**1. Introduction**

South Africa is a critical player in the global commodities. It is the largest producer (exporter) of global platinum [1]. In addition, South Africa is the eighth global producer of gold and, notably, the third largest in terms of global gold reserves, after Australia and Russia. Furthermore, the rand is one of the most traded emerging market currencies and, globally, among the top 20 [2, 3]. Understanding how it is integrated with commodities is crucial, considering the sovereign risk premium. Subsequently, this chapter studies the network connectedness of the dollar-denominated commodity (gold and platinum) prices, South Africa's rand exchange rate, and sovereign risk premium proxied by credit default swaps spread (CDS).

The connectedness of financial variables has increasingly become more critical for financial market participants and policymakers alike, especially after the 2007**–**2009 global financial crisis (GFC). Antonakakis et al. [4] echo this view and highlight that investigating the propagation of the financial crisis into the economy has been at the epicentre of academic research in recent years, especially in the aftermath of the GFC. Therefore, to provide insights into the financial market participants (and policymakers alike) interested in investing in South Africa, it is critical to understand the dynamic connectedness of commodities and currency accounting for the country's risk premium, as proposed here. Mensi et al. [5] argue that the growing cross-market information transmission in economic activities has amplified the contagion spillover and vulnerability of markets to external shocks. This provides context as to why studying the dynamic connectedness of financial networks is crucial.

To investigate how integrated our network is, we apply a Time-Varying Parameter–Vector Autoregressive (TVP-VAR) Connectedness Approach advanced by Antonakakis et al. [4] which was initially proposed by Diebold and Yilmaz [6]. This approach allows us to capture better the connectedness dynamics within the network of our selected financial variables. Several empirical studies have assessed the connectedness of commodities and exchange rate markets. A recent study by Mensi et al. [5], which extends from Antonakakis and Kizys [7], investigates volatility spillover, connectedness, and quantile dependence between precious metals and developed market currencies. Generally, their results reveal dynamic spillovers across precious metals and currencies and their respective connection with significant world events. However, their study does not include South Africa.

The most relevant recent study by Sayed and Charteris [3] investigates whether metals, grains, and energy commodities influence the South African rand and its volatility. They find a high degree of interconnectedness between the rand and commodities. Nevertheless, while their study is a good reference for our study, it does not account for the sovereign risk premium, which can provide further details in analysing the connectedness of the rand and commodities. In addition, our study is framed within an asymmetric time-varying setting, while their study uses a Dynamic Conditional Correlation-Generalised Autoregressive Conditional Heteroskedastic setting. Therefore, our study's contribution reflects both the time-varying connectedness approach and the inclusion of the CDS in the commodity-currency connectedness analysis in the case of South Africa.

The CDS spread can be used as a proxy for the sovereign credit risk profile, inferring those large spikes in CDS spread imply increased riskiness of sovereign assets, as investors demand high-risk premia to compensate for increased default risk [8]. Numerous studies have considered the connectedness of financial assets and CDS. Studies on the determinants of CDS have advanced two compelling reasons: firstly, financial integration and demand growth for sovereign CDS and secondly, investors' perception of higher credit risk and variability in emerging rather than developed markets [9]. Using Principal Component Analysis, de Boyri and Pavlova [9] find that global financial market factors are important drivers of sovereign CDS variability for the BRICS and MIST countries. Simonyan and Bayraktar [10] study the asymmetric dynamics in sovereign CDS pricing for a group of 11 emerging markets, including South Africa, and find some specific idiosyncratic and external factors (VIX and oil price) crucially affect CDS asymmetrically. Therefore, the CDS spread is crucial for financial market participants (and policymakers alike); so, we include it in our commodity-currency connectedness study for South Africa.

#### *Asymmetric TVP-VAR Connectedness Approach: The Case of South Africa DOI: http://dx.doi.org/10.5772/intechopen.107248*

We focus on South Africa exclusively because of its global position within the African continent and its importance in the global commodities markets. In addition, we concur with Sayed and Charteris [3] that a single country study can provide great insights not identifiable in multiple country studies. Generally, we find a higher degree of interconnectedness within the network throughout the sample period. In addition, dynamic total connectedness is visibly responsive to some critical economic events, notably between 2014, 2016 and 2018.

The remainder of this chapter is organised in the following way: Section 2 provides a brief literature review on interconnectedness between financial variables, and network connectedness. Section 3 provides brief information about the data and details about the asymmetric TVP-VAR econometric model. Section 4 presents empirical findings, and Section 5 provides conclusions, as well as policy recommendations.
