Downside Risks to the Japanese Outlook for 2002
by Adam S. Posen, Peterson Institute for International Economics
Presented at a conference on the "Global Economic Prospects"
Institute for International Economics
April 3, 2002
© Peterson Institute for International Economics
After more than a decade of economic stagnation and minimal structural change, Japan stands on the brink of outright financial crisis—the only debate is whether the Japanese government can dodge its imminent economic threats for another six months at most, or whether the government still can throw money at these problems with decreasing effectiveness until the public debt becomes unsustainable (which should be no later than the start of 2005). Either way, volatility in Japanese asset markets will be extremely high for the next 36 months, with significant declines on average in asset prices and the yen.
It is true that Japanese savers remain highly risk-averse, which puts a drag on change in general as well as provides a buffer for policy mistakes—however, their passivity should not be exaggerated but they cannot be depended upon to stave off crisis1. In the last six months, in anticipation of deposit insurance limits and bank failures, over ¥150 trillion in household savings has moved from weaker banks to cash, gold, offshore accounts, and the four too-big-to-fail merged banks. Given the ongoing drain of liquidity from Japanese asset markets, the movement of even just a few percent more of household savings, along with what mobile international capital remains in Japan, could produce large declines in asset prices. Such movements would rapidly intensify both the current credit crunch for Japanese small- and medium-sized enterprises and the Japanese government's mounting fiscal difficulties.
This paper argues that the Japanese economy is likely to tumble into crisis sometime before the Diet's supplemental budget process begins in September 2002. There are three elements to this argument:
1. the underlying forces increasing Japan's vulnerability to shocks;
2. what a crisis in Japan would look like; and
3. potential triggers for crisis in the coming months
Crisis in Japan is not inevitable. A comprehensive policy package of bank closures and recapitalization, money-financed Bank of Japan [BoJ] purchases of long Japanese government bonds [JGBs], and replacement of public works spending with tax cuts could still save the Japanese economy from this fate; increased explicit pressure from the US and other governments could somewhat increase the chances of such a policy package being adopted2. This paper makes the realistic assumption, however, that there will be no meaningful change in Japanese economic policy for the horizon of this forecast.
Increasing Vulnerability to Shocks
Japanese economic policy is running out of room for error. President Bush's February visit to Tokyo did not result in any new initiatives by Prime Minister Koizumi to clean up the bad loans—in fact, the recent announcement by the Financial Services Agency (FSA) that results of this year's "special inspections" of banks would be disclosed only by bank category in mid-April, without revealing specific banks' situations, confirms that supervisory scrutiny is moving backwards3. The mid-February firing of former Foreign Minister Tanaka has burst the bubble of Koizumi's popularity; the initial 30-plus percent drop in his approval ratings has been sustained, and according to some polls fallen even further. Diet opposition both within and outside of the LDP-coalition is sufficient to block any Koizumi initiatives, and has increased since Tanaka's firing.
This combination makes it very unlikely that the Koizumi government will respond to Japan's economic problems for the remainder of his time in office (which is the foreseeable future). There is a game of chicken amongst the "responsible" ministries (BoJ, FSA, Ministry of Finance [MoF]) stalemating any progress on policy, and which neither the Cabinet Office nor Koizumi can break up in the absence of external pressure or overwhelming popular support. The lack of any clear alternative to Koizumi for Prime Minister makes it likely he will hold on to the office, be spared explicit bashing by the United States, and remain ineffectual, until an outright crisis hits.
Yet, in the absence of pre-emptive policy action, two powerful economic forces are pushing the Japanese economy toward outright crisis. One is the vicious spiral of debt-deflation, which has now accelerated in Japan. Every succeeding month of deflation raises the real burden of outstanding debt, and decreases consumer demand, both of which increase the number of bad loans; every additional bad loan decreases credit available to viable businesses, and further depresses asset prices (especially real estate), reinforcing deflation (figure 1 shows the declining Japanese price level).
While always theoretically possible, debt-deflation has become a palpable reality in Japan—each succeeding month of deflation increases the number of bad debts outstanding, and every additional debt falling into arrears or default decreases demand, putting further downward pressure on prices. Even on the government's vastly understated estimates of nonperforming loans, classification IV ("risk management" or defaulted) loans have been rising more quickly than banks can provision (given their diminished capital and profits), an imbalance unseen even in 1998. Figure 2 shows the ratio of new classification IV loans to banks' write-offs of bad debt, which now is higher than one—this is an explosive process and ultimately unsustainable for more than a few quarters without government intervention in the form of bank closures and recapitalizations4. That recapitalization, however, even if attempted may not turn out to be easily managed.
That is because the other force at work is the well-known accumulation of public debt in Japan, having reached over 130 percent of GDP (gross) in FY2001 (see figure 3). Admittedly, the threat to solvency this presents for the Japanese government is often exaggerated5: net public debt is much lower than gross debt, even discounting most government assets; gross household savings remains in excess of 200 percent of the value of GDP; only 5-6 percent of JGBs are held by foreigners; and all Japanese public debt is denominated in yen.
There remains, however, the risk of a liquidity crisis in the JGB market from several sources, including pre-emptive downgrades by bond ratings agencies; a fall off in tax revenues associated with debt-deflation6; a collapse of a bankrupt local or prefectural government, and the associated banks, requiring a public bailout7; and the need to suddenly inject a large amount of capital into the banks or the deposit insurance system in response to a bank run.
Any of these, almost certainly accompanied by a sudden rise in rates on long JGBs, would tip the government's debt dynamics into as unsustainable a process as the private debt, where new debt accumulates faster than principal and interest can be paid. With the annual government deficits currently running at 8 percent of GDP, the current relatively low share of interest payments in Japanese government expenditures (shown in figure 3) would rise quickly with increasing interest rates. Even in the absence of interest rate increases, the Japanese government can afford at most three additional years of slow or negative growth before those explosive debt dynamics kick in, and 2002 will already use up one of those.
It is reasonable to argue that neither of these underlying dynamics constitutes an "action-forcing event" in the near term—that is a shock that will move faster than Japanese authorities can respond, in and of themselves. And the noted passivity of Japanese savers does give the government some leeway. Still, the underlying dynamics and eroding confidence act in the other direction, diminishing the Japanese government's room for error, and a pick-up in exports to the United States will only provide limited additional growth cushion8. Meanwhile, the list of potential if not likely shocks that could accelerate these dynamics is long (see section on "Potential Triggers for the Crisis" below).
In any event, no one should be fooled by the recent run-up in Japanese equities and in the yen against the dollar into thinking that the economy has strengthened. The recent equity/exchange rate appreciation in Japan is the result of two factors unlikely to have lasting impact: the annual repatriation of cash ahead of March 31 fiscal year-end by indebted Japanese banks and companies holding foreign assets, and the good news on the US economy updating people's expectations of exports to the United States9. These transient effects are being amplified by aggressive tactics of the MOF and FSA, including:
Various investors and businesses in Japan will make their own assessments of when the debt-deflation and/or the public debt load is getting out of hand. Until that time, markets will be tentative at best about yen-denominated assets, but continuing to unload the better and more liquid of them. Brief sales of foreign (US) assets by Japanese banks to acquire cash will only feed further capital outflow with a lag when the banks are unwilling to lend repatriated funds domestically, and reach their limits on JGB purchases-the result will be a downward trend punctuated by occasional rallies, feeding volatility. The passage of time will only increase the actual and perceived likelihood of eventual crisis, bringing forward the day "eventually" comes to pass.
What a Crisis in Japan Would Look Like 11
Obviously, even an outright crisis in Japan would not be Argentina redux. Japan remains too wealthy, retains too large a creditor position, and its government institutions remain too entrenched, for there to be a repeat of the type of collapse seen in the southern cone. Yet, Japan has far greater international implications given its size and creditor status, and still has the potential for dauntingly negative economic outcomes.
In essence, a crisis in Japan would be a repeat of what happened there in end-1997/early-1998, but on a larger scale, and lasting more than a year. That would entail:
The deservedly unsympathetic international response would be likely to exacerbate matters. There would be extreme volatility in international capital markets, with flight of money out of Asia generally and cascading regional depreciations against the US dollar. This would put enormous political pressure on the US Congress from exposed industries, and on the US government from Beijing, Seoul, and other East Asian capitals, to respond with trade threats at a minimum13. Of course, there would also be some direct disruption to US markets from financial crisis in Japan, but the resulting economic effect would probably be limited—although US banks' exposure to Japan remains sizable ($33 billion as of September 2001), Japan's share in those banks' total outstanding foreign exposure has declined (especially their exposure to banks), and US banks' ability to withstand shocks is higher now than in the 1980s or early 1990s14.
China in particular would play any crisis in Japan opportunistically, seeking to demonstrate its leadership role in Asia (by maintaining its fixed exchange rate peg against the dollar) while criticizing Japan. Given the Japanese government's own nationalism and mounting insecurity vis-à-vis China, at least initially Japan would be prone to escalate rather than negotiate. Middle classes and politicians elsewhere in emerging Asia will respond angrily to Japan for exporting adjustment when they view themselves (justifiably or not) as having suffered in order to reform. This could destabilize some regimes, worsen trade conflicts with Japan, and shift additional exports to the United States (further complicating American attempts to build domestic support for the Doha trade round, particularly any efforts to put antidumping on the table).
Looking to the longer-term, Japan "hitting the wall" would be a two-part process, whenever it begins. The first part would be the realization of contingent liabilities, disintermediation, and capital flight, as described. The second hit would come once a large part of the Japanese financial system (not just banks) has to pass through public ownership, as a result of the crisis. This is standard for major financial crises, even in developed economies (such as Sweden in the early 1990s). There would then be an avoidable but likely mishandling of the resale and privatization of those assets by the Japanese government, particularly as far as foreign acquisition would be concerned. Such mishandling in that situation would deprive Japan of capital once again and put off the day of recovery. That bridge would only have to be crossed, however, another year or two out from when the first wall is hit. Over the medium-term of 3-6 years, one could expect Japan to finally get its financial system stabilized, with large-scale foreign participation15.
Potential Triggers for the Crisis
Recent market manipulation and good luck (with respect to the US recovery) will get the Japanese government through the March 31 fiscal year-end mark-to-market without horrible incident. At some point in the near future, however, a negative shock can be expected to force the Japanese government to take on what is now a contingent liability as an explicit (large) expenditure on the government's balance sheet. The most likely candidates are: bankruptcy of a regional or prefectural government, with collapse of the attending mid-size bank; pension fund obligations of a failed construction or retail company having to be assumed by the government; or a loss of savings to a number of Japanese citizens in a bankrupt life insurer. Alternatively, the government could be faced with such a claim and have to publicly renege upon it.
Either way, when that occurs, there will be a sudden spike in JGB rates, as well as increased scrutiny and fear about the rest of the contingent liabilities in that class (other local governments, other pension funds)16. There will also be some combination of capital flight out of Japan, from the more liquid parts of the market, and further withdrawal of domestic Japanese funds from the banking system into cash, gold, and postal savings. This will cause investment to contract. The crisis scenario spins out from there along the lines described above.
Even healthy economies have unexpected setbacks, and the ability of policymakers to respond to them determines their ultimate impact upon the economy. Consider the list of likely policy-driven shocks to the Japanese economy in coming months, which are also factors likely to lead to the mishandling and magnification of the effect of any exogenous negative shocks that will occur:
This would not be an easy set of hurdles to clear, even for an economy and a government in peak condition. The assumption that nothing will go wrong, not even an earthquake, seems unduly rosy. In fact, Japan has been quite lucky for the past few years, with a growing world economy, a relative lack of protectionist pressures against it, and, even by Japanese standards, an absence of direct political opposition. This is unlikely to last. And the underlying vulnerabilities of the Japanese economy from banking fragility, debt-deflation, and fiscal indebtedness make it ill-prepared to withstand even small setbacks.
1. Adam S. Posen, Restoring Japan's Economic Growth, Washington: Institute for International Economics, 1998, see Chapter 4.
2. This potential solution is discussed in Adam S. Posen, "Macroeconomic Prospects and Policy Options for Japan," January 15, 2002, revision of speech given at "Where Does the Japanese Economy Go From Here?" Conference, Columbia Business School, November 1, 2001: http://www.iie.com/publications/papers/posen0102.htm
3. See "FSA to Release Bank Inspection Results On Combined Basis: Yanagisawa," Nihon Keizai Shimbun, March 19, 2002.
4. Of course, if total official bad loan numbers, let alone more realistic independent estimates of outstanding bad loans were used, rather than the minimum bound of Classification IV loans, the ratio would be many times higher.
5. Posen, 1998, op cit, Chapter 3.
6. The drag from recession on tax revenue is the real source of Japan's rising deficits, rather than public works spending. See Kenneth N. Kuttner and Adam S. Posen (2002), "Passive Savers and Fiscal Policy Effectiveness in Japan," IIE Working Paper, forthcoming (to be published in Journal of the Japanese and International Economies).
7. Almost all local and prefectural Japanese governments are in significant debt, and fund their debt through (partly coerced) bank loans from local banks, rather than by issuing bonds.
8. Only once since 1984 has either exports or net exports made a positive growth contribution of 1% or more to Japanese GDP. See Richard Jerram, "Exports Set to Lead the Recovery," Economic News, ING-Barings, March 20, 2002.
9. There seems to have been an overly optimistic reaction on this latter score. All available evidence suggests a steadily and significantly declining elasticity of Japanese exports to yen depreciation. See Satoru Horibe, "Can the Yen's Depreciation Help to Cure the Japanese Economy?", Tokyo-Mitsubishi Review, 7(3): February 2002, and Mikihiro Matsuoka, "Is yen depreciation a panacea? Evidence of declining support to corporate profits," Deutsche Bank Group Economic Research, February 25, 2002. Of course, there are also political and economic limits to how large one can expect US net imports and the trade deficit to increase from present levels. See Catherine L. Mann, "How Long the Strong Dollar?", IIE Policy Brief, forthcoming.
10. On March 22, 2002, the Asahi Shimbun reported that " 'many traders suspect the gains are just the result of excessive intervention by the financial authorities'…verbal orders and threats made privately by officials have stopped traders from making legitimate sales. Fundamentally what FSA did was to spring a bear trap, the paper said, forcing the shorts to scramble to cover their positions. To reinforce it, Asahi added, officials have 'persistently intervened in day-to-day activities of market players." Translation in The Daily Japan Digest, March 22, 2002, p. 1.
11. I am indebted to Fred Bergsten, Mike Mussa, Marc Noland, and Ted Truman for discussions thrashing out this scenario, but imply no responsibility of them for this forecast.
12. For a discussion of the credit crunch in Japan, see section 4 of Kenneth N. Kuttner and Adam S. Posen, "The Great Recession: Lessons for Macroeconomic Policymakers from Japan," Brookings Papers on Economic Activity, 2001:2, pp. 93-185. The IBJ bond spread shown in figure 5 is representative—all traded major bank bonds require large premiums over JGBs to be held at present. In other words, despite the disappearance of the traditional Japan premium (on the three-month LIBOR-based yen), Japanese banks are again paying a premium to borrow. James Fiorillo, "Japan Premium Back with A Vengeance," Industry News, ING-Barings, February 20, 2002.
13. In the U.S., the industries most heavily hit would be the usual suspects of auto parts, electronics, and steel, the latter of which has certainly demonstrated its protectionist clout. For an assessment of the likely trade effects of a yen depreciation, see Marcus Noland, Sherman Robinson, and Zhi Wang, (1999), "The Continuing Asian Financial Crisis: Global Adjustment and Trade," IIE Working Paper No. 99-4: http://www.iie.com/publications/wp/1999/99-4.htm
14. I am grateful to Ted Truman for this assessment.
15. The argument for a long-term convergence between U.S. and Japanese financial sectors, and the implications of that for trans-Pacific relations, is given in Adam S. Posen (2002), "Finance," in U.S.-Japan Relations in a Changing World, Steven Vogel, ed., Brookings Institution Press, forthcoming, http://www.iie.com/publications/wp/2001/01-8.pdf
16. Morris Goldstein, The Asian Financial Crisis: Causes, Cures, and Systemic Implications, IIE, 1998, argues that a similar "wake up call" to investors regarding a class of debt contributed to the Asian crisis.
17. Some ¥27trillion (equivalent to over 5 percent of GDP) had left uninsured accounts in the year to January.
18. According to our rough estimates, if Koizumi's debt cap is maintained, and growth is 1.5 percent less than government forecasts in FY2001 (that is growth in line with most private sector forecasts), the needed public sector contraction will take nearly another 1 percent off of GDP over two years. See Kuttner and Posen (2002), op cit.