Non-Performing Loans As a Result of Deflation

Yutaka HARADA (Vice President, Policy Research Institute, Ministry of Finance)

Few Empirical Studies on Cause and Effect

There are various arguments on relationships between non-performing loans (NPLs) in the banking sector and deflation or long-term economic stagnation in Japan. Some argue that Japan's long-term stagnation since the 1990s has been caused by the delay in the disposition of NPLs and, therefore, solving the NPL problem will lead to Japan's economic recovery. Others insist that creation of new NPLs due to deflation is the main cause for the delay in the disposition of NPLs, and that the NPL problem could not be solved without stopping deflation.

Either argument is logically possible, but there has been virtually no empirical study to determine which is the case, due to the lack of long-term data on NPLs. Even if it is true that deflation delays the disposition of NPLs, there would be no consensus as to how to deal with deflation.

The purpose of this paper is to show empirically whether NPLs, defined as firms' excess debts relative to their cash flows, are causing deflation and recession or vice versa. NPLs are firms' excess debts relative to their cash flows (profit plus depreciation). A firm's debt, however large it might be and however small its collateral might be, is not an NPL so long as there is enough profit for the firm to pay it off within a reasonable period of time.

Having defined NPLs this way, we can consider possible channels in which deflation causes NPLs and vise versa. First, let us take up a possible channel in which deflation cause NPLs.

A firm's profit is its revenue minus material cost, capital cost and labor cost. When revenue declines due to deflation, profit will decrease by the same percentage, even if all the costs decline at the same rate as deflation. If the rate of deflation is 2%, then profit will decline by 2%.

Furthermore, there are reasons why profit is likely to decrease faster than revenue. First of all, the amount of principals for debts does not decrease by deflation. Second, it is often the case that interests do not decline by the same percentage as deflation. This is because interests tend to be fixed for a few years and also because the nominal interest rate cannot go below zero. Third, it is quite difficult to cut wages because nominal wages have downward rigidity. As profit is normally a very small percentage of total revenue, profit could be reduced drastically for the three reasons cited above.

Thus, we have identified the mechanism in which deflation leads to a decline in profit and an increase in excess debt. Furthermore, we can think of the effect of the decline in profit, which leads to a decrease in corporate investment, affecting aggregate demand due to the multiplier effect.

Significant Effect of Deflation on Excess Debts

Next, let us consider a theoretical channel in which excess debts cause deflation. There may be a channel in which NPLs weaken banks' function of financial intermediation, leading to decreases in investment and total demand, and resulting in deflation. However, there are other ways of raising funds than through banks, such as through the securities market. If that works, the possibility that NPLs cause deflation seems to be rather small.

What has been said so far remains only a logical possibility. Let us now estimate the magnitude of various effects. First, we have estimated quarterly data of excess debts, based on MOF's quarterly corporate statistics regarding debts (D) and cash flows (C), by identifying excess debts as the "excess" ratio of D to C over the "normal" ratio of D to C during the "normal" period (1975-84) before the bubble appeared.

According to the estimated data on excess debts, it increased drastically from the bubble period to the bubble-bursting period, and then increased further in recessionary periods and decreased during recovery periods in the 1990s.

Next, we have constructed a model of the Japanese economy including the excess debt variable. It consists of seven variables, namely, excess debt balance, real public capital formation, money supply (M2 plus CD), exchange rate, domestic wholesale price level, real wage rate, and real GDP. We have estimated this model by using quarterly data for the period from 1982 through 2001, when the value of excess debts remained positive after the oil crisis period, and tested whether deflation leads to NPLs or vice versa.

According to our estimation result on the effect of NPLs on deflation, a 1% decrease in the excess debt balance would lower the domestic wholesale price level only by 0.003%. This is a negligible effect. In contrast, the effect of deflation on NPLs has turned out to be much larger. A 1% increase in the domestic wholesale price level would reduce the excess debt balance by 21.1%.

We have estimated what would have happened to the excess debt levels if there had been no change in the domestic wholesale price level since the April-June 1991 period, while in reality the wholesale price level has been declining since then. As shown in the graph, the result is that the excess debt balance would have declined to one fourth of the actual level if deflation had been avoided in the 1990s.

With regard to the relation between excess debts and economic growth, we have found that the effect of a reduction in NPLs on real GDP is statistically insignificant. On the other hand, a 1% increase in real GDP would lead to a 2.8% decrease in the excess debt balance.

In summary, we can maintain that the effect of excess debts on deflation is negligible compared to the effect of deflation on increases in excess debts, and there is no evidence found for any negative effect of excess debts on the real economy, whereas there is clear evidence that a recession increases excess debts. However, to reduce the excess debt balance, the effect of a real economic recovery is smaller than the effect of elimination of deflation. That means that stopping deflation is shown to be most effective in reducing excess debts.

Monetary Policy to Reverse Price Trends

The question is how to stop deflation. It is important to adopt appropriate monetary policy in accordance with the basics of macroeconomics, where inflation and deflation are monetary phenomena from the long-run viewpoint.

It does not seem correct to argue that economic development in China would bring about deflation, since Japan is the only major country that is in deflation, while China is exporting to so many countries in the world. For example, the ratio of Chinese imports to GDP is 1.0% for the U.S. and 1.2% for Japan. That is not much different from the U.S., but only Japan is in deflation.

Furthermore, there is no other country in the world that has left deflationary trends unchecked. The U.S., Germany, France, and the U.K. all adopted significant monetary easing policies by increasing money supply in order to avoid the formation of deflationary expectations when faced with decreases in the rate of consumer price increase below 1 or 2% in the 1990s.

Money supply can ultimately be increased by BOJ's market operations to purchase more government bonds on a permanent basis. There is an argument that money supply could not be increased since BOJ's purchase is falling short of its planned amount in the market. But this only implies that there are general expectations in the market that government bond prices would not decrease (interests would not increase) in the long run and BOJ's deflationary policy might well continue in the future.

Therefore, BOJ must express its strong determination to stop deflation, while trying to purchase more government bonds in the market. Then it will succeed in its market operations, increasing money supply and stopping deflation.

Stopping deflation could not possibly lead to hyperinflation. The reason why Japan suffered from hyperinflation after the end of the Showa Great Depression was that the Japanese military ordered huge amounts of fiscal expenditures and money supply at that time. There is no such force what so ever in today's Japan.

It is extremely urgent that quantitative easing of money be adopted to stop deflation without hesitation. Stopping deflation is a necessary first step to the solution of the excess debt problem in Japan.

(Translation of the original Japanese article that appeared in the February 19, 2002 issue of Nihon Keizai Shinbun)