Advanced DSGE Macroeconomic Modelling for Policy Analysis Training Course

Course overview

Policy teams, central banks, finance ministries and research units need macroeconomic models that do more than describe historical relationships. They must quantify the effects of interest-rate changes, fiscal packages, commodity-price shocks, exchange-rate movements and financial disturbances before decisions are made. This course addresses the practical challenge of building, calibrating and interpreting Dynamic Stochastic General Equilibrium (DSGE) models that produce defensible policy scenarios, while making assumptions, uncertainty and model limitations visible to senior decision-makers.

Participants work from core DSGE structure through to policy-ready applications: household and firm optimisation, nominal rigidities, monetary-policy rules, fiscal blocks, open-economy channels, shock processes and Bayesian estimation. They learn to express equilibrium conditions in a model file, solve linearised systems, calibrate parameters, estimate selected parameters, generate impulse-response functions, conduct historical decompositions and compare alternative policy rules. Attention is given to model diagnostics, identification, prior sensitivity and communicating results without overstating precision.

Instructor-led demonstrations are paired with guided coding labs using Dynare, MATLAB and Python. Participants progressively develop a policy model around a realistic small open economy, test specified shocks and present findings as an economic policy briefing. They leave with a documented DSGE model template, reproducible code, calibrated scenario outputs, charts of impulse responses and a structured policy-analysis note that can be adapted to their institution's forecasting or research workflow.

The course is designed for economists and quantitative analysts who already understand macroeconomic relationships and want to move from reduced-form analysis or standard forecasting models into structural, scenario-based policy modelling. It is particularly valuable where staff need to challenge model assumptions, commission external modelling work or explain macroeconomic transmission mechanisms to non-modelling stakeholders.

Course objectives

By the end of this course, participants will be able to:

• Specify household, firm and policy-maker optimisation conditions for a baseline DSGE model
• Log-linearise equilibrium conditions and assemble the model’s state-space representation
• Build and solve a New Keynesian DSGE model in Dynare using MATLAB-compatible syntax
• Calibrate preference, technology, price-rigidity and policy-rule parameters from economic evidence
• Estimate selected DSGE parameters using Bayesian priors, posterior diagnostics and observed macroeconomic data
• Generate and interpret impulse-response functions, forecast-error variance decompositions and historical decompositions
• Compare monetary and fiscal policy rules through counterfactual shock simulations
• Produce a reproducible policy briefing with model assumptions, scenario charts and decision-relevant caveats

Benefits of attending

Target competencies

Who should attend

• Central Bank Economists — who assess monetary-policy transmission and prepare scenario material for policy committees
• Macroeconomic Policy Analysts — who need structural evidence for fiscal, inflation and growth-policy recommendations
• Finance Ministry Economists — who evaluate budget measures, debt risks and macro-fiscal shock scenarios
• Economic Research Managers — who commission, review or quality-assure DSGE-based analysis
• Financial Stability Analysts — who examine macro-financial shocks and their transmission to the real economy
• Quantitative Economists and Econometricians — who want to extend reduced-form forecasting work into structural modelling

Requirements and prerequisites

Participants should be comfortable with intermediate macroeconomics, including inflation, output gaps, interest rates, exchange rates and fiscal policy; and should understand basic econometrics, probability and regression interpretation. Familiarity with constrained optimisation, matrix algebra and difference equations is strongly recommended because the course uses first-order conditions and linearised systems. Participants should have some experience reading or editing code in MATLAB, Python, Julia or a similar analytical language. Prior Dynare experience is helpful but not required. This is not a suitable first exposure to macroeconomics, programming or econometric modelling, and advanced general-equilibrium theory is not assumed.

Training methodology

The programme combines instructor-led model derivation with supervised coding labs and policy-case workshops. Each day participants extend a common small open-economy DSGE model in Dynare, inspect solution output in MATLAB and use Python for data preparation or charting. Short technical demonstrations are followed by pair-based debugging, parameter-choice discussions and interpretation exercises. Cases cover inflation persistence, commodity-price disturbances, fiscal expansion and external demand shocks. The final day uses a policy-committee simulation in which teams defend a model-based recommendation, state uncertainty explicitly and create an application plan for their own institution.

Course outline

Day 1: DSGE foundations and model architecture

• Purpose and limits of structural macroeconomic policy models
• Recursive optimisation and intertemporal household choice
• Firm profit maximisation and production technology shocks
• Market-clearing conditions and resource constraints
• Steady states, balanced growth paths and stationarisation
• Log-linearisation around deterministic steady states
• Dynare model-file structure, declarations and equation blocks

Workshop: Participants code a baseline real-business-cycle model in Dynare and produce a verified steady-state and first simulation output.

Day 2: Nominal rigidities and monetary transmission

• Calvo price setting and the New Keynesian Phillips curve
• Consumption Euler equations and the dynamic IS curve
• Taylor rules, interest-rate smoothing and policy shocks
• Inflation targeting versus output-stabilisation objectives
• Rational expectations and Blanchard-Kahn determinacy conditions
• Impulse-response functions for demand, cost-push and productivity shocks
• Dynare solution diagnostics and common model-coding errors

Workshop: Participants add sticky prices and a Taylor rule to their model, then compare inflation and output responses to three monetary-policy shocks.

Day 3: Calibration, estimation and model validation

• Economic calibration using national accounts and empirical literature
• Selection of observables and measurement equations
• Bayesian priors, likelihood construction and posterior distributions
• Metropolis-Hastings estimation workflow in Dynare
• Convergence assessment, posterior moments and parameter correlations
• Identification analysis and prior-posterior comparison
• Posterior predictive checks and model-fit limitations

Workshop: Participants estimate selected parameters from a supplied quarterly macroeconomic dataset and prepare a calibration-and-diagnostics memo.

Day 4: Open-economy and fiscal policy applications

• Small open-economy consumption and risk-premium mechanisms
• Exchange-rate dynamics and uncovered interest parity
• Imported inflation, terms-of-trade and commodity-price shocks
• Government spending, taxes and fiscal-rule specification
• Public debt dynamics and fiscal sustainability constraints
• Counterfactual simulations of monetary and fiscal policy mixes
• Historical shock decomposition and narrative event mapping

Workshop: Teams model an external commodity-price shock and compare inflation, exchange-rate and debt outcomes under alternative fiscal responses.

Day 5: Policy communication and model governance

• Forecast-error variance decomposition for policy prioritisation
• Scenario design, baseline construction and shock-path assumptions
• Sensitivity analysis for key structural parameters
• Model-risk registers and assumption documentation
• Reproducible code, version control and results audit trails
• Translating impulse responses into policy-committee charts
• Communicating uncertainty, caveats and model boundaries

Workshop: Participants complete a policy-committee case by submitting a reproducible scenario pack and delivering a concise model-based recommendation.

Tools & standards covered

Dynare, MATLAB, Python, GNU Octave

A typical training day

Live online deliveries follow the same structure in the East Africa Time zone, with shorter screen blocks and longer breaks.

What the fee includes

How you can take this course

Certification

Participants who complete the full five days receive the Skillset Development Certificate of Completion, stating the course title, course code, dates and delivery format — suitable for professional-development records and employer reimbursement.

Frequently asked questions