Japan: Selected Issues
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar
Close

This Selected Issues paper analyzes macroeconomic developments and prospects for Japan during the 1990s. Following a surge in activity during 1996 and early 1997, the economy fell into recession in the second quarter of 1997. Real GDP fell by 3¾ percent during the four quarters ended March 1998; the unemployment rate reached historical highs; and deflationary pressures reemerged. The downturn was largely unexpected, and most forecasters had projected growth of about 2 percent in 1997. This paper also examines fiscal policy issues for Japan.

Abstract

This Selected Issues paper analyzes macroeconomic developments and prospects for Japan during the 1990s. Following a surge in activity during 1996 and early 1997, the economy fell into recession in the second quarter of 1997. Real GDP fell by 3¾ percent during the four quarters ended March 1998; the unemployment rate reached historical highs; and deflationary pressures reemerged. The downturn was largely unexpected, and most forecasters had projected growth of about 2 percent in 1997. This paper also examines fiscal policy issues for Japan.

VII. The Asia Crisis and Japan1

1. The financial crisis in Asia and the economic slowdown in Japan are having substantial and ongoing spillover effects. The main transmission channels for these effects are:

  • Trade linkages: As a result of increasing economic integration within the Asia region over the past decade, Japan and the rest of Asia account for substantial shares of each other’s exports and imports. Consequently, declines in domestic demand in one country have serious adverse effects on external demand in other countries in the region, and exchange rate changes have an important impact on relative prices.

  • Financial linkages: Driven by the relocation of Japanese production capacity to Asia, Japanese capital flows played an important role in Asia’s external financing in recent years. By contrast, the increasing fragility of the banking system in Japan is now dampening financial flows to Asia, while the contraction of economic activity in Asia is in turn leading to a deterioration in the asset quality of Japanese banks and reducing Japan’s foreign investment income.

  • Other market linkages: Reflecting these trade and financial relations, foreign exchange and stock markets in Asia and Japan are closely linked, allowing rapid transmission of disturbances across markets.

2. This chapter first provides an overview of Japan’s increasing integration in the Asia region since the mid-1980s and then seeks to quantify the recent interaction through these linkages. The focus is on relations with the countries most affected by the Asia crisis: Indonesia, Korea, Malaysia, the Philippines, and Thailand (the Asia-5).

A. Increasing Economic Integration within the Asia Region

3. Japan and Asia have become increasingly integrated over the past decade, as increased financial flows from Japan to Asia fostered the expansion and transformation of trade flows. Japanese investment in Asia reflected the relocation of production capacity from Japan to Asia (the “hollowing out” of Japan), a process driven by both comparative advantage and macroeconomic factors. The share of Japan’s outward direct investment going to the Asia-5 more than doubled from about 5 percent in the mid-1980s to about 11 percent in the mid-1990s (Figure VII.1). Lending by Japanese banks to all of Asia, including Hong Kong SAR and Singapore, reached almost $300 billion by the mid-1990s and accounted for more than half of total BIS-area bank lending to this region (Table VII.1).2

FIGURE VII.1.
FIGURE VII.1.

SHARE OF ASIA-5 IN FDI OUTFLOWS

Citation: IMF Staff Country Reports 1998, 113; 10.5089/9781451820447.002.A007

Table VII.1.

Japan: BIS-Area Bank Lending to Asia, 1993–97 1/

article image
Source: Bank for International Settlements, The Maturity, Sectoral and Nationality Distribution of International Bank Lending.

Data starting in December 1995 are not strictly comparable to earlier periods because of the revision of the Japanese balance of payments reporting system.

Includes Hong Kong SAR and Singapore.

4. Much of Japan’s direct investment in Asia was in the manufacturing sector,3 in contrast to Japan’s direct investment in the United States, which focussed on the financial services and real estate sectors. In the Asia-5, inward direct investment from Japan accounted for the largest share of total inflows (about one-fifth) during the 1990s. The establishment of production facilities in Asia allowed the combination of inexpensive skilled labor, which Asia had in relatively abundant supply, and imported capital and technology, in which Japan had a comparative advantage. At the same time, Japan’s direct investment in Asia helped to shift Asia’s comparative advantage in labor-intensive manufactured products toward higher value-added goods.

5. The relocation of production was facilitated by the globalization of production processes, through outsourcing, especially in high-technology and information-related industries. These industries are characterized by: (i) a variety of processing stages with different technological-, capital-, and labor-intensities; (ii) large economies of scale, for example in the production of memory chips; and (iii) international standardization for many parts, including microprocessing units, memories, and liquid crystal displays. Japanese firms in these industries took advantage of the increasing integration of the world economy to locate the labor-intensive stages of production in countries where wages were relatively low. As a result, trade with Asia in information-related products increased sharply—for example, the share of office machines in the Asia-5’s exports to Japan rose from ½ percent in 1987 to 8 percent in 1997.

6. The timing of the shift in production capacity largely reflected the increased openness to trade and investment of Asian developing countries, the yen appreciation, and the low cost of capital in Japan. The adoption of adjustment and liberalization policies in Asian countries in the mid- to late 1980s created the appropriate domestic conditions for the rapid expansion of foreign investment. The appreciation of Japan’s real exchange rate, especially following the Plaza Accord in 1985 and again during 1993–95, provided Japanese companies with both the incentive and the resources to locate production abroad (Figure VII.2). The easy availability of capital in Japan, especially during the period of high growth and low interest rates during the late-1980s (the so-called “bubble period”), also contributed to the expansion of investment abroad by Japan.

FIGURE VII.2.
FIGURE VII.2.

EFFECTIVE EXCHANGE RATES, 1985–98

June 1997=100

Citation: IMF Staff Country Reports 1998, 113; 10.5089/9781451820447.002.A007

7. The relocation of production capacity, along with the acceleration of economic growth in Asia, led to the rapid expansion of trade. The establishment of factories overseas increased Japanese exports of capital goods, and subsequently reduced Japanese exports of consumer products, as overseas production came online. Japan’s exports to the Asia-5 more than doubled in U.S. dollar terms between the mid-1980s and the mid-1990s while the Asia-5’s share of Japan’s exports jumped to about 20 percent (Figure VII.3). Exports of capital goods (including machinery and motor vehicles) increased to about 60 percent of total exports to the Asia-5, while the share of consumer durables (mostly manufactured goods) fell to about 20 percent (Chart VII.1).

FIGURE VII.3.
FIGURE VII.3.

SHARE OF EXPORTS TO ASIA-5

In percent

Citation: IMF Staff Country Reports 1998, 113; 10.5089/9781451820447.002.A007

CHART VII.1
CHART VII.1

JAPAN COMPOSITION OF TRADE WITH ASIA-5

Citation: IMF Staff Country Reports 1998, 113; 10.5089/9781451820447.002.A007

Sources: Nikkei Telecom, WEFA, and staff estimates.

8. The shift of production raised Japanese imports of manufactured goods and reduced imports of raw materials from Asia, reflecting the substitution of domestic production by overseas production (so-called “reverse imports”). As these reverse imports from Asia increased, the share of Japan’s imports of consumer nondurables (including food and beverages) and intermediate goods (including raw materials and iron and steel) from the Asia-5 declined sharply, while the share of consumer durables and capital goods (including personal computers and other office machines) rose about 55 percent (Chart VII.1). Over the same period, the Asia-5’s share of Japan’s total imports remained broadly unchanged (Figure VII.4). The relocation of Japanese manufacturing facilities abroad helps to explain why standard trade models were underpredicting Japanese imports of manufactured goods from Asia in the early 1990s.4 Empirical work confirms that flows of Japanese FDI temporarily increased Japan’s exports to Asia (while factories were being built), while the stock of Japanese FDI increased Japan’s imports.5

FIGURE VII.4.
FIGURE VII.4.

SHARE OF IMPORTS FROM ASIA-5

In percent

Citation: IMF Staff Country Reports 1998, 113; 10.5089/9781451820447.002.A007

9. As the increase in bilateral trade flows between Asia and Japan did not keep pace with the growth of Asia’s total exports and imports, Asia’s direction of trade actually shifted away from Japan. Over the past decade, both the share of the Asia-5’s exports destined for Japan and the share of imports originating in Japan have fallen somewhat (Figure VII.5). For the Asia-5 as a whole, Japan is less important as an export destination than the United States, which has a share of about 20 percent, but is the largest market for Indonesia, where it accounts for more than one-quarter of total exports.

FIGURE VII.5.
FIGURE VII.5.

SHARE OF JAPAN IN ASIA-5’S TRADE

In percent

Citation: IMF Staff Country Reports 1998, 113; 10.5089/9781451820447.002.A007

B. Spillover Effects Between Japan and East Asia

Impact of East Asia on Japan

10. The Asia crisis has already had a substantial adverse impact on Japan. In addition to the sharp fall in Japan’s exports to Asia, the quality of Japanese claims on Asia has deteriorated. The depreciation of the yen against other G7 currencies since mid-1997 reflected in part the expected impact of these adverse shocks (Figure VII.6). Similarly, the weakening of stock prices in Japan from mid-1997 reflected in part concerns about the impact of the Asia crisis on the profitability of both Japanese financial and nonfinancial firms.6

FIGURE VII.6.
FIGURE VII.6.

NOMINAL EFFECTIVE EXCHANGE RATES, 1995–98

June 30, 1997=100

Citation: IMF Staff Country Reports 1998, 113; 10.5089/9781451820447.002.A007

11. The contraction of Japan’s exports to Asia has been exacerbated by the collapse of investment in Asia. Japan’s exports to the Asia-5 fell by almost 30 percent in U.S. dollar terms in the first five months of 1998 compared to the same period in 1997, and the share of Japan’s exports destined for the Asia-5 decreased sharply (Chart VII.2). The decline in Japan’s exports reflected the abrupt slowdown in real fixed investment in 1998 in the Asia-5 (a drop of some 30 percent). As a result, Japan’s export of automobile-related goods, information-related goods, and capital goods and parts, have been particularly affected.

CHART VII.2
CHART VII.2

JAPAN RECENT TRADE DEVELOPMENTS

Citation: IMF Staff Country Reports 1998, 113; 10.5089/9781451820447.002.A007

Sources: Nikkei Telecom, WEFA, and staff estimates.1/ In U.S. dollars; seasonally adjusted.

12. The Asia crisis is putting downward pressure on prices in Japan, through lower import prices, but the effect has been small thus far. The prices of Japan’s imports from East Asia fell by about 10 percent in yen terms (about 15 percent in U.S. dollar terms) between early 1997 and early 1998, only moderately more than the overall decline in Japan’s import prices (Chart VII.2). Given the large depreciations of the Asia-5 currencies against the yen (about 35 percent), this implies that the exchange rate pass-through has been limited so far, possibly reflecting the small shares of East Asian exporters in Japan and the expectation that some of the depreciation may be temporary.

13. The Asia crisis is also putting broader pressure on the earnings of nonfinancial firms in Japan. Japanese corporates are exposed to Asia not only through trading relationships, but also through loans, equity investments, long-term receivables, and guarantees in various projects. Thus, the negative shock to Japanese firms’ external demand is exacerbated by a decline in investment income and a deterioration in asset quality. While these effects are difficult to quantify, especially given the depressed economic conditions in Japan, the Asia crisis has been cited by credit ratings agencies as an important factor in downgrading Japanese firms.

14. The crisis in Asia has added to concerns about the asset quality of Japanese banks. Aggregate BIS data show that, as of December 1997, Japan’s banks had loans to the Asia-5 of about $87 billion (¥11 trillion), equivalent to about 2 percent of bank loans and 55 percent of the Tier-1 capital of Japan’s major banks—a higher fraction than in any other major industrial country (Table VII.1).7 This figure may overstate Japanese banks’ vulnerability to the extent that about one-third of these loans were to foreign affiliates of Japanese companies, which in many cases can count on strong parent company support. Against this, a part of Japanese bank loans to Hong Kong SAR ($76 billion at end-1997) and Singapore ($59 billion) were on-lent to the crisis countries, while some lending by European subsidiaries of Japanese banks may be mistakenly counted as lending by European banks. Although the absence of published detailed balance-sheet data (and the complete lack of published off-balance-sheet data) makes it difficult to provide a careful assessment of exposures and vulnerabilities, some analysts estimate that impaired loans of Japanese banks from exposure to Asia could be about $30 billion,8 and have cited this vulnerability as an important factor behind the downgrading of their ratings.

Impact of Japan on East Asia

15. Economic difficulties in Japan have exacerbated the Asia crisis. The Asia-5’s U.S. dollar value of exports to Japan are falling sharply (Table VII.2), while exports to other industrial countries are growing strongly. Moreover, capital flows from Japan to the Asia-5 have slowed, reflecting the fragility of Japanese banks as well as the financial problems of many Japanese corporates. The yen’s weakness against non-Asian currencies has put downward pressure on Asian currencies, reflecting the impact of direct trade and financial linkages with Japan, as well as increased competition in third markets.

Table VII.2.

Japan: Asian Country Trade Growth, 1997Q1–1998Q1 1/

article image
Source: WEO database.

The countries consist of Korea, Thailand, Malaysia, and the Philippines.

16. The divergence between Asia’s export performance to Japan versus other industrial countries is broadly consistent with the different cyclical positions and exchange rate movements across countries, but cannot be fully accounted for by standard trade equations. This divergence was explored by comparing Asia’s export volume growth to Japan with that to the United States (Table VII.2). Standard trade models explain real export growth in terms of the growth of foreign demand (“income”) and changes in real exchange rates (“prices”).9 Although price indices for bilateral trade are not yet available, the available evidence suggests that the Asia-5’s export prices10 declined by about 15 percent in U.S. dollar terms in the first quarter of 1998 compared to the same quarter a year ago, reflecting in particular declines in commodity prices.11 While Japan experienced a sharp decline in real GDP through the first quarter of 1998, the United States recorded growth of 4 percent. Even with the assumption of high short-run income elasticity (say of 1½ and 2, respectively, for Japan and the United States—in line with staff estimates), the implied price elasticity needed for consistency with the Asia-5’s observed export volume growth rates to Japan and the United States would have to be twice as high for the United States as Japan (about 1½ and ¾, respectively).12

17. The discrepancy between the implied price elasticities for Asian exports to Japan versus the U.S. may reflect underlying structural factors that are not fully captured by reduced-form trade equations. In general, the commodity composition of exports to various trade partners may give rise to different price elasticities. For example, Asia’s exports to the United States consist primarily of electrical and electronic goods, which compete directly with each other and may therefore be more price sensitive, while exports to Japan consist more of commodities, which may be more sensitive to the weakness of business fixed investment in Japan.13 Also, standard trade elasticities may not fully incorporate the impact of the growth of Japanese firms’ production facilities in Asia following many years of large flows of foreign direct investment.14

18. Financial flows from Japan to Asia have slowed sharply since the onset of the crisis, reflecting both a decline in bank lending and a fall in foreign direct investment. The fragility of Japanese banks has reduced their ability and willingness to extend loans. The fall in BIS-area bank lending to the Asia-5 between mid- and end-1997 ($16 billion) is mostly accounted for by the decline in lending by Japanese banks ($11 billion, Table VII.1). Although part of the decrease may reflect the depreciation of the yen against the U.S. dollar, valuation changes alone cannot explain the decline.15 Also, valuation changes would tend to have the same effect on Japanese banks’ lending across countries, while in fact lending to Thailand actually increased and lending to Malaysia fell most sharply.

19. Similarly, direct investment by Japanese firms in Asia has slowed, reflecting the reduced availability of capital in Japan. Data from Japanese sources show that direct investment in Indonesia fell 35 percent in U.S. dollar terms in the second half of FY1997 (September 1997 to March 1998) from the year before. The growth of direct investment in Thailand declined significantly from 80 percent in the first half of the fiscal year to minus 3 percent in the second half.

C. Quantifying the Spillover Effects

20. Different approaches to quantifying the cross-country impact of the crisis suggest that adverse spillover effects are in the range of 1–1¼ percent of GDP in Japan and somewhat larger in East Asia. The common feature of all of these methods is the construction of simulation models (of varying degrees of complexity) to isolate the crosscountry effects. The simplest approach is to use trade elasticities to calculate the adverse effects of trade shocks on economic activity. A more sophisticated approach is based on a regional model of aggregate demand and trade for East Asia and Japan. Finally, a dynamic model of the world economy (an extension of MULTIMOD) is used to incorporate financial flows and other non-crisis countries. While each of these models sheds some light on the magnitude of the spillover effects, each one has important drawbacks.

21. A major difficulty faced by all of the approaches is how to calibrate the magnitudes of the shocks associated with the crisis. The method used in each case was to infer the sizes of the shocks from the changes in the projections from the May 1997 WEO to the staff’s latest outlook. While this is a simple and transparent technique of parametrizing the shocks, it means that these approaches cannot distinguish between exogenous shocks and the endogenous responses to them, and is subject to whatever biases may be included in the staff’s outlook.

Trade Elasticities

22. Simple calculations using trade elasticities suggest that Japan’s output loss from spillover effects in 1998 will be about 1¼ percent of GDP. These calculations incorporate: (i) the impact of foreign demand on exports; (ii) the impact of real exchange rate changes on imports; and (iii) the impact on foreign investment income (for Japan). However, these calculations ignore financial linkages, including bank lending, foreign direct investment, and stock market effects. These calculations also do not take into account feedback relationships, i.e., the fact that a drop in domestic demand in Asia depresses output in Japan, which in turn reduces Japan’s imports from Asia.

23. The output loss in Japan reflects the export effect (minus 1 percent of GDP), the import effect (minus ¼ percent of GDP), and the earnings effect (minus 0.1 percent of GDP):

  • Export effect: Projected import volumes in Japan’s Asian trading partners in 1998 are lower by about 15 percent in the latest WEO projections than in the May 1997 WEO projections.16 Given that roughly 40 percent of Japan’s exports are destined for Asia, Japan’s projected export volume would be reduced by about 6½ percent. Since exports represent about 13 percent of GDP in Japan, and using a multiplier of about 1¼ (broadly in line with standard MULTIMOD results), Japan’s output will be lower by about 1 percent of GDP.17

  • Import effect: The real depreciations of the Asian currencies would lower these countries’ export prices and therefore increase in Japan’s imports. Using import weights, Asian currencies depreciated by about 11 percent against the yen between mid-1997 and mid-1998, reflecting the large depreciations of many Asian currencies but also the offsetting appreciations of the Chinese yuan and the Hong Kong SAR dollar—which are pegged to the U.S. dollar—against the yen. As roughly one-third of Japan’s imports are from Asia, this implies an effective appreciation of the yen of about 4 percent. Assuming that half the depreciation is eventually passed through in the form of lower prices, and an import price elasticity of ¾ (in line with the analysis discussed above), Japan’s import volume would rise by about 2 percent. Given that imports account for about 12 percent of GDP and the multiplier effect, Japan’s GDP would be reduced by about ¼ percent.

  • Earnings effect: The reduced profitability of operations based in Asia would lower Japan’s investment income from abroad. Given that interest payments and equity earnings from abroad together amount to about 1 percent of GDP, and assuming that earnings from Asia—which accounts for about 10 percent of the stock of Japan’s outward direct investment—declines to zero, Japan’s GDP would fall by about 0.1 percent.

Regional Interaction of Trade and Aggregate Demand

24. A regional model of trade and aggregate demand suggests that the output losses from spillover effects will be about 1 percent of GDP in Japan, 1¼ percent of GDP in Korea, and 1½ percent of GDP in Southeast Asia.18 This approach explicitly incorporates feedback effects by linking three areas: Japan, Korea, and Southeast Asia (SEA). For each area, real domestic demand is assumed to be a function of real GDP and real stock prices (to take into account the effect of financial market turbulence), and real import demand is assumed to be a function of the real effective exchange rate and the level of real domestic demand. The areas are linked by the import equations, which endogenously influence other areas’ exports. Real GDP is then the sum of domestic demand and net exports.

25. The model is parameterized using simple regression analysis, as well as with reference to previous studies.19 The magnitudes of the shocks are chosen to be broadly consistent with recent developments in the three areas. The exchange rate and stock market shocks are roughly equal to the real changes in these variables between mid-1997 and mid-1998. The shocks to domestic demand are roughly in line with the changes that have been made to WEO projections over the past year. As noted above, determining the appropriate sizes of the shocks is complicated by the fact that observed changes include the endogenous response to the crisis. The type and magnitude of the assumed shocks are summarized in the tabulation below.20

Assumed Shocks to Baseline

(Percent deviation from baseline)

article image

26. The simulation results suggest that each area’s own shocks have had the largest effect on its output, though spillover effects are important (tabulation). In Japan, the total output loss is projected to be 5¾ percent, of which just under 5 percent results from shocks to domestic variables and 1 percent (about one-sixth of the total output loss) from shocks to other areas. In Korea and Southeast Asia, the total output losses are much larger, at 11¼ percent and 10¾ percent, respectively, of which 1¼percent and 1½ percent (about one-tenth of the total output losses) are due to spillover effects. Within each block, the positive output effects of the exchange rate shocks are much smaller than the negative output effects of the domestic demand shocks, reflecting the Keynesian nature of the model and the high weight of domestic demand in GDP in each area. The magnitudes of the output losses stemming from demand shocks in other countries are generally similar to the effects of the exchange rate shocks.21

Regional Impact of Asia Crisis

(In percent of baseline 1998 GDP)

article image

27. The results of this exercise are subject to a number of important caveats. First, although the model takes account of the effect of stock market developments on domestic demand, other financial linkages are ignored, including the impact of interest rates or capital flows on real activity, and the effects of cross-country financial market interaction. This is potentially an important drawback given concerns that capital flows from Japan to East Asia have been constrained, reflecting the “credit crunch” in Japan. Second, the exclusion of China, where growth is also expected to slow substantially as a result of the recent turmoil, may understate the effects of the crisis. However, the exclusion of the rest of the world (to which exports would be expected to rise) likely overstates the effects.

Global Interaction of Trade and Financial Flows

28. An extension of the Fund’s MULTIMOD model suggests that the spillover effects of the crisis could eventually be smaller than calculated in the two previous approaches, because the reversal of capital flows reduces interest rates in other countries, which stimulates the interest-sensitive components of demand throughout the world. MULTIMOD is a dynamic, multi-country model of the world economy that explicitly includes a global capital market and forward-looking behavior of economic agents, and was designed in part to study the transmission of shocks across countries.22 The model, which already includes blocks for each of the seven largest industrial countries and an aggregate block for other industrial countries, developing countries, and transition countries, has been augmented with explicit country blocks for Korea, Indonesia, and an aggregate block for Malaysia, the Philippines, and Thailand.23 Using MULTIMOD, the magnitude of the shocks in East Asia and Japan are gauged using the change between the May 1997 (pre-crisis) WEO projections and the staff’s initial projections for the summer 1998 WEO round.

29. Simulation results from MULTIMOD suggest that, while the overall output declines are similar to those discussed in the regional trade model, the spillover effects are only small fractions of GDP (Table VII.3). The key factor behind this result is that capital flows, which tend to flee crisis countries, reduce interest rates in the rest of the world and thus stimulate the interest-sensitive components of aggregate demand elsewhere. In Japan, while the projected level of real GDP is 8 percent below its baseline in 1999, this is entirely the result of the Japan-specific shocks. East Asia has a small negative impact on activity in Japan in 1998, through trade flows, but thereafter the favorable response of interest-sensitive components of aggregate demand (including those outside Asia) to the decline in the real long-term interest rate outweighs the direct trade effect.

Table VII.3.

Japan: MULTIMOD Simulation Results

article image
Source: Laxton and Sarel (1998).

30. Similarly, Indonesia, Korea and Southeast Asia are projected to suffer large declines in activity, but mostly due to their own shocks. Japan has an adverse impact on East Asia of about minus 0.5 percent of GDP, reflecting its large share of these countries’ exports. While the dominant transmission mechanism from Japan to East Asia is trade, as reflected in the deterioration of the current account balances, the Japan-specific shocks have a small (favorable) effect on real long-term interest rates in Southeast Asia, which partly mitigates the trade effect over time.

31. These results, like those of the other approaches, are subject to important caveats. First, and most important, MULTIMOD does not explicitly model bilateral financial relationships, so this approach cannot capture the effect of the “credit crunch” in Japan on bank lending to East Asia. Second, this approach cannot address “pure” contagion effects, such as the effect of an increase in the risk premium in one country on the risk premium in another or cross-country linkages between stock market prices.

APPENDIX

Description of Regional Model24

32. The model is composed of equations that endogenously explain aggregate demand for three country blocks: Japan (J), Korea (K), and Southeast Asia (A). In this case, Southeast Asia is defined to include Hong Kong SAR, Indonesia, Malaysia, the Philippines, Singapore, Taiwan Province of China, and Thailand.

For example, the Japan block is composed of the following three equations:

GDP Identify: Y _ J = D _ J + ( XW _ J α K _ J M _ K α A _ J M _ A ) M _ J
Domestic Demand : log ( D _ J ) = δ 0 + δ 1 log ( Y _ J ) + δ 2 log ( S _ J ) log δ 3 log ( D _ J 1 )
Import demand : log ( M _ J ) = μ 0 + μ 1 log ( D _ J ) + μ 2 ( e _ J ) + μ 3 ( M _ J 1 )

All variables are expressed in real terms where:

Y_J = GDP in Japan

D_J = Domestic demand

M_K = Korea’s imports of goods and services (NIA basis)

αK_j = Share of Korea’s imports from Japan

M_A = Asia’s imports of goods and services (NIA basis)

αA_j = Share of Asia’s imports from Japan

XW_J = Japanese exports to the rest of the world (NIA basis)

S_J = The Japanese stock price index deflated by the CPI

e_J = Japan’s real effective exchange rate index (CPI basis)

33. The first equation is an identity requiring that GDP equals the sum of domestic demand and net exports. Exports are the sum of imports from the other two blocks, multiplied by Japan’s trade shares, plus Japan’s exports to the rest of the world. For the purpose of this exercise, constant trade shares are assumed, whereas actual trade shares are time varying. In order to ensure that the national accounts identity holds, a residual is calculated for the first equation.

34. The second equation represents a domestic demand function (i.e., a behavioral relationship determining the sum of private consumption, investment, and government demand), which is determined by GDP, the real stock price index, and the lagged dependent variable. The third equation represents an import demand function, which is determined by domestic demand, the real effective exchange rate, and the lagged dependent variable.

35. The parameters of the domestic demand and import volume equations are imposed, but are broadly consistent with the results of OLS regressions, which were estimated for the three blocks individually.25 The long-run elasticities are summarized below:

Long-Run Elasticities

article image

Thus, the three blocks represent nine equations that determine nine unknowns: imports, domestic demand, and GDP in Japan, Korea, and Southeast Asia. The exogenous variables that determine the equilibrium values of these variables are: world demand for exports from each of these regions, the real level of the stock market, and the real effective exchange rate.

1

Prepared by James Morsink.

2

Within the Asia-5, Japanese banks’ share of total lending was especially high in Indonesia and Thailand (about 55 percent).

3

See Linda Goldberg and Michael Klein, “Foreign Direct Investment, Trade, and Real Exchange Rate Linkages in Southeast Asia and Latin America,” NBER Working Paper No. 6344, December 1997.

4

See Bankim Chadha, “External Adjustment in Japan: Recent Developments and the Medium-Term Outlook,” in Japan—Selected Issues, IMF Staff Country Report No. 96/114, October 1996.

5

See Tamim Bayoumi and Gabrielle Lipworth, “Japanese Foreign Direct Investment and Regional Trade,” in Aghevli, Bijan, Tamim Bayoumi, and Guy Meredith (eds.), Structural Change in Japan: Macroeconomic Impact and Policy Challenges, (Washington DC, International Monetary Fund: 1998).

6

However, these market linkages are difficult to quantify given the inherent volatility of asset markets.

7

Total Japanese bank lending to Asia, including Hong Kong SAR and Singapore, was $250 billion at end-1997.

8

This number reflects the application of a 10–15 percent ratio (in line with banks’ recent self-assessment exercises) to Japanese banks’ total exposure to Asia ($250 billion as of end-1997). U.S. and European banks have typically provisioned at a rate of 10–15 percent, reflecting the higher risk of their exposures to Asia.

9

See, among others, Takatoshi Ito, Peter Isard, Steven Symansky, and Tamim Bayoumi, “Exchange Rate Movements and their Impact on Trade and Investment in the APEC Region,” IMF Occasional Paper No. 145 (December 1996), and Carmen Reinhart, “Devaluation, Relative Prices, and International Trade: Evidence from Developing Countries,” IMF Staff Papers, Vol. 42 (June 1995).

10

This particular calculation excludes Indonesia, because Indonesia’s oil exports complicate the assessment of trade adjustment.

11

Over the same period, Japanese domestic prices fell by about 3½ percent in U.S. dollar terms, while the U.S. CPI rose by about 1 percent.

12

Standard trade equations also performed poorly in the aftermath of the Mexico crisis. Following the real depreciation of the peso, the price elasticity of exports to the U.S. was closer to 1½, rather than the unit elasticity used in standard trade equations.

13

Japanese imports of construction materials are presently very weak, reflecting the severity of the current decline in residential investment and public works spending.

14

For example, declines in domestic demand in Japan could have an amplified effect on Asian exports because Japanese production facilities in Asia are the marginal suppliers for Japanese firms.

15

Assuming that 20 percent of private sector external debt is yen-denominated and the remainder U.S.-dollar denominated, based on the only available information on currency composition (for Thailand), would imply that the 12 percent depreciation of the yen between mid- and end-1997 would reduce U.S. dollar denominated debt about 2½ percent, or less than $3 billion.

16

For the purpose of this calculation, Asia is defined as China, Hong Kong SAR, Indonesia, Korea, Malaysia, Philippines, Singapore, Taiwan Province of China, and Thailand.

17

As Asia accounts for a much larger share of exports in Japan than in the United States or Germany, the Asian crisis will have a much greater adverse effect on Japan.

18

See Appendix I for a description of the model. In this case, Southeast Asia is defined as Thailand, Malaysia, Indonesia, Philippines, Hong Kong SAR, Singapore, and Taiwan Province of China.

19

Although the estimated empirical relationships were not always robust to changes in the specification or the sample period, the multipliers used are broadly consistent with observed developments and earlier studies (see Appendix I).

20

The numbers refer to deviations from the baseline projections during the 1997–98 simulation period.

21

Recent estimates by the Bank of Japan would suggest even greater spillover effects from both domestic demand and exchange rate shocks. See Kamada, Koichiro, Yasutaka Oenoki, and Katsunori Watanabe, “A Local Model of Asian Economies,” IMES Working Paper 98–5, Bank of Japan (June 1998).

22

Douglas Laxton, Peter Isard, Hamid Faruquee, Eswar Prasad, and Bart Turtelboom, MULTIMOD Mark III: The Core Dynamic and Steady-State Models, IMF Occasional Paper No. 164, (Washington DC: International Monetary Fund, 1998).

23

Douglas Laxton and Michael Sarel, “The Impact of the Asian Crisis and the Slowdown in Japan: A MULTIMOD Simulation Approach,” mimeo, 1998.

24

This appendix was prepared by Christian Thimann and Christopher Towe.

25

For Japan, import equations presented by B. Chadha in Japan—Selected Issues, IMF Staff Country Report No. 96/114 (October 1996) provided useful estimates, which suggested that in the long run import volumes are roughly unit elastic with respect to relative prices, but that the elasticity might be larger than unity in the case of domestic demand. Estimates of Korean import volume equations by L. Giorgianni and Gian Maria Milesi-Ferretti in “Determinants of Korean Trade Flows and the Geographical Destination,” IMF Working Paper WP/97/54 (April 1997), suggest a near zero relative price elasticity and an elasticity for domestic absorption of around 1.3.

  • Collapse
  • Expand
Japan: Selected Issues
Author:
International Monetary Fund