The effect of the increase in monetary base on Japan's economy at zero interest rates : an empirical analysis

In this paper, we quantify the effect of so-called “quantitative monetary easing” which the Bank of Japan adopted in March 2001. Now that short-term interest rates are almost zero and monetary base growth is over 20 percent year on year, active debate continues with respect to the effectiveness of monetary policy at zero interest rates. Taking into account the regime change in monetary policy and the possible non-linearity of money demand at low (or near zero) interest rates, we use a Bayesian VAR, a VAR with time-varying coefficients, to extract the effect of the increase in monetary base at zero interest rates. The result of a Bayesian VAR indicates that while an increase in monetary base previously had a positive impact on prices, it does not now at zero interest rates. In order to investigate the possible reason for this result, we then estimate a money demand function, and test whether a satiation level in demand for monetary base exists at zero interest rates. The key finding here is that the null hypothesis of the non-existence of the satiation level can be statistically rejected. This means that there may remain room for an increase in monetary base to stimulate the economy at zero interest rates. Despite the existence of the satiation level of money demand, why does the Bayesian VAR result suggest that an increase in monetary base does not have a positive impact on economic activity at zero interest rates? One way to consistently interpret these two results is that the effect of the increase in monetary base is highly uncertain and very small, if any. We confirm this view by estimating models that include both aggregate demand and aggregate supply functions and testing whether monetary base enters these equations significantly. Finally, we discuss reasons why the expansion of monetary base at zero interest rates has such a limited and uncertain effect on the economy.