A Tale of Two Cities:
Jakarta and New York

Koichi Hamada

Introduction

To me, Yale is a treasure, a source of valuable information and inspiration. New Haven was where I met my teachers and friends who taught me everything in economics, and it is now where I interact with students. The merit of this profession is that the terms of trade in knowledge exchange improve with age in my favor. I learn more from my students than they learn from me.

Being a student of “an ivory tower economist,” I tended to be “an arm chair economist,” who more often discussed the problems of economic development than directly observed and tried to solve them. Thanks to the Asian currency crisis, I have had chances to visit many financially ailing countries.

Part I: Jakarta

Cities in Asian capitals remind me of the benefit of growth by increasing input even without an increase in the total factor productivity. In spite of Paul Krugman’s criticism, growth with increased output would be an extremely envious situation for policy makers in Africa and CIS countries. Grand views of the sky scrapers in Jakarta and other Asian cities remind me of the fast pace and benefit of the growth process before the crisis, but the unfinished part of them showed the fact that expectations were betrayed. All the efforts based on the optimistic views on the future economic conditions were set back because their forecasts had not materialized.

In crisis countries, the balance of current account is a source of worry, that is the source of worry in the flow dimension. Many countries such as Indonesia, Malaysia, Korea, and particularly Thailand, had suffered from a large deficit. At the same time, during the time of crisis, a sudden, huge reversal of capital flow (Radelt and Sachs) took place. This can be interpreted as the shift of portfolio preference from the asset denominated in the currency of an Asian country to the asset denominated in the US dollar. In other words, when we consider the economic mechanism of a currency crisis, we have to consider both the flow aspect and the stock aspect of the critical process.

The dislocation of stock played a more abrupt and crucial role in the currency crisis than the dislocation of flow relationship.  For example, Indonesia’s current account deficit was not as serious as that of Thailand or Malaysia, but the currency depreciation was most precipitous in Indonesia. The Philippines experienced large current account deficits but managed to escape becoming one of crisis countries. What made the Indonesian case nearly catastrophic was the sudden shift of asset preference ignited by the loss of confidence in political and economic stability.

In economics, the relationship between stock equilibrium and flow equilibrium has long been discussed. The flow relationship indicates how quickly economic variables change, as savings, investment and the balance of payments affect the speed of asset accumulation in a national economy and its international credit and debt position. The imbalances in government deficits and the current account deficits are imbalanced in flow dimensions. On the other hand, the asset balance is the equilibrium in stock dimensions in such a way that outstanding assets -- here we emphasize the distinction of assets denominated in various currencies -- are held exactly at the ongoing economic variables. From this point of view, the IMF policy recommendation to balance the budget of the current account balance is the attempt to remedy the flow side of the problem and its recommendation to adjust the asset market by a high interest rate policy is the attempt to remedy the stock side of the problem.

There are many explanations for the causes of the Asian currency crisis. The fundamentals of monetary policy and external balances, the speculative international bank runs based on conjectural expectations among speculators, the moral hazard in lending to developing countries due to the explicit or implicit assurance by the governments, the mismatch of short and long term demand and supply in the credit market, and herding behavior of investors -- each of these explanations is relevant in many actual situations. Whatever the causes may be, a sudden dislocation is what did happen as phenomena. Thus, the stock problem, a sudden dislocation of asset demands, is crucial to the Asian currency crisis, though it cannot be denied that flow imbalances are also related to stock problems.

This conceptual framework is nothing but the portfolio (or asset) approach to the determination of exchange rates. Many Yalies, for example, Pentti Kouri, and in particular Dale Henderson here, studied years ago the combination of stabilizing flow relationship and the destabilizing stock relationship.

There are various versions in the formulations of the portfolio approach to the exchange rate determination. The model can be complicated depending on whether the agents have rational expectations, whether residents have their own currency biases, and whether agents hold other countries’ assets in a cross-country way. I will present here a simple version taking the currency of Indonesia, Rupiah as the example.

Suppose that Indonesians hold only the asset denominated in Rupiah, but that the rest of the world holds the asset denominated in Rupiah and that denominated in dollar. Let us denote the exchange rate of the Rupiah in terms of the dollar as e, in such a way that e = 1/16,000 or 1/8,000, meaning a Rupiah is worth 1/16,000 dollar. Let the total asset that the rest of the world possesses be Z Rupiah. Then the balance of payments of Indonesia is a function of the exchange rate e, and the amount of indebtedness Z. The balance of payments is a decreasing function of the exchange rate e and an increasing function of the indebtedness Z.  In terms of the increase in Z, that is the negative of the balance of payments of Indonesia, one obtains,

where f_Z 0, and f_e < 0.

The portfolio balance equation is written in such a way that people in the rest of the world holds a higher proportion of the Indonesian asset in their portfolio if the expected rate of appreciation of the value of the Indonesian currency is higher. That is,

If we impose the assumption of rational expectations such that E[(de/dt)/e] = (de/dt)/e, we obtain from equation (2), the following

where h_z> 0, h _e>0.

In Figure 1, we show the phase diagram of the simultaneous equation system, (1) and (3).

CC indicates the combination of e and Z that keeps the current account of Indonesia in balance, or that maintains the value of Indonesian asset held by the rest of the world constant. This is an intrinsically stable relationship and the value of Z increases in the left side of CC, and decreases in the right. PP indicates the combination of e and Z that keep the portfolio balance of the rest of the world. This is an intrinsically unstable relationship so that e increases above PP and decreases below PP.

First, the rest of the world is willing to hold a large amount of Indonesian debt. Then the equilibrium is like A, where Indonesian debt is large and the value of Indonesian Rupiah is high. Then, the asset demand for the asset in Rupiah declines precipitously, and the new equilibrium is like B. Since Z can move only slowly, only e jumps and the path of variables takes the trajectory like A through B’ to B.

Thus the model predicts that first a sudden overshooting depreciation in an Asian currency by the dislocation of demand for the currency and that then the current account of the Asian country gradually improves. The prediction of this model surprisingly applies well to the experiences of Asian countries, and perhaps to those of Latin American countries. Table 1 shows the changes in exchange rates after the dislocation of currency demand and the following slow adjustment in current accounts. In all the countries, one can detect jumps in exchange rates and the reversal of the current account.

This is the analysis of a floating exchange rate regime where the exchange rate is determined freely. This regime is what the IMF often prefers to recommend to ailing nations. In many cases including Indonesia, the IMF attempts to shift CC schedule to above by balancing the government budget. Or it attempts to shift PP schedule to above by a high domestic interest rate that could attract international capital. Both measures had limited successes. We should not take this regime as the only or the standard regime. The other regimes can be analyzed as well. The capital control changes the slope of portfolio relationship and also the speed of adjustment. The proportionate Tobin tax on all the capital transactions will substitute g((1 - rho)pi) for g(pi), where rho indicates the rate of capital transaction tax. It is easy to see the arrows indicating directions in the phase diagram become steeper. In this portfolio asset model, capital transaction taxation will increase rather than decrease the volatility of exchange rates.

The fixed exchange rate including the Currency Board System fundamentally changes the nature of differential equations. The exchange rate is no longer an endogenous, forward-looking variable, but a policy variable to be determined by the monetary authorities. Accordingly, exchange rates are no longer jumping variables, either.

Depending on the regime, the adjustment mechanism thus differs. So does the amount of adjustment cost to crisis countries. In other words, adjustment costs to employment, price stability and income distribution are hidden behind equations (1) and (2). Once the dislocation of the asset demand has come, some adjustment costs are borne inevitably. We have to cultivate an open-minded view on the relative cost and benefit of adjustment mechanisms. Rather than arguing for or against the IMF scheme, we have to compare several alternatives such as the Currency Board System, the Tobin tax, and capital controls of Chilean or Malaysian way by the criterion how these systems economize the adjustment costs that is anyway needed after the dislocation of capital demand.

Part 2: New York

Prosperity, historic peak of the Dow Jones, and the trust in the American business, what are in New York appear completely different from that in Jakarta. However, if we reconsider a little more objectively, do many things in New York not show isomorphic images to those in Jakarta before the crash? The flow relationship in the United States is very consumption oriented and extravagant. On the other hand, the whole world is willing to invest in the United States and into the assets denominated in the dollar as just before the Asian crisis.

The net debt is large as a percentage of GDP in the United States. To the relief of this audience I would add that the international debt to GDP ratio in the United States is approximately 15 per cent at the end of 1997, while those figures before crises run around much higher percentages for Asian countries. My back of the envelope calculation shows, the ratios are around 36% for Thailand, 38% for Malaysia, and 45% for Indonesia at the outset of the crises. For Brazil a similar figure is 18%.

Anyway, the equation system as well as Figure will apply if we reinterpret e as the exchange rate of the dollar by other currencies, e.g., yen per dollar, and if we reinterpret Z as the amount of foreign investment in the United States from the rest of the world.

What happens if the asset preference for the dollar were to decline suddenly, by some impairment of credibility in the American economy or by the emerging new market to invest rather than in the Wall Street. The dollar and probably the stock market in New York would suddenly collapse, and the current account balance of payments would improve gradually to achieve a less indebted equilibrium position of the United States economy in the world. Could we dismiss the possibility of this scenario just as a dream?

Figure 1
The Portfolio Approach to Exchange-Rate Determination.

Rupiah exchange rate in terms of dollar
Z: Net accumulated assets held by the Rest of the World
What if e is the yen-dollar rate and Z is the net accumulated debt of the United State?