Recent economic developments in Japan suggest that there is a growing interest in how monetary policy should be conducted under low inflation and nominal interest rates. In this paper, we first confirm the effectiveness of conventional wisdom against a liquidity trap, namely taking aggressive responses around the zero bound and variant of price level targeting so that it can intensify the policy duration effect, in the large scale dynamic general equilibrium model, the JEM (Japanese Economic Model). We then show that the optimal tangible policy rules around the zero bound are possibly state-contingent at the same time. J. Japanese Int. Economies20 (3) (2006) 364–379.
Recent economic developments in Japan suggest that there is a growing interest in how monetary policy should be conducted under low inflation and nominal interest rates. Although various different approaches have been taken to identify monetary and/or fiscal policy strategies able to reduce the economic burden imposed by the zero lower bound, the end prescription of all these approaches is always similar: namely, to increase expected inflation and lower expected future interest rates.1 Achieving this may not be very difficult in a theoretical rational expectation model, but finding a credible means of doing so in practice is a whole new ball game. Among several proposed solutions, the conventional wisdom seems to be that the central bank should commit to lower future rates through some variation on price-level targeting. Indeed, the policy of the Bank of Japan seems to subscribe broadly to this line of thinking of committing future monetary policy. On March 19, 2001, the Monetary Policy Meeting of the Bank of Japan released the “new procedures for money market operations and monetary easing.” It was stated that the new procedures for money market operations would continue to be in place until the year-on-year change in the core consumer price index (excluding perishables, on a nationwide basis) registers stably zero percent or higher."2 The “new procedures for money market operations” marked, of course, the adoption of the quantitative easing policy, namely the switching of the operating target for money market operations from the uncollateralized overnight call rate, which was to be maintained close to zero, to the outstanding balance of current accounts at the Bank of Japan.3 It may reasonably be concluded that the quantitative easing policy has been successful in flattening the yield curve and therefore boosting the economy to some extent. This policy is sometimes referred to as a policy-intensifying “policy duration effect.” According to Okina and Shiratsuka (2003), “Even though short term interest rates decline to virtually zero, a central bank can produce further easing effects by a policy commitment. A central bank can influence market expectations by making an explicit commitment as to the duration it holds short-term interest rates at virtually zero. If it succeeds in credibly extending its commitment duration, it can reduce long-term interest rates. We call this mechanism the ‘policy duration effect,’ following Fujiki and Shiratsuka (2002).” In this paper, we first reconfirm the effectiveness of the conventional wisdom, namely taking aggressive responses around the zero bound and variant of price level targeting so that it can intensify the policy duration effect using our large-scale dynamic general equilibrium model, the JEM: Japanese Economic Model.
We then show that the optimal tangible policy rules around the zero bound are possibly state-contingent at the same time. As stressed in Eggertson and Woodford (2003), Svensson, 2001 and Svensson, 2003b and Jeanne and Svensson (2004), it is of great importance for the central bank to commit to a certain policy scheme, especially to some forms of price-level targeting, in a liquidity trap. At the same time, the central bank needs to consider the balance between invulnerability and intelligibility in commitment since the dimension of uncertainty is higher than usual.4 Therefore, the mere adoption of price level target may not be enough for an economy to reach the normal state. Considering the non-linearity in a dynamic system caused by the zero bound, it is highly likely that the optimal (but only within the limited class) as well as intelligible linear instrument rule around the liquidity trap is highly state-dependent.
This paper is organized as follows. In Section 2 two we briefly summarize the Japanese Economic Model (JEM). Then, in Section 3, we conduct simulations to see how modified Taylor rule can mitigate severity caused by the liquidity trap. Section 4 discusses the state-contingency concerning the tangible monetary policy around the liquidity trap. Finally, Section 5 concludes.
We have first reconfirmed the effectiveness of the conventional wisdom, namely taking aggressive responses around the zero bound and variant of price level targeting so that it can intensify the policy duration effect using our large-scale dynamic general equilibrium model, the JEM. We then show, however, that a single linear instrument rule with aggressive responses—which is desirable in the neighborhood of the zero bound—may not be effective in both “precautionary loosening” and “intensifying the policy duration effect” respectively and is therefore not credible. They may be highly state-contingent. Considering the fact that adherence to a single linear instrument rule around the zero bound is not desirable, the current policy scheme—which is state-contingent and consistent with the simple taxonomy of the exit policy—may be considered a good alternative.
