Using Dynamic Stochastic General Equilibrium Models in a Policy Environment

Good macroeconomic frameworks and models provide an important basis for economic decision making. The IMF is engaged in several capacity development activities on macroeconomic frameworks and forecasting policy analysis systems in the Caucasus, Central Asia, and Mongolia (CCAM) region. 

More and more central banks, as well as ministries of finance and economy, intend to upgrade their analytical tools, with a strong interest in dynamic stochastic general equilibrium (DSGE) models. Forecasting and policy analysis tools have become even more important during the current covid-19 pandemic and will be key as countries are exiting from the pandemic and dealing with heightened global uncertainty. The evolving nature of shocks calls for a coherent assessment of alternative economic policy scenarios. 

On March 1-3, the IMF's Caucasus, Central Asia, and Mongolia Capacity Development Center, the IMF Institute for Capacity Development (ICD), and the IMF Monetary and Capital Markets Department organized a regional workshop on DSGE modelling. The workshop aimed at (i) sharing experience in the region, (ii) reviewing some of the frontier thinking; and (iii) discussing how CCAMTAC can support peer-learning. 

The approaches and tools to analyze macroeconomic developments have evolved over time. This is also reflected in the IMF’s modernization of its financial programming framework. In the 1950s, the focus was on the key components of national accounts and on ensuring accounting consistency between the real, fiscal, external, and monetary sectors. The next generation of models included more factors, such as supply (production function), expectations (mostly adaptive) and dynamics (done via Error Correction Models). Subsequently, the emphasis moved to thinking about gaps – deviations from long-term trends in output, fiscal and financial variables.

In these semi-structural models, the discussions about economic policies were geared towards closing these gaps in the medium-term. Falling in this category, forecasting policy analysis systems (FPAS) with quarterly projection models (QPMs) have played an important role in central banks that target inflation. During the next wave, applications moved to micro-founded models with a focus on structural shocks: dynamic stochastic general equilibrium models.

While more challenging to build and maintain, DSGE models are very suitable to analyze specific economic policy questions in the fiscal, monetary and exchange rate areas. More recent work has focused on integrating financial and macro-financial factors in these models. The IMF’s Integrated Policy Framework (IPF) falls in this category. 

Lessons from Experience

In the CCAM region, the IMF has helped some ministries of finance (Armenia, Georgia) to set up DSGE models, as part of a project supported by the Netherlands, and also worked with a number of central banks to set up forecasting policy analysis systems (Armenia, Georgia, Kazakhstan).

Emerging lessons are: 

• No model fits all circumstances. All generations of models and approaches have their advantages and disadvantages. All approaches still have some merits today.  
• DSGE models have been very helpful to construct policy scenarios and analyze the implications of key structural reforms, including changes to the tax system, expenditure policies, debt, exchange rates and financing strategies.  
• While DSGE models can serve a wide range of purposes, getting a DSGE model should not be an objective in itself—it should respond to clear and specific needs and policy questions for which these models are the preferred modeling option. The time and resource requirements to develop and implement them should not be underestimated. After putting substantial effort in their development, globally some institutions have (temporarily) abandoned their use, for varying reasons: (i) they were not the most suited model for the intended purpose; (ii) the time and energy needed to develop and manage the models was challenging in a resource-constrained context; and (iii) the management of the model was overly dependent on a small group of staff, leading to risks associated with staff turnover. 

Frontier of thinking 

Traditional economic thinking was often framed in the context of one target (e.g., inflation) and one instrument (e.g., policy interest rate). In practice, advanced and emerging market central banks have used a mix of instruments to achieve one or multiple targets. The toolbox includes the policy interest rate, foreign exchange interventions, macroprudential policies, and capital flow management measures. 

• During the past 2-3 years, IMF staff has made important strides to develop a framework and models taking greater account of real-life frictions and vulnerabilities. 
• Recent models incorporate additional elements, such as the financial sector; frictions, including a lack of credibility, foreign currency mismatches, balance sheet effects; and strengthen the analysis of key policy trade-offs. 

How can CCAMTAC Help?

CCAMTAC has a “macroeconomic framework, financial programming, forecasting and policy analysis” resident advisor and wants to promote peer learning. Considering available resources, the scope for country-specific support is limited, given the medium-term commitment and resource needs for such projects.

The seminar suggested several not mutually exclusive options for country authorities and CCAMTAC, considering that some countries in the region have already substantial modelling experience.  

• Making full use of available training opportunities: for newcomers this provides an excellent opportunity to get familiar with some of the needed technical skills for economic policy analysis and DSGE models. This includes both relevant online courses offered by the IMF as well as virtual / face-to-face courses offered by the training centers – the Joint Vienna Institute (JVI) and the Singapore Training Institute (STI).  
• Fostering peer learning: CCAMTAC is ready to organize peer presentations by country authorities to receive feedback on their modelling work. These could be supplemented on occasion by regional peer workshops. 
• Analytical work: In an advanced stage, country authorities could consider joint research and analytical work.

The workshop provided an excellent opportunity to bring experts and specialists from regional country authorities together. The active participation of officials from central banks, ministries, and agencies helped to cross the institutional aisle and underline the alternative use of models in the various institutions.  

Useful links:

• CCAMTAC website
• IMF Monetary and Capital Markets Department's Capacity Development Strategy 2022-2025—"Supporting Financial Resilience Through Disruption and Change”
• IMF Capacity Development web pages
• IMF online courses