Table 1.
Overview of the selected previous studies which analyze the return and volatility spillovers.
| Authors | Data and Period | Methods | Findings |
|---|---|---|---|
| A. The spillover literature with the various asset classes | |||
| Beirne et al. (1) | Data: the stock indices in 41 emerging market economies in Asia, Europe, Latin America, and the Middle East; Period: 1993/9–2008/3 | Tri-variate VAR-BEKK-GARCH(1,1)-in-mean model | Spillovers from regional and global markets to local markets exist in the majority of EMEs. The nature of cross-market linkages varies across countries and regions |
| Allen et al. (2) | Data: stock indices in China, USA, Australia, Hong Kong, Japan, and Singapore; Period: period 1 (1991/8–1992/6); period 2 (1992/7–1996/12); period 3 (1997/1–2006/12); period 4 (2007/1–2010/11) | GARCH, VARMA–GARCH, VARMA–AGARCH models | Volatility spillovers exist across these markets in the pre-GFC periods, but there is little evidence of spillover effects from China to related markets during the GFC |
| Gilenko and Fedorova (3) | Data: stock indices in BRIC, SP500, DAX, Nikkei, and EMI; Period: pre-crisis period (2003/4–2007/12); crisis period (2008/1–2009/3); recovery (post-crisis) period (2009/3–2012/7) | 4-dimensional BEKK-GARCH-in-mean model | There were some lagged mean-to-mean spillovers between the BRIC stock markets. Volatility-to-volatility spillovers between these stock markets are largely present |
| Kitamura (4) | Data: euro, the pound, and the Swiss franc; Period: 2008/7-2009/7 | Varying-correlation model of multivariate GARCH | Return volatility in the euro spills into the pound and the Swiss franc, and these markets are highly integrated with the euro |
| Dean et al. (5) | Data: Australian equity and government bond; Period:1992/1–2006/11 | Bivariate DCC-GARCH model and BEKK-GARCH model | Negative bond market returns spillover into lower stock market returns. Bond market volatility spills over into the equity market but the reverse is not true |
| Sadorsky (6) | Data: Oil (WTI), Stock (WilderHill Clean Energy Index, ECO; and the NYSE Arca Technology Index, PSE); Period: 2001/1–2010/12 | Multivariate GARCH models (BEKK, diagonal, constant conditional correlation, and dynamic conditional correlation) | The DCC model also presents evidence of evidence of a statistically significant short-term persistence volatility spillover from oil to stock (ECO) |
| Smales (7) | Data: Oil (WTI), Stock (S&P500), and geopolitical risk (GPR) index; Period: 1986/1–2018/5 | Multivariate GARCH models (BEKK, diagonal, constant conditional correlation, and dynamic conditional correlation) | This DCC model shows short- and long-term volatility persistence for oil and stock prices, together with spillover effects that run from oil to stock returns |
| Yousaf and Hassan (9) | Data: Stock (China, India, Korea, Indonesia, Pakistan, Malaysia, Philippines, Thailand, and Taiwan), Oil (Brent); Period: 2000/1–2018/6 including the US subprime crisis period and the Chinese stock market crash period | Ling and McAleer's (8) VAR-GARCH model |
|
| Cevik et al. (10) | Data: Oil (WTI and Brent), Stock [Borsa Istanbul 100 (BIST) index]; Period: 1990–2017 | Univariate EGARCH model with Hong's (11) causality-in-mean test and causality-in-variance test | Return spillover: Oil→Stock; Volatility spillover: doesn't exist |
| Yang and Doong (12) | Data: the stock indices and exchange rate in the G-7 countries; Period: 1979/5–1999/1 | Bivariate VAR with CCC-EGARCH-X model | Return spillover: Stock→FX; Volatility spillover: Stock→FX |
| Kumar (13) | Data: the stock indices and exchange rate in India, Brazil, and South Africa; Period: 2000/1–2011/1 | VAR framework with the spillover index of Diebold and Yilmaz and multivariate BEKK-GARCH model | Return spillover: Stock→FX; Volatility spillover: stock→FX |
| Su (14) | Data: the stock indices and exchange rate in UK, Switzerland, Japan, South Korea, Singapore, Taiwan, and India; Period: 2001–2012 | Univariate AR(1)-EGARCH(1,1)-X model | Return spillover: FX→Stock; Volatility spillover: FX→Stock |
| Sui and Sun (15) | Data: the stock indices and exchange rate in BRICS; Period: 2005–2014 | VAR, variance decomposition, and impulse response functions | Return spillover: FX→Stock |
| Erdogan et al. (16) | Data: the Islamic stock indices and exchange rates in India, Malaysia, and Turkey; Period: 2013–2019 | The Granger causality test and the causality-in-variance test of Hafner and Herwartz (17) | Return spillover: Stock→FX; Volatility spillover: Stock→FX in Turkey |
| Uzonwanne (18) | Data: Stock (CAC40, DAX, FTSE, Nikkei, S&P500), Cryptocurrency (bitcoin); Period: 2013/3–2018/3 | Multivariate VARMA-AGARCH model | Return spillover: doesn't exist; Volatility spillover: Stock→Cryptocurrency |
| Su (19) | Data: Stock (Dow Jones, Nasdaq, and S&P500), Oil (WTI, GasNyh, and Heating), FX (UDI); Period: 2003/10–2015/8 | Bivariate VAR with BEKK-GJR-GARCH-MX-t model with a structural break |
|
| B. The spillover of the covid-19 pandemic impact | |||
| Dutta et al. (20) | Data: Stock (S&P 500), Oil (WTI), gold, and Climate Bond; Period: 2017/3–2020/6 | Bivariate VAR-ADCC-GARCH model | There is a bidirectional volatility linkage between climate bonds and the three indexes under study, whereas return linkages are marginal |
| Yousaf and Ali (21) | Data: Bitcoin, Ethereum, and Litecoin; Period: the pre-COVID-19 period (2019/1–2019/12) and the COVID-19 period (2020/1–2020/4) | VAR-DCC-GARCH model |
|
| Yousaf et al. (22) | Data: oil, gold, and Bitcoin; Period: the pre-COVID-19 period (2019/5–2019/12; the COVID-19 period (2020/1–2020/5) | VAR-DCC-GARCH model |
|
| Arfaoui and Yousaf (23) | Data: S&P 500, WTI oil, Bitcoin, gold; Period: before COVID-19 (2015/1–2019/12) and during COVID-19 (2020/1–2021/8) | Multivariate VAR asymmetric BEKK GARCH model | Return spillover: oil market is the most affected market in the before COVID-19 period but gold is the major receiver in the COVID-19 period; Volatility spillover: oil market is very sensitive to gold and US stock markets, especially during the COVID-19 outbreak |
| Zaremba et al. (24) | Data: the term spread (TERM), the change in the number of COVID-19 infections, the relative rate of change in central bank total assets, broad Government Response Index, Containment and Health Index, and economic response Index; Period: 2020/1–2020/9 | Panel regressions | First, the expansion of the disease significantly affects sovereign bond markets. Second, the growth of confirmed cases significantly widens the term spreads of government bonds. Third, an increase in the relative rate of change in the central bank balance sheet total assets exerts a negative effect on the term spread |
| Aharon et al. (25) | Data: the yield curve of G-7 countries and MCI (Media Coverage Index); Period: 2020/1–2021/8, covering the entire COVID-19 crisis | TVP-VAR methodology | The MCI and USA are the leading transmitters of spillover across all the yield curves in the G-7 countries. Moreover, Japan is a consistent receiver of risk from the G-7 countries |
| Gubareva et al. (26) | Data: the emerging market (EM) bond with the investment grade (IG) and high yield (HY); Period: 2020/1–2021/12 | TVP-VAR methodology | The option-adjusted spreads (OAS) of the IG and HY financials have recovered to the pre-COVID levels by the end of the year 2020, while for the HY sovereigns and corporates the OAS remain twice as wide as before the pandemic |
| Umar et al. (27) | Data: five Non-Fungible Tokens (NFTs) (Art, Collectibles, Games, Metaverse, and Utilities) and Media Coverage Index (MCI); Period: 2020/1–2021/12 | TVP-VAR methodology | Metaverse and Collectibles appear to be recipients of spillover for returns, whereas Art appears to be a net recipient of spillover for volatility. On the other hand, MCI appears to be a net transmitter for both return and volatility |
| Umar et al. (28) | Data: Seven high short interest indices (consumer, energy, financials, healthcare, industrials, real estate investment trusts, and technology), RavenPack Coronavirus MCI and Panic Index (PI); Period: 2020/2–2021/6 | TVP-VAR methodology | The returns spillovers are more vigorous than the volatility spillover. Moreover, stocks of companies belonging to the energy and healthcare sectors are net recipients of returns and volatility spillover from the MCI |
| Umar et al. (29) | Data: spot price index of S&P GSCI gold, silver, platinum, and palladium, RavenPack COVID-19 induced global panic index (GPI); Period: 2020/1–2020/7 | TVP-VAR methodology | First, the panic induced by COVID-19 is a shock transmitter to precious metals market. Second, we found silver to resist to these shocks while gold was a net receiver for almost all the period of analysis. Third, platinum and palladium on the other hand show a switching time varying patterns of connectedness to COVID-19 panic |
| Umar et al. (30) | Data: cryptocurrencies (Bitcoin, Ethereum, and Ripple), the fiat currencies (euro, GBP, and Chinese yuan), and the RavenPack Coronavirus MCI; Period: 2020/1–2020/12 | TVP-VAR methodology | The media coverage index and the cryptocurrencies are the net transmitters of shocks while the fiat currencies are the net receivers of shocks |
| Umar et al. (31) | Data: the bond indices for the EM High-Yield, EM Investment Grade, and the US Treasuries, and RavenPack Coronavirus MCI; Period: 2020/1–2020/12 | TVP-VAR methodology | Our results show a significant increase in the dynamic connectedness between media coverage, emerging market bonds, and US bonds, as well as between the respective volatilities, especially during the early phases of the COVID-19 pandemic, with the highest values observed in March 2020 |
| Umar et al. (32) | Data: the volatility of the S&P GSCI spot commodity indices and the Ravenpack Coronavirus Panic Index (PI); Period: 2020/1–2020/7 | Wavelet coherence methodology | There are intervals of low coherence across various time and frequency scales for these indices. The low coherence intervals show that diversification benefits |
| Ali et al. (33) | Data: the infectious disease-related equity market volatility (IDEMV) and bond indices (US, UK, Japan, Switzerland, Canada, Australia, Sweden, China, and Europe); Period: 2000/1–2021/2 | Wavelet coherence methodology | The results show no significant co-movement between these bond indices and IDEMV, thus confirming that they serve as a hedge against IDEMV |
| Umar and Gubareva (34) | Data: the Bloomberg Galaxy Crypto Index (BGCI), fiat currencies (EUR, GBP, and RMB), and the Ravenpack Coronavirus PI; Period: 2020/1–2020/5 | Wavelet coherence methodology | All the PI-currency pairs display similar patterns along the time and frequency scales in the respective heatmaps implying high coherence and interdependence around the apogee in the mid-March of the COVID-19 panic |
1. TVP-VAR is the abbreviation of Time-varying parameter vector autoregression. 2. The symbol “Stock→FX” denotes there exists a spillover from the stock market to the exchange rate market (or currency market) so are the other symbols “Oil→Stock” and “Stock→Cryptocurrency.” 3. The symbols “→(←)” in column “oil-Bitcoin” at panel “Return spillover for pre-COVID-19 (COVID-19)” in Yousaf et al. (22) denotes that there exists a return spillover from oil to Bitcoin in the pre-COVID-19 period but from the bitcoin to oil in the COVID-19 period. At the same inference process, the symbols “←(x)” in column “oil-stock” at panel “Volatility spillover for pre-SB (post-SB)” in Su (19) denote there exists a volatility spillover from stock to oil in the pre-SB period but no spillover exists between the stock and oil markets in the post-SB period.