Lessons from Japan's Banking Crisis

3.1 Reasons for the Delay in Decisive Action

There are several reasons for the delay in crafting and implementing decisive policy action. First, the initial approach adopted in 1991–1997 was based on the expectation that a resumption of economic growth would restore the financial health of banks and their client borrowers. Land prices were also expected to bottom out soon and resume rising again. The asset price bubble experienced in the late 1980s was the first in post-war Japanese history, and thus it took time to even recognize the situation as a bubble. It was in 1993 that the government provided the detailed analysis that characterized the event as a “bubble” in its white paper on the economy. Until the bursting of the bubble, a strong belief—called the “land myth”—had prevailed, that land prices would never decline.

Once the bursting of the asset price bubble began to damage bank balance sheets, there was still a delay among policymakers in recognizing the severity of the impacts of the asset price collapse on bank NPLs and of bank NPLs on the real economy. It was only after the economy faced a systemic banking crisis in 1997–1998 that the authorities began to take decisive measures as described in the previous section. Until then, the authorities had underestimated the seriousness of the impact of the declining real estate prices on financial institutions and macroeconomic conditions, and the powerful “adverse feedback loops” between the financial sector and the real economy. As a result, lingering fears about the solvency of banks, which eventually proved founded, persisted in the market.

Second, despite the long stagnation of the real economy in the early 1990s, there was no significant domestic pressure (due to high savings, low inflation, relatively low levels of unemployment, no fiscal crisis, and no social unrest) nor external constraints (due to large foreign exchange reserves, a large net external asset position, no capital flight, no balance of payments difficulty, and no currency crisis) which otherwise could have prompted the government to accelerate the resolution of banking sector problems. Possessing enough fiscal space allowed the government to resort to Keynesian fiscal policy in order to support aggregate demand and help insolvent corporations survive, particularly in the construction sector. If domestic and external constraints had bound the government, then it would have been forced to address the crisis earlier and more decisively.⁸

Third, the crisis itself had developed slowly and gradually, because the problem was confined to bank loans, and the accounting and disclosure standards were slow to reflect a change in economic value. There were indeed imperfections in accounting and disclosure standards, which enabled financial institutions to avoid recognizing loan losses. Partly because of this, banks were insufficiently incentivized to promptly address the NPL problem. Nonetheless, the government could have sent a clear message encouraging the quick write-off of troubled assets through either tax incentives or other devices. Rather, a forbearance policy was introduced, allowing banks to report the costs, rather than the market price, of equities and real estate assets on their balance sheets, which may well have further postponed taking decisive actions.

Fourth, the authorities lacked the legal framework to resolve troubled, large financial institutions, which fuelled their delay in adequately addressing problems at these institutions. In order to ensure the timely resolution of troubled financial institutions, it is crucial that the authorities possess a legal framework for resolution, specified operational procedures, and sufficient funds to cover capital shortages. It was only in 1998, after a series of failures of large financial institutions, that the full-fledged safety net framework was put in place.

It is also important to be prepared for crisis situations by revising bankruptcy and foreclosure laws, because often the effective disposal of troubled assets requires the resolution and restructuring of the borrower firms. In addition, it is critical to convince the politicians how important it is to contain the problem at an early stage, particularly when a policy package involves public funds.

3.2 Lessons from Japan's Experience: Banking Sector Policy Issues

To contain a systemic banking sector crisis, comprehensive policy measures should be designed and implemented, including rigorously assessing major banks' balance sheets, removing NPLs from them, and recapitalizing such banks. These measures should ideally be based on providing appropriate incentives for private banks, but market-based incentives may not be available under severely stressed market conditions. Sometimes instituting quantitative guidelines and/or regulatory mandates is necessary to force banks to undertake decisive actions.

Four lessons can be learned from Japan's banking crisis experience. First, in order to address a banking crisis properly, prompt action to gauge the exact amount of loan losses is a critical initial step, although this is not an easy task. In many cases, problems arising from insufficient liquidity and high risk aversion obscure the intrinsic value of the troubled assets. Assets are regarded as troubled either because of the expected loss of future cash flows or the sudden loss of liquidity on the part of borrowers. While economic conditions may dictate that the loss of future cash flows may not be recovered, the liquidity value may at least partly be recovered once normal access to markets is restored. Practice may dictate revising estimates as the crisis develops. If the crisis is global, then international financial institutions may estimate the total amount of loss; the national authority also should undertake this task for the purpose of designing recapitalization and other policies.

In the 1990s, there were no well-functioning markets for credit risk products that might provide a good measure of the market price of credit risk. From the time of the bursting of asset price bubbles in Japan, the extent of deterioration in bank asset quality posed the greatest uncertainty. The regulators did not even clearly define NPLs until 1998 when financial reconstruction schemes and prompt corrective measures were introduced. Today, a system for timely disclosure of NPL levels is in place; however, some ambiguity about the valuation of disclosed bank assets still exists when the market is under severe stress. This is especially true for newly introduced innovative products.

Second, a government recapitalization operation that involves taxpayer funds is the most direct policy measure to contain the acute phase of market turmoil.⁹ Public recapitalization can be effective if the size of the operation exceeds a certain threshold. In the case of Japan, the first capital injection of JPY1.8 trillion in March 1998 was considered to be very meager as compared to the total amount of NPLs, which amounted to JPY22 trillion at that time. With hindsight, the second capital injection, of JPY7.5 trillion, made in March 1999 was more effective. Table 4 [ PDF 49.7KB | 1 page ] shows the relative importance of government capital injections in maintaining sufficient capital adequacy levels for banks.¹⁰

In undertaking these recapitalization operations, a strict examination of bank assets was needed, and so the Financial Reconstruction Commission examined the state of banking sector health. However, its reports to the National Diet were unsatisfactory, as they failed to reveal the state of bank balance sheets. As a result, uncertainty about the magnitude of possible loan losses and NPLs added turmoil to markets.

In theory, a strict asset assessment should be conducted before recapitalization operations. However, in reality, assessments are generally not conducted beforehand, because of the rapid pace of market developments. Furthermore, as observed in Japan, it requires months, if not years, for the authorities to fully assess banks' balance sheets. Even recently in the US, the process of stress-testing required several months. When the market is under extreme stress, the resultant urgency may not allow adequate time for extensively scrutinizing bank books before undertaking recapitalization operations.

In addition, in a democracy, bank recapitalization often requires parliamentary approval. So the success and speed of the operation depend on how well politicians understand the problem and are willing to support it.

Third, the removal of impaired assets from banks' balance sheets is essential to the restoration of bank health. A government initiative to purchase bank assets is often necessary to restructure bank balance sheets during a crisis, as when markets lose their ability to determine prices, the government is better able to maintain flexibility in timing and so could realize higher values for those troubled assets. To share the upside benefits, one of the possible approaches is to simultaneously purchase impaired assets from, and the preferred stocks of, the troubled but viable institutions. To the extent that the troubled banks restore their financial health, taxpayer funds could be retrieved either by higher asset prices or by dividends. This illustrates a complementary role of recapitalization and asset purchases.

The pricing rule makes it difficult to design an effective public asset purchase program. If the purchase price set by a government-sponsored asset management company is too low, then no bank would be willing to sell. If the purchase price is too high, then the program involves a transfer of taxpayer funds to banks which made bad decisions.

In Japan, initially the Resolution and Collection Corporation (RCC) purchased assets from failed banks and, beginning in October 1998, from solvent banks as well. The total value of its purchases between October 1998 and March 2005 was JPY353 billion, with a book value of JPY4 trillion. According to Hoshi and Kashap (2008), Japan's experience with asset management was, at best, mixed because of its contracts limitation and the small scale of operations. They argue that, most importantly, the purchase of NPLs did not fix the capital shortage problem, as the size of asset purchases was not large enough to restore sufficient capital adequacy ratios. The RCC's utility can be found in its provision of an opportunity for banks to remove troubled assets from their balance sheets even in the absence of demand in the market.

Fourth, economic stagnation can cause new NPLs to emerge rapidly, and deplete bank capital. As discussed by Fukao (2007), capital injections without economic recovery are not effective. Even if the initial market turmoil is contained through providing liquidity and recapitalizing banks, banking sector troubles could recur without the recovery of the real economy. Figure 3 [ PDF 52.8KB | 1 page ] shows the relationship between real GDP growth and changes in the outstanding amount of NPLs— as measured by the size of “risk management” assets of all banks—on an annual basis. In years when GDP grew less than 1%, the outstanding NPLs rose, and in years when GDP grew more than 1%, NPLs declined except in FY2000, the only year in the lower right quadrant.

This suggests that macroeconomic policy plays an important role in support of economic activity, which could both minimize future losses of banks and encourage the flow of risk funds into the capital markets.

Download this Paper [ PDF 238.6KB| 23 pages ].

Comment(s)

There are [0] comment(s) for this entry. Post a comment.