Research Article
Print
Research Article
Macroeconomic consequences of Exchange rate fluctuations for BRICS countries
expand article infoKashif Hasan Khan‡, Iman Bastanifar§, Zahra Mohammadi§
‡ Paragon International University, Phnom Penh, Cambodia
§ University of Isfahan, Isfahan, Iran
Open Access

Abstract

This paper investigates the asymmetric macroeconomic effects of sustained exchange rate disturbances in the BRICS economies — Brazil, Russia, India, China, and South Africa — within an enhanced New Keynesian DSGE framework. Exchange rate disturbances are modeled as latent structural shocks entering the exchange rate determination equation, while empirical exchange rate persistence is used only as a calibration discipline rather than for structural identification of the shock process. The analysis considers three distinct macroeconomic environments: a long-term horizon encompassing major global crises (2000–2023, including the 2007–2008 financial crisis and the 2020–2021 COVID-19 pandemic), a comparable period excluding the crisis years, and a phase of heightened geopolitical tensions (2014–2023). Results indicate that shock transmission mechanisms depend critically on structural features and external conditions. Crisis-ridden and geopolitically stressed environments exhibit higher persistence, stronger price pass-through, deeper contractions in real activity and liquidity tightening, accompanied by prolonged deviations from steady state. By contrast, more stable periods display greater resilience, moderated volatility and faster convergence dynamics. These findings highlight the importance of exchange rate regime design, coordinated monetary policy and strategic import substitution in enhancing macroeconomic stability in the BRICS economies. Strengthening reserve buffers and expanding domestic capacity for essential goods can mitigate the long-lasting effects of adverse exchange rate disturbances.

Keywords

BRICS, Exchange rate, foreign reserves, DSGE model

JEL: F31, E44, E52, F51, C63.

1. Introduction

Exchange rate dynamics have long been at the heart of academic research and policy discussions, particularly in emerging economies that are highly vulnerable to global financial volatility, geopolitical uncertainty and external shocks. Among these economies, the BRICS group (Brazil, Russia, India, China and South Africa) occupies a distinctive position. Together, they account for a significant proportion of global output and trade, while also working to reform the international financial system with the aim of reducing dependence on the US dollar. As these economies intensify their efforts to de-dollarize, expand the use of local currencies and deepen economic cooperation within the group, a thorough understanding of exchange rate movements and their underlying structural factors becomes increasingly important for monetary autonomy and overall financial stability.

The increasing push towards de-dollarization refers to the efforts of the BRICS countries to reduce their dependence on the US dollar in global trade and finance (Liu et al., 2022; Bastanifar et al., 2025). The establishment of alternative financial institutions and the strengthening of intra-BRICS trade and investment mechanisms expose them to unique vulnerabilities and reactions in foreign exchange markets. Therefore, studying exchange‑rate disturbances in the BRICS context helps illuminate how geopolitical realignments and currency reform efforts intersect with traditional financial stability concerns.

In 2009, Brazil, Russia, India and China established the BRIC association with the objective of deepening economic cooperation and reducing reliance on Western-dominated financial systems (Kuszewska & Khan, 2020). The accession of South Africa in 2010 not only broadened the group’s geopolitical and economic reach but also reinforced its collective ambition to challenge the dominance of the U. S. dollar in international trade and finance. Since then, the BRICS countries have undertaken a range of initiatives aimed at strengthening monetary cooperation, including the promotion of local currencies in bilateral trade, the establishment of new financial institutions such as the New Development Bank, and the development of alternative payment mechanisms, including BRICS Pay. Together, these initiatives signal a strategic shift towards de-dollarization and the gradual emergence of a more multipolar global financial architecture (Dyer, 2023; Marx & Khan, 2025).

The motivations underlying this transition are multifaceted. They include reducing vulnerability to U. S. dollar volatility, limiting exposure to unilateral sanctions and enhancing domestic monetary policy autonomy. By fostering greater financial independence and reinforcing economic linkages among member states, BRICS seek to recalibrate the global economic order in a manner more favorable to emerging economies. At the same time, exchange rate dynamics within BRICS economies remain closely tied to both macroeconomic fundamentals and political stability (Bastanifar et al., 2025). Figures 1 and 2 illustrate the observed movements of BRICS national currencies against the U. S. dollar and the euro, respectively. In both figures, the Y‑axis denotes official exchange rates, while the X‑axis represents the corresponding time periods. These descriptive patterns serve to motivate the subsequent model‑based analysis rather than to directly identify structural exchange‑rate shocks.

Figure 1. 

Trend of official exchange rates of BRICS in terms of the US Dollar. Source: Compiled by the authors based on World Bank (2024), available at: https://data.worldbank.org/indicator/PA.NUS.FCRF

This study addresses a critical gap in the literature by examining the macro­economic consequences of exchange rate fluctuations in the BRICS economies. Figure 1 illustrates the trend of official exchange rates of BRICS economies in terms of the US dollar over the study period. Specifically, it analyzes how structural exchange‑rate disturbances affect key macroeconomic variables — including GDP, inflation, investment and exchange rates — while accounting for external constraints such as economic sanctions and trade barriers that shape foreign‑exchange market conditions.

This study contributes to the literature in three ways. First, it develops a DSGE framework tailored to the BRICS economies in which exchange rate movements are driven by a latent disturbance process rather than directly identified structural shocks. Second, it introduces regime‑specific calibration scenarios based on empirical exchange rate persistence observed under crisis, non‑crisis, and geopolitical stress environments. Third, it provides a comparative analysis of the real and nominal macroeconomic responses in these different macroeconomic settings.

The research aims to explore the influence of persistent exchange rate disturbances within the model framework on the key macroeconomic variables — both real (GDP, consumption, investment, capital formation, real money balances) and nominal (prices, currency valuation, monetary base, and reserves) — across the BRICS economies under three distinct macro‑financial settings: a long‑term period including global crises, a comparable period excluding those crises, and the recent phase of geopolitical pressure. To attain this goal, the study employs a dynamic stochastic general equilibrium (DSGE) model that incorporates a latent exchange rate disturbance. Rather than being directly estimated as a structural shock, the persistence of this disturbance is calibrated using empirical evidence on exchange rate persistence obtained from descriptive autoregressive analysis (see Figure 2).

Figure 2. 

Trend of official exchange rates of BRICS in terms of Euro. Source: Compiled by the authors based on Eurostat (2024), available at: https://ec.europa.eu/eurostat/databrowser/view/ert_bil_eur_a/default/table?lang=en

The significance of this study is its relevance to economic policymaking and global financial stability. As the BRICS countries intensify efforts to develop monetary alternatives and challenge dollar dominance, understanding macroeconomic responses to exchange rate disturbances becomes crucial. The study contributes to ongoing discussions about the viability of a multipolar financial order and the vulnerability of emerging markets to external shocks. It provides valuable insights for policymakers, central banks, investors and academic researchers interested in the evolving intersection of geopolitics, monetary policy and global financial governance.

The divergence in exchange rate policies among the BRICS members highlights the need for greater coordination. While a fully fixed regime may limit monetary autonomy, a fully floating exchange rate regime could amplify currency instability in economies exposed to external shocks and geopolitical uncertainty. By contrast, a managed floating exchange rate framework offers a more balanced alternative. It preserves macroeconomic flexibility while limiting vulnerability to sudden shocks, particularly during periods of heightened geopolitical tension, economic sanctions or sudden changes in capital flows. This study contends that such a regime may be better suited to the structural and institutional realities of the BRICS economies as they advance toward deeper financial integration and greater monetary autonomy. Against this backdrop, the research addresses a notable gap in the literature by analyzing the macroeconomic effects of exchange rate fluctuations in the BRICS countries. It examines how exchange rate movements affect key economic indicators, including GDP, inflation, investment and trade balances, while also assessing the influence of external factors, most notably sanctions and trade barriers, on the exposure of BRICS currencies to volatility and instability (Baig et al., 2025).

To capture the dynamics and persistence of these effects, the analysis is conducted under three distinct time-span scenarios: a long-term period including global crises (2000–2023, covering the 2007–2008 financial crisis and the 2020–2021 pandemic), the same long-term horizon with the crisis years excluded (2000–2023 without the crises), and a recent geopolitical phase (2014–2023). This division enables a systematic comparison of real and nominal variable behavior across crisis, stable, and geopolitically stressed environments.

The central research questions are: (1) How do changes in exchange rates affect real macroeconomic variables such as real GDP, investment, consumption, capital formation and real money balances in the BRICS economies? (2) How do these changes influence nominal variables, including currency valuation, foreign reserves, the monetary base and domestic prices?

Research Hypotheses

The analysis is guided by the following hypotheses regarding the macroeconomic transmission of persistent exchange rate disturbances in the BRICS economies:

H1. Persistent exchange rate shocks generate heterogeneous responses from macroeconomic variables in the BRICS economies, depending on the macro‑financial scenarios.

H2. Persistent exchange rate disturbances increase exchange rate pass‑through to domestic prices, with stronger effects under crisis and geopolitical‑pressure scenarios than under the non‑crisis scenario.

H3. Exchange rate shocks lead to a decline in real money balances, reflecting liquidity erosion in the domestic monetary system.

H4. Macroeconomic variables are more likely to return to their pre‑shock steady state in the non‑crisis scenario, whereas crisis and geopolitical‑pressure scenarios are associated with incomplete equilibrium reversion and the emergence of new steady‑state levels.

To capture these dynamics, the analysis employs a DSGE framework in which exchange rate movements are driven by a latent disturbance process. Empirical observations on exchange rate persistence are used only as calibration guidance for the persistence parameter of this shock.

This specific choice of the 2014–2023 period is driven by its characterization as the peak phase of geopolitical tensions affecting the BRICS economies. Since 2014, several major events have shaped the external environment: the imposition of broad economic sanctions on Russia following the Ukraine crisis (Gurvich & Prilepskiy (2015)), the escalation of the United States–China trade war that began in 2018 (Bown (2019)), and pronounced volatility in global oil and commodity prices (Bucevska & et al., 2023). These events significantly affected the external environment and were associated with increased volatility and pressure in foreign exchange markets. Focusing on this turbulent and non‑standard environment allows for a more precise calibration of shock persistence parameters (ρ) and a detailed analysis of macroeconomic responses under sustained external stress.

The findings of this study reveal the heterogeneous and asymmetric effects of exchange rate shocks on the BRICS economies when analyzed using a calibrated DSGE framework across three distinct time scenarios: (1) the full period from 2000 to 2023, including major global crises; (2) the same period excluding crisis years (2007–2008 and 2020–2021) to capture structural resilience; and (3) the recent geopolitical phase spanning 2014 to 2023. Our impulse response functions confirm that in all scenarios, a shock generally triggers a decline in real variables (real GDP, investment and consumption) and an escalation in nominal variables (domestic prices and significant currency depreciation).

This study contributes to the growing body of literature on emerging markets, macroeconomic resilience and financial governance. By modeling the interaction between exchange rate movements and key macroeconomic variables within a unified DSGE framework, it offers practical insights for policymakers, central bankers and researchers. Its relevance is heightened by ongoing debates about the future of the global financial architecture, the risks associated with exchange rate misalignments, and the push by the BRICS countries to establish more stable and autonomous financial systems. The structure of this paper is as follows: Section 2 provides a comprehensive review of the literature on exchange rate volatility and macroeconomic performance in emerging markets. Section 3 outlines the empirical methodology, including model calibration and data sources. Section 4 presents the results from impulse response functions (IRFs), which trace the effects of exchange rate shocks on both real and nominal variables. Finally, Section 5 concludes by summarizing the key findings and providing policy recommendations for enhancing exchange rate stability and financial coordination in the BRICS context.

2. Litrature review and theoretical foundation

Exchange rate fluctuations remain a central concern in macroeconomic research, particularly for emerging economies such as the BRICS countries — Brazil, Russia, India, China and South Africa. These economies operate at the intersection of global financial integration and domestic structural transformation, at the same time pursuing strategies aimed at reducing dependence on the US dollar. In this evolving context, it has become increasingly important to understand not only the determinants of exchange rate movements, but also how they are transmitted to the macroeconomy. The existing literature on exchange rate dynamics in emerging economies can be broadly grouped into three interrelated strands. The first strand focuses on the empirical determinants and volatility of exchange rates. Studies such as Ghauri et al. (2024) identify key macroeconomic drivers — including inflation, fiscal imbalances, trade openness and capital account liberalization — as central to exchange rate fluctuations. Bastanifar et al. (2025) highlight the role of external shocks and geopolitical pressures in amplifying currency volatility, particularly in the BRICS context. This body of work provides valuable insights into observed exchange rate behaviour. However, as it predominantly relies on reduced-form approaches, it offers limited insight into the underlying structural mechanisms driving these dynamics.

The second strand emphasizes the role of exchange rate regimes and policy responses. Ball and Reyes (2008) document the phenomenon of “fear of floating” in developing economies, where concerns about inflation pass-through and financial instability lead policymakers to intervene in foreign exchange markets. Frenkel and Rapetti (2012) argue that managed floating regimes provide a pragmatic balance between stability and flexibility. This policy-oriented literature highlights the importance of institutional and monetary frameworks in shaping exchange rate outcomes but does not fully capture the dynamic transmission of exchange rate disturbances across macroeconomic variables.

The third strand consists of structural macroeconomic models, particularly Dynamic Stochastic General Equilibrium (DSGE) frameworks, which conceptualize exchange rate movements as the result of underlying structural shocks. Within this framework, exchange rate disturbances are treated as latent processes that spread throughout the economy. Jesus et al. (2020) demonstrate that DSGE models are effective in capturing the asymmetric transmission of such disturbances for policy analysis. Despite these advances, an important gap remains Unlike volatility-based models such as the Generalised Autoregressive Conditional Heteroskedasticity (GARCH) model, the DSGE approaches provide a theoretically consistent environment in the literature. Empirical studies tend to focus on observable exchange rate volatility, while DSGE-based analyses model exchange rate disturbances as unobservable structural shocks. Consequently, there is a discrepancy between the empirical characterization of exchange rate persistence and its depiction in structural macroeconomic models. This is particularly relevant for the BRICS economies, where exchange rate dynamics are heavily impacted by geopolitical tensions, sanctions and changing monetary policies.

Recent contributions have further emphasized the importance of these external factors. Feder-Sempach et al. (2025) and Hussain et al. (2024) show that the BRICS currencies exhibit heterogeneous responses to global financial stress. Luo (2024) highlights differences in hedging efficiency across the BRICS currencies. Gilal and Ismail (2025) demonstrate that geopolitical shocks generate asymmetric exchange rate responses, often accompanied by inflationary pressures. Akhyar and Rahmi (2024) emphasize that macroeconomic resilience depends not only on monetary policy but also on institutional quality. Taken together, these studies suggest that exchange rate dynamics in the BRICS economies cannot be fully understood without integrating structural, institutional and geopolitical dimensions.

Within DSGE frameworks, the persistence of exchange rate disturbances is governed by structural parameters that are not directly observable and must therefore be calibrated or estimated indirectly. However, existing studies rarely use observed exchange rate persistence in a systematic way, often conflating it with the identification of structural shocks. This limits the empirical relevance of such models. The present study addresses this gap by explicitly combining empirical and structural approaches. It models exchange rate disturbances as latent structural processes within a DSGE framework. It uses empirically observed exchange rate persistence solely for calibration purposes, rather than as a direct measure of structural shocks. By doing so, the analysis maintains theoretical consistency while enhancing empirical relevance, particularly in the context of BRICS economies under varying macroeconomic and geopolitical conditions.

3. Methodology and results

Currency crises and exchange rate volatility have been extensively analyzed within small open economy DSGE frameworks, which emphasize the interaction between monetary policy, external balances and trade openness. In particular, the dynamic propagation of exchange rate shocks and their transmission to macroeconomic aggregates can be systematically studied using optimizing open‑economy models with forward‑looking agents. In line with the standard approach of Galí and Monacelli (2002), exchange rates directly impact the intertemporal budget constraints of households and the resource constraints of the economy through external trade and foreign asset positions. In this class of models, foreign currency prices and exchange rates determine the valuation of external assets and liabilities, while monetary policy and nominal rigidities shape the macroeconomic responses to exchange rate disturbances.

3.1. Research design

The key feature of the model is the inclusion of real money balances in the household utility function within a standard Money‑in‑the‑Utility (MIU) framework, while no cash‑in‑advance constraint is imposed on consumption. This specification provides a tractable way to capture liquidity effects in the DSGE environment. Within this framework, we examine how exchange rate disturbance shocks propagate through the economy and affect key macroeconomic variables. In addition, the model incorporates a latent exchange rate disturbance shock whose persistence parameters are calibrated using empirical moments derived from observed exchange rate dynamics in BRICS economies.

The calibrated model is then employed to analyze impulse response functions (IRFs), with technical details reported in the appendix A.

To ensure a consistent interpretation of the impulse response functions, several evaluation criteria are used. Heterogeneous effects are identified by comparing differences in the magnitude and persistence of responses across scenarios and economies. Stronger exchange‑rate pass‑through is inferred when domestic prices respond more strongly or more persistently to the exchange rate disturbance relative to the non‑crisis scenario. Liquidity erosion is identified when real money balances display a negative and sustained response following the exchange rate shock. Finally, incomplete return to equilibrium is identified when variables do not converge back to their pre‑shock steady state within the simulation horizon and instead stabilize at a different level.

This paper’s methodology is based on the approach of dynamic macroeconomic equilibrium modelling, which applies modern macroeconomic theory, which is rooted in the logic of macroeconomic fluctuation models, to forecast the macroeconomic impact of time series data (Smets & Wouters, 2007).

In order to capture different macroeconomic environments and the potential role of crisis periods in exchange rate dynamics, the analysis is conducted under three distinct time-span scenarios: (i) a full long-term horizon from 2000 to 2023, including major global crises (the 2007–2008 financial crisis and the COVID‑19 pandemic( 2020 ,2021) ); (ii) the same long-term horizon, but with crisis years removed to isolate structural effects; and (iii) the core sample period of the paper, 2014–2023, corresponding to intensified geopolitical tensions and the formal inclusion of BRICS data in harmonized datasets. This three-scenario design allows for comparative DSGE simulations with two sets of shock persistence parameters: one that accounts for the influence of crises and one that excludes it. This provides a richer understanding of the transmission of exchange rate shocks in the BRICS economies.

3.1.1. Exchange rate determination in the DSGE model

In the DSGE framework, the nominal exchange rate is an endogenous variable determined by the model’s equilibrium conditions in a small open economy setting. Exchange rate dynamics arise from the interaction of monetary policy, external asset positions and price rigidities, consistent with standard open economy DSGE models (Galí and Monacelli, 2002; Barzani et al., 2026).

Fluctuations in the nominal exchange rate are driven by a latent exchange rate disturbance process, which enters the exchange rate equation as an exogenous innovation. This disturbance captures unobserved factors affecting exchange rate dynamics that are not explicitly modeled within the structural blocks of the economy.

The persistence of the latent exchange rate disturbance process is governed by an autoregressive parameter. This parameter is not identified directly from the observed exchange rate series. Instead, it is calibrated using empirical evidence on the persistence of observed exchange rate dynamics, as described in Section 3.2.3. All remaining structural relationships governing exchange rate behavior are derived from the DSGE model and solved under rational expectations.

Formally, the latent exchange rate disturbance follows a first‑order autoregressive process within the DSGE framework, where the persistence parameter governs the propagation of the shock while the innovation term represents unexpected external disturbances affecting exchange rate dynamics.

3.2. Measures

In order to analyze the macroeconomic propagation of structural disturbances, the model is calibrated using a set of structural parameters and externally specified shock processes.

3.2.1. Data collection for calibration and estimation procedures

The parameterization of the DSGE framework is organized by applying parameters through the two main approaches: building on prior research and deriving them explicitly. Some parameters are drawn from prior studies, while others are calculated using relevant ratios of key macroeconomic variables. To inform the calibration of selected structural parameters, the analysis relies on empirical characteristics of observed macroeconomic and exchange rate series. Stationarity assessments are conducted on observed exchange rate series to characterize their empirical time-series properties, which are used as calibration targets for specifying the persistence of latent exchange rate disturbance shocks, in line with the approach discussed by Haralambides et al. (2024).

In this study, the authors draw on various sources to estimate parameters. Table 1 presents the estimated ratios based on stable consumption‑to‑output ratios (cy)for the BRICS countries. These ratios are calculated using World Bank data across three distinct time‑span scenarios: Scenario 1 (2000–2023, including crisis years), Scenario 2 (2000–2023, excluding crises), and Scenario 3 (2014–2023, the period of intensified geopolitical pressures). Tables 8 to 12 report the calibrated values of the parameters that govern the persistence of the latent exchange rate disturbance processes. These values are informed by the empirical characteristics of the observed exchange rate dynamics. In accordance with Haralambides et al. (2024), the results of the test for stationarity are provided in Appendix B of the paper.

Table 1.

Parameters estimated based on stable ratios for BRICS countries (cy)

Country Scenario 1: 2000–2023 (with crises) Scenario 2: 2000–2023 (without crises) Scenario 3: 2014–2023
South Africa 0.81 0.81 0.84
China 0.55 0.55 0.54
Russia 0.68 0.68 0.71
India 0.71 0.71 0.70
Brazil 0.81 0.81 0.83

In this study, parameter values in Table 1 are used to calibrate the DSGE simulations under three distinct time‑span scenarios:

  1. Full long‑term period (2000–2023, including crises) – incorporating the 2007–2008 global financial crisis and the COVID‑19 pandemic to capture the complete dynamics of consumption/output ratios under extreme macroeconomic conditions.
  2. Long‑term period without crisis years (2000–2023, excluding crises) – removing the years 2007–2008 and 2020–2021 to isolate structural trends and reduce volatility in c/y ratios.
  3. Baseline article period (2014–2023) – focusing on the phase of intensified geopolitical tensions and consistent availability of harmonized BRICS macroeconomic data.

This scenario division ensures that the calibration process captures both structural and cyclical dimensions of consumption/output behavior, allowing for comparative shock response analysis across different economic environments.

3.2.2. Performance of the model

To assess the validity of the model, the analysis was conducted over three distinct time horizons:

  1. Long-term including crises (2000–2023): This period encompasses major economic shocks, including the global financial crisis of 2007–2008 and the COVID-19 pandemic period (2020–2021).
  2. Long-term excluding crises (2000–2023 with crisis years removed): The years 2007, 2008, 2020, and 2021 are excluded to examine parameter stability under relatively calm economic conditions.
  3. Article’s focus period (2014–2023): This is the period in which harmonized data for BRICS countries is available and geopolitical pressures — particularly sanctions against Russia and China — reached their peak. This selection provides a dataset with more stable macroeconomic ratios for other members and enables a concentrated analysis of exchange rate shocks under heightened external tensions.

The first two scenarios enable the calibration of alternative values for the persistence parameters governing latent exchange rate disturbance shocks for identical chronological spans, isolating the pure structural effects from the compounded crisis-driven volatility. This design demonstrates how removing crisis episodes substantially reduces shock persistence and facilitates faster macroeconomic recovery.

According to DSGE methodology, structural shocks such as the exchange rate shocks can be modeled over short-, medium-, and long-term horizons (Ca’ Zorzi et al 2016). In this study, the exchange rate responds endogenously to a latent structural disturbance within the DSGE framework, and the model’s moments are compared with the annual moments of economic time series variables for the BRICS countries.

To broaden international coverage, nominal EUR exchange rate series for the BRICS countries were used alongside USD-based exchange rates (source: Eurostat). The combined use of USD and EUR exchange rates allows the study to capture the effects of external shocks, sanctions and global financial developments more comprehensively. As shown in Appendix B, stationarity tests confirm that the cyclical component of the exchange rate series is stationary, indicating that autoregressive models are used solely to characterize the empirical persistence of observed exchange rate series, which serve as calibration targets for specifying the persistence of latent exchange rate disturbance processes in the DSGE model.

To ensure complete exchange rate coverage for the BRICS countries, two primary sources were employed:

  • USD-based exchange rates (LCU/USD): Annual nominal exchange rates were obtained from the World Bank database for Brazil, Russia, India, South Africa and China. For certain currencies, coverage begins well before 2000, allowing the construction of a long time series.
  • EUR/USD rates : Annual average values were retrieved from the European Central Bank (ECB) Data Portal — “External sector and exchange rates” > “Reference rates” > “US dollar/Euro, Daily” — with the frequency set to annual averages for the period 2000–2014. This rate is identical across countries, as it represents the Euro–US dollar parity.

For the 2014–2023 period, harmonized EUR exchange rate series (LCU/EUR) were directly extracted from the Eurostat database (dataset code: ert_bil_eur_a) for the ruble, rand, real, rupee and yuan.

Since official LCU/EUR data from Eurostat are available only from 2014 onward, the missing earlier values for 2000–2014 were computed indirectly using the following identity:

LCU/EUR=LCU/USDEUR/USD (1)

This procedure applies the country-specific nominal LCU/USD rates from the World Bank to the common EUR/USD rate from the ECB, ensuring consistency across the BRICS group. In the case of China, where no direct Eurostat LCU/EUR series exist even for recent years, the same indirect calculation method was applied for the entire time span, using yuan/USD from the World Bank and the EUR/USD conversion from ECB.

It should be emphasized that the use of the euro as a common reference currency does not imply a single common currency or an identical exchange rate across BRICS. Since the LCU/USD component is country-specific, the resulting LCU/EUR series also remain country-specific for each economy.

By integrating both direct and indirectly computed series, a continuous LCU/EUR panel was established for all the BRICS countries from 2000 to 2023, allowing the model to capture exchange rate shocks consistently across different currencies and periods.

3.2.3. Data and calibration strategy for exchange rate shock persistence

The observed nominal exchange rate series for the BRICS countries are analyzed first to characterize their empirical time series properties. In order to avoid spurious persistence arising from long‑run trends, the series are decomposed using the Hodrick–Prescott (HP) filter, and the cyclical component is extracted. Autocorrelation and partial autocorrelation functions are then examined for the cyclical component under each scenario (see Figures 3, 4, and 5 for Scenarios 1, 2, and 3, respectively) in order to inform the specification of persistence parameters governing a latent exchange rate disturbance process in the DSGE model. Stationarity tests applied to the cyclical component (reported in Appendix B) confirm that the series used for calibration are stationary.

Figure 3. 

Correlogram of exchangerate disturbance shocks in BRIC Economies (Scenario1). Source: Compiled by the authors

Figure 4. 

Correlogram of exchangerate disturbance shocks in BRIC Economies (Scenario2). Source: Compiled by the authors

Figure 5. 

Correlogram of exchangerate disturbance shocks in BRIC Economies (Scenario3). Source: Compiled by the authors

Rather than modeling a common currency, the analysis employs a common calibrated persistence parameter across BRICS for each scenario. This parameter is chosen to reflect the average empirical persistence of the observed exchange rate dynamics and is used to govern a latent exchange rate disturbance process within each country’s DSGE simulation.

The estimated autoregressive coefficients are interpreted solely as descriptive measures of empirical persistence in observed exchange rate dynamics. These coefficients are not treated as structural parameters of the DSGE model. Importantly, the persistence parameter used in the DSGE simulations is calibrated from the stationary cyclical component of the exchange rate series rather than from the level series. This ensures that the disturbance process represents temporary macroeconomic fluctuations rather than the near‑permanent movements in exchange rates. Instead, they serve as empirical reference points for calibrating the persistence parameter governing a latent exchange rate disturbance process within the theoretical framework. External factors such as sanctions, trade restrictions and global financial disturbances influence observed exchange rate dynamics in the data; their effects are therefore reflected only indirectly through the empirical persistence patterns used to guide the calibration of the disturbance process.

Tables 2–4 report empirical estimates of the persistence of observed nominal exchange rate series under alternative specifications and sample periods. These estimates are used to guide the calibration of persistence parameters for latent exchange rate disturbance processes in the DSGE simulations.

It is important to note that episodes of Economic sanctions and trade restrictions — common external shocks for the BRICS economies in recent years — directly affect trade flows, capital movements and external financial relations and thus have a marked impact on exchange rate dynamics. Episodes of economic sanctions, trade restrictions and financial tensions are important elements of the macroeconomic environment faced by BRICS; they may influence observed exchange rate dynamics through various channels, including trade flows and capital movements. In this study, however, such events are not modeled as distinct structural shocks within the DSGE framework. Instead, the autoregressive coefficient estimated from the cyclical component of the exchange rate series is used solely as a descriptive measure of empirical persistence in the observed data. The persistence parameter employed in the DSGE simulations is therefore calibrated to reproduce the average persistence of exchange rate fluctuations, without attributing this persistence to any specific geopolitical or policy event.

Given the autocorrelation structure of the exchange rate series for the BRICS countries, which shows a gradually declining pattern in the autocorrelation function and a dominant spike at the first lag in the partial autocorrelation function, the Box–Jenkins methodology suggests the use of an AR(1) specification. The estimation results are presented in Tables 2–4.

Table 2.

AR (1) model estimation results for exchange rate price data in the BRICS countries (Scenario 1)

Variable Coefficient Std.Error t-Statistic Prob
AR(1) 0.484508 0.080739 6.000934 0.0000
R-squared Adjusted R-Squared 0.241668 0.234957 Akaike info criterion Schwarz criterion Hannan-Quinn criter -1.406231 -1.358493 -1.386854
Table 3.

AR (1) model estimation results for exchange rate price data in the BRICS countries (Scenario 2)

Variable Coefficient Std.Error t-Statistic Prob
AR(1) 0.470265 0.079933 5.883250 0.0000
R-squared Adjusted R-Squared 0.234483 0.227708 Akaike info criterion Schwarz criterion Hannan-Quinn criter -1.460526 -1.412788 -1.441150
Table 4.

AR (1) model estimation results for exchange rate price data in the BRICS countries (Scenario 3)

Variable Coefficient Std.Error t-Statistic Prob
AR(1) 0.459568 0.138292 3.323173 0.0000
R-squared Adjusted R-Squared 0.204344 0.185841 Akaike info criterion Schwarz criterion Hannan-Quinn criter -1.162847 -1.082551 -1.132914

The correlogram analysis suggests that there is moderate persistence in exchange rate dynamics. Across the three scenarios, the estimated AR(1) coefficients range from around 0.46 to 0.48, indicating that exchange rate fluctuations demonstrate short-term persistence, though they do not exhibit near-unit root behaviour.

Consequently, the persistence parameter used in the DSGE simulations reflects the dynamics of short‑term exchange rate fluctuations rather than the near‑unit‑root behavior observed in the level series.

Tables 5 to 9 show the calibrated structural parameters used in the DSGE model for each BRICS country. While most parameters are taken from existing literature, the persistence parameters governing the exchange rate disturbance process were calibrated by the authors.

Parameter β represents the household discount factor. Parameters α (share of capital in output) and δ (capital depreciation rate) characterize the production structure of the enterprise sector. The parameters ρS1, ρS2, and ρS3 denote the persistence parameters of the latent exchange rate disturbance process under Scenarios 1, 2, and 3 respectively.

Table 5.

Calibrated parameters for India

Parameter Value Description Sources
β 0.9823 discount factor Khan et al., 2025
α 0.33 share of capital in output, intermediate goods Anand & Khera, 2016
δ 0.025 rate of depreciation Khan et al., 2025
ρS1 0.48 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 1 Calibrated by the authors
ρS2 0.47 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 2 Calibrated by the authors
ρS3 0.45 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 3 Calibrated by the authors
Table 6.

Calibrated parameters for Brazil

Parameter Value Description Sources
β 0.9882 discount factor Carvalho & Valli (2011)
α 0.3 share of capital in output, intermediate goods Carvalho & Valli (2011)
δ 0.025 rate of depreciation Pereira (2022)
ρS1 0.48 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 1 Calibrated by the authors
ρS2 0.47 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 2 Calibrated by the authors
ρS3 0.45 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 3 Calibrated by the authors
Table 7.

Calibrated parameters for South Africa

Parameter Value Description Sources
β 0.99 discount factor Khan et al., 2025
α 0.5 share of capital in output, intermediate goods Khan et al., 2025
δ 0.025 rate of depreciation Du Plessis et al, 2014
ρS1 0.48 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 1 Calibrated by the authors
ρS2 0.47 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 2 Calibrated by the authors
ρS3 0.45 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 3 Calibrated by the authors
Table 8.

Calibrated parameters for Russia

Parameter Value Description Sources
β 0.90 discount factor Martyanova & Polbin (2023)
α 0.35 share of capital in output, intermediate goods Martyanova & Polbin (2023)
δ 0.11 rate of depreciation Dubrovskaya et al., 2022
ρS1 0.48 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 1 Calibrated by the authors
ρS2 0.47 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 2 Calibrated by the authors
ρS3 0.45 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 3 Calibrated by the authors
Table 9.

Calibrated parameters for China

Parameter Value Description Sources
β 0.966 discount factor Seneca (2010)
α 0.30 share of capital in output, intermediate goods Seneca (2010) & the authors’ calculation
δ 0.1 rate of depreciation Liu & Ou (2024)
ρS1 0.48 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 1 Calibrated by the authors
ρS2 0.47 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 2 Calibrated by the authors
ρS3 0.45 Persistence parameters governing the latent exchange rate disturbance process under Scenarios 3 Calibrated by the authors

The persistence parameters reported in Tables 5–9 are calibrated using empirical persistence estimates obtained from the stationary cyclical component of the exchange rate series, as discussed in Section 3.2.3.

All structural parameters are assigned prior distributions based on benchmark values from the DSGE literature. These values serve as prior means and are updated using observable data through Bayesian estimation to obtain the posterior distributions.

The model assumes a common structural framework for the BRICS economies and introduces cross-country heterogeneity through the calibration of country-specific parameters. This approach enables the comparability of impulse responses to be maintained while accounting for variations in macroeconomic conditions and policy environments.

Table 10 shows the parameters that require calibration, based on World Bank data for the BRICS economies. The dataset covers the period from 2000 to 2023 for the three scenarios described: 2000 to 2023 with crisis, 2000 to 2023 without crisis, and 2014 to 2023. The variables presented in this table include the growth rate of money (θ), the growth rate of production (n) and the inflation rate (π).

Table 10.

Calibrated parameters

Country Scenario 1: 2000–2023 (with crises) θ n π Scenario 2: 2000–2023 (without crises) θ n π Scenario 3: 2014–2023 θ n π
China 0.14 0.12 0.02 0.14 0.16 0.01 0.07 0.06 0.01
Russia 0.21 0.11 0.09 0.20 0.13 0.10 0.06 0.01 0.05
South Africa 0.1 0.05 0.05 0.10 0.06 0.05 0.05 0.005 0.04
Brazil 0.14 0.06 0.06 0.14 0.08 0.06 0.05 0.01 0.06
India 0.14 0.09 0.06 0.13 0.11 0.06 0.1 0.06 0.04

3.3. IRFs techniques and the results

In DSGE models, impulse response functions (IRFs) measure the dynamic effects of shocks on model variables. The economic plausibility of the impulse responses was carefully examined to ensure that the model generates theoretically consistent dynamics. In particular, it is expected that a depreciation shock will increase domestic prices via exchange rate pass-through, reduce real money balances due to higher inflation, and put contractionary pressure on real variables such as output, consumption and investment. The simulated responses reported in Figures 6–20 follow this expected transmission mechanism, suggesting that the model produces economically consistent dynamics across the BRICS economies.

In all figures, the horizontal axis shows the number of years after the shock, while the vertical axis represents the deviation of each macroeconomic variable from its steady‑state level. The summary statistics of the responses are presented in Tables 11–13.

It is important to clarify that the cross‑country differences observed in the impulse response functions do not stem from differences in the persistence of the exchange rate shock, which is held constant within each scenario. Instead, these heterogeneous responses reflect underlying structural differences across BRICS economies, including variations in calibrated macroeconomic parameters, inflation dynamics, monetary conditions and growth processes. Consequently, the observed IRFs capture genuine economic heterogeneity rather than purely mechanical effects driven by model parameterization.

3.3.1. Impulse response function results – Scenario 1 (2000–2023 with crises)

Scenario 1 includes the full sample period (2000–2023) and therefore captures major global crises such as the 2007–2008 financial crisis and the COVID‑19 pandemic (2020–2021). The resulting impulse responses for the key macroeconomic variables analyzed in this scenario are presented in Figures 6, 7, 8, 9, and 10. This scenario allows the analysis to examine how exchange rate shocks propagate during periods of heightened macroeconomic volatility.

Figure 6. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in South Africa (Scenario 1). Source: Compiled by the authors

Figure 7. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in China (Scenario 1). Source: Compiled by the authors

Figure 8. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in Russia (Scenario 1). Source: Compiled by the authors

Figure 9. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in India (Scenario 1). Source: Compiled by the authors

Figure 10. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in Brazil (Scenario 1). Source: Compiled by the authors

3.3.2. Impulse response function results – Scenario 2 (2000–2023 without crises)

Scenario 2 excludes the crisis years (2007–2008 and 2020–2021) in order to examine the structural effects of exchange rate shocks under relatively stable macroeconomic conditions. The corresponding impulse response functions for this scenario are illustrated in Figures 11, 12, 13, 14, and 15.

Figure 11. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in South Africa (Scenario 2). Source: Compiled by the authors

Figure 12. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in China (Scenario 2). Source: Compiled by the authors

Figure 13. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in Russia (Scenario 2). Source: Compiled by the authors

Figure 14. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in India (Scenario 2). Source: Compiled by the authors

Figure 15. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in Brazil (Scenario 2). Source: Compiled by the authors

3.3.3. Impulse response function results – Scenario 3 (2014–2023)

Scenario 3 focuses on the more recent period (2014–2023), characterized by intensified geopolitical tensions and improved data consistency for BRICS economies. Figures 16, 17, 18, 19, and 20 depict the impulse response functions observed under this specific macroeconomic setting.

Figure 16. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in South Africa (Scenario 3). Source: Compiled by the authors

Figure 17. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in China (Scenario 3). Source: Compiled by the authors

Figure 18. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in Russia (Scenario 3). Source: Compiled by the authors

Figure 19. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in India (Scenario 3). Source: Compiled by the authors

Figure 20. 

Impulse response functions (IRFs) of exchange rate shocks on macroeconomic model variables in Brazil (Scenario 3). Source: Compiled by the authors

The impulse response functions suggest that exchange rate disturbances lead to persistent adjustments in several macroeconomic variables. In many cases, variables converge slowly toward a new equilibrium level rather than returning immediately to their pre‑shock state. This persistence indicates that sustained exchange rate disturbances may shift macroeconomic variables away from their targeted levels for an extended period. Therefore, persistent exchange rate shocks may cause macroeconomic variables to deviate significantly from their target values.

The estimated impulse responses reveal that both output and consumption experience a short‑run decline following the exchange rate shock but gradually return to their steady‑state levels after roughly ten periods. This dynamic pattern confirms that the shock is transitory rather than permanent, implying that the model is stable. The degree of persistence differs across scenarios, reflecting structural factors rather than statistical artifacts. Specifically, persistence tends to be stronger in environments with weaker monetary policy credibility, higher exchange rate pass‑through and tighter balance‑sheet constraints, all of which delay the adjustment of real activity.

Tables 11, 12 and 13 present the application of the exchange rate standard deviation as a shock to selected macroeconomic variables across three scenarios. The magnitude and direction vary among the respective countries in each scenario.

Table 11.

Response of macroeconomic variables to a 1% exchange rate shock scenario 1

Countries Real Gdp(y)% Investment (i)% Consumption (C)% Domestic Price Capital (k) % National Currency (M)% Real balance of money (m)% Money Base (MB)
India -0.07 -0.04 -0.5 +0.12 -0.04 -0.5 -4 -0.1
Brazil -0.05 -0.025 -0.35 +0.09 -0.025 -0.3 -3 -0.1
China -0.026 -0.012 -0.18 +0.08 -0.012 -0.12 -2 -0.05
Russia -0.16 -0.12 -1.1 +0.25 -0.12 -1.5 -6 -0.5
South Africa -0.045 -0.04 -0.28 +0.1 -0.035 -0.3 -2.2 -0.1
Table 12.

Response of macroeconomic variables to a 1% exchange rate shock scenario 2

Countries Real Gdp(y)% Investment (i)% Consumption (C) % Domestic Price Capital (k) % National Currency (M)% Real balance of money (m)% Money Base (MB)
India -0.07 -0.04 -0.48 +0.12 -0.04 -0.5 -4 -0.18
Brazil -0.05 -0.025 -0.35 +0.1 -0.025 -0.3 -3 -0.1
China -0.016 -0.008 -0.12 +0.06 -0.007 -0.065 -1.8 -0.02
Russia -0.26 -0. 2 -1.6 +0.31 -0.18 -2 -10 -0.6
South Africa -0.045 -0.04 -0.28 +0.1 -0.035 -0.3 -2.2 -0.18
Table 13.

Response of macroeconomic variables to a 1% exchange rate shock scenario 3

Countries Real Gdp (y) % Investment (i) % Consumption (C) % Domestic Price Capital (k) % National Currency (M) % Real balance of money (m)% Money Base (MB)
India -0.032 -0.018 -0.21 +0.08 -0.016 -0.15 -2.1 -0.09
Brazil -0.05 -0.025 -0.35 +0.09 -0.025 -0.25 -3 -0.1
China -0.016 -0.008 -0.12 +0.06 -0.007 -0.065 -1.8 -0.02
Russia -0.036 -0.02 -0.25 +0.09 -0.025 -0.2 -2.5 -0.1
South Africa -0.03 -0.02 -0.18 +0.09 -0.02 -0.15 -2 -0.1

4. Discussion (IRF Analysis)

The DSGE results and IRFs are interpreted jointly to identify the main transmission mechanisms of exchange rate disturbances in the BRICS economies, providing support for еру hypotheses. Rather than discussing each figure as an isolated result, this section integrates the responses of real and nominal variables across scenarios in order to explain how exchange rate depreciation generates inflationary pressures, triggers monetary adjustment, weakens liquidity conditions, and ultimately affects output and consumption. The comparison across scenarios further clarifies how macroeconomic stability and structural resilience determine the persistence of these effects.

4.1. Equilibrium return dynamics

The IRFs indicate that the persistence of exchange rate disturbances differs markedly across the three scenarios and plays a central role in shaping macroeconomic adjustment. In Scenario 1 (2000–2023 including global financial and health crises) and Scenario 3 (2014–2023 under heightened geopolitical pressures), most variables do not return to their pre-shock equilibrium levels within the simulation horizon. This suggests that exchange rate disturbances in periods of crisis or geopolitical tension are transmitted more persistently through domestic prices, liquidity conditions, and policy reactions, thereby shifting the economy toward a new equilibrium. By contrast, in Scenario 2 (2000–2023 excluding crisis years), most variables gradually converge toward a new equilibrium level, indicating that in a more stable macroeconomic environment the effects of exchange rate disturbances are more effectively absorbed, allowing the economy to adjust more smoothly towards a stable post‑shock equilibrium.

4.2. Effects on real variables

The responses of real variables point to a contractionary effect of exchange rate disturbances across all BRICS economies. This effect reflects the transmission of the shock from currency depreciation to higher domestic prices, which reduces real purchasing power and dampens domestic demand. Real GDP declines in all countries, generally ranging between approximately −0.02% and −0.26%, with the largest contraction observed in Russia. Investment reacts more moderately, although the decline is more pronounced in Russia, suggesting greater sensitivity of capital formation to financial and external shocks. Consumption generally shows a stronger decline than GDP and investment, although the magnitude varies considerably across countries, with the largest contractions observed in Russia, reflecting the erosion of real purchasing power following higher domestic prices and tighter monetary conditions. Taken together, these responses suggest that the exchange rate disturbance affects the real economy mainly through inflationary pressure, reduced household demand, and weaker overall domestic absorption.

4.3. Effects on Nominal Variables

The nominal responses reveal the main transmission channel of the exchange rate disturbance. In all scenarios, the shock leads first to a depreciation of the national currency, although the magnitude differs across countries and periods. This depreciation is then transmitted into domestic prices through import price pass-through, generating inflationary pressures that are particularly strong in Russia, with moderate increases observed in India, Brazil and South Africa. As inflation rises, the real balance of money contracts sharply, indicating a substantial erosion of liquidity. This decline is especially pronounced in China and India, where the simulations show that exchange rate disturbances are associated with a much stronger monetary adjustment. The contraction in the monetary base reinforces this mechanism and is consistent with a policy response aimed at limiting inflationary pressures and stabilizing macroeconomic conditions. Taken together, the nominal results suggest that exchange rate disturbances are not limited to currency valuation effects. They propagate through domestic price formation, reduce real liquidity and trigger contractionary monetary dynamics, thereby amplifying the adverse effects on the broader economy. These monetary and price adjustments explain the contraction observed in real variables, such as GDP, consumption and investment, discussed in the previous section.

4.4. Policy implications

While the exchange rate disturbances analyzed in this study are modeled as latent structural shocks within the DSGE framework, the policy implications follow directly from the transmission mechanisms identified in the IRF analysis. The results show that exchange rate depreciation feeds into domestic inflation, erodes real liquidity, and weakens real economic activity, particularly in scenarios characterized by crisis conditions or geopolitical tension. From this perspective, reducing vulnerability to exchange rate disturbances requires a combination of exchange rate management, stronger foreign reserve buffers and lower dependence on imported strategic goods in critical sectors. As illustrated in Figures 1 and 2, the period of the pandemic (2020–2022) further highlights that economies with greater exposure to external currency movements, such as Russia and India, may be more vulnerable in sectors that depend on imported goods and services, including healthcare-related inputs. These observations reinforce the broader finding that exchange rate stability and strategic domestic capacity are essential components of macroeconomic resilience. Enhanced regional monetary cooperation and long-term planning for essential goods and services could therefore improve resilience to future global shocks.

5. Conclusion

Using a calibrated DSGE framework, this study shows that persistent exchange rate disturbances have a significant macroeconomic impact on the BRICS economies. These disturbances propagate through currency depreciation, higher domestic prices due to exchange rate pass‑through, and a contraction in real money balances, ultimately leading to declines in real output, investment and consumption.

Overall, the findings indicate that vulnerability to exchange rate shocks is shaped less by the size of the disturbance itself and more by underlying structural and macroeconomic conditions. Economies with stronger macroeconomic stability, higher policy credibility, adequate foreign reserve buffers and lower dependence on imported strategic goods tend to exhibit milder and more manageable macroeconomic adjustments.

In essence, the results highlight the importance of strengthening resilience through sound policy frameworks, effective exchange rate management, strategic domestic production capacity and deeper regional cooperation. Improving these dimensions would enable the BRICS economies to better absorb external shocks and maintain long-term macroeconomic stability.

References

  • Akhyar, M., & Rahmi, R. (2024). National economic resilience in the face of global economic crises: Macroeconomic perspective. Golden Ratio of Social Science and Education, 4(2), 190–200. https://doi.org/10.52970/grsse.v4i2.284
  • Anand, R., & Khera, P. (2016). Macroeconomic impacts of product and labor market deregulation in India. IMF Working Paper No. 16/47. International Monetary Fund.
  • Baig, S., Shidhani, R. A., & Khan, K. H. (2025). The Impact of Geoeconomic and Political Factors in the Middle East on the Development Prospects of IMEC. Asian Journal of Middle Eastern and Islamic Studies, 19(2), 115-129. https://doi.org/10.1080/25765949.2025.2605884
  • Barzani, M. V., Khan, K. H., Bastanifar, I., & Mohammadi, Z. (2026). Macroeconomic implications of international regional settlement corridors: Toward a new financial regional integration in Eurasia. Journal of Economic Integration, 41(1), 71-95. https://doi.org/10.11130/jei.2025029
  • Bastanifar, I., Khan, K. H., & Koch, H. (2025). Understanding BRICSization through an economic geopolitical model. Journal of Open Innovation: Technology, Market, and Complexity, 11(1), 100440. https://doi.org/10.1016/j.joitmc.2024.100440
  • Bucevska, V., Gjelevski, B., & Matevska, L. (2023). Oil prices and their long-term relationship with macroeconomic and financial indicators. Economic Review: Journal of Economics and Business, 21(1), 3–24. https://doi.org/10.51558/2303-680X.2023.21.1.3
  • Ca’ Zorzi, M., Kolasa, M., & Rubaszek, M. (2016). Exchange rate forecasting with DSGE models. ECB Working Paper No. 1905. European Central Bank.
  • Carvalho, F. A., & Valli, M. (2011). Fiscal policy in Brazil through the lens of an estimated DSGE model. Working Paper No. 240. Central Bank of Brazil.
  • Dixit, A. K., & Stiglitz, J. E. (1977). Monopolistic competition and optimum product diversity. The American Economic Review, 67(3), 297–308.
  • Du Plessis, A., Lomax, K., & Steinbach, R. (2014). A medium-scale DSGE model for South Africa: Calibrated and estimated perspectives. South African Reserve Bank Working Paper.
  • Feder-Sempach, E., Szczepocki, P., & Bogołębska, J. (2025). BRICS currencies as safe havens and hedges: An empirical analysis of their potential for G7 stock markets before and during the COVID-19 pandemic. Economic and Political Studies, 13(1), 70–97. https://doi.org/10.1080/20954816.2025.2453352
  • Frenkel, R., & Rapetti, M. (2012). Exchange rate regimes in the major Latin American countries since the 1950s: Lessons from history. Revista de Historia Económica, 30(1), 157–188. https://doi.org/10.1017/S0212610911000292
  • Galí, J., & Monacelli, T. (2002). Monetary policy and exchange rate volatility in a small open economy. NBER Working Paper No. 8905. National Bureau of Economic Research.
  • Galí, J., & Rabanal, P. (2005). Technology shocks and aggregate fluctuations: How well does the RBC model fit postwar US data? In M. Gertler & K. Rogoff (Eds.), NBER Macroeconomics Annual 2004. MIT Press.
  • Ghauri, S. P., Ahmed, R. R., Streimikiene, D., Qadir, H., & Hayat, A. (2024). Macroeconomic factors driving exchange rate volatility and economic sustainability: Case study of Pakistan. Amfiteatru Economic, 26(66), 612–628.
  • Gilal, A. A., & Ismail, N. A. (2025). Economic policy uncertainty and foreign reserves: Empirical evidence from BRICS countries. International Journal of Economics & Business Administration, 13(2), 3–23.
  • Haralambides, H., Bastanifar, I., Khan, K. H., & Shahryari, Z. (2024). Asymmetric distance and business cycles (ADBC): A new understanding of distance in international trade models through the example of Iran’s trade corridors. The Journal of Economic Asymmetries, (30), e00389. https://doi.org/10.1016/j.jeca.2024.e00389
  • Hussain, M., Bashir, U., & Rehman, R. U. (2024). Exchange rate and stock prices volatility connectedness and spillover during pandemic-induced crises: Evidence from BRICS countries. Asia-Pacific Financial Markets, 31(1), 183–203. https://doi.org/10.1007/s10690-023-09411-0
  • Jesus, D. P. D., Besarria, C. D. N., & Maia, S. F. (2020). The macroeconomic effects of monetary policy shocks under fiscal constraints: An analysis using a DSGE model. Journal of Economic Studies, 47(4), 805–825. https://doi.org/10.1108/JES-01-2019-0011
  • Katagiri, M. (2024). Systematic foreign exchange intervention and macroeconomic stability: A Bayesian DSGE approach. International Journal of Central Banking, 20(2), 291–342.
  • Kozlovtceva, I., Ponomarenko, A., Sinyakov, A., & Tatarintsev, S. (2019). Financial stability implications of policy mix in a small open commodity-exporting economy. Bank of Russia Working Paper No. 42.
  • Kuszewska, A., & Khan, K. H. (2020). The significance of India’s (re) connectivity strategy in Central Asia: an introduction. KW Publishers Pvt Ltd
  • Liu, C. H., & Ou, Z. (2024). Has fiscal expansion inflated house prices in China? Evidence from an estimated DSGE model. International Review of Economics & Finance, (96), 103541. https://doi.org/10.1016/j.iref.2024.103541
  • Liu, Z. Z., & Papa, M. (2022). Can BRICS de-dollarize the global financial system? Cambridge University Press.
  • Luo, H. (2024). Currency volatilities of BRICS countries: The impact of commodity prices, interest rates and geopolitical risks. University of the Witwatersrand.
  • Martyanova, E. V., & Polbin, A. V. (2023). General equilibrium model with the entrepreneurial sector for the Russian economy. Russian Journal of Economics, 9(2), 109–133.
  • Marx, A., & Khan, K. H. (2025). Transnational Economic Corridors. A New Form of Global Governance? In Strategic Navigation through Economic Corridors: Intersecting Geoeconomics and Geopolitics (pp. 15-33). Springer Nature Singapore.
  • Pereira, R. M. (2022). Estimating a stylized fiscal policy DSGE model for Brazil. Discussion Paper No. 241. Institute of Applied Economic Research (IPEA).
  • Seneca, M. (2010). A DSGE model for Iceland. Central Bank of Iceland Working Paper No. 50.
  • Smets, F., & Wouters, R. (2007). Shocks and frictions in U. S. business cycles: A Bayesian DSGE approach. European Central Bank Working Paper No. 722.

Data Availability and Supplementary Materials: Due to the extensive nature of the technical appendices, the full derivation of the DSGE model, additional calibration tables, and supplementary results are also available online at https://drive.google.com/file/d/1loPB5xBGbbdrPq9EcVMSFM9CaFVEteJW/view?usp=drivesdk

Appendix A

A1. Households

In this framework, the economy consists of a single representative household whose preferences depend on three distinct components: domestic consumption ct, leisure 1 – ht, and real money balances (mt). The intertemporal utility function is expressed as:

MaxEt∑t=0∞βt[log⁡(ct)+ψlog⁡(1−ht)+log⁡(mt)] (2)

Here, β ∈ (0,1) represents the subjective discount factor measuring the weight assigned to future versus present utility. The parameter ψ captures the relative importance of leisure in household welfare. Real money balances mt reflect the quantity of domestic currency that households in BRICS economies require to preserve purchasing power in the presence of inflationary trends. Variable ct denotes domestic consumption at time t, and ht is the time devoted to labor supply.

The household accumulates physical capital according to the standard law of motion:

kt+1=(1−δ)kt+it (3)

Households supply labor to production sectors, rent capital to firms, and hold real money balances. Their optimal decisions over ct, ht, mt, and it are determined by maximizing utility subject to the intertemporal budget constraint, taking market wages and returns on capital as given. Through this setup, the household sector provides the fundamental link between consumption, savings, investment and monetary demand in the DSGE environment (Galí,& Monacelli, 2002).

At the end of each period, the household updates its capital stock according to the standard capital accumulation law. In this framework, kt denotes the capital stock in the current period, while kt + 1 represents the capital stock in the next period. Depreciation at rate δ reduces the value of the existing capital stock, while investment it increases the capital available for the following period. Accordingly, the capital stock in the next period equals the undepreciated portion of current capital plus current investment.

To capture the interplay between consumption, investment, and external transactions, the representative household’s intertemporal budget constraint can be expressed as:

ptct+ptit+ExtTrt+Mt=[ p t r t k t+ p t w t h t+ Π t]+iitBt+M t − 1−Tt−B t − 1 (4)

Where:

ct — household consumption

it — investment expenditure

Ext — bilateral exchange rate (USD or EUR per unit of domestic currency) for each BRICS member, with historical trends presented in Figures 1 and 2

trt — volume of foreign exchange transactions, influenced by national foreign reserves

ExtTr — value of foreign trade costs faced by the household; movements in Ext​, which reflect the realization of an underlying exchange rate disturbance process.

pt — domestic price index for consumption and investment goods

Mt — nominal money holdings for transactions

rt — rental rate of capital

wt — real wage

Пt— profits distributed by firms

iit — interest rate on household savings or credit

Bt — savings or loans held by households

Tt — lump-sum tax liabilities

By substituting Equation (3) into Equation (4), we obtain the following expression:

ptct+pt(kt+1−(1−δ)kt)+Extrt+Mt=[ptrtkt+ptwtht+Πt]+iitBt+Mt−1−Tt−Bt−1 (5)

To express the household budget constraint in real terms, all nominal variables are divided by the aggregate price level pt. This transformation converts nominal magnitudes into real variables and allows the model to be written in stationary real terms. Accordingly, defining real money balances as mt=Mtpt and expressing all other nominal variables in real terms, the household’s intertemporal budget constraint can be written as follows:

ct+(kt+1−(1−δ)kt)+Exttrt+mt==[rtkt+wtht+Πt]+iitbt+mt−11+πt−Tt−bt−11+πt (6)

The Lagrangian associated with the household’s optimization problem can therefore be written as follows:

L=Et∑t=0∞βt[log⁡(ct)+ψlog⁡(1−ht)+log⁡(mt)]−−Etβtλf[ct+(kt+1−(1−δ)kt)+Exttrt+mt−[rtkf+wtht+Πf]−iitbt−mt−11+πt+Tt+bt−11+πf] (7)

The household chooses c, m, bt, ht and kt + 1, to maximize the utility stream given by Eq. (2), subject to the capital evolution in Eq. (3) and the budget constraint above (Eq. 4). This optimization yields first-order conditions that define the optimal trajectories of consumption, money balances, bond holdings, labor supply, and capital over time.

Maximizing the household’s utility function subject to the budget constraint (Eq. 4) yields the following first‑order conditions:

1ct=λt (8)

ψ1−ht=λtwt (9)

1mt=λt−β⋅Etλt+11+πt (10)

λt=λt+1⋅βEt[(rt+1)+1−δ] (11)

λt⋅iit=λt+1⋅Etβ1+πt (12)

Using the first‑order conditions, the equilibrium relationships can be derived. By substituting Equation (8) into Equation (9), the labor supply condition is obtained, which characterizes the optimal trade‑off between leisure and consumption for the representative household.

ψ1−ht=wtct (13)

By substituting Equation (8) into Equation (10), we obtain:

1mt=1ct−β⋅Et(1+πt)⋅ct+1 (14)

This condition represents the optimal money demand equation. It shows that households choose real money balances by equating the marginal utility of holding money with the opportunity cost of holding money in terms of forgone consumption and expected inflation. Higher expected inflation reduces the demand for real money balances because it increases the cost of holding money.

By substituting Equation (8) into Equation (11), we obtain:

1ct=1ct+1⋅βEt[(rt+1)+1−δ] (15)

This expression represents the consumption Euler equation, which characterizes the household’s optimal intertemporal consumption decision. It shows that households allocate consumption over time by comparing the marginal utility of consumption today with the expected discounted marginal utility of consumption tomorrow, adjusted by the expected return on capital.

By substituting Equation (8) into Equation (12), the optimal bond‑holding condition can be written in terms of consumption. This condition represents the Euler equation for bonds. It states that households allocate their savings between periods by comparing the marginal utility cost of purchasing bonds today with the expected discounted marginal utility benefit from bond returns tomorrow, adjusted for expected inflation. In equilibrium, households choose their bond holdings in a way that satisfies the trade-off between current and future consumption.

iitct=Etβ(1+πt)ct+1 (16)

Here, λt represents the Lagrange multiplier associated with the household’s intertemporal budget constraint, capturing the marginal utility of wealth at time t.

These conditions jointly describe the optimal trajectories for consumption, labor supply, money holdings, capital and investment.

A2. Final-good producing firm

The final good is assembled from a continuum of differentiated intermediate goods indexed by j∈[0,1]. Under the assumption that these goods are imperfect substitutes, aggregation follows a CES framework with elasticity of substitution θ>1. The Dixit–Stiglitz (1977) composite is given by:

yt=|∫01yjtθ−1θ|θθ−1 (17)

Here:

yt — quantity of the final output

yjt — quantity of intermediate good j

θ— demand elasticity across varieties

Pt — price index for the final good

pjt — price of intermediate variety j

The final‑good firm chooses purchase quantities of each yjt to maximize the profit:

max(ptyt−∫01pjtyjtdj) (18)

The corresponding optimality condition yields the demand curve for variety j:

yjt=(pjtPt)−θyt (19)

which implies the following aggregate price index:

Pt=[∫01pj,t1−θ]11−θ (20)

A3. Intermediate-good producing firm

Intermediate‑good producers operate under monopolistic competition, hiring capital kt and labor lt to produce output according to a constant‑returns‑to‑scale Cobb–Douglas technology:

yj,t=zlkj,lαlj,t1−αα∈(0,1) (21)

Where:

α — capital share in production

zt — total factor productivity (TFP) shock

Price dynamics feature nominal rigidity captured by the Rotemberg (1982) quadratic adjustment cost formulation, à la Dib (2003):

Ψ2(pj,tpj,t−1(1+πss)−1)2yj,t (22)

Here, ψ ≥0 controls the degree of price stickiness:

ψ =0: fully flexible prices

ψ >0: sticky prices

Parameter definitions:

θ: elasticity of substitution between intermediate varieties (θ>1)

α: capital share in output

ψ: Rotemberg adjustment cost parameter

zt: total factor productivity shock

πss: steady‑state inflation rate

Intermediate-good producers choose prices to maximize the expected discounted stream of profits. Each firm takes aggregate variables and demand for its variety as given and solves:

maxEt∑s=0∞βs⋅Λt,t+s[pj,t+s⋅yj,t+s−MCt+s⋅yj,t+s−ψ2(pj,t+spj,t+s−1(1+πss)−1)2yj,t+s] (23)

where MCt denotes real marginal cost and ​Λt,t + s is the stochastic discount factor of households. Firms face the demand schedule derived from the final‑good producer:

yjt=(pjtPt)−θyt (24)

The optimal pricing condition implied by this problem generates the New Keynesian Phillips Curve under Rotemberg price adjustment costs.

The Rotemberg price adjustment cost parameter (ψ) is calibrated based on the value reported in Furlanetto and Seneca (2012), with a benchmark value of ψ = 174.7. In the Bayesian estimation, this value is used as the prior mean, allowing the data to update the parameter around this benchmark.

A4. Calibration of the latent exchange rate disturbance process

To characterize the empirical persistence of observed nominal exchange rate series, we estimate a pooled AR(1) process using panel data on BRICS currency pairs against the USD and EUR. These estimates do not identify structural shocks. Rather, they summarize the time‑series properties of observed exchange rate movements and are used solely as calibration targets within the DSGE framework.

Within the DSGE model, exchange rate fluctuations are driven by a latent exchange rate disturbance process. The persistence parameter of this unobserved process is calibrated so that the model‑implied exchange rate dynamics match the empirical persistence observed in the data, as summarized by the AR(1) estimates.

ztex=ρzt−1ex+εtexεtex∼N(0,σ2) (25)

Where:

ρ — persistence parameter calibrated using AR(1) estimates from the BRICS panel

ztex — latent exchange rate disturbance process

ρ — persistence parameter of the latent disturbance, calibrated using AR(1)

εtex — innovation to the latent disturbance process

The persistence parameter ρ is calibrated separately for each historical scenario considered in the analysis. Specifically, three AR(1) coefficients are estimated from exchange rate data corresponding to: (i) the full sample period including crisis episodes (2000–2023), (ii) a non‑crisis sample excluding the global financial crisis and pandemic years (2007–2008, 2020–2021), and (iii) the geopolitical tension period (2014–2023). Each scenario therefore employs its own calibrated persistence parameter in the DSGE simulations.

Although the persistence parameter (ρ) of the exchange rate disturbance is calibrated as a common value across BRICS economies within each scenario, the model does not impose identical economic structures across countries. Cross‑country heterogeneity is explicitly introduced through country‑specific calibration of key structural parameters, including the discount factor (β), capital share (α), depreciation rate (δ), monetary growth (θ), inflation (π) and output growth (n), as reported in Tables 5–9.

Adopting a pooled persistence parameter allows the analysis to isolate how a comparable exchange rate disturbance propagates through structurally different economies, thereby enhancing the cross‑country comparability of impulse response functions. Moreover, given the limited time dimension of the panel, this strategy mitigates over‑parameterization and estimation instability without constraining country‑specific economic dynamics.

The AR(1) estimates reported in this section characterize the persistence of observed exchange rate series. They are not structural shocks, but serve as calibration targets for the persistence of the latent exchange rate disturbance process used in the DSGE simulations.

The eigenvalue structure of the linearized system satisfies the Blanchard–Kahn conditions, ensuring local stability. Consistent with the simulated IRFs, all endogenous variables return to their steady‑state values over time following an exchange rate shock. Hence, the shock is transitory, and the persistence observed in the impulse responses arises from nominal and real frictions rather than model instability.

A 5: Central bank balance sheet and shock transmission mechanism

The interaction between central bank foreign exchange (FX) interventions, reserves management, and monetary policy has become a central topic in open-economy macroeconomics, especially for emerging markets facing external shocks, sanctions, or financial volatility. Many recent DSGE models—for example, Katagiri (2024)—analyze how systematic FX interventions can stabilize exchange rates and macroeconomic fluctuations using policy rules, typically implemented through a loss function and feedback mechanisms. In Katagiri’s model, the central bank sets an FXI rule responding to deviations in the exchange rate and manages the ratio of FX reserves to GDP, without explicitly tracking the direct financial mechanics of the central bank’s balance sheet.

However, existing models often abstract from the explicit balance sheet identity of the central bank. The direct transmission channel from changes in FX reserves to the monetary base — and, thus, to money supply and inflation — is rarely modeled in a fully structural, endogenous fashion. This gap is particularly relevant when analyzing economies under sanctions or with limited access to foreign capital, where the central bank’s ability to inject or withdraw liquidity via its FX assets is a key driver of domestic monetary conditions.

The core innovation of this research is the development of a tractable DSGE model that features the central bank’s balance sheet as an explicit, dynamic constraint within the model’s structural core. Unlike many existing approaches that treat monetary aggregates and FX reserves in a reduced-form or exogenous manner, this model directly connects the evolution of the monetary base to both foreign reserves and domestic assets, making the transmission channels of policy and external shocks fully endogenous.

To operationalize this framework, we construct a dynamic stochastic general equilibrium (DSGE) model in which the central bank’s balance sheet identity is featured at the core of the monetary block.

In what follows, we present the structure of the model, detailing the behavioral equations, the central bank’s balance sheet constraint, and the shock transmission mechanisms.

A5.1. Explicit central bank balance sheet modeling

In this section, we formally introduce the central bank’s balance sheet as a distinct and policy-relevant sector in the model. This explicit structure is critical for tracing how external shocks — especially those amplified by sanctions and foreign asset shortages — are transmitted to key domestic macroeconomic variables such as the monetary base, liquidity, inflation and output.

Balance sheet structure

  • Assets:
  • Foreign exchange reserves (R t ): primarily USD, EUR
  • Other assets ( D t ​ ): government securities, gold, domestic bank claims

Liabilities:

Currency in circulation

Required and excess reserves of banks with the central bank

Based on asset approach to the monetary base at time t:

MBt = Rt + Dt (26)

A5.2. Foreign reserves dynamics

Inspired by the work of Kozlovtceva et al. (2019) as well as balance sheet of a central bank, authors present new dynamic foreign reserves. The dynamic evolution of foreign reserves is given by:

Rt = Rt–1 – Ext × Pwt × IMt – ICBt (27)

where1:

Rt: Foreign reserves at time t

Rt–1: Previous reserves

Ext: exchange rate

Pwt: World price of imports

IMt: Import volume

ICBt: Central bank’s FX intervention

By substituting the foreign reserves dynamics equation (Equation 27) into the monetary base equation (Equation 26), we obtain the following relationship:

MBt = Rt–1 – Ext × Pwt × IMt – ICBt + Dt (28)

Substituting the foreign reserves dynamics equation into the monetary base identity makes it clear that an increase in the nominal exchange rate (such as a rise in the value of the US dollar or euro) leads to a depletion of central bank foreign reserves and, consequently, a reduction in both the nominal and real monetary base. This process not only tightens domestic liquidity but also elevates the cost of imports, putting direct upward pressure on the overall price level and accelerating inflation. Therefore, the dependence of the monetary base on foreign reserves means that any adverse realization of exchange rate disturbances, caused for example by exchange rate regime switching, will impact exchange rate volatility, especially under unstable external conditions such as sanctions or trade tariffs. This will lead to higher inflation and increase household budget constraints in vulnerable emerging market economies.

A5.3 Exchange rate determination

Exchange rate dynamics in the model are governed by a forward‑looking asset‑pricing condition. The nominal exchange rate is treated as an endogenous variable that reflects agents’ expectations of future exchange rate movements as well as unobserved external disturbances.

Formally, the exchange rate satisfies:

Ext=EtExt+1+ztex (29)

where Et denotes the rational expectations operator and ztex represents a latent exchange rate disturbance process.

The disturbance process ztex is exogenously specified and calibrated as described in Section A4. Importantly, this process is not identified from the exchange rate equation itself, nor does the model estimate structural exchange rate shocks from observed exchange rate data. Instead, the empirical properties of observed exchange rate series are used solely to discipline the persistence of the latent disturbance entering the model.

Through this equation, latent external forces — such as shifts in expectations, capital flow pressures, or geopolitical and policy‑related news — affect the exchange rate and are subsequently transmitted to the rest of the economy via the model’s equilibrium conditions.

The model does not explicitly include a trade openness parameter or a CES aggregation between domestic and imported consumption goods. Instead, the external sector operates through the exchange rate and foreign reserve dynamics, which affect the monetary base and domestic liquidity conditions. This specification allows the model to focus on the transmission of exchange rate disturbances through the central bank balance sheet rather than through trade composition channels.

A5.4 Monetary policy rule (Taylor rule)

Monetary policy is represented by a Taylor-type interest rate rule following Taylor (1993). The central bank adjusts the nominal interest rate on bonds, denoted by iit, in response to inflation and output fluctuations.

The policy rule is specified as:

iit=ρ⋅iit−1+(1−ρ)(∅π⋅πt+∅y⋅ytf) (30)

where iit denotes the nominal interest rate on bonds, πt is the inflation rate, and ytf represents the output gap. The parameter ρ captures interest rate smoothing, while ∅π and ∅y measure the responsiveness of monetary policy to inflation and output fluctuations.

The parameters of the Taylor rule are calibrated following the empirical estimates reported in Galí and Rabanal (2005), which are also employed in Furlanetto and Seneca (2012). In particular, the benchmark values are set to ρ = 0.69, = 1.35, and = 0.26. Since the main focus of this study is on the transmission of exchange rate shocks rather than the estimation of country‑specific monetary policy rules, the same policy parameters are applied across all BRICS economies to ensure comparability within the multi‑country framework. In the Bayesian estimation, these values are used as prior means, allowing the data to update the parameters around these benchmarks.

A 6: Equilibrium conditions

II=δ×k (31)

ii=β1+π (32)

r=Aαk1−αl1−α (33)

k=(rAαh1−α)1−1+α (34)

c=wψ1−h (35)

w=c⋅ψ1−h (36)

Ex=((1−τ)(h×w+r×k)+i×b+m(1+π)−T−c−i−m (37)

cm=π1+π (38)

yt=zkαl1−α (39)

MB=R+D (40)

EX=−ICBPw×IM (41)

MBt=R−Ex×Pw×IM−ICB (42)

Appendix B

Appendix B presents the tables and analytical results pertaining to the stationarity properties of the exchange rate time series. It reports and discusses the outcomes of multiple unit root and stationarity tests conducted specifically on the exchange rate data across three different time scenarios. These three scenarios represent distinct time periods that have been separately examined to ensure the data meet the necessary conditions for estimating AR models within each interval.

Table B1.

Results of panel unit root tests on exchange rate data (Scenario 1)

Method Statistic Prob. ** Crosssections
Levin, Lin & Chu t* −7.37514 0.0000 5
ADF – Fisher Chisquare 62.3335 0.0000 5
PP – Fisher Chisquare 51.6441 0.0000 5
Table B2.

Results of panel unit root tests on exchange rate data (Scenario 2)

Method Statistic Prob. ** Crosssections
Levin, Lin & Chu t* −7.06596 0.0000 5
ADF – Fisher Chisquare 59.6772 0.0000 5
PP – Fisher Chisquare 55.4842 0.0000 5
Table B3.

Results of panel unit root tests on exchange rate data (Scenario 3)

Method Statistic Prob. ** Crosssections
Levin, Lin & Chu t* −5.40573 0.0000 5
ADF – Fisher Chisquare 33.0920 0.0003 5
PP – Fisher Chisquare 24.5187 0.0063 5

Panel unit root tests were conducted for the exchange rate series. The Levin–Lin–Chu (LLC), Fisher‑type ADF, and Fisher‑type PP tests consistently reject the null hypothesis of a unit root at the 5% significance level (p < 0.05) across the three scenarios. These results provide strong evidence that the transformed exchange rate series used in the analysis are stationary.

Appendix C .Bayesian estimation for Brazil

This appendix presents the Bayesian estimation results of the DSGE model for Brazil. A set of observable macroeconomic variables was used to discipline the model and identify key structural parameters.

Bayesian estimation combines prior information about the parameters with the likelihood implied by the data. Priors are assigned based on standard values in the DSGE literature, and the posterior distributions are obtained using Markov Chain Monte Carlo (MCMC) methods.

Figure C1. 

Prior and posterior densities for Brazil’s parameters. Source: Compiled by the authors

Figure C1 depicts the prior (dashed) and posterior (solid) densities for the main structural parameters of the Brazilian DSGE model. The parameters include the standard deviation of the exchange rate shock (SE_e), the discount factor (β), the capital share (α), and the depreciation rate of capital (δ). In addition, the parameter (Q) denotes the Rotemberg price adjustment cost parameter (ψ), while (Q_ pi) and (Q_y) correspond to the Taylor-rule coefficients on inflation ∅π and ∅y output , respectively. The shift from priors to posteriors reflects the information contributed by the data through Bayesian updating. The parameter ρ captures interest rate smoothing in the monetary policy rule. The parameters ρ_s1, ρ_s2, and ρ_s3 represent the persistence coefficients of the latent exchange rate disturbance process calibrated for Scenario 1, Scenario 2, and Scenario 3, respectively. The shift from priors to posteriors reflects the information contributed by the data through Bayesian updating.

Figure C2. 

Logposterior and loglikelihood profiles of selected structural parameters for Brazil. Source: Compiled by the authors

Figure C2 displays the log‑posterior and log‑likelihood kernels of selected parameters. The blue line represents the log‑posterior density and the brown line the log‑likelihood component. The alignment of their peaks indicates well‑behaved parameter identification around the posterior mode (marked by the vertical line).

Figure C3. 

Posterior density plots for Brazil’s parameters. Source: Compiled by the authors

Figure C3 illustrates the Markov Chain Monte Carlo (MCMC) trace plots for the main parameters across three parallel chains (Interval, m2, and m3). The stable and well‑mixed patterns of the chains confirm convergence and efficient sampling of the posterior distribution.

Figure C4. 

MCMC trace plots for selected structural parameters in the Bayesian estimation for Brazil. Source: Compiled by the authors

The panels show the evolution of the simulated chains for the standard deviation of the exchange rate shock (SE_e), the discount factor (β), the capital share parameter (α), and the depreciation rate of capital (δ). In addition, Q denotes the Rotemberg price adjustment cost parameter (ψ), while Q_pi and Q_y correspond to the Taylor-rule coefficients on inflation (∅π) and output (∅π), respectively. The plots present the behavior of these parameters across iterations under three parallel chains (Interval, m2, and m3), illustrating the stability and convergence of the posterior simulations. The parameter ρ captures interest rate smoothing in the monetary policy rule. The parameters ρ_s1, ρ_s2, and ρ_s3 represent the persistence coefficients of the latent exchange rate disturbance process calibrated for Scenario 1, Scenario 2, and Scenario 3, respectively. The shift from priors to posteriors reflects the information contributed by the data through Bayesian updating.

Table C1.

Prior and posterior summary statistics of selected structural parameters for Brazil

Table C1 reports the prior and posterior summary statistics for the estimated structural parameters. It includes the prior means, posterior means, and 90% highest posterior density (HPD) intervals for β, α, and δ. The prior distributions follow standard beta or normal specifications commonly used in the DSGE literature.

Appendix D. Bayesian estimation for Russia

This subsection reports the Bayesian estimation results of the DSGE model for Russia. The estimation follows the same procedure described for Brazil. Prior distributions are assigned based on commonly used values in the DSGE literature, and the posterior distributions are obtained using Markov Chain Monte Carlo (MCMC) methods.

Figure D1. 

Prior and posterior densities for Russia’s parameters. Source: Compiled by the authors

Figure D1 presents the posterior distributions of selected structural parameters of the DSGE model estimated for Russia. The figure displays the posterior densities for the standard deviation of the exchange rate shock (SE_e), the discount factor (β), the capital share parameter (α) and the capital depreciation rate (δ). In addition, the parameter Q represents the Rotemberg price adjustment cost parameter (ψ), while Q_pi and Q_y denote the Taylor-rule coefficients on inflation (∅π) and output (∅π), respectively. The distributions illustrate the uncertainty surrounding the estimated parameters and reflect the information contained in the data combined with the specified prior distributions. The parameter ρ captures interest rate smoothing in the monetary policy rule. The parameters ρ_s1, ρ_s2, and ρ_s3 represent the persistence coefficients of the latent exchange rate disturbance process calibrated for Scenario 1, Scenario 2, and Scenario 3, respectively. The shift from priors to posteriors reflects the information contributed by the data through Bayesian updating.

Figure D2. 

Logposterior and loglikelihood profiles of selected structural parameters for Russia. Source: Compiled by the authors

Figure D2 reports the log‑posterior and log‑likelihood kernel functions for selected structural parameters. The vertical line indicates the posterior mode. The close correspondence between the two curves suggests that the likelihood plays an important role in shaping the posterior distributions and that the parameters are well identified.

Figure D3. 

Posterior density plots for Russia’s parameters. Source: Compiled by the authors

Figure D4. 

MCMC trace plots for selected structural parameters in the Bayesian estimation for Russia. Source: Compiled by the authors

Figure D3 presents the estimated posterior densities of the selected structural parameters. The unimodal and well‑behaved distributions indicate stable estimation results and support the identification of the model parameters.

Table D1.

Prior and posterior summary statistics of selected structural parameters for Russia

Figure D4 displays the Markov Chain Monte Carlo (MCMC) trace plots for the estimated parameters. The chains fluctuate around stable means and show no visible trends, indicating satisfactory convergence and good mixing of the MCMC simulations.

Table D1 reports the prior means, posterior means and 90 percent highest posterior density (HPD) intervals for the estimated parameters. The data provide meaningful information for parameter estimation, as reflected in the differences between prior and posterior moments and the relatively narrow HPD intervals.

Appendix E. Bayesian estimation for India

This section presents the Bayesian estimation results of the DSGE model for India, following the methodology described for Brazil and Russia. Prior distributions are assigned to structural parameters based on standard DSGE literature values, and the posterior distributions are obtained via Markov Chain Monte Carlo (MCMC) sampling.

Figure E1. 

Prior and posterior densities for India’s parameters. Source: Compiled by the authors

Figure E1 shows the posterior distributions of selected structural parameters for India, including the standard deviation of the exchange rate shock (SE_e), the discount factor (β), the capital share (α) and the depreciation rate of capital (δ). In addition, the parameter Q represents the Rotemberg price adjustment cost parameter (ψ), while Q_pi and Q_y denote the Taylor-rule coefficients on inflation (∅π) and output (∅y), respectively. The posterior densities are well behaved and centered around economically plausible values, indicating that the data are informative about the parameters and that the Bayesian estimation provides stable results for the Indian economy. The parameter ρ captures interest rate smoothing in the monetary policy rule. The parameters ρ_s1, ρ_s2, and ρ_s3 represent the persistence coefficients of the latent exchange rate disturbance process calibrated for Scenario 1, Scenario 2, and Scenario 3, respectively. The shift from priors to posteriors reflects the information contributed by the data through Bayesian updating.

Figure E2. 

Logposterior and loglikelihood profiles of selected structural parameters for India. Source: Compiled by the authors

Figure E3. 

Posterior density plots for India’s parameters. Source: Compiled by the authors

Figure E4. 

MCMC trace plots for selected structural parameters in the Bayesian estimation for India. Source: Compiled by the authors

Figure E2 illustrates the log‑posterior and log‑likelihood kernel profiles for the selected parameters. The vertical lines indicate the posterior modes. The proximity of the log‑posterior and log‑likelihood peaks suggests robust parameter identification.

Table E1.

Prior and posterior summary statistics of selected structural parameters for India

Figure E3 presents the posterior density estimates, with dashed vertical lines marking posterior modes. The well-shaped unimodal distributions confirm stable estimation and good parameter identification.

Figure E4 displays MCMC trace plots for the main parameters across multiple chains. The absence of trends and stable fluctuations demonstrate satisfactory convergence and effective mixing.

Table E1 summarizes the prior means, posterior means, and 90% highest posterior density (HPD) intervals for parameters. The close alignment between posterior estimates and economically plausible ranges shows the effectiveness of the Bayesian updating.

Appendix F. Bayesian estimation for China

This section presents the Bayesian estimation results of the DSGE model calibrated for China, focusing on a set of key structural parameters governing households’ intertemporal behavior, production structure, capital dynamics, and the volatility of the exchange rate disturbance. Prior distributions are selected based on the DSGE literature and empirical studies on emerging economies. Posterior distributions are obtained using Markov Chain Monte Carlo (MCMC) simulations.

Figure F1. 

Prior and posterior densities for China’s parameters. Source: Compiled by the authors

Figure F2. 

Logposterior and loglikelihood profiles of selected structural parameters for China. Source: Compiled by the authors

Figure F1 displays the prior and posterior densities for the main structural parameters. The posterior densities are concentrated around well-defined values, indicating robust parameter identification and stable estimation results.

Figure F3. 

Posterior density plots for China’s parameters. Source: Compiled by the authors

Figure F4. 

MCMC trace plots for selected structural parameters in the Bayesian estimation for China. Source: Compiled by the authors

Figure F2 compares the log‑posterior and log‑likelihood kernels near the posterior modes, indicated by vertical lines. The close alignment of the curves suggests well‑behaved estimation and reliable parameter identification.

Table F1.

Prior and posterior summary statistics of selected structural parameters for China

Figure F3 presents unimodal posterior density estimates with dashed vertical lines marking posterior means. The distributions confirm stable estimates and economic plausibility.

Figure F4 shows the MCMC trace plots across chains. The stable trajectories and absence of trends or drifts indicate convergence and good mixing, supporting the credibility of the posterior estimates.

Table F1 summarizes the prior means, posterior means, and 90% highest posterior density (HPD) intervals for the parameters. The differences between prior and posterior estimates confirm the informativeness of the data and the refinement of prior assumptions.

Appendix G. Bayesian estimation for South Africa

This section reports the Bayesian estimation results of the DSGE model calibrated for South Africa. The estimation follows the same procedure applied to the other BRICS economies in order to ensure cross‑country consistency and comparability of the structural parameters.

Figure G1 presents the prior and posterior densities of the main structural parameters estimated for South Africa. The figure illustrates the distributions associated with key parameters governing intertemporal preferences, production structure, capital dynamics, and the volatility of the exchange rate disturbance. The posterior densities are concentrated around well-defined values, indicating reasonable parameter identification.

Figure G1. 

Prior and posterior densities for South Africa’s parameters. Source: Compiled by the authors

Figure G2 compares the log‑posterior and log‑likelihood kernel around the posterior mode for the selected parameters. The vertical lines denote posterior modes. The close alignment between the curves suggests that the likelihood information is consistent with the posterior distribution and supports the stability of the estimated parameters.

Figure G2. 

Logposterior and loglikelihood profiles of selected structural parameters for South Africa. Source: Compiled by the authors

Figure G3 presents the posterior density estimates for the structural parameters, with dashed vertical lines indicating posterior means. The unimodal and concentrated distributions suggest well‑identified parameters and economically plausible estimates.

Figure G3. 

Posterior density plots for South Africa’s parameters. Source: Compiled by the authors

Figure G4 displays the MCMC trace plots for the estimated parameters across multiple chains. The trajectories fluctuate around stable values with no visible trends or structural breaks, indicating satisfactory convergence and good mixing of the Markov chains.

Figure G4. 

MCMC trace plots for selected structural parameters in the Bayesian estimation for South Africa. Source: Compiled by the authors

Table G1 reports the prior means, posterior modes, and standard deviations for the estimated parameters. The posterior estimates remain within economically plausible ranges while differing from their priors, suggesting that the data are informative and that the Bayesian estimation refines the prior assumptions.

Table G1.

Prior and posterior summary statistics of selected structural parameters for South Africa

Appendix H: Sensitivity analysis of impulse response functions

To assess the robustness of the results, a sensitivity analysis of the impulse response functions is conducted for Brazil. The persistence parameter of the Taylor rule (ρ) and the monetary policy response coefficients to inflation (∅π) and output (∅y) are varied around their baseline values to examine whether the main dynamic responses of the model remain qualitatively unchanged.

Figure H1. 

IRF sensitivity to the persistence parameter (ρₛ₁) in Brazil. Source: Compiled by the authors

Figure H2. 

IRF sensitivity to the monetary policy output response parameter (∅y) in Brazil. Source: Compiled by the authors

Scenario 1 is selected as the representative case for the sensitivity analysis because it reflects the baseline specification of the model and provides a clear benchmark for evaluating the robustness of the impulse response results.

Figure H3. 

IRF sensitivity to the monetary policy inflation response parameter (∅π) in Brazil. Source: Compiled by the authors

Figure H1 illustrates the sensitivity of the impulse response functions to alternative values of the persistence parameter (ρ₁) for Brazil under Scenario 1. The baseline value corresponds to the posterior mean (ρ₁ = 0.484), while the alternative values (0.47 and 0.50) lie within the estimated HPD interval. The results indicate that the qualitative dynamics of the model remain broadly unchanged across these parameter variations.

Figure H4. 

IRF sensitivity to the persistence parameter (ρₛ₁) in Russia. Source: Compiled by the authors

Figure H5. 

IRF sensitivity to the monetary policy output response parameter (∅y) in Russia. Source: Compiled by the authors

The baseline value is the posterior mean (∅y=0.26), while the alternative values (0.10 and 0.42) lie within the estimated HPD interval. The impulse responses remain qualitatively similar, indicating robustness of the results to variations in ∅y.

The baseline value corresponds to the posterior mean (∅π=1.34), while the alternative values (1.18 and 1. 50) lie within the estimated HPD interval. The impulse responses remain qualitatively similar, indicating that the results are robust to variations in the monetary policy response to inflation.

Figure H6. 

IRF sensitivity to the monetary policy inflation response parameter (∅π) in Russia. Source: Compiled by the authors

The baseline value corresponds to the posterior mean estimate (ρ₁=0.485), while the alternative values (0.47 and 0.50) lie within the estimated HPD interval. The impulse responses remain qualitatively similar, indicating that the model results are robust to variations in the persistence parameter.

Figure H7. 

IRF sensitivity to the persistence parameter (ρₛ₁) in India. Source: Compiled by the authors

The baseline value corresponds to the posterior mean estimate (∅y=0. 25), while the alternative values (0.09 and 0.41) are selected from the estimated HPD interval. The impulse response functions exhibit only minor quantitative differences, suggesting that the results are robust to changes in the monetary policy response to output in Russia.

Figure H8. 

IRF sensitivity to the monetary policy output response parameter (∅y) in India. Source: Compiled by the authors

The baseline value corresponds to the posterior mean estimate (∅π =1. 35), while the alternative values (1.19 and 1.51) are drawn from the estimated HPD interval. The impulse responses display only slight quantitative variations, confirming that the model’s results remain robust to changes in the monetary policy response to inflation in Russia.

Figure H9. 

IRF sensitivity to the monetary policy inflation response parameter (∅π) in India. Source: Compiled by the authors

The baseline value corresponds to the posterior mean estimate (ρs1=0.484), while the alternative values (0.472 and 0.50) are selected from the estimated HPD interval. The impulse response functions show only minor quantitative changes, indicating that the model’s results for India are robust to variations in the persistence parameter.

Figure H10. 

IRF sensitivity to the persistence parameter (ρ₁) in China. Source: Compiled by the authors

The baseline value corresponds to the posterior mean estimate (∅y=0.25), while the alternative values (0.09 and 0.40) are chosen from the estimated HPD interval. The impulse responses display only limited quantitative differences, suggesting that the model’s results for India are robust to variations in the monetary policy response to output.

Figure H11. 

IRF sensitivity to the monetary policy output response parameter (∅y) in China. Source: Compiled by the authors

The baseline value corresponds to the posterior mean estimate (∅π=1.35), while the alternative values (1.18 and 1.51) are selected from the estimated HPD interval. The impulse response functions exhibit only slight quantitative changes, indicating that the model’s results for India are robust to variations in the monetary policy response to inflation.

The baseline value corresponds to the posterior mean estimate (ρ₁ = 0.484), while the alternative values (0.47 and 0.50) are selected from the estimated HPD interval. The impulse response functions exhibit only minor quantitative changes, indicating that the model’s results for China are robust to variations in the persistence parameter.

Figure H12. 

IRF sensitivity to the monetary policy inflation response parameter (∅π) in China. Source: Compiled by the authors

The baseline value corresponds to the posterior mean estimate of the output response coefficient (∅y = 0.25), while the alternative values (0.09 and 0.40) are selected from the estimated HPD interval. The impulse response functions display only limited quantitative differences across these parameter values, and the overall qualitative dynamics remain stable. This indicates that the model’s results for China are robust to plausible variations in the monetary policy response to output.

Figure H13. 

IRF sensitivity to the persistence parameter (ρₛ₁) in South Africa. Source: Compiled by the authors

Figure H14. 

IRF sensitivity to the monetary policy output response parameter (∅y) in South Africa. Source: Compiled by the authors

Figure H15. 

IRF sensitivity to the monetary policy inflation response parameter (∅π) in South Africa. Source: Compiled by the authors

The baseline value corresponds to the posterior mean estimate of the inflation response coefficient (∅π=1.35). The alternative values (1.18 and 1.51) are selected from the estimated HPD interval. The impulse response functions show only moderate quantitative differences across these values, while the overall dynamics and qualitative patterns remain unchanged. This suggests that the model’s results for China are robust to reasonable variations in the monetary policy response to inflation.

The baseline value corresponds to the posterior mean estimate (ρ₁ = 0.484), while the alternative values (0.47 and 0.50) are selected from the estimated HPD interval. The impulse response functions show only minor quantitative differences across these values, while the qualitative dynamics remain unchanged, indicating that the results for South Africa are robust to variations in the persistence parameter.

The baseline value corresponds to the posterior mean (∅y=0.26), while the alternative values (0.09 and 0.42) are taken from the HPD interval. The impulse responses change only slightly, indicating robustness of the results.

Table K1.

Bai–Perron multiple structural break test for the exchange rate (Brazil)

The baseline value corresponds to the posterior mean (∅π=1. 35), while the alternative values (1.19 and 1.53) are taken from the HPD interval. The impulse responses show only minor quantitative differences, indicating robustness of the results.

Appendix K. Structural break test for the exchange rate

The Bai–Perron (2003) multiple structural break test is employed to detect possible regime shifts in exchange rate dynamics across BRICS economies.

Table K2.

Bai–Perron multiple structural break test for the log exchange rate (Russia)

To examine whether the exchange rate series exhibits structural instability, the Bai–Perron multiple breakpoint test was applied to the logarithm of the exchange rate for Brazil over the sample period 2000–2023. The test allows for the detection of multiple structural breaks in the mean of the series.

Table K3.

Bai–Perron multiple structural break test for the log exchange rate (India)

Table K4.

Bai–Perron multiple structural break test for the log exchange rate (China)

Using a trimming parameter of 0.15 and allowing for a maximum of two breaks, the sequential F‑statistic indicates the presence of two statistically significant structural breaks at the 5% significance level. The estimated break dates are 2015 and 2020. These breakpoints suggest that the dynamics of the exchange rate changed during the sample period, with the latter breakpoint coinciding with the global COVID‑19 shock.

Table K5.

Bai–Perron multiple structural break test for the log exchange rate (South Africa)

This test is reported as a robustness check. In the main DSGE framework, exchange rate disturbances are modeled as latent shocks, allowing the model to capture periods of heightened volatility without explicitly introducing regime‑switching mechanisms.

The Bai–Perron multiple breakpoint test was applied to the logarithm of the exchange rate for Russia over 2000–2023. The test identifies two statistically significant structural breaks (2015 and 2020), corresponding respectively to the financial sanctions period and the global COVID‑19 shock.

The Bai–Perron multiple breakpoint test applied to the logarithm of the exchange rate for India over 2000–2023 identifies two statistically significant structural breaks at the 5% level, occurring in 2003 and 2015.

The Bai–Perron multiple breakpoint test applied to the logarithm of the exchange rate for China over the period 2000–2023 identifies one statistically significant structural break in 2004 at the 5% significance level.

The Bai–Perron multiple breakpoint test applied to the logarithm of the exchange rate for South Africa over 2000–2023 identifies two statistically significant structural breaks at the 5% level, occurring in 2003 and 2015.

Appendix L. Model validation

To evaluate the empirical performance of the model, a simple moment‑matching exercise is conducted. The unconditional moments of real GDP obtained from the DSGE model simulations are compared with those computed from the empirical data for the BRICS economies. Specifically, the mean and standard deviation of HP‑filtered log‑transformed GDP are calculated for both the data and the simulated series under the three alternative scenarios (S1–S3). Table L1 reports the comparison between empirical moments and model‑implied moments. The results indicate that the model broadly captures the average behavior of output across countries under different scenarios.

Table L1.

Empirical vs. modelsimulated GDP moments for BRICS economies

Country Variable Data Mean Model Mean (S1) Model Mean (S2) Model Mean (S3) Data SD Model *SD (S1) Model SD (S2) Model SD (S3)
Brazil GDP 0.2793 0.2005 0.2005 0.1966 0.0059 0.000 0.000 0.000
Russia GDP 0.2773 0.2188 0.2226 0.2046 0.0069 0.000 0.000 0.000
India GDP 0.2799 0.2111 1.4356 0.2047 0.0058 0.000 0.000 0.000
China GDP 0.2937 0.2052 0.2025 0.2025 0.0074 0.000 0.000 0.000
South Africa GDP 0.2644 0.2005 0.2005 0.1966 0.0072 0.000 0.000 0.000

1 imports financed through central bank foreign assets
login to comment