| The Taylor rule is an instrument rule implementing a nominal short-term interest rate in response to inflation and the output gap. The rule is based on the Taylor Principle that a central bank should raise (reduce) the policy rate more than one-to-one, relative to an increase (decrease) in inflation. Nonetheless, the rule has never been committed to by any central bank.; In Chapter 1, I explore the main aspects of the Taylor rule, beginning with an overview of monetary policy and where the Taylor rule fits into the monetary policy framework. I also discuss empirical studies, the rule's limitations and implications on discretionary policymaking. In addition, I survey the empirical work on the US as well as other developed economies. Most studies from other countries show that the Taylor rule in most cases explains their monetary policy conduct well.; Moreover, I discuss problems inherent to the Taylor rule in practice, for example, the real-time data---the data available at the time that the policy is made---versus the final or revised data can result in the different conclusion being drawn, the zero bound constraint on a nominal interest rate since a nominal interest rate cannot be negative and finally the assumption of a constant natural real interest rate which contradicts economic theory suggesting that it can vary across time.; Lastly, it may be the case that the successful monetary policy does not depend on the policy rule, as seen during the Greenspan chairmanship of the Federal Reserve. In this period, he implemented a risk management strategy, which can be considered as a discretionary policy.; In Chapter 2, I examine one of many restrictions, i.e. the zero bound constraint on the nominal interest rate. Based on a simple backward-looking model used by Rudebusch and Svensson (1999), Chapter 2 shows that the rule continues to perform well if the constraint binds but the recession is not too severe. The major effect of the zero bound constraint is that the distributions of inflation and the output gap deviate from the unconstrained ones, while the effect on inflation is more pronounced. The risk of hitting the zero bound and falling into the deflationary trap is less the higher the inflation target and the less aggressive the rules than with a lower inflation target and more aggressive rules.; Finally, Chapter 3 examines another problem with the Taylor rule, the constant natural real interest rate. When the original Taylor rule was introduced in 1993, the natural real interest rate (NRR) was assumed constant at 2%. However, in the longer-term, the NRR should vary with real factors and structural changes of an economy. This chapter examines the effect of replacing the constant NRR by the time-varying one in the Taylor rule. I first present three pieces of evidence showing that the time-varying NRR is the correct interest measure to use. Then, I show that the time-varying NRR has more predictive power on the output gap and inflation than a constant one, as illustrated by simple and partial correlations. With the historical approach, the Taylor rule with the time-varying natural real interest rate tracks the actual federal funds rate more closely than that with a constant one, even the during the period of changing in the Federal Reserve's operating procedure. This is because the rule with time-varying NRR embraces changes in economic structure. Lastly, the constant natural real interest rate is dominated by the time-varying one both in real-time and latest available data, as shown in the stochastic simulations. |
