The intricate relationship between oil prices and exchange rates is a central issue in economic research. The interconnectedness of these markets has significant implications for systemic risk, as evidenced by the amplified impact of the 2007-2008 subprime crisis. This study builds upon existing research, which demonstrates a fragmented understanding of the dynamic interplay between these crucial variables. Exchange rates are critical indicators of economic health, reacting swiftly to economic and political shifts. Crude oil, a vital global commodity, significantly influences macroeconomic and monetary factors, with price volatility impacting economic downturns, trade balances, and inflation. The study aims to analyze the general and directional interconnectedness between foreign exchange and crude oil markets using two approaches: Diebold and Yilmaz's generalized connectedness technique and the time-varying parameter vector autoregression (TVP-VAR) extended joint connectedness approach. Three theoretical channels – the trade channel, the wealth effect channel, and the portfolio reallocation channel – are considered to explain the impact of oil price fluctuations on exchange rates. Existing literature reveals inconsistencies in findings regarding the relationship, with varying results depending on factors such as the period studied, countries involved, dataset, and methodologies employed. The study seeks to provide a more comprehensive understanding of this complex relationship.

A significant body of research examines the interconnectedness of oil prices and exchange rates. Studies by Reboredo (2012), Noman et al. (2023), and Chang (2020) highlight the negative impacts of rising oil prices on economies. Conversely, the volatility of crude oil prices, exacerbated by geopolitical factors and health crises, significantly impacts other financial variables, including exchange rates, especially for oil-importing countries (Uche et al., 2022a; Ali et al., 2022). Previous studies using connectedness approaches have yielded varied results, depending on the period, countries analyzed, datasets, and methodologies. For example, Wen and Wang (2020) investigated volatility connectedness among major currencies, while Singh et al. (2018) examined the dynamic links between crude oil prices and exchange rate volatility. Malik and Umar (2019) explored the strengthened correlation between oil price shocks and exchange rate movements since the global financial crisis. Other studies focused on the impacts of trade policy uncertainty (Huynh et al., 2020) and dynamic interconnectedness among G7 currencies (Wan and He, 2021; Wang et al., 2024; Gohar et al., 2022c). Shang and Hamori (2021) highlighted WTI crude oil's effect on exchange rates, especially in oil-importing countries. Nekhili et al. (2021) investigated time-frequency interactions between currencies and commodities, emphasizing oil's role in currency market volatility. Adekoya and Oliyide (2020) and Asadi et al. (2022) examined the correlation between the US dollar, oil, and other markets during the COVID-19 pandemic. Ahmad et al. (2020) and Alam et al. (2019) explored causal relationships between oil prices and exchange rates using high-frequency data. Fasanya et al. (2021) investigated interactions amid unpredictable US economic policies, while Albulescu and Ajmi (2021) and Uche et al. (2022b) and Hashmi and Chang (2023) examined oil's impacts across time and frequency domains. This review demonstrates a lack of consistent findings and highlights the need for a more advanced methodology.

This study employs two connectedness approaches: Diebold and Yilmaz's (2012, 2014) generalized connectedness method and a time-varying parameter vector autoregression (TVP-VAR) extended joint connectedness approach developed by Balcilar et al. (2021). The TVP-VAR approach, refined using insights from Lastrapes and Wiesen (2021) and Antonakakis et al. (2020), offers advantages in handling the high volatility of the variables involved. It accurately captures dynamic changes in coefficients over time, is less sensitive to outliers, and doesn't require arbitrary selection of rolling windows. The methodology includes a theoretically derived normalization approach, improving upon the traditional DY method. The choice of the TVP-VAR model is justified by the substantial volatility and intricate dynamic patterns in the data. The model's adaptability and ability to provide precise interpretations of shifting interrelations between variables are crucial for the study's depth and rigor. The dataset includes exchange rate data (against the US dollar) from five oil-exporting countries (Canada, Mexico, Russia, Brazil, and Norway) and six oil-importing countries (UK, South Korea, Japan, China, Eurozone, and India). Monthly exchange rates were calculated by averaging daily rates from the Federal Reserve Economic Data (FRED), and oil prices were tracked using the S&P GSCI Crude Oil Index from DataStream. A log-return transformation was used to address the unit root problem. Descriptive statistics (mean, variance, skewness, kurtosis, Jarque-Bera test, and Fisher and Gallagher (2012) weighted portmanteau test) were calculated. Unit root tests (Stock et al., 1996) confirmed the stationarity of the return series. The generalized connectedness method was applied to calculate connectedness, and the TVP-VAR extended joint connectedness approach was used to assess time-varying connectedness between exchange rates and oil prices. The methodology includes calculating the generalized forecast error variance decomposition (GFEVD), net pairwise directional spillovers, net total directional connectedness, and the total connectedness index (TCI). For robustness, an asymmetric TVP-VAR model was also estimated.

The study's key findings are presented in Table 2, which shows average connectedness using the TVP-VAR extended joint connectedness method. Crude oil significantly influences the currency rates of Norway and Russia, while Norway and Canada's exchange rate fluctuations significantly drive oil price volatility. The 'From' column indicates crude oil's substantial contribution to overall network variance (28.3%), while the 'To' row shows its cumulative effect on exchange rates (27.41%). The 'Net' row identifies the Canadian dollar as the most significant net shock sender and the Russian ruble and Japanese yen as the main net shock recipients. The total connectedness index (TCI) averages 44.51%, indicating substantial interconnectedness. A comparison with the Diebold and Yilmaz (2012, 2014) generalized connectedness approach reveals differences, particularly regarding the role of the Euro and Norwegian Krone as shock transmitters/receivers. Figure 1 illustrates dynamic total connectedness (TCI), showing high values during the 2007-2008 financial crisis, the European sovereign debt crisis, Brexit, and the COVID-19 pandemic. Figure 2 depicts dynamic net total directional connectedness, highlighting the fluctuating roles of variables as net transmitters/receivers. Figure 3 shows dynamic net pairwise directional connectedness between exchange rates and crude oil, revealing a time-varying bidirectional relationship. The asymmetric TVP-VAR model (Appendix 2) yielded similar results to the overall return analysis.

The findings demonstrate a strong, time-varying relationship between oil prices and exchange rates, particularly in oil-dependent countries. The bidirectional causality challenges previous research that mainly suggested a unidirectional influence from oil markets to currency markets. The heightened connectedness during periods of economic stress highlights the importance of considering these dynamics in managing systemic risk. The TVP-VAR extended joint connectedness method provides a superior analytical framework compared to the generalized connectedness approach, offering more nuanced insights into the dynamic interrelationships. The differences in results between the two methods underscore the importance of selecting appropriate methodologies based on the data's characteristics. The study's findings have significant implications for investment strategies, portfolio management, and policymaking. The observed interdependence suggests that simple diversification across these markets may not adequately mitigate risk. Understanding the dynamic roles of assets as net transmitters or receivers can inform tailored risk management strategies and investment decisions.

This study utilizes a time-varying parameter VAR extended joint connectedness approach to analyze the relationship between oil prices and exchange rates, offering methodological advancements over previous research. The findings reveal a significant and time-varying interconnectedness between these markets, with stronger linkages in oil-exporting countries and heightened connections during crises. The TVP-VAR method provides a superior analytical framework, highlighting the importance of choosing suitable methodologies for analyzing complex market dynamics. The results have significant implications for investors, policymakers, and market participants in managing risk and making informed decisions. Future research could expand the dataset to include a broader range of countries and explore the impact of renewable energy sources on the studied relationship.

The study primarily focuses on major oil-exporting and -importing countries, potentially limiting the generalizability of the findings to other nations. While the total connectedness index is a key metric, exploring additional indices or methodologies could provide further insights. The dynamic nature of global economic interactions requires continuous monitoring and research.