Skip to main content

Forward-looking Models and Monetary Policy in Japan

April 2000
Koichiro Kamada
Ichiro Muto

Views expressed in Working Paper Series are those of authors and do not necessarily reflect those of the Bank of Japan or Research and Statistics Department.
Questions and opinions on the working paper should be e-mailed to each author whose address is indicated in the document.

Click on cwp00e07.pdf (339KB) to download the full text.


The role of expectations has long been recognized in economic theory. For instance, stock prices depend not only on present performance of firms and interest rates, but also on the expectations about future performance and interest rates. Private fixed investment is affected not only by present profitability of a firm, but also by expectations for future profitability. Despite this recognition, there is little consensus on how expectations should be incorporated in macro-econometric models. In practice, some econometric models incorporate very simple mechanisms of expectation formation, such as adaptive expectations. However, since adaptive expectations are based on present and past information, they cannot take into account events that are expected to occur in the future. Nonetheless, it is easy to imagine that future events affect present activity. Suppose for instance that a government announces that it will have a tight fiscal policy several years in the future. If private economic agents expect this policy to be adopted, and change their economic behavior accordingly, there may be changes in economic variables, such as long-term interest rates, even though there is no actual reduction in fiscal expenditure and even though the policy is only announced. This mechanism of expectation formation is called rational expectations in general, and "model-consistent expectations" in macro-econometric models. The models that incorporate such expectations are called "forward-looking models" and are distinguished from "backward-looking models," in which expectations are based on current and past information, as are adaptive expectations. In this paper, we build a forward-looking model by incorporating rational expectations into a backward-looking model and then investigate the properties of the forward-looking model.

Forward-looking models have many applications. Among them, the analysis of monetary policy rules is the most relevant for central bankers.1 Suppose that a central bank adopts a particular policy rule. In reaction to changing economic situations, the rule requires a specific policy, and the central bank implements that policy. Meanwhile, private economic agents choose their present behavior, based on their expectations of future monetary policy by the central bank. Taking into consideration expectation formation and behavioral change among private economic agents, the central bank should choose an optimal policy rule to achieve its goals. Forward-looking models allow for easy incorporation of monetary policy rules into macro-econometric models and enable us to investigate the implications of rational expectations in the models.2 In practice, many central banks devote resources to intense analysis of monetary policy rules that uses rational expectations.3 In contrast, in Japan, there are virtually no analyses based on forward-looking models and there has been little accumulation of empirical analyses for monetary policy rules.

Monetary policy analysis based on forward-looking models emphasizes the impact of a change in a policy rule on the dynamic processes by which output and prices revert to long-run equilibrium values. Since forward-looking models take long-run equilibrium as given, they provide no way to investigate how the long-run equilibrium is affected by a change in a policy rule. Forward-looking models are merely tools with which we analyze how an economy reverts to long-run equilibrium. Analysis that uses a forward-looking model is still under development. Hence, the focus is currently on a qualitative analysis of the model's properties under the assumptions of a particular mechanism of expectation formation and a particular rule of monetary policy. The reason is that the monetary authorities lack quantitative information on how rational private economic agents are and on how helpful a central bank's credibility is when it comes to economic stabilization. Consequently, forward-looking models are too underdeveloped to calculate the short-run effects of monetary tightening and easing and to compare a rule targeting interest rates with a rule targeting something else.

This paper will give an introductory explanation of forward-looking models, build and estimate a small-sized forward-looking model for the Japanese economy, and discuss the properties of the forward-looking model through various simulations. The remaining part of this paper is constructed as follows: Chapter 2 gives a theoretical explanation of forward-looking models. We assume that in asset markets, participants form rational expectations for long-term interest rates and foreign exchange rates. Chapter 3 investigates the properties of a forward-looking model by comparing it with a backward-looking model. We show that when a future expansion in demand and acceleration in inflation are expected, present long-term interest rates rise, which exerts negative pressure on the present economy before the actual future expansion occurs. Chapter 4 uses real data and estimates a forward-looking model for the Japanese economy. Chapter 5 uses that estimated model and analyzes policy rules by implementing simulations. Chapter 6 uses stochastic simulations, in which models are under random shocks, and presents analysis of policy rules under uncertainty. Appendix A explains how to implement stochastic simulations. Appendix B investigates the role of the zero-percent constraint on nominal interest rates.

  1. Much of the literature uses small models for policy rule analysis; for example, see Fuhrer and Moore [1995] and McCallum and Nelson [1999]. The Board of Governors of the Federal Reserve System (FRB) continues to develop large models, such as FRB/US (Brayton and Tinsley [1996]) and FRB/Global (Levin et al. [1997]).
  2. Impetus for using forward-looking models comes from the Lucas critique that argues simulations with fixed parameters are useless, since changes in policy parameters may change behavioral patterns of economic agents.
  3. In the countries whose central banks adopt inflation-targeting policy, such as the Bank of England (the Bank of England [1999] and Batini and Haldane [1999]), the Bank of Canada (Donald et al. [1996] and Black and Rose [1997]), and the Reserve Bank of New Zealand (Black et al. [1997]), analysis of macro-econometric models that explicitly incorporate rational expectations is very popular. The FRB is also developing a large forward-looking model (see Reifschneider et al. [1997] for the relationships between monetary policy and macro-econometric models).