No. 34 — September 1, 2000


Weekly Review

--- by Arthur J. Alexander

The Bank of Japan's August 11 vote to boost the interest-rate target for uncollateralized overnight interbank funds to 0.25 percent from zero (see JEI Report No. 32B, August 18, 2000) ended a unique 18-month policy that had dropped the cost of money to levels not seen since 14th-century Italy, according to some monetary historians. Shortly before the decision to reverse course was made, the rate on the benchmark Japanese government bond was 1.7 percent and bank deposits yielded a scant 0.05 percent. One largely unexplored issue is how the central bank executed its February 1999 move toward zero interest rates and then undid it.

In recent years, the central banks of most advanced economies have implemented monetary policy by announcing a target interest rate, usually for funds lent overnight between banks. A central bank then adds or subtracts funds from the overnight market to bring the market-determined rate into line with its policy target. Overnight market funds consist mainly but not exclusively of money in excess of the amounts that banks need to meet central bank reserve requirements, defined as a fraction of total bank deposits. As is typical in open-market operations, BOJ injects money into the market by buying all kinds of securities, although most are government issues; it drains liquidity from the system by selling securities.

Economists are prone to draw sweeping curves showing the supply and the demand for money, giving little regard to the mechanics of how central banks actually work the supply side of the equation. Central bankers as well as participants in the money market, however, have to pay close attention to the details.

Every day, any one government undertakes millions of transactions that add to or subtract from the economy's supply of money. Tax payments by individuals and companies take money out of circulation. Salaries paid to civil servants and checks written to government contractors put money into play. Government bonds floated to cover deficits and old bonds redeemed as they come due also affect the money supply. A central bank must counter all these transactions just to keep the money supply on an even keel and, in addition, inject or withdraw funds to implement its current monetary stance.

Japan's central bank is able to monitor government payments and receipts because it acts as Tokyo's banker. All transactions go through its coffers. Over the years, BOJ money-desk officials have developed a sense of the seasonal nature of the flows, but they still monitor contracts and other transactions for any unusual transfers that might require action.

At the beginning of each month, BOJ publishes its estimate of the supply and the demand for funds in the coming 30 days. This forecast is updated daily, and the central bank then alerts the money market as to what to expect the next day. The bank enters the Tokyo market at least twice a day to implement its plans, which are adjusted during the course of that time in response to actual transactions.

Every day after 5:30 p.m., BOJ publishes the results of its daily operations. These may be somewhat different from the projected actions because of the less-than-perfect predictability of the millions of transactions to which BOJ must be responsive. Interest rates, therefore, may deviate from the target simply due to the technical difficulties of matching the supply and the demand for funds. These daily uncertainties were part of the reason that it took about four days for the uncollateralized overnight interbank rate to hit BOJ's 0.25 percent target. Every morning, the bank gingerly tested reactions by gradually reducing the funds it supplied.

In addition to the routine job of maintaining the overnight interbank interest rate at or near the target level, BOJ directs monetary policy. To implement the zero interest-rate policy that it adopted in February 1999 (see JEI Report No. 7B, February 19, 1999), the bank injected ¥1 trillion ($9.1 billion at ¥110=$1.00) beyond what banks needed to meet their reserve requirements into the money market every morning. These funds earned zero returns but cost almost nothing to borrow. Commercial banks could have obtained this money and lent it out at a higher interest rate, thereby earning a profit.

However, companies with access to financial markets were raising needed funds directly or were tapping internal cash flows. Given Japan's then-stagnant economy, small and midsize firms that rely on banks for their financing requirements collectively had little use for the extra money being supplied by the central bank. Moreover, banks were trying to reduce their loan portfolios to get their balance sheets into better shape after a decade of disastrous lending practices. As a result of the economywide weak demand for funds, even at zero cost, most of the extra ¥1 trillion ($9.1 billion) pumped daily into the overnight market ended up back in BOJ's vaults each evening.

Even though this daily round-trip to nowhere might seem to have had little economic significance, the availability of so much money did have consequences. First, it kept the overnight interbank rate close to zero for 18 months. Moreover, longer-term rates were dragged down by the extremely low short-term rates. At least according to conventional measures, monetary policy was extremely accommodative of Japan's struggling economy.

With rates at virtually zero, money disappeared from the overnight market. For example, life insurers, which formerly had kept some of their assets there, deposited these funds in banks where they at least earned 0.05 percent.

Tanshi (money brokers), the six firms that have specialized in the short end of the money market since the end of the 1800s, were hard hit by the drying up of the overnight market. Although they are BOJ's exclusive agents in conducting open-market operations in overnight funds, their earnings fluctuate with the volume of transactions. As the market shrank, profitability declined. Moreover, participants in the market for instruments with maturities of up to 90 days, which once worked through tanshi, now trade directly.

Three of the six tanshi recently merged, but the new company still is significantly smaller than most securities firms. Their protected position as specialists in a disappearing product have left them without the managerial know-how or the capital to operate successfully in a deregulated financial industry. Standard & Poor's Corp., the credit-rating company, recently warned that tanshi were facing harsh times.

With the post-August 11 rise in short-term rates, money has begun to move out of bank deposits and back into the overnight market. These transfers no doubt will force banks to reconsider their extremely low deposit rates. Returns on 10-year government bonds also are moving higher, approaching 1.9 percent from the previous 1.7 percent. Although a recovering economy and expectations of increased government demand for new borrowing already were putting upward pressure on long-term bond rates, the increase in the short-term end of the market added to these forces.

BOJ's mid-August withdrawal of the extra ¥1 trillion ($9.1 billion) in excess reserves might have been expected to reduce the monetary base by the same amount. However, these funds do not meet the definition of this aggregate. Thus, their withdrawal had little direct effect on the money supply. Moreover, because the extra money that BOJ had been supplying did not go toward new lending or additional economic activity, it had no primary stimulative effect on the economy. Consequently, its absence also should make little difference. At least that is the thinking or the hope of BOJ economists.

BOJ's 18-month experiment with zero interest rates was a helpful exercise. Even if it failed to stimulate growth as much as some American experts had hoped, the recent conduct of monetary policy did not exacerbate the economy's problems. The central bank's new policy may not be very restrictive, but it does signal the end of a chapter that has few parallels in history.

The views expressed in this report are those of the author
and do not necessarily represent those of the Japan Economic Institute

Issue Index aaaa 2000 Archive Index aaaa Subscriber Area aaaa Home