Published: 22.03.2022 Updated: 10.07.2024

Latvijas Banka maintains and develops economic modelling tools in order to ensure high-quality and independent analytical vision of monetary and economic policy making.

Economic models are based on economic theory, and they are calibrated to the structure of Latvia’s economy.

Models of Latvijas Banka are used on a regular basis to make macroeconomic projections, analyse current economic developments and evaluate economic policy proposals.

The short-term inflation projections (STIP) model is used by Latvijas Banka to forecast future inflation in Latvia and assess developments in various inflation components.

The STIP model exploits statistical correlations between the harmonised index of consumer prices and its major determinants, i.e. wages, global food prices and oil prices. The STIP model is an important tool used by Latvijas Banka in the forecasting process.

More about the STIP model

Examples of STIP model usage

The computable general equilibrium (CGE) model contains detailed information on sectors and goods, taking account of the input-output relationships between various sectors of the economy.

Latvijas Banka uses the CGE model for analysing the direct and indirect effects of various macroeconomic developments and fiscal policy changes on the overall economy and selected sectors on a regular basis.

The current version of the model comprises 63 sectors and 63 goods using data from input-output as well as supply and demand tables for 2015. The number of equations in this model is close to 30 thousand.

While developing Latvia's CGE model, special attention was paid to the fiscal block: the model consists of the major government expenditure types and five revenue sources, including four major taxes such as the personal income tax, mandatory social security contributions, value added tax and excise tax.

Latvijas Banka has merged the EUROMOD and CGE models into a single CGE-EUROMOD model, which is used to analyse the impact of macroeconomic changes on income of various social groups as well as to assess the indirect effect of changes in taxes and benefits.

For further information about the computable general equilibrium (CGE) model

Examples of CGE model usage

EUROMOD is a European tax-benefit microsimulation model. It enables analysis of the changes in tax and benefit rules and calculates the effect of these changes on disposable income of different social groups. Thus, EUROMOD is an appropriate tool to assess the effect of changes in taxes and benefits on income inequality and poverty rates.

The model has been developed and maintained by the Institute for Social and Economic Research at the University of Essex in cooperation with national expert teams. The Latvian national expert team is based at the Baltic International Centre for Economic Policy Studies (BICEPS). EUROMOD is a static model based on the data obtained from the survey on income and living conditions (EU-SILC). 

Latvijas Banka has merged the EUROMOD and CGE models into a single CGE-EUROMOD model, which is used to analyse the impact of macroeconomic changes on income of various social groups as well as to assess the indirect effect of changes in taxes and benefits.

More about the EUROMOD microsimulation model

Examples of EUROMOD usage

Latvijas Banka employs the following models to forecast GDP growth

  • The barometer compiles information on approximately 60 monthly economic indicators having the highest correlation with the target (GDP) time series. The barometer utilised for tracking Latvia's economic activity adopts a similar framework, as developed for the KOF Swiss Economic Barometer in Abberger et al. (2014, 2017).
  • The dynamic factor model (DFM) is based on the Kalman filter. The DFM allows for simultaneous modelling of economic time series observed at different frequencies (quarterly and monthly). The model built for the use with Latvian data adopts the framework of the EuroSTING indicator of Camacho and Perez-Quiros (2010). Siliverstovs (2012, 2016) applies the same framework for modelling of the Swiss economy.
  • The suite of nowcasting models consists of several short-term econometric (bridge, factor and VAR) models. It is used to nowcast short-term economic activity.
  • LATIN economic activity indicator is obtained via the regularised multivariate direct  filter using 220 economic time series of the Latvian economy, and to a lesser extent, of the Lithuanian, Estonian and euro area economy.

Model-related publications

  • Abberger, K., Graff, M., Siliverstovs, B., Sturm, J.E. (2017). Using Rule-Based Updating Procedures to Improve the Performance of Composite Indicators, vol. 68, issue C., Elsevier, pp. 127–144, Economic Modelling
  • Abberger, K., Graff, M., Siliverstovs, B., Sturm, J.E. (2014). The KOF Economic Barometer, Version 2014. KOF, Working Papers, 353/2014, p. 51, KOF Swiss Economic Institute
  • Camacho, M. and Perez-Quiros, G. (2010). Introducing the Euro-STING: Short-term Indicator of Euro Area Growth, vol. 25, issue 4, pp. 663-694, Journal of Applied Econometrics
  • Siliverstovs, B. (2016). The Franc Shock and Swiss GDP: How Long Does It Take to Start Feeling the Pain?, vol. 48, issue 36, pp. 3432–3441, Taylor & Francis Journals
  • Siliverstovs, B. (2012). Are GDP Revisions Predictable? Evidence for Switzerland, vol. 58, 4/2012, pp. 299–326, Applied Economics Quarterly
  • Bušs, G. (2016). Real‐Time Signal Extraction with Regularized Multivariate Direct Filter Approach, vol. 35, issue 3, pp. 206–216, Journal of Forecasting

The DSGE model for Latvia is a dynamic, stochastic general equilibrium model for Latvia as a small open economy in the euro area. It belongs to the class of New-Keynesian models.

DSGE models are used by monetary, fiscal and macroprudential institutions to analyse policy scenarios, explain historical developments and carry out forecasting.

DSGE models are based on general equilibrium theory and microeconomic principles. Unlike real business cycle models, New-Keynesian models include various (inter alia, price) frictions and market failures. Meanwhile, in contrast to CGE (computable general equilibrium) models, DSGE models are dynamic, and expectations of market participants play an important role in their results. In the wake of the global financial crisis, DSGE models were supplemented with financial market frictions. Currently, DSGE models represent the predominant type of structural models used in the theoretical economics literature and economic policy institutions.

Fiscal DSGE model for Latvia

  • G. Bušs, P. Grüning, O. Tkačevs (2024). Choosing the European fiscal rule, Baltic Journal of Economics, Vol. 24(1), DOI: 10.1080/1406099X.2024.2340402
  • G. Bušs, P. Grüning (2023). Fiscal DSGE Model for Latvia, Baltic Journal of Economics, Vol. 23(1), DOI: 10.1080/1406099X.2023.2173915

DSGE model for Latvia (suitable for forecasting)

Examples of DSGE model usage

To assess the economic effects of the transition to a greener economy in Latvia, Latvijas Banka uses two types of models: computable general equilibrium (CGE) models and environmental dynamic stochastic general equilibrium (E-DSGE) models. While the particular strength of CGE models is the explicit introduction of many economic sectors with heterogeneous production technologies and the evaluation of detailed sector-specific economic effects in the short run, the DSGE models are well suited to investigate the dynamic aspects of the transition period and to take account of uncertainty along the green transition.

First, the CGE model developed at Latvijas Banka is based on the Input-Output table for Latvia and consists of more than 60 thousand equations, which allows to assess how changes in costs, foreign and domestic supply and demand, taxes, productivity, and macroeconomic conditions affect economic activity in each industry (e.g., production, export, import, prices, wages) and the government budget. The model has recently been updated to improve the modelling of the energy sector and energy use by different industries, as well as to calculate CO2e emissions from production and costs related to carbon pricing. In particular, the costs faced by producers due to carbon taxes and the European Union Emissions Trading Scheme (EU ETS) are now modelled explicitly. This enables the model to simulate the economic effects of changes to carbon taxation rates and rules (coverage) and the prices of EU ETS carbon emission quotas, including the introduction of a new EU ETS for housing and transport planned for 2026.

Second, in the realm of E-DSGE models, there are currently two models developed at Latvijas Banka. The first one is a medium-scale small open economy model, calibrated to Latvia’s economy, with a carbon-intensive (brown) sector utilizing fossil fuels imported from abroad and an emission-free (green) sector utilizing domestically produced renewable energy. The key difference to the classic DSGE model for Latvia is thus the explicit introduction of two production sectors that differ in their environmental impact. This basic production structure is complemented by explicitly adding green fiscal policy considerations in the fiscal sector (i.e., emission caps, carbon taxes, and green subsidies) and a banking sector that provides loans to both the green and the brown sector but faces loan type-specific bank regulation policies. The second model is a multi-sector New Keynesian closed economy model, calibrated using aggregate and sectoral data for the Euro Area including input-output tables, that allows for input-output linkages, sector-specific abatement and emission intensities, and investment networks at the NACE 1 and 2 levels. The current model version has 37 sectors (NACE 2 level for the manufacturing sector and NACE 1 level otherwise). It is used to investigate the economic and environmental effects of asset stranding risk (proxied by shocks to capital utilization rates) and various fiscal policies (investment taxes and subsidies, consumption taxes, carbon taxes).

Related publications

  • Benkovskis, K., Jaunzems, D., Matvejevs, O. (2023). A Purpose-Based Energy Substitution Structure for CGE: Working Papers, 2023/07, Latvijas Banka.
  • Grüning, P. (2022). Fiscal, Environmental, and Bank Regulation Policies in a Small Open Economy for the Green Transition: Working Papers, 2022/06, Latvijas Banka.
  • Grüning, P., Kantur, Z. (2023). Stranded Capital in Production Networks:
  • Implications for the Economy of the Euro Area: Working Papers, 2023/06, Latvijas Banka.