circumcribed if that central bank provides liquidity that wealth-holders are not content to hold. A central bank in this categoryfor example, the Bank of Brazilcan issue its currency (reals) only insofar as it is willing to convert those reals freely into dollars (or whatever other currency may be demanded by those in a position to sell reals). This process works both ways. A government which issues a currency that is in demand by wealth-holders has a special privilege, essentially the ultimate seignorage right. This is the Because of intra-national capital flows, an economy as a whole can occupy one row, while a region within it occupies another. Japan, for example, chronically sits in the top row, while Tokyo falls into the bottom row. 10
5

case with the US in the current period. We have noted that big government spending has shrunk as a response to financial disorder. This implies that inflating away debt is no longer (or is less frequently) a feasible response to a Minsky crisis (a build-up of debt relative to asset values). This can mean that a debt overhang will simply persist. Fortunately for the US, the special circumstance of its dollar (and the ability of the US government to credibly underwrite asset-market risks) makes it easy to bundle and sell dollar-denominated debt to wealth-holders. Could the savings-and-loan crisis have been resolved so speedily without this special situation? A final set of constraints makes it more difficult to contain financial instability in the open-economy setting set out in Table 4. The presence of overseas asset owners who can engage in flightwhether a crawl or a sprint--away from local financial assets (and thus generate Minsky crises) reduces local authorities room for maneuver. This flight, if large enough, can bring on a Minsky crisis and trigger a debt-deflation regardless of what domestic asset owners do, and regardless of the state of the local economic cycle. And because the local central bank lacks the ability to act as lender-of-last-resort by issuing dollars, the complications thicken. Indeed, the collapse of asset prices as foreign investors retreat will have a double effect since it will also destabilize currency values. The next two sections focus on the emergence of asset bubbles in Japan and Korea in the late 1980s, and the relationship of these bubbles to these nations subsequent crises in the 1990s. To what extent are Minsky cycles and Minsky crises among the root causes of these crises? Section 6 will discuss the thorny question of what policy steps might be taken to ease these financial and macroeconomic crises. 4. Boom and bubble and the crisis in Japanese growth The post-1984 Japanese asset bubble involved both stock-market and land prices. Figure 2 shows annual change in land prices in Japan in the Tokyo metropolis. This figure demonstrates the rapid emergence of a bubble in the late 1980s. It also shows that Tokyo land prices have risen faster in upswings and declined faster in downswings. In turn, the Nikkei and Tokyo stock-price averages moved up even more sharply than did Tokyo land prices in the late 1980s, spiked in 1989, moved rapidly downward until 1992, and recovered some value in subsequent years.6 One interesting feature of Japans experience is that the bubble in its market for land persisted. One reason for this may be Japans historic restrictions on equity-market participation. Consumers savings have had relatively few places to go, so real-estate prices might be expected to move with other consumer savings outlets. The inflation-adjusted increase in housing prices has been virtually identical, on average, to the return on time deposits and the cost of housing loans from 1971 to the present. Housing prices have, however, been extremely variable; this suggests, in effect, that Japanese households able to afford homes have to weigh safe, stable time-deposit returns against a risky, time-varying return on a home. With this observation in mind, note the prediction of Table 2 vis--vis Japan: due in part to capital-flow pressures, a landprice bubble was likely, and that its most likely locus is in housing prices in Tokyo. Behind the equity-price bubble, in turn, lie two other structural factors: the Plaza Accord, which turned many To understand the intensity of trends in Tokyo, note that while the new housing price/annual income ratio was 4.4 in Japan in 1989, in Tokyo that year it was 7.4 (Kanemoto 1997: 615). 11
6

Japanese intermediaries investment behavior inward; and the Japanese banking practice of lending primarily on collateral value. What is the consequence of an asset bubble that hangs on without crashing? The people who live in Japan must bear it, a burden that comes at considerable cost. Figure 3 investigates two aspects of this cost. It first depicts the estimated percentage of housing price financed, on average. A trend line fit to these data points shows that this percentage has been hovered around 58% from 1971 to 1996. This figure also depicts the average ratio of housing price to the average income of home buyers. There is a clear upward trend here; the housing-price/income ratio drifts upward from just over 2.0 in the early 1970s to just under 3.5 in the mid-1990s. There is evidence here of the growing budgetary stress on households trying to buy homes. Figure 4 further supports this conclusion with data on the ratio of savings and of liabilities to income, 1960 to 1996. This figure shows savings-to-income and liabilities-to-income ratios for nonworker households and for worker households with liabilities. For non-worker households, the savings-to-income ratio is substantially higher than the liability-to-income ratio, and the savings/liability gap grows over time. But for worker households, the liability-to-income ratio is at least as high as the savings-to-income ratio--and it may be much higher. In other words, liability burdens are increasingly heavy on households who own homes, far outpacing savings. It seems that most younger Japanese households either face futures without secure housing, or futures in which provisions for housing create considerable financial stress. Calculations not shown here suggest lower-income Japanese households are especially exposed to financial fragility of this sort. In sum, the striking fact about the Japanese asset bubble of the 1980s is its continuing grip on Japans economy. Resolving this widespread Minsky crisis requires measures either to stabilize and increase asset values and/or to reduce the effective weight of debt loads. Neither step has been taken. In California, a combination of population/wealth in-migration and industrial innovation have permitted asset-value recovery and renewed income growth. Such crisis-resolution mechanisms have not been deployed in Japan, for several reasons. Rural-urban migration is largely spent. Japans population is graying, and any in-migration in recent years has largely involved temporary service-sector workers. This leaves measures to increase spending. One possibility is industrial renewal; but with Japans industrial groups ossified, the stimulus for such a renewal would have to come from start-up and independent firms. Such firms have been among the most affected by the bubble: in many cases their collateral assets have declined in value and compromised their creditworthiness; and where such firms have remained viable the banks on which they depend for credit are unable to lend due to loan losses and shrunken capital value.7 With financially fragile banks unable to support industrial innovation, recent Japanese government reform plans have focused on two elements: first, using large-bank mergers combined with securitization of bad-loan portfolios to recapitalize banks and renew their lending capacity; second, using a fiscal stimulus to lift incomes and trigger a Keynesian multiplier process. Thus far, a two-pronged economic recovery package of this sort has not proven politically viable. The fiscal-policy initiatives proposed have involved modest stimuli, and temporary Tier-one capital in Japanese banks consists in part of equity holdings. A rise in the Nikkei thus increases bank capital and loan-making capacity, and vice versa for a fall in the Nikkei. 12
7

measures. This has left the Japanese public without confidence in these stimulus packages efficacy and caused them to become still more conservative in their spending habits, putting more downward pressure on demand and pushing asset prices toward deflation rather than the required reflation. Meanwhile, government efforts to rescue banks by permitting mergers and/or bad-loan sales has met political resistance; policies permitting big businesses to work out their problems are regarded with skepticism by Japanese voters (whose post-bubble plight is illustrated in Figure 4). The household sector itself has its hands tied. While housing prices have declined, so have home-buyers incomes. Financial fragility has increased steadily for Japanese households, even as job security has lessened and unemployment increased. In the past four years, the yen has steadily fallen against the dollar as the crisis has lingered. It is as if the prior value of the yen itself and the previous glowing assessments of the Japanese model were themselves a bubble that has now burst. In sum, the Japanese economy is in the midst of intertwined general economic and Minsky crises, whose end is not in sight. 5. Boom and bubble and the Korean crisis If the Japanese financial crisis has been a long, grinding crisis playing out over a period of years, the Korean financial crisis appeared like a tsunami that broke over the heads of the Korean people in late 1997. This crisis had many of the hallmarks of a classic currency crisis. On the surface, it would seem that foreign investors realized that currency values were unsustainable given the Korean economys persistent trade deficit and declining reserves; their withdrawal triggered a precipitous decline in the value of the won from its former bubble level. Such an assessment of Koreas financial crisis should be resisted, however, for several reasons. For one thing, the pace of overseas long-term investments in Korea slowed in the past several years, as Figure 5 shows; but at the same time the volume of short-term investments in Korea exploded. The currency-crisis model doesnt handle this case. For another, Korea has been a classic boom economy; as in Japan, there has been a flow of population and wealth into Seoul from other areas in Korea. And as in the other two bubble economies considered here, an asset bubble in both equity prices and land prices emerged--but not at the end of the 1990s. Figure 6 presents the basic data for Korean land and equity-market prices in the period 1982-98, using the same scale as for Figure 2. The Korean equity market experienced a bubble which peaked in 1989 after gaining 250% in four yearseven larger than the Japanese asset bubble. After losing about half its value by 1992, the Korean equity market rose again to another peak in 1994. Since then, this market has lost value in four consecutive years; the 1998 average real value of Korean equities (using the KOSPI) has below the level of 1982. In sum, there was hardly anything resembling a bubble that needed bursting in the period preceding Fall 1997. While the events of recent months demand our attention, Koreas entire experience from the 1989 bubble on deserves analysis. For whereas Japans economy experienced a late-1980s chill from which it never recovered, Koreas did not. One key is that Koreas bubble was less severe. While its equity markets hit a higher peak, these markets were less developed than Japans; so the effects of this price collapse were less widespread; further, land prices experienced a smaller bubble, which peaked two years after 1989. Koreas economy was certainly as susceptible to bubble and fragility problems. In the framework developed here, and given Koreas generally negative trade balance in the recent past, Korea clearly should be considered as a boom economy, especially Seoul and its environs. We have emphasized that the 13

special challenge for boom economies is to channel capital (and labor) inflows so that they do not bid up the prices on existing capital assets and generate rent-seeking opportunities. Whereas California historically relied on its powerful commercial banks to accomplish this conversion of financial claims into real capital, Korea became renowned for its public and corporate sectors success in building a world-class industrial economy. In Table-1 terms, Korea had an efficient set of institutions for converting intra- and international wealth inflows into real asset accumulation. Part of Koreas fascination for outsiders was its ability to accomplish this nominal/real conversion and accomplish rapid or late industrialization (Amsden, 1989) with minimal use of arms-length market relationsto govern the market, in Wades (1990) apt phrase. The land bubble of the early 1990s was one aspect of a profound challenge for Korea the shortage of affordable housing, health services, and other amenities for Korean families, especially the many lower- and middle-income families in the burgeoning urban areas. The very success of Koreas growth strategy, which had avoided extremes of wealth and poverty, became a threat to the widespread provision of adequate and affordable housing and other services, as Yoon (1994) and Kim (1997) describe in detail. Pushed in part by an energetic popular movement, Koreas government pursued an aggressive housing policy; a massive construction effort was begun, together with progressive taxes on the sale of existing homes which socialized some of the gains from housing-price appreciation. Real government expenditures on housing and community development and on social welfare increased after the mid-1980s (admittedly no faster than did expenditures on economic services and defense). This effort to build homes and encourage ownership effectively countered the land (and housing) price bubble by putting new real assets in place. Thousands of new housing units have been put in place; thousands more are under construction. The economic situation of Korea was inexorably altered in the mid-1990s as it deregulated financial markets and took other steps to gain admission into the OECD.8 A new set of non-bank intermediaries (the merchant banks) arose. In part these intermediaries speculated on the arbitrage opportunities arising from regional interest-rate mismatches; in part these intermediaries channeled off-shore funds to chaebols seeking external finance to fuel their expansion plans. In any event, the Korean public sector lost control of the growth machine, and Korean dynamics flipped from boom-without-bubble to bubble economy. The funds flowing into Korea were almost entirely short-term (Figure 5). Using short-term off-shore funds to support currency speculation and chaebol investment, since these funds required continual rollover, invited either a currency collapse, an attack on Korean industry, orin the event--both. Since late 1997, the withdrawal of foreign investment and the collapse of asset and currency values (Figure 6) has created unsustainable debt burdens for many businesses and households. Bankruptcies are occurring at historically unprecedented levels, spurred by banks need to reduce lending so as to build up capital-asset ratios. Firmly under the thumb of the IMF, the Korean economy is now caught in an excruciating squeeze. Real money-supply growth has turned sharply negative after 1996; real domestic credit has declined sharply as of 1997. Price inflation has rocketed upward as the devaluation of the won has taken hold ; the real rate of GDP growth has fallen below zero in 1998; and the debt/GDP ratio has trended steadily upward A full accounting of Koreas liberalization and its consequences is beyond our scope here; see Chang, Park, and Yoo (1998), Wade and Veneroso (1998), and Crotty and Dymski (1998). 14
8

(except for a respite in 1998 attributable to the high bankruptcy rate). After Korea avoided Minskian fragility for years through careful economic management, a Minsky crisis has now arrived with a vengeance; and it is strangling businesses and households caught in Koreas arrested economic development. The principal strategies being pursued in the face of this collapse are increased ease of entry for overseas investors and, as in Japan, large bank mergers. 6. Conclusion and policy implications Two views of Asias crisis have predominated until now. In the IMFs view, the Korean crisis can be traced to the Korean governments excessive intervention in credit flows in the Korean financial sector. As Koreas financial markets became more open, some of Koreas privileged inside players made excessively risky off-shore investments; when these went bad, money managers and investors outside of Korea reacted to Koreas unsound financial architecture by (rationally) withdrawing their funds. Once Koreas financial practices eliminated advantages for domestic players and ease cross-border commitments, credit and capital will flow back to Korea from abroad. Japans financial problems, too, can be traced to its historic antipathy to overseas capital and its non-market criteria for allocating financial resources. This view rests on the premise that investment capital is the globally scarce resource, whose presence or absence determines national welfare; then any nations micro and macro policies are wrong whenever they either inhibit capital mobility or the prerogatives of offshore bankers and investors. A contrary view has emerged based on the twin ideas of incentive incompatibility and asymmetric information in credit markets. The idea here is that asset bubbles and financial crises result from behavioral pitfalls to which human economic systems are chronically prone. Asset bubbles and financial crises then comprise a pattern repeated throughout history, mistakes made and made again. This paper has suggested a new contributing factor in the Asian crisis: asset bubbles can result from imbalances between inflows of financial capital and the growth rate of real capital assets. Far from being an unusual episode in economic dynamics, imbalances of this sort are a chronic tendency in boom economiesthat is, in nations or regions that characteristically import capital and/or people from elsewhere. In a boom economy, what stands between stable asset prices and an asset-price collapse is an intermediation sector and/or a capital-controls policy dedicated to channeling inflows into appropriate real-asset accumulation. When these institutions are weakened, an asset price bubble and collapse become more likely. Indeed, we show that Minsky crises can be triggered by cross-border shifts of capital, not just by Minsky cycles. The interpretation of asset bubbles developed here spatializes Minskys notion of financial fragility that is, it shows the special balance-sheet tensions that emerge when financial positions extend across economic borders. Interestingly, this extension of Minskys well-known financialinstability model makes use of Minskys own earlier, neglected insights into the importance of cross-border relations in economic growth. This revised Minskian perspective throws new light on the recent experience of Japan and Korea, boom economies especially their recent asset bubble problems. We find that there is no one asset-bubble story that is told and retold over the millenia, no one lesson that can immunize policy-makers from the pitfalls of asset bubbles. Quite the opposite point is made here: asset bubbles are a chronic tendency of boom economies, and boom economies are recurring structural phenomena. One might argue that bubbles and busts are more likely given behavioral shifts 15

within the banking sector. In particular, an increasing number of firms within this sector are committed to making income by position-taking on asset projects initiated by others rather than by investing in extensive or intensive development. The no-bubble condition becomes less and less feasible the more wealth inflows are used to acquire existing assets on the premise that their monopoly rents will increase. As Minsky used to put it, firms and wealth-owners increasingly take position to make position; but expectations that asset prices will either continue to rise (the Japanese bubble of the late 1980s), or at least hold their value while cash-flow margins are exploited (the Korean situation of the late 1990s), invariably run afoul of structural constraints implicit in the Minsky cycle and/or in cross-border real-financial relations. Several policy implications follow from the model developed here. First, a nation characterized by large and sustained inflows of foreign exchange has a higher degree of financial fragility than a nation which does not, independent of the character of those nations internal balance-sheet relations. It is only prudent economic policy to insure that such inflows do not create asset bubbles by seeking out scarce real assets without being put to work in augmenting real-asset capacity. That is, a policy of inward capital-flow regulations is a prudent means of avoiding asset-price booms and collapses. Second, in a boom economy, a key mechanism for avoiding asset bubbles (and subsequent collapses) is the maintenance of a banking sector capable of directing and coordinating financial flows, instead of only reacting to flows. Loss of capacity of this sort suggests the importance of reestablishing it; and institutions that depend on position-taking should be reconsidered carefully in terms of their ability to affect resource flows within the overall structure of economic activity. Resolving the Asian crisis: Japans dilemma. The broader question addressed here is whether the Asian crisis can be ameliorated or even blocked through the application of the bigbank/big-government policies formulated by Minsky for the case of financial fragility. We have already suggested that these policies have grown weaker even for the US since the deregulation era took hold in the 1980s. There are reasons to believe that these policies are even more problematic for the Asian crisis per se. Japan is the obvious candidate for reviving East Asias growth through the application of such policies. However, several problems impede Japans application of these policies, even if she were willing to try them. The use of big bank policies involves several problems. First is that the Japanese must be willing to support the use of the yen as a reserve currency. This in turn means losing some control over the placement of yen-denominated assets, over the locus of yen holding, and over the terms and conditions of yen-denominated assets. All these steps will be controversial in Japan. Further, applying big-bank policies to the Asian crisis involves a multiple-currency problem. The asset bubbles in Asia grew up in a variety of countries; so intervention to stop the worst effects of debt deflation involves not one currency, but a chain of currencies. The problem of supporting any one currency (say, the baht or the ringgit) with yen is that this implicitly represents support for the government issuing this currency. This raises a moral-hazard issue that is lacking for a central bank that takes action to defend against financial meltdowns within one countrys borders.

16

Another problem is that the problem of debt overhang in Japan must be resolved before it is free to act. This in turn means either floating the debt away or inflating it away. Either course of action runs counter to deeply-seated attitudes in Japan. Floating it away requires opening up Japans financial markets to outside participation; so much bad debt has been built up in the Japanese bubble that offshore bond-holders participation will be required to offload significant amounts of this debt. This will require bundling, securitizing, and underwriting these debt claims. Experience with securitization in other nations has shown that this is readily done by experienced investment bankers if appropriate guarantees are extended (and if prospective holders are identified for whatever toxic waste remains after various cash tranches are stripped off). Of course, the provision of such guarantees requires public subsidies; and this has been a political sticking point. Once the debt is floated away to offshore holders, renewing economic growth in Japan requires revitalizing aggregate demand. This brings us to the other choice, inflating away the debt. Inflating away the debt requires identifying a sector willing and able to increase aggregate expenditure. The government sector has engaged in some increased expenditure, much of it focused on large-scale public works projects; this spending has been criticized as ineffective. Government spending would be more effectively used to address the problems of housing affordability and child-care availability which now bedevil Japanese households. A shift of fiscal policy toward such objectives would, however, run counter to long-standing social conventions concerning gender roles and the limits of state involvement. That the current arrangements are not working is evident in the ongoing marriage strike by women of child-bearing age, which has resulted in a plummeting Japanese birth rate (and in predictions of a declining Japanese population). What is required now is a way out of the years-long political stalemate which has straitjacketed Japan, long after the end of the era of income doubling has made a welfare-state system a social necessity. If government will not spend, households and corporations are the other sectors capable of increased aggregate demand. Household spending, however, has slowed dramatically in the current Japanese recession, especially on consumer durables. The increased insecurity associated with the ending of life-time employment arrangements hardly puts people in the spending mood. And corporate investment expenditure on Japanese goods and services will go only so far. If a sector willing to carry Japanese aggregate-demand expansion on its back is identified, further problems remain. A significant boost in aggregate demand will weaken Japans export balance. Ministry of Finance (MOF) officials characteristically view a trade surplus as a means of supporting a fiscal deficit: the idea is that the trade surplus earns the capacity for government deficit spending. A significant shift in thinking on the part of the MOF is thus required as a component of a new dedication to aggregate-demand expansion. In sum, any course of action that permits Japan to ease out of its Minsky crisis, and hence to play a more active role in ameliorating Minsky crises throughout Asia, will require significant shifts on the part of important players in Japans economic and political establishment. The capacity of Japan to pursue Minskys policies, and the effectiveness of these policies in the current global neo-liberal climate, is very much in question. Unless and until means of unwinding the Minsky crises in East Asia are found, the Asian financial crisis will remain a stone around the neck of world economic growth. Bibliography 17

Akyuz, Yilmaz. 1998. The East Asian financial crisis: Back to the future? Geneva: United Nations Commission on Trade and Development. Amsden, Alice. 1989. Asias Next Giant. Oxford and New York: Oxford University Press. Buchanan, Daniel C., 1973. One Hundred Famous Haiku. San Francisco: Japan Publications. Chang, Ha-Joon, Hong-Jae Park, and Chul Gyue Yoo. 1998. Interpreting the Korean crisis: financial liberalisation, industrial policy, and corporate governance. Cambridge Journal of Economics, 1998 (vol. 22, no. 6). Crotty, James. 1993. Are Keynesian uncertainty and macrotheory compatible? Conventional decision making, institutional structures, and conditional stability in Keynesian macromodels, pages 105-42, in Gary Dymski and Robert Pollin, editors, New Perspectives in Monetary Macroeconomics: Explorations in the Tradition of Hyman P. Minsky. Ann Arbor: University of Michigan Press. Crotty, James, and Gary Dymski. 1998. Can the global Neoliberal regime survive victory in Asia? The political economy of the Asian crisis. International Papers in Political Economy 1998(2). Dymski, Gary A. 1997. Deciphering Minskys Wall Street paradigm. Journal of Economic Issues 31(2), June: 501-508. Dymski, Gary A., and Robert Pollin. 1993. Hyman Minsky as hedgehog: The power of the Wall Street paradigm. In Financial Conditions and Macroeconomic Performance. Edited by Steven Fazzari and Dmitri Papadimitriou. Armonk, NY: M.E. Sharpe, 1993. Grant, James. 1992. Money of the Mind: Borrowing and Lending in America from the Civil War to Michael Milken. New York: Noonday. Gray, H. Peter, and Jean M. Gray, Minskian fragility in the international financial system, pages 143-68, in Gary Dymski and Robert Pollin, editors, New Perspectives in Monetary Macroeconomics: Explorations in the Tradition of Hyman P. Minsky. Ann Arbor: University of Michigan Press. Keynes, John Maynard. 1936. The General Theory of Employment, Interest, and Prices. London: Macmillan. Kim, Woo-Jin. 1997. Economic Growth, Low Income, and Housing in South Korea. London: Macmillan Press. Kregel, Jan. 1998. Yes, it did happen again A Minsky crisis happened in Asia. Mimeo, Jerome Levy Economics Institute, May. Krugman, Paul. 1998. What happened to Asia? MIT Department of Economics, January. Minsky, Hyman. 1965. Commercial banking and rapid economic growth in California. In California Banking in a Growing Economy: 1946-75., edited by Hyman Minsky. Berkeley, CA: Institute of Business and Economic Research, University of California. Pages 79-134. Minsky, Hyman. 1975. John Maynard Keynes. New York: Columbia University Press. Minsky, Hyman. 1982. Can It Happen Again?: Essays on Instability and Finance. Armonk, NY: M.E. Sharpe. 18

Minsky, Hyman. 1986. Stabilizing an Unstable Economy. New Haven: Yale University Press. Stiglitz, Joseph E; and Marilou Uy. 1996. Financial markets, public policy, and the East Asian miracle. World Bank Research Observer 11(2), August: 249-276. Thirlwall, A.P. 1980. Regional problems are balance of payments problems. Regional Studies 14(5). Wade, Robert, and Frank Veneroso. 1998. The Asian crisis: The high debt model vs. the Wall Street-Treasury-IMF complex. New Left Review No. 228, March-April. Yoon, Il-Seong. 1994. Housing in a Newly Industrialized Economy: The Case of South Korea. Brookfield VT: Avebury Publishing Co.

19

Figure 1: A stylized picture of a Minsky crisis

Asset price collapse Output growth Capacity utilization P P=P /P ratio ! Y=0 PK=PI
KK II

Sl aughter of

Debt/income

Stages: Robust

Fragile

Ponzi

Collapse

NOTE: The variables shown are measured against cyclical trend, with time elapsing from left to right in the diagram.

20

Table 1: Relationships between Regional Economic Growth Rates and Asset Bubbles
Is an asset bubble likely to form in one economy within a region? No Yes No wealth or labor Moral hazard problems transfers, dispersed in banking, inadequate technology and supervision productivity Higher innovation rate Faster wealth inflows or productivity in into growing region: The high-growth area boom economy case No capital flight, or Uneven economy-byoffsetting capital economy pace of capital outflows and inflows flight from region Capital flight Higher innovation rate from the region, failure or productivity in of overseas lenders to high-growth area renew short-term loans

Balanced regional growth Patterns in regional economic growth rates: One economy grows faster in a growing region Balanced regional slowdown or stagnation One region or nation grows faster (or declines less) than others in a stagnant or declining region

Table 2: Cross-border wealth flows and asset price pressures

International border condition: Net foreign savings > 0 (wealth inflows) Strong upward asset-price pressure: More overseas and domestic claims on available real assets (1980s California,1990s Seoul) Indeterminate asset-price pressure: More overseas claims, but fewer domestic claims, on available real assets (1980s US Midwest) Net foreign savings < 0 (wealth outflows) Possible upward asset-price pressure: Net acquisition of foreign assets, more domestic claims on available real assets (1980s Tokyo) Downward asset-price pressure: Net acquisition of foreign assets, fewer domestic claims on available real assets (other urban areas in Japan)

Intranational border condition:

Wealth inflows into this domestic region from other domestic regions

Wealth outflows from this domestic region to other domestic regions

NOTE: The asset price pressures indicated here do not take into account other influences on asset prices, such as income growth. 21

Table 3: Determinants of the Pace of Transformation of Financial Claims into Net New Real Assets
Type of capital inflows into domestic region: Foreign direct investment (domestic asset construction by overseas capital) Intra-national lending and capital-emitting capacity: Strong: robust banks, High financial-to-real asset borrower-lender relations, transformation with local and capital sources multiplier effects Weak: undercapitalized banks, High financial-to-real asset fragile borrower-lender links, transformation with small and few capital sources multiplier effects Cross-border capital-flow policies: Active: ex ante policies High and stable financial-toencouraging long-term real asset transformation investment, punishing or preventing short-term flight Passive: no conditions or High sensitivity of local controls vis-a-vis capital investment to external inflows or outflows conditions, govt policies Portfolio investment (domestic asset purchase by overseas capital) Strength of financial-to-real asset transformation depends on domestic allocation system Low financial-to-real asset transformation

Enhanced financial-to-real asset transformation by discouraging short-termism Low real-asset creation potential

Table 4: Bordered economies, Minsky crises, and Minsky cycles

No asset bubble: Financial prices and real prices are aligned Asset bubble: Financial asset prices rise above real asset prices in an expansion Possible if overseas financial-asset bubbles occur and pull domestic asset prices along Minsky cycle generates an asset bubble A bubble economy can generate an asset bubble due to capital inflows Asset-price collapse: Financial asset prices fall below real asset prices Minsky crisis if a collapse in overseas asset prices pulls down domestic asset prices Minsky cycle generates a Minsky crisis Reductions in capital inflows or sudden capital flight can generate Minsky crisis

Foreign savings<0: an asset-buying economy

Likely case if financial markets are tightly controlled Early stages of Minsky cycle: recovery from prior recession Boom without bubble

Closed economy case Boom economy: foreign savings>0, a capital-absorbing economy

22

Figure 2: Land and Stock Price Index values, Japan

350

300 Index v alue, 1985=100 250

200

150

100

50 1982 1984 1986 1988 1990 1992 1994 1996

Source: Akira Matsum oto, Ehim e University.

All figures are detrended by the Japanese CPI. Tokyo land v alues Nikkei av erage Tokyo stock av g

Figure 3: Japanese housing finance trends, 1971-96: Price/income ratio and borrowing/cost ratio
4.00 3.50 3.00 2.50 40.00 2.00 30.00 1.50 1.00 0.50 0.00 20.00 10.00 70.00 60.00 50.00

Left axis: Housing price/income ratio

1971 1974 1977 1980 1983

Right axis: Borrowing as % of housg cost

1986 1989 1992 1995

0.00

Pct financed = borrowing / (borrowing + own funds)

Price/income Pct financed Price/income trendline Pct financed trendline

Price/income records the ratio of average house price to average home-buyer income; percent financed records the percentage of housing price externally financed, on average.

23

Figure 4: Financial savings and liabilities relative to income, average for Japanese households with liabilities, 1960-96
3.5 3 2.5 2 1.5 1 0.5 0
19 60 19 65 19 70 19 75 19 80 19 82 19 83 19 84 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96

Left axis: Ratio of sav ings or liabilities to incom e

Source: 1996 Family Savings Survey, Govt of Japan

Sav ings/incom e, worker hshds Liab's/inc, workers w/liab's (B) Liab's/incom e, non-workers w/liab's Liab's/inc, workers w/liab's (A) Sav ings/incom e, non-worker hshds

NOTE: Figures for non-worker hshds assum e average incom es are the sam e for those w ith and without liabilities. The sam e assum ption is m ade for w orker hshds under assum ption A. Assum ption B assum es worker hshds with liabilities have average incom e levels double those of hshds w ithout liabilities.

Figure 5: Capital accounts, Korea, 1979-April 1998 Millions of US $ (excluding errors and omissions)
$16,000 $12,000 $8,000 $4,000 $0 -$4,000 -$8,000 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997

Direct investment Source: Asian Development Bank.

Portfolio investment

Other capital flows

24

Figure 6: Land and Stock Price Index values, Korea, 1982-August 1998
350
Index value: 1985 = 100

300

250

200

150

100

50 1982 1984 1986 1988 1990 1992 1994 1996 1998

Sources: land prices, Land Taxation in Korea (Ro, 1997); 1990-96 stock prices, Korean Stock Exchange; 1982-89 stock prices, author's calculations.

Figures are detrended by the Korean CPI.

Urban land prices Stock mkt average

25