The study's central theme revolves around the impact of the COVID-19 pandemic on the interconnectedness of BRICS nations' interest rate term structures. A core principle in finance emphasizes portfolio diversification to minimize risk by investing in assets with low correlations across sectors and countries. BRICS nations, with their diverse economies, offer potential diversification benefits for international portfolios. However, globalization increases market interlinkages and interdependence, potentially diminishing diversification advantages, especially during economic crises. Previous research on BRICS interdependencies during past crises (like the Global Financial Crisis and the European Sovereign Debt Crisis) has yielded mixed results regarding the degree of dependence and spillover effects. While some studies show a dominant role for China and Russia in cross-market spillovers, others highlight the different market structures and relative independence of BRICS economies. The COVID-19 pandemic, an unprecedented global crisis, presents a unique opportunity to examine the impact and spillovers amongst BRICS nations. Its disruptive effect on economies worldwide, particularly in emerging markets like those in the BRICS, necessitates an analysis of their interdependencies. The study is motivated by the need to understand BRICS interconnectedness, particularly in a crisis scenario, and to provide investors with insights into asset risk profiles. The BRICS were chosen for their rapid economic growth, potential for diversification benefits in their stock markets, and the large share of the world's GDP and population that they represent. Existing research focusing on the COVID-19 crisis impact on emerging markets primarily examines stock or foreign exchange markets, with limited attention to sovereign debt markets and their term structures. The term structure of interest rates (TSIR), however, is a reliable indicator of economic downturns. Decomposing the TSIR into level, slope, and curvature components allows for a more nuanced analysis of short-, medium-, and long-term interest rate dynamics. The current study aims to bridge this gap and analyze the connectedness between BRICS yield curves and the uncertainty surrounding the COVID-19 crisis, using the coronavirus media coverage index (MCI), and by decomposing the yield curve into its factors.

The literature review is structured into three sections: studies on the COVID-19 crisis's impact on BRICS economies; studies on connectedness within BRICS countries; and studies on the connectedness of TSIRs. Regarding the COVID-19 impact, several studies highlight negative effects on emerging markets, particularly concerning bond market liquidity and stock market volatility. Studies using the coronavirus MCI index illustrate the role of media coverage in shaping market sentiment. The second branch of the literature review covers research on BRICS connectedness. This includes studies focusing on interdependencies between oil shocks and government bonds, volatility spillovers in BRICS and global markets during periods of economic crisis and the connectedness between BRICS markets and other assets like Bitcoin and precious metals. Finally, regarding TSIR connectedness, many studies explore bond market integration through TSIRs, often decomposing them into level, slope, and curvature components. Some studies examine the connectedness between sovereign bonds and oil prices, highlighting the maturity-dependent nature of bond market integration and increased co-movements during crises. The research also highlights papers analyzing interrelationships between yield curves in different geographic areas.

The study employs a two-step methodology. First, it decomposes each country's yield curve into level, slope, and curvature factors using the dynamic Nelson-Siegel model (Diebold & Li, 2006). This model provides a parsimonious representation of the yield curve, with the level factor representing long-term interest rates, the slope factor reflecting short-term rates, and curvature capturing the medium-term dynamics. The dynamic version of this model uses time-varying parameters to capture shifts in the yield curve over time. Second, the study uses Antonakakis and Gabauer's (2017) time-varying parameter vector autoregression (TVP-VAR) methodology within a dynamic connectedness framework (Diebold and Yilmaz, 2009, 2012, 2014) to model the factor dynamics of the BRICS yield curves and their connectedness with the coronavirus MCI. The TVP-VAR approach is chosen for its ability to handle short data series and its advantage over traditional rolling window approaches by allowing variances to vary via a stochastic volatility Kalman filter estimation with forgetting factors. This allows the model to effectively capture changes in the underlying structure of the data. The methodology involves estimating a TVP-VAR model for the time series of the yield curve components of the five BRICS countries, along with the coronavirus MCI. The model then allows for the calculation of generalized forecast error variance decompositions (GFEVDs) to quantify the degree to which shocks in one variable are transmitted to others. From the GFEVDs, several connectedness measures are constructed, including the total connectedness index, directional connectedness measures ('to' and 'from'), and net connectedness measures. These measures allow for a comprehensive analysis of the direction and magnitude of spillover effects between the different yield curve components and the coronavirus MCI.

The study's key findings are presented in tables and figures. Table 1 presents descriptive statistics and unit root tests for the yield curve components, revealing that all series are stationary after differencing. Figure 1 shows the dynamic total connectedness of the BRICS term structure using the coronavirus MCI. Total connectedness fluctuates over time, peaking during the first wave of the pandemic and, to a lesser extent, during the later waves. The level and curvature components show more persistent connectedness compared to the slope component. Figure 2 examines the dynamic contribution 'to' the system, showing that Brazil, South Africa, and Russia were dominant transmitters, particularly during the first wave. Figure 3 displays the dynamic contribution 'from' the system, indicating greater volatility in the level component, with China showing the lowest connectedness 'from' the system. Figure 4 illustrates the net dynamic total connectedness, highlighting the largest differences between BRICS countries before and during the first wave. Brazil and South Africa show net transmitting positions, while China and India are net receivers. The differences in net connectedness reduce significantly after the first wave, but increase again towards the end of the sample period, coinciding with the later waves of the pandemic and the Russian invasion of Ukraine.

The findings address the research question by demonstrating the time-varying nature of connectedness within the BRICS bond markets during the COVID-19 pandemic. The increased connectedness during the first wave, and to a lesser extent the later waves, is consistent with the hypothesis that diversification benefits diminish during periods of global economic distress. The identification of Brazil and South Africa as net transmitters of shocks, and China and India as net receivers, provides valuable insights into the transmission channels within the BRICS network. The results suggest that shocks originating in Brazil and South Africa have a significant impact on the other BRICS economies. The reduced connectedness following the first wave reflects potentially increased resilience of BRICS countries as the crisis evolved, but it's vital to consider the limitations of this interpretation. The increased differences in net connectedness during the final wave underline the impact of specific geopolitical events, particularly the impact of Russia's invasion of Ukraine. This warrants further investigation. The findings highlight the importance of considering different yield curve components when assessing market interconnectedness.

This study offers novel insights into the connectedness of BRICS term structure components during the COVID-19 pandemic. The time-varying nature of the total connectedness and the identification of net transmitters (Brazil, South Africa) and receivers (China, India) are significant contributions. The study emphasizes the importance of the coronavirus MCI in quantifying the uncertainty impact. The limitations of the study are discussed, along with suggestions for future research, including a more detailed examination of geopolitical factors, using high-frequency data and exploring the implications for policy and investment.

The study's limitations include the use of daily data potentially neglecting higher frequency dynamics, the reliance on the coronavirus MCI as a proxy for uncertainty, and the focus on the BRICS without direct comparisons to other emerging or developed markets. The limited sample period related to the availability of data from the Russian market after February 28, 2022, and the lack of analysis of specific policy responses may also restrict the interpretations.