March 28, 2003
Despite all the chatter about how to stop deflation, there has been insufficient analysis of where Japan's deflation comes from. Supply and demand factors do not explain the whole story. Deflation is also a monetary phenomenon. And if the causes of deflation are monetary in nature, the policies to combat it must take these causes into consideration.
Deflation implies the amount of money in the economy is not growing. Because the money supply is not increasing, prices fall and the country experiences deflation. Money supply is the sum of cash, ordinary deposits, and time deposits (M2) plus convertible deposits (CD). Normally, if the central bank boosts the supply of high-powered money, credit is created in the banking sector and as a result M2 expands by about ten times the amount of high-powered money supplied.
But since the 1990s, the extent to which the amount of high-powered money supplied by the Bank of Japan is transformed into the money multiplier (M2 plus CD) has been decreasing. Since the end of 2000 in particular, when the BOJ began to increase the amount of high-powered money aggressively through quantitative easing measures, the money multiplier shrank at a rapid pace, as if to nullify the increase in such base money.
Why does an increase in high-powered money no longer boost money supply? I suspect that the under-capitalization of banks has led to the decline in the money multiplier and the lack of growth in money supply.
If the banking sector falls short of capital, the risk of bank failures increases. Although deposits are protected by deposit insurance in case the banks collapse, a degree of inconvenience, such as in deposit withdrawals and account transfers, cannot be avoided. When such conditions prevail, trust in bank deposits is shaken, and, as a result, depositors withdraw their deposits and increase their cash holdings. In other words, M2 decreases because the amount of money in bank deposits available for settlements falls compared to the amount of available cash.
This mechanism is not merely a theoretical possibility. A study group led by Professor John Boyd of the University of Minnesota and Professor Bruce Smith of the University of Texas-Austin analyzed 23 countries that have experienced banking crises in recent years. A banking crisis was defined as a situation where banks had non-performing loans of at least five to ten percent of their total lending, as once the ratio of non-performing loans exceeds this level, their net worth effectively falls below zero.
The study confirmed that when such a crisis occurs, M2 declines sharply and inflation rates fall. In an economy where the inflation rate is low, there is a positive correlation between inflation and economic growth, and Smith says that economies likely slow during banking crises because of a decrease in M2.
Looking at data for Japan, we can confirm that the amount of cash held by the non-financial sector has been increasing. Up to the 1980s, virtually all of the cash supplied by the BOJ was absorbed by the banking sector. Yet since 1990, cash increases have mostly flowed out of banks and into non-financial firms. This is a clear indication of a loss of trust in bank deposits. Incidentally, where a decrease in M2 (deposit liabilities) is shown on bank balance sheets, a decline in cash is offset by a decline in securities holdings, and lending does not show any marked decrease. These observations match those of Boyd's group: in recent banking crises, bank loans do not fall.
These results show that while banks are not entirely responsible for deflation, the under-capitalization of banks and the loss of trust in the settlement system that ensues is a cause of deflation.
So re-capitalizing the banking system may help combat deflation. Shoring up banks' capital will help them regain depositor confidence, and the public will deposit more of their assets at banks. This in turn will lead to an increase in M2, and allow Japan to eradicate deflation. This is the same mechanism by which the United States was able to rid itself of deflation in 1933, during the Great Depression.
Boyd's group also reports that for every one percent of GDP spent on revitalizing the banking system, through means such as liquidating failed banks and re-capitalizing banks, the rate of M2 growth increases by five percentage points. In other words, in countries with excessive non-performing loans, reinforcing the capital of the banking system to improve its soundness helps increase money supply and induce inflation.
It is important for banks to receive capital injections at an early stage if deflation is to be tackled, so recent efforts by Japanese mega banks to procure large amounts of capital should be praised. Nevertheless, in times of banking crises, problems such as the asymmetry of information make it difficult for banks to procure sufficient capital, as uncertainty about the banks discourage private investors from putting up capital.
If political difficulties could be ignored, deflation could be eradicated by an injection of public funds into the banks at an early stage, and then selling the bank equity holdings to private investors once trust in the banking system were restored.
But simply declaring that public funds will be injected into banks "when deemed necessary," as is now the case, is insufficient to restore depositor trust in the banking system. In fact, it will more likely aggravate the economic downturn and deflation, as banks that do not want capital infusions will rush to condense their assets, thus leading to a credit crunch and aggressive loan collection.
An effective policy to fix the banking system and at the same time cure deflation would be to inject public funds into banks and have the Bank of Japan finance this policy by purchasing government bonds. That way, public trust in bank deposits could be restored, while high-powered money should increase significantly, resulting in an increase in money supply and a rise in the inflation rate.
The side effect of such a policy would be that, in the short term, government debts would expand sharply, heightening concerns about the future of the nation's finances. Uncertainty will lead to anxiety about the future and create shrinkage in personal consumption and capital investment.
Therefore, consumption and capital investment will recover if such uncertainties were dispelled through the prompt presentation of a schedule for future tax hikes and spending cuts. The government should swiftly decide on a social security system and a schedule for fiscal reconsolidation that includes drastic reforms in the expenditure structure, such as reductions in civil servant personnel costs as well as reforms to the overall tax system. As a precondition, the government should at the same time approve fiscal outlays to boost bank capital.
Finally, inducing inflation through unconventional monetary measures, such as the purchase of stocks and property, might be the second best policy if bank re-capitalization is not making headway due to political and regulatory constraints. But it is a mistake to believe that inflation can be induced without fiscal burden. If the BOJ were to purchase stocks and property with low profitability, it would effectively be putting non-performing assets on its balance sheet, and mild inflation of two or three percent would not be enough to absorb such latent losses. Those losses will eventually have to be repaid through tax revenue. So, it should be recognized that unconventional monetary policies are in fact fiscal policies that have been put on the backburner.
A further conclusion of the above study was that countries that solved their banking problems through inflation saw repeated recurrences of non-performing loans and banking crises. This should also be kept in mind.
(This essay was reprinted from the Nikkei Shimbun, Jan. 28, 2003.)
Author, Keiichiro Kobayashi
Fellow
Research Institute of Economy, Trade and Industry (RIETI)
Editor-in-Chief, Ichiro Araki
Director of Research
Research Institute of Economy, Trade and Industry (RIETI)
e-mail: araki-ichiro@rieti.go.jp
tel: 03-3501-8248 fax: 03-3501-8416
RIETI invites you to visit its English website
[http://www.rieti.go.jp/en/index.html].
The opinions expressed or implied in this paper are solely those of the author, and do not necessarily represent the views of the Ministry of Economy, Trade and Industry (METI), or of the Research Institute of Economy, Trade and Industry (RIETI).
