FRBSF Economic Letter
1996-23 | August 9, 1996
In September of 1995 the Bank of Japan (BOJ) reduced its discount rate to 0.5 percent, a post-war low not only for Japan, but for the entire OECD. The usual interpretation of interest rate movements would imply that the BOJ has been engaged in an aggressively expansionary monetary policy, as lower interest rates tend to stimulate demand. However, other conventional gauges of monetary policy suggest the opposite. At the same time interest rates have fallen to historic lows, money growth and inflation have dropped as well. Inflation is virtually nonexistent, and money growth (as measured by M2) has been only about 2 to 3 percent per year. These figures suggest that Japanese monetary policy has not been expansionary at all.
How are we to resolve this ambiguity? The key is to realize that by itself an observed change in a price or quantity can never reveal the underlying source of the change. Just as an increase in the price of gasoline cannot tell us whether the demand for gas has increased or whether the supply has decreased, lower interest rates cannot tell us whether the supply of money has increased or whether the demand for it has decreased. Instead, we must consider the simultaneous movements of both prices and quantities if we are to disentangle the effects of supply and demand. This Letter discusses how this is done in the context of monetary policy, and then reviews the results of a recent study by Kasa and Popper (1995), who apply these procedures to measure the policies and operating procedures of the Bank of Japan during the period 1975-1994.
Figure 1 and Figure 2 depict hypothetical supply and demand curves of money, with the quantity of money (m) on the horizontal axis and the interest rate (i) on the vertical axis. The demand curve is negatively sloped since the quantity of money demanded declines as interest rates rise. For example, as interest rates rise individuals transfer funds out of their checking accounts and put them into higher-yielding assets like Treasury Bills. On the other hand, the supply of money is positively sloped because when interest rates rise it becomes more costly for banks to hold non-interest-bearing reserves, and therefore they extend more loans and issue more demand deposits.
Now, monetary policy acts by shifting the money supply curve. For example, an “easing” of monetary policy would shift the curve down and to the right, as depicted in Figure 1. The position of the demand curve, on the other hand, is determined by the interactions among banks, businesses, and households in the credit markets. Hence, shifts in this curve primarily reflect changes in current or anticipated business conditions. For example, expectations of a weakening economy tend to depress business loan demand, which shifts the schedule down and to the left, as depicted in Figure 2.
From these Figures we see that the recent combination in Japan of low interest rates and low money growth is suggestive of a demand contraction rather than a supply expansion. That is, recent Japanese experience is more consistent with Figure 2 than with Figure 1. In this particular case it is easy to identify the demand shift. Namely, the recession in Japan since 1991 has reduced Japanese firms’ demand for money and credit. However, the situation is probably a little more complicated. At the same time loan demand has been shrinking, there have likely been opposing forces acting on the supply curve. On the one hand, the continuing bad loan problem in the Japanese banking sector has tended to reduce credit supply as banks attempt to restore their capital ratios. On the other hand, expansionary efforts by the BOJ (e.g., through discount rate reductions) have probably counteracted this. With these offsetting forces acting on the money supply, a downward shift in demand appears to have predominated.
Of course, this is a highly stylized and simplistic account of recent events in Japan. In practice, policymakers are not magically endowed with knowledge of supply and demand curves. All they observe is a confusing sequence of price/quantity pairs that must somehow be interpreted. This can be quite difficult since during any given period many things typically happen that cause both curves to shift. Moreover, the world is not as static as this textbook story suggests. Adjustment costs and other rigidities are pervasive in real world markets, and this makes demand and supply curves dynamic, and therefore subject to the vagaries of shifting expectations. Despite the difficulties involved, it is nonetheless essential that policymakers be able to distinguish the underlying supply and demand factors driving the market. After all, an inflation caused by falling money demand (perhaps due to technological changes) calls for an entirely different policy response than one caused by the over-extension of bank credit.
In a recent paper, Kasa and Popper(1995) applied newly developed econometric methods to the problem of distinguishing supply and demand factors in Japanese monetary policy. These methods were developed by Bernanke and Mihov(1995), who used them to study U.S. monetary policy. Three main results emerged from our analysis. First, the process of Japanese financial market liberalization that began in the late 1970s has produced a convergence of operating procedures between the BOJ and the U.S. Federal Reserve. Second, no single target or operating procedure seems to describe the behavior of the BOJ. Instead, the BOJ seems to care about both money supply volatility and interest rate volatility. Third, we find little evidence to support the conventional view that BOJ policy independently produced the so-called “asset-price bubble” of the late 1980s, and then subsequently “burst” it. The remainder of this Letter will explain these findings in more detail.
In the United States, monetary policy is conducted within a well-developed and relatively unregulated financial market. For example, when the Federal Reserve wants to lower interest rates it simply buys government securities on the “open market.” This adds to bank reserves and drives down the federal funds rate (i.e., the interest rate banks charge each other on overnight loans). In contrast, until 1979, when the process of capital market liberalization began in earnest, the government securities market in Japan was not well developed, particularly at the shortend of the maturity spectrum. For example, six-month government bonds were first issued in 1986, and three-month bonds were not issued until 1989. Moreover, high transactions costs due to taxes have inhibited the trading of near-maturity long-term bonds. As a consequence, the BOJ had to resort to more direct methods when it wanted to change interest rates or the money supply. In particular, the BOJ relied on the “discount window” to influence credit market conditions. That is, rather than buy or sell securities on the open market, the BOJ lent money directly to commercial banks, typically at below-market rates. Naturally, the possibility of borrowing money from the BOJ at low interest rates and then lending it out at higher rates led to a state of chronic excess demand at the discount window. The conventional view of Japanese monetary policy is that the BOJ exercised control over the money supply and interest rates by altering the degree to which it accommodated demand at the discount window. This sort of rationing policy is sometimes referred to as “moral suasion.”
Our results support the notion that the BOJ has at times engaged in moral suasion. We capture the degree of excess demand at the discount window by the spread between a market-determined Certificate of Deposit (CD) rate and a more regulated interest rate (i.e., the Tegata rate). (CDs did not exist in Japan until 1979, so for earlier years we use the Gensaki (repurchase) rate which, like the CD rate, was relatively unregulated.) The idea is that a widening of the CD/Tegata spread indicates growing excess demand and a tightening of BOJ policy. Interestingly, we found that both the use and the effectiveness of moral suasion have declined over time. In particular, we split our sample into two halves, one running from 1975 through 1984, and the other running from 1985 through 1994. Only in the first half of the sample was moral suasion significant. We attribute the declining role of moral suasion to the gradual liberalization of Japanese financial markets. As markets have become both broader and deeper, the BOJ seems to be relying less on direct quantity controls, like discount window rationing, and more on the market mechanism.
Our second result relates to the targets and operating procedures of the Bank of Japan. Of course, like most other central banks the BOJ ultimately cares about growth and price stability. The real question, however, is how best to achieve these goals, and here opinions differ. “Monetarists” advocate the targeting of monetary aggregates as the surest route toward maximum growth with price stability. “Keynesians”, on the other hand, typically advocate an interest rate target. (See Poole (1970)).
One of the advantages of the Bernanke-Mihov methodology is that it allows one to determine econometrically which target the central bank follows, while imposing few restrictions on the ultimate goals or effects of monetary policy. In their analysis of U.S. monetary policy, Bernanke and Mihov find that the Fed’s target has changed over time, but that currently it seems to be targeting the federal funds rate. Perhaps not surprisingly, we find similar results for Japan. During the first half of the sample the BOJ seemed to weight both money supply stability and interest rate stability. During the second half of the sample the BOJ seemed to focus more on limiting interest rate fluctuations (although it continued to respond to both). This suggests that simple univariate characterizations of Japanese monetary policy are likely to be misleading.
Finally, the defining event of recent Japanese economic history was the run-up in land and equity prices that occurred between 1986 and 1990, and the crash that followed. The legacy of bad debt this episode produced still plagues recovery efforts in Japan. Some have blamed the BOJ for both the bubble and the subsequent crash. The claim is that the BOJ was too expansionary during the late 1980s, then overreacted in its efforts to rein in the escalation of asset prices. In response, the BOJ has argued that at the time it faced international pressure to bring down the value of the yen through a policy of low interest rates. At the same time, it was worried about the potentially adverse effects that an overly strong yen would have on Japan’s important export industries.
Our analysis is useful because it sheds indirect light on the BOJ’s arguments. In particular, the methodology we use distinguishes between “policy as usual” (i.e., policy that is consistent with past policies given the current state of the economy), and “policy shocks,” which represent an unanticipated departure from past policies. Perhaps surprisingly, we find that Japanese monetary policy during this period was for the most part consistent with past policies given the state of the economy (e.g., the value of the yen, the inflation rate, and the level of industrial production). That is, policy during this period was indeed first expansionary and then turned abruptly toward contraction, but these policy moves appear to be consistent with the way the BOJ had been responding in the past to the state of the economy. From this perspective, the apparently sharp monetary expansion during the late 1980s probably reflected a combination of an outward money demand shift brought about by a buoyant economy (remember Figures 1 and 2!), and an expansionary BOJ policy occasioned by yen appreciation. Of course, this does not imply that had the BOJ taken different actions the whole episode would not have been avoided.
Santa Clara University
Bernanke, Ben S., and Ilian Mihov. 1995. “Measuring Monetary Policy.” NBER Working Paper No. 5145.
Kasa, Kenneth, and Helen Popper. 1995. “Monetary Policy in Japan: A Structural VAR Analysis.” Pacific Basin Working Paper 95-12, Federal Reserve Bank of San Francisco.
Poole, William. 1970. “Optimal Choice of Monetary Policy Instruments in a Simple Stochastic Macro Model.” Quarterly Journal of Economics, pp. 197-216.
Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.
Please send editorial comments and requests for reprint permission to
Attn: Research publications, MS 1140
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120