Assisting central banks in making optimal monetary decisions

Posted on: 4 January 2024 in Research

Professor Oliver de Groot presents a toolkit to assist central banks’ monetary policy decisions with optimal projections.

What is the best mix of central bank policy measures?  How can computational challenges be addressed to facilitate the most appropriate decisions? Is there a  model that simplifies the process while still factoring  in multiple policy instruments and constraints?

In collaboration with economists at the European Central Bank (ECB), Oliver has developed new methods for studying the ideal mix of central bank policy instruments.

In the paper, 'A toolkit for computing Constrained Optimal Policy Projections (COPPs)', Oliver presents a toolkit for studying optimal policy with a minimal set of inputs: baseline forecast, a policy objective and impulse response functions to anticipated policy instrument shocks.

With these three inputs, the toolkit derives optimal policy projections in numerous scenarios and policy constraints, and alternative methods for solving the forward guidance puzzle.

The goal is to inform policymakers and help improve monetary policy decisions in an environment with multiple instruments and powerful non-linear forces.

The underlying research is currently being used to impact policy making at the ECB, and engage economists at other central banks, resulting in widespread benefits in the form of improved monetary policy across several economies.

Addressing monetary policy challenges

Before 2008, monetary policy decisions revolved almost exclusively around the setting of the nominal shortterm interest rate, and policy advice invariably came from stylised linear models with a single policy instrument.

However, since then, the setting of monetary policy has become more challenging.

With interest rates across the globe  at historically low levels, policy makers have been constrained in their use of the nominal interest rate to set policy and have resorted to a set of nonstandard measures, including forward guidance and large-scale asset purchases (ie quantitative easing).

While these measures are aimed at stimulating the economy and escaping the liquidity trap, the use of multiple policy instruments complicates the setting of monetary policy, as the central bank has to decide an appropriate combination of these  non-standard measures.

Furthermore, non-standard measures face constraints themselves, such as limits to asset purchases arising from legal or market-functioning considerations.

At the same time, the financial crisis of 2008 and the COVID-19 crisis have highlighted powerful non-linear forces that operate on economies.

The combination of multiple policy instruments and important nonlinearities pose computational challenges for the academic study of monetary policy and practical challenges for policymakers.

Oliver’s computational toolkit for COPPs offers a Dynare¹ compatible solution to model optimal monetary policy in the presence of multiple instruments and constraints.

Simplifying monetary policy decisions

While optimal policy and the tools to compute it are well established, De Groot’s toolkit for COPPs simplifies this task along a variety of dimensions:

• First, instead of requiring a specification of structural equations, the toolkit only needs a minimal set of inputs: a baseline projection for target and instrument variables, and impulse responses of those variables to policy shocks
• Second, it solves optimal policy projections under different levels  of commitment from the central bank: commitment, limited-time commitment and discretion
• Third, it handles multiple policy instruments and can be used  to study problems outside the monetary policy realm (eg  optimal fiscal or optimal macroprudential policies)
• Fourth, it handles multiple constraints on policy instruments (eg lower bound on policy rate, upper bound to asset purchases)
• Fifth, it allows alternative approaches to address the forward guidance puzzle (overestimation of the effects of central banks' communications), generating empirically plausible responses to policymaker announcements. 

Impact and future research The toolkit for COPPs is currently used by ECB staff when briefing and advising ECB Council Members and Eurosystem National Central Banks.

The methods and tools developed by Oliver and his collaborators, gained significant interest from various central banks following University of Liverpool’s 2021 Conference on “Optimal Monetary Policy: Computational Advances  and their Applications”.

The event brought together academic scholars and central bank practitioners, from across the international central banking community including the Bank of Canada, Bank of Ireland, Bank of  Italy, ECB, Federal Reserve Board, and Norges Bank.

Oliver and ECB Advisor, Dr Roberto Motto, also guest edited a December 2021 Special Issue of the Journal of Macroeconomics, facilitating the link between rigorous academic research and practical policy impact.

Oliver continues to collaborate with the ECB and is in discussions with other national central banks in Europe and North America, regarding the use of these new tools for advising monetary policymakers.

Further advances and development of the tools are planned for the coming years.

1 Dynare is a macroeconomic modelling software platform, which handles a wide class of economic models, in particular dynamic stochastic general equilibrium (DSGE) and overlapping generations (OLG) models.

Professor Oliver De Groot Chair in Macroeconomics and Head of the Management School's Economics Group