How the Japanese Government Bond Market Has Responded to the Zero Interest Rate Policy

Yosuke Takeda
Policy Research Department Chief economist Yasuhide Yajima

Abstract

Since February 1999， including a brief intermission， the Bank of Japan has consistently pursued a zero interest rate policy. This paper examines how the Japanese Government Bond (JGB) market -- the benchmark indicator of the long-term interest rate -- has responded to the zero interest rate policy. We study changes in the JGB market from the perspective of the two goals of the zero interest rate policy -- supplying liquidity， and dispelling deflationary concerns.

First， after testing the expectations hypothesis for determination of long-term interest rates， we conduct an empirical analysis of the effect of the zero interest rate policy on the JGB market， focusing on the forward rate. The results are as follows.

Overall， the expectations hypothesis holds for the JGB market.
At the same time， we confirmed the existence of a liquidity premium over the long term. That is， the term premium is an increasing function of maturity. This means that the liquidity premium hypothesis rejected by Kuroda (1982) has been validated in recent years.
During the period of the zero interest rate policy， policy changes toward further easing had the effect of reducing the marginal term premium. In particular， the reinstatement of the zero interest rate policy on March 19， 2001 successfully reduced the marginal premium of government bonds of various maturities by a statistically significant amount. A similar result has been confirmed regarding the BOJ's declared commitment to pursuing the zero interest rate policy (the policy duration effect) by Shiratsuka and Fujiki (2001) in their analysis of money markets.
However， as Ueda (2001) points out， after August 14， 2001， when the BOJ decided to increase direct purchases of long-term government bonds， the marginal premium increased， albeit temporarily. This point might suggest that market participants had been anticipating the imminent onset of inflation due to loss of fiscal discipline.
Finally， alongside the zero interest rate policy， we must not ignore the effect of JGB credit downgrades on the marginal premium. Moody's downgrades of long-term JGBs have been immediately followed by an increase in the marginal premium. In particular， the Aa2 downgrade on September 8， 2000， and Aa3 downgrade on December 4， 2001 caused statistically significant increases.

From the above empirical results， we assess the zero interest rate policy based on its two goals of supplying liquidity and alleviating deflationary concerns as follows.

The zero interest rate policy has been successful in supplying liquidity and thereby offsetting the liquidity premium. In particular， the reinstatement of the zero interest policy was quite effective in this regard.
Although the inflation premium cannot be isolated due to the absence of inflation-indexed bonds in Japan (Kitamura， 1997)， our findings support Ueda's (2001) concern -- that the increase in marginal premium immediately after the BOJ decided to increase direct purchases of long-term JGBs can be construed as an inflation premium. However， the effect of this non-traditional operation is temporary; indeed， we found the effect of the JGB credit downgrading on the premium to be more pronounced. Thus confidence in the central bank -- a prerequisite to introducing inflation targeting -- is not yet evident in the JGB market.

* The authors wish to acknowledge Teruyoshi Suzuki of NLI Research Institute for his assistance in measuring forward rates.
** Y. Takeda is associate professor of economics at Sophia University， and visiting researcher at NLI Research Institute. Y. Yajima is an economist at NLI Research Institute.

Yasuhide Yajima