HEDGING STOCKS IN CRISES AND MARKET DOWNTURNS WITH GOLD AND BONDS: INDUSTRY ANALYSIS

Anirut Pisedtasalasai

Department of Banking and Finance, Faculty of Commerce and Accountancy Chulalongkorn University, Thailand.

Keywords:Hedging stock return, Stock–gold correlations, Stock–bond correlations, Flight-to-quality, Safe haven, Crises, Covid-19, DCC-GARCH.

JEL Classification: C01; G11; G12.

ARTICLE HISTORY: Received:28 October 2020, Revised:23 November 2020, Accepted:14 December 2020, Published:30 December 2020

Contribution/Originality: In this paper, the dynamic stock–gold and stock–bond correlations are extensively investigated at industry level to further understand diverse behavior of individual industries relative to the overall market. Furthermore, the impact of crises on the correlations is also investigated individually since they may not be homogeneous across all crises.

1. INTRODUCTION

Understanding the nature of time variation in correlations between asset returns has always been a crucial issue in finance because a cross-correlation carries theoretical implications for understanding the formation of asset prices as well as practical implications for asset allocation and risk management (Connolly, Stivers, & Sun, 2005). Hedging stock markets with gold has received a great deal of attention in financial markets, particularly in the wake of the 2008 global financial crisis. Over the past two decades, the literature revealed that gold serves as a hedge and a safe haven for stocks during economic catastrophes or stock market stress. Hillier, Draper, & Faff (2006) documented that gold, together with other precious metals, are a hedge for stocks, especially during periods of abnormal stock market volatility. Using US, UK and German daily data, Baur & Lucey (2010) found that gold is a hedge for stocks in the US and the UK with the two assets having zero or negative correlations. Furthermore, a safe haven property of gold was also detected in all three markets. Correlations between gold and stocks become negative during stock market turmoil since investors sell risky stocks and evacuate funds to safe assets such as gold. This phenomenon is known as “flight-to-quality” or “flight-to-safety”.

Baur & McDermott (2010) further extended the issue by covering more countries with both developed and emerging markets and by using different data frequencies, including daily, weekly and monthly observations. They found that gold is a hedge and a safe haven for the US and major European stock markets, but not for Australia, Canada, Japan and large emerging markets. The hedging and safe haven properties of gold are strongest in daily data, implying that the impacts were short-lived. Hood & Malik (2013) compared the ability of gold with that of the VIX index in hedging and serving as a safe haven for the US stock market and they found that VIX was a superior hedging tool and served as a better safe haven than gold during the sample period.

Gurgun & Unalmis (2014) confirmed the hedge and safe haven properties of gold in some emerging and developing countries. Chkili (2016) employed A-DCC to analyze stock–gold time-varying correlations for BRICS countries (Brazil, Russia, India, China and South Africa). The result suggests that the stock–gold correlations were low to negative during both the global financial crisis of 2007–2008 and the European debt crisis of 2010–2011. Shahzad, Raza, Shahbaz, & Ali (2017) showed that gold provides a strong hedge for most markets investigated during normal and bullish conditions. However, this hedging ability disappears during extreme bear markets.

Bonds are another asset class that has been extensively explored regarding whether they act as a hedge or a safe haven for stocks. However, the empirical findings for bonds are mixed. Many studies (Baur & Lucey, 2009; Connolly, Stivers, & Sun, 2005; d’Addona & Kind, 2006; Kim, Moshirian, & Wu, 2006; Li & Zou, 2008; Li, Zheng, Chong, & Zhang, 2016; Yang, Zhou, & Wang, 2009) show that the flight-to-safety phenomenon is also typically observed in a stock–bond relationship. As investors reorganize their portfolios by replacing risky assets with safe assets, bond and stock market returns become negatively correlated (Kim, Moshirian, & Wu, 2006). Baur & Lucey (2009) studied stock–bond correlations in eight developed markets and revealed that the flight-to-safety phenomenon occurred during many crises and it commonly took place across countries. Hence, they suggest there is a link between the occurrence of the flight-to-safety phenomenon and a cross-country contagion. Skintzi (2019) document flight-to-safety in the Eurozone countries. Interestingly, the result showed that uncertainty in domestic stock markets drove stock–bond correlations in core EU countries during normal times, whereas uncertainty in global markets influenced the stock–bond correlations in peripheral EU countries during volatile periods.

On the other hand, some studies document positive relationships between stock and bond returns. The primary argument presented is that stocks and bonds share the same economic factors that influence their future cash flows and discount rates. Li (2002) suggests that uncertainty around expected inflation and real interest rates is the key driver of stock–bond correlations. Ilmanen (2003) found that during high inflationary periods, changes in common discount rates dominated changes in cash flow expectations and caused the stock–bond co-movements to be positive. Yang, Zhou, & Wang (2009) examined time-varying stock–bond correlations in the US and the UK over the past 150 years. Although they found that stock–bond correlations during recessions were lower than those during economic expansions in the US, the result in the UK indicated greater stock–bond correlations during economic expansions. Park, Zhongzhen, & Young (2019) an examined stock–bond correlation in Korea and the flight-to-safety phenomenon during the global-risk driven crises of 2007–2012 was reported. However, stock and bond returns declined together during the local risk-driven crisis of 1997–1999 forcing stock–bond correlations to be positive. Flavin & Lagoa-Varela (2019) examined the relationship between stock and long-term government bond returns in the crisis-hit Eurozone peripheral economies, including Greece, Ireland, Italy, Portugal and Spain, and concluded that the stock market and government bond returns moved even more in lockstep during the crisis period. Interestingly, they found that the key driver of the increased correlations between stock and bond returns was largely attributable to the financial sector.

In this paper, the hedging and flight-to-safety properties of gold and bonds during substantial stock market downturns and crises using daily data from the Stock Exchange of Thailand (SET) is examined. This paper offers contributions to the existing literature in several ways. First, existing literature concentrates more on investigating this issue in developed markets, whereas there is a lack of findings observed in emerging markets and even less so in the comparison of gold and bonds as hedge instruments for stocks. Second, previous studies largely consider stocks as a single asset class. Hence, the relationship between stock and bond returns as well as the relationship between stock and gold returns are estimated based on a stock index level that represents the overall stock market. As motivated by Flavin & Lagoa-Varela (2019), stock–bond correlations are not homogenous across industries and a finding based on the market level could only potentially be driven by some industries. Baele & Londono (2013) also documented that the cross-sectional dispersion in industry betas is larger during recessions. Therefore, this paper extends the issue by investigating at the industry level to establish an understanding around diverse behavior of individual industries relative to the overall market.

Third, to examine hedging and flight-to-safety properties of gold and bonds during events of market stress or turmoil, previous studies typically proxied the events by using periods of extreme stock market volatilities, large negative stock return environments and crises. In this paper, periods of crises and stock market downturns cover a longer time frame than those used in previous studies. For example, Baur & Lucey (2009) covered 20 trading days (i.e., one calendar month) for a crisis period. However, each stock market pullback in this paper could take several months to complete. Having a longer time frame could lead to a more general conclusion than employing a shorter time frame. This is because flight-to-safety behavior does not only take place during crisis days but also occurs during periods leading to the crises and other substantial market downturns as the sentiment of holding stocks turns poor. Finally, the effect of crises and stock market downturns on stock–gold correlations are separately investigated for each individual crisis since different crises could cause distinct effects on the correlations.

The rest of the paper proceeds as follows: Section 2 describes the data used in the study and presents descriptive data analysis; Section 3 outlines the methodology employed; Section 4 discusses the empirical findings; and the final section offers some conclusions.

2. DATA DESCRIPTIVE

The data used in this study consist of daily observations of the Thai stock market and industry indexes, bond indexes and gold price covering the period between January 1st, 2004 and March 23rd, 2020. The total return indexes of the SET and the SET’s industries were adopted for computing stock returns in the Stock Exchange of Thailand. The industry classification is based on the SET and covers the following eight industries: agriculture and food (Agro & Food), consumer products (Cons), financials (Fin), industrial (Ind), property & construction (Prop & Con), resources (Res), services (Serv) and technology (Tech).

The daily London Bullion Market Association (LBMA) gold bullion prices are converted to Thai baht and are applied instead of spot gold prices in the local Thai market since the spot gold prices released by the Gold Trader Association in Thailand are not available in terms of daily time-series; the time-series of daily prices of gold futures contracts traded on the Thailand Futures Exchange (TFEX) do not cover all the periods in this study. Furthermore, disregarding the movement of the exchange rate of THB/USD, the local gold price in Thailand tends to follow the LBMA gold price. Even the TFEX relies on the LBMA to price the underlying asset for its gold futures contracts. All of the total return indexes for the Thai stock market and industries are collected from Refinitiv Datastream. The total return indexes of Thai government bonds and corporate bonds are used to compute daily rate of returns of bonds in Thailand. These bond indexes are reported by the Thai Bond Market Association (ThaiBMA), which is the primary market for trading bonds in Thailand. The data for bonds are collected from Bloomberg. For the period of data used in this study, five crises and stock market downturns are identified in Table 1.

In Figure 1, the daily rebased time-series of the stock market index in relation to the government and corporate bond indexes and the gold price are illustrated. The highlighted areas indicate the periods of the crises and stock market downturns. Figure 1 shows that when the stock market substantially declines, gold prices tend to increase and provide a hedge for stocks, particularly during the global financial crisis in 2008, the stock market meltdown in 2015 and the Covid-19 pandemic in 2020. Both government and corporate bond indexes tend to be stable and less correlated with the stock index during crises and downturns.

Figure 1. Performance of stocks, bonds and gold during crises and market downturns.

Table 2 reports the mean and standard deviation of annualized daily returns on the stock market and industry indexes, the bond indexes and gold for the overall observation and observations for each crisis and market slide. For the full sample, the average daily returns on the stock market index (i.e., the SET index) are 7.1% per annum with a standard deviation of 20.1% per annum. Considering industry performances, services, and agriculture and food considerably outperformed the market, whereas consumer products, financials, industrials, and property and construction underperformed. In the same period, gold offered a slightly higher return (7.7% per annum) and a lower risk (SD of 18% per annum) than stocks. Government and corporate bonds provided lower returns and risks, as expected. Unsurprisingly, the stock market index experienced a large negative return of 95.7% per annum in crisis and stock market downturn periods. This average negative return was primarily prompted by the global financial crisis in 2008, the European debt crisis in 2011 and the Covid-19 pandemic in 2020. Industry-wise, financial, industrial, property and construction, and resource stocks performed particularly poorly during the crises. On the other hand, gold provided a positive average return and even performed slightly better during the crises and downturn periods. This evidence indicates that gold can be a hedge for stocks during stock market stress. The average government and corporate bonds return during crises and downturns are still positive and are generally similar to those of the full sample period.

3. RESEARCH METHODOLOGY

To examine hedging and flight-to-safety properties of gold and bonds during substantial stock market retreats, two approaches are employed.

3.1. Examining Dynamic Correlations during Crises and Stock Market Downturns

The DCC-GARCH model has been used by a number of studies to model time-varying stock–bond correlations and stock–gold correlations (e.g., Barunik, Kocenda, & Vacha, 2016; Basher & Sadorsky, 2016; Ciner, Gurdgiev, & Lucey, 2013). It is important to inspect whether these correlations are stable over time and behave differently across stock market conditions, especially declines.

The DCC-GARCH model is estimated in two steps. First, we estimate a univariate GARCH for each return series to obtain a time-varying standard deviation matrix and standardized residuals, which will be required to construct the dynamic conditional covariance matrix in the second step. Equations 1 and 2 present the specification of the GJR-GARCH (1,1) as follows:

Once time-varying stock–gold correlations and stock–bond correlations are obtained from the DCC-GARCH process, these correlations are then examined to determine whether gold and bonds can serve as a hedge or a safe haven for stocks.

3.2. Hedging Stock Market and Industry Returns Using Gold and Bond Returns

To test whether gold and bonds act as a hedge or engender a flight-to-safety phenomenon for the stock market and its individual industries, I employ the approach initially used by Baur & Lucey (2009) and Baur & Lucey (2010). Their approach has also been used or adapted in several later studies (e.g., Baur & McDermott, 2016; Ciner, Gurdgiev, & Lucey, 2013; Hood & Malik, 2013). This approach estimates the correlation between the returns of a stock market and those of another asset class by measuring the sensitivity of the asset class returns (i.e., gold and bond) with the stock returns. The main advantage of this approach is that the correlations observed in the form of betas and the marginal effects of crises and stock market downturns on the betas can be statistically tested. The testing model is described in Equations 8 and 9 as follows.

The β0 coefficient captures the correlation between gold or bonds and stocks. To control for heteroscedasticity in the analysis, a conditional variance is modeled by using GJR-GARCH (1,1).

To extend the analysis to test whether there are any changes in the stock–gold and stock–bond correlations during the crises and stock market downturns, the regressions presented in Equations 10 and 11 are estimated.

In Equation 10, Crisis is a dummy variable that captures the all crises and stock market downturn periods, including the global financial crisis in 2008, the European debt crisis in 2011, Thailand’s political turmoil in 2013, the stock market crash in 2015 and the Covid-19 pandemic in 2020. The β1 coefficient measures an additional sensitivity or correlation between gold returns (or bond returns) and stock returns during those crisis periods, while β0 measures correlations for non-crisis periods. A negative β1 would indicate a safe haven property of gold or bonds for stocks since it implies that investors sell risky stocks and shelter the funds in safe assets such as bonds or gold during stock market volatility.

Park, Zhongzhen, & Young (2019) documented that stock–bond correlations are not stable across crises. Therefore, in this study, the combined crisis dummy was split into five individual crisis dummies as shown in Equations 12 and 13.

In Equation 12, Crisis08, Crisis11, Crisis13, Crisis15 and Crisis20 are dummy variables that capture each individual crisis and stock market downturn. The crisis betas (i.e., β11,β12,β13, β14 and β15) represent an additional impact that each crisis has on the correlations.

4. EMPIRICAL FINDINGS

4.1. Estimates of the Dynamic Correlations of Stocks, Gold and Bonds

Figure 2 illustrates the dynamic daily correlation of the stock market returns with gold, government and corporate bond returns estimated using the DCC-GARCH model. The highlighted areas indicate the periods of crises and stock market downturns. In Panel A, the daily correlations of stock and gold returns vary over time, both in small positive and negative territories. However, the correlations become predominately negative in all crises and downturn periods. This evidence supports the property of gold as a hedge and a safe haven for stocks during stock market turbulence as investors flee from risky stocks to safe assets, such as gold. On the other hand, the results for government and corporate bonds reported in Panels B and C, respectively, are somewhat mixed. The stock–government bond and stock–corporate bond correlations during crises and downturns oscillate between negative and positive zones. Furthermore, the correlations behaved differently in each crisis. The stock–bond correlations were noticeably positive during the global financial crisis in 2008 and the political turmoil in 2013, but they were clearly negative in the European debt crisis in 2011.

Figure 2. Dynamic correlations of stock market, gold and bond returns.

Table 3 reports the average daily correlations of the stock market and industry returns with gold and bond returns from the DCC-GARCH model. In Panel A, the result shows that stock–gold correlations are generally negative and become even more negative during crises and downturn periods, particularly during the stock market upheaval in 2015. This finding is consistent with the result illustrated in Figure 2, indicating gold as a hedge and a safe haven for stocks. Considering individual industries, the result suggests that the negative stock–gold correlations during stock market turmoil are mainly propelled by the property and construction, financials and technology industries, while the correlations of the other industries do not seem to change significantly from the full sample. In contrast, the result of the stock–bond correlations presented in Panels B and C downplay the role of both government and corporate bonds as a hedge and a safe haven for stocks since the correlations are, on average, slightly higher during the crises and downturns even though the correlations remain low, which may suggest that they still possess some diversification capability. Furthermore, the stock-bond correlations observed are inconsistent across crises. They are negative in the European debt crisis in 2011 and the Covid-19 pandemic in 2020, but they turn out to be positive in the other three crises and downturns. This result coincides with findings by Park, Zhongzhen, & Young (2019), who discovered that the stock–bond correlations are not homogenous across crises; the results are somewhat different across industries. Interestingly, the increasing stock–government and stock–corporate bond correlations during crises are more pronounced in the property and construction, financials and technology industries.

4.2. Hedging Stock Market and Industry Indexes Using Gold

In Table 4, Models 1, 2 and 3 report the result of stock–gold correlations estimated from Equations 8, 10 and 12, respectively.

The β0 coefficient in Model 1 is negative but insignificant. It suggests that, in general, the sensitivity or correlation between gold and the stock market is close to zero, indicating gold as a hedge for equities. Model 2 further investigates the impact of crises and market downturns on the correlations. The β0 coefficient now captures the correlations in non-crisis periods, while β1 captures an increase or decrease in the correlations caused by crises and downturns. The β0 coefficient continues to be insignificantly negative. Similarly, β1 is also insignificantly negative, which suggests that the flight-to-safety phenomenon between gold and stocks during crises and downturns does not generally or significantly exist in Thailand. However, when individual crises and downturns are examined, the results in Model 3 show that the stock market rout in 2015 (represented by β14) is negatively significant at a 10% level. This finding suggests that the flight-to-safety phenomenon between stocks and gold was marginally observed in the 2015 downturn period where the stock–gold correlations became even more negative during the stock market distress.

Table 5 presents the results of the stock–gold correlations for individual industries. For brevity, only the results of Models 2 and 3 estimated from Equations 10 and 12, respectively, are reported.

The β0 coefficient is insignificant for all industries except the agriculture and food industry, which reports a strong negative correlation at a 1% significance level in both Models 2 and 3. This suggests that the ability of gold in hedging stocks is more pronounced in the agriculture and food industry. Considering the impacts of crises and downturns on the correlations presented in Model 2, β1 is not significant for all industries. However, when individual crises were investigated (see Model 3), β14, which represents the 2015 downturn variable, is strongly negatively significant for four industries, namely agriculture and food, consumer products, property and construction and resources, whereas those of the other industries are also negatively significant but at a 10% level. Therefore, the results presented in Tables 4 and 5 indicate that gold can serve as a hedge for the stock index and individual industries but the effects can vary to a certain extent across industries, as documented by Flavin & Lagoa-Varela (2019). Consistent with Park, Zhongzhen, & Young (2019), the flight-to-safe phenomenon, where the stock–gold correlations become even more negative, was only found to be significant in some crises.

4.3. Hedging Stock Market and Industry Indexes Using Government Bonds

Table 6 reports the result of stock–government bond correlations. The β0 coefficient is significantly positive at a 10% level in all three models, implying that the stock market–government bond correlations are generally positive, albeit close to zero.

The β1 coefficient in Model 2 is insignificant, which suggests that the correlations do not significantly change during the crises and downturns. However, when individual crises and downturns are investigated in Model 3, the correlations are not affected by each individual crisis and downturn except in the political turmoil in 2013. The correlations in this crisis significantly increased by approximately 5%. This finding implies that government bonds provide less of a hedging effect, when such an effect is most needed, and clearly contradicts the findings in several prior studies (e.g., Baur & Lucey, 2009; Kim, Moshirian, & Wu, 2006; Skintzi, 2019) that support the presence of the flight-to-safety phenomenon between stocks and bonds. Nevertheless, the result is consistent with Ilmanen (2003), Yang, Zhou, & Wang (2009) and Park, Zhongzhen, & Young (2019) in that it shows increases in the stock–bond correlations in times of stock market turbulence as market players turn to consider bonds as equally risky as stocks.

Table 7 reports the results of stock–government bond correlations based on individual industries. The correlations in non-crisis periods, β0, stay close to zero in all cases. The crisis variable in Model 2 is not significant, which indicates that, overall, the crises and downturns do not influence the correlations between industry and government bond returns.

However, when individual crises and downturns are analyzed, the political turmoil in 2013 significantly increased the correlations in all industries. These results are also consistent the results previously reported in Table 6 suggesting less of a hedging ability of bonds during crises. Conversely, during the Covid-19 pandemic in 2020, the flight-to-safety phenomenon was only observed in some industries, including agriculture and food, industrials, property and construction and technology, which is consistent with findings by Park, Zhongzhen, & Young (2019). It suggests that the correlations are not stable across crises and can be unpredictable.

4.4. Hedging Stock Market and Industry Indexes Using Corporate Bonds

Table 8 reports the results of stock–corporate bond correlations. The results generally led to similar conclusions as the previous result on the stock–government bond correlations; the stock–corporate bond correlations stayed close to zero. Crises and downturns did not tend to affect the correlations except during the political turmoil in 2013. This crisis increased the correlations by 1.66% at a 10% significance level.

Table 9 reports the results of stock–corporate bond correlations based on individual industries. Consistent with the results of the overall market, the correlations in non-crisis periods were not significantly different from zero in all cases. Only the political turmoil in 2013 saw significantly enlarged correlations in every industry except technology.

5. CONCLUSION

This paper examined whether gold and bonds can serve as a hedge or a safe haven asset for stocks during crises and stock market downturns in Thailand. First, in comparison to bonds, gold provides a better hedge for stocks since stock–gold correlations are predominantly negative. Meanwhile, despite boasting positive correlations on average, the relationships between stocks and bonds often stay close to zero. Second, gold can act as a safe haven for stocks during stock market turmoil although this finding was only observed in some crises. Stock–bond correlations, on the other hand, increased slightly in some crises suggesting that bonds provide less hedging protection to stocks when it is most needed. This striking finding is also consistent with the results when individual industries were analyzed. More importantly, it contradicts a common belief that bonds are a safe haven for stocks. A possible explanation for this discovery is that since stocks and bonds share similar sources of companies’ cash flows, bonds are viewed as risky as stocks in times of extreme market volatility and hence market players treat them as one and the same. Finally, the correlations were not stable across crises and industries of stocks investigated.

REFERENCES

Baele, L., & Londono, J. M. (2013). Understanding industry betas. Journal of Empirical Finance, 22(C), 30-51.Available at: https://doi.org/10.1016/j.jempfin.2013.02.003.

Barunik, J., Kocenda, E., & Vacha, L. (2016). Gold, oil, and stocks: Dynamic correlations. International Review of Economics and Finance, 42(C), 186-201.Available at: https://doi.org/10.1016/j.iref.2015.08.006.

Basher, S. A., & Sadorsky, P. (2016). Hedging emerging market stock prices with oil, gold VIX, and bonds: A comparison between DCC, ADCC and GO-GARCH. Energy Economics, 54(C), 235-247.Available at: https://doi.org/10.1016/j.eneco.2015.11.022.

Baur, D. G., & Lucey, B. M. (2009). Flights and contagion—An empirical analysis of stock–bond correlations. Journal of Financial stability, 5(4), 339-352.

Baur, D. G., & McDermott, T. K. (2010). Is gold a safe haven? International evidence. Journal of Banking & Finance, 34(8), 1886-1898.Available at: https://doi.org/10.1016/j.jbankfin.2009.12.008.

Baur, D. G., & McDermott, T.K. (2016). Why is gold a safe haven? Journal of Behavioral and Experimental Finance, 10(C), 63-71.Available at: https://doi.org/10.1016/j.jbef.2016.03.002.

Baur, D. G., & Lucey, B. M. (2010). Is gold a hedge or safe haven? An analysis of stocks, bonds, and gold. The Financial Review, 45(2), 217-229.Available at: https://doi.org/10.1111/j.1540-6288.2010.00244.x.

Chkili, W. (2016). Dynamic correlations and hedging effectiveness between gold and stock markets: Evidence for BRICS countries. Research in International Business and Finance, 38(C), 22-34.Available at: https://doi.org/10.1016/j.ribaf.2016.03.005.

Ciner, C., Gurdgiev, C., & Lucey, B. M. (2013). Hedges and safe havens: An examination of stocks, bonds, gold, oil and exchange rates. International Review of Financial Analysis, 29(C), 202-211.Available at: https://doi.org/10.1016/j.irfa.2012.12.001.

Connolly, R., Stivers, C., & Sun, L. (2005). Stock market uncertainty and the stock-bond return relation. Journal of Financial and Quantitative Analysis, 40(1), 161–194.Available at: https://doi.org/10.1017/S0022109000001782.

d’Addona, S., & Kind, A. H. (2006). International stock–bond correlations in a simple affine asset pricing model. Journal of Banking & Finance, 30(10), 2747-2765.Available at: https://doi.org/10.1016/j.jbankfin.2005.10.007.

Flavin, T. J., & Lagoa-Varela, D. (2019). On the stability of stock-bond comovements across market conditions in the Eurozone periphery. Global Finance Journal, 100491.Available at: https://doi.org/10.1016/j.gfj.2019.100491.

Gurgun, G., & Unalmis, L. (2014). Is gold a safe haven against equity market investment in emerging and developing countries? Finance Research Letters, 11, 341-348.Available at: https://doi.org/10.1016/j.frl.2014.07.003.

Hillier, D., Draper, P., & Faff, R. (2006). Do precious metals shine? An investment perspective. Financial Analyst Journal, 62, 98–106.Available at: https://doi.org/10.2469/faj.v62.n2.4085.

Hood, M., & Malik, F. (2013). Is gold the best hedge and safe haven under changing stock market volatility? Review of Financial Economics, 22, 47-52.Available at: https://doi.org/10.1016/j.rfe.2013.03.001.

Ilmanen, A. (2003). Stock-bond correlations. The Journal of Fixed Income, 13, 55-66.Available at: https://doi.org/10.3905/jfi.2003.319353.

Kim, S.-J., Moshirian, F., & Wu, E. (2006). Evolution of international stock and bond market integration: Influence of the European Monetary Union. Journal of Banking & Finance, 30(5), 1507-1534.Available at: https://doi.org/10.1016/j.jbankfin.2005.05.007.

Li, L. (2002). Macro factors and the correlations of stock and bond returns. Working Paper, Yale International Center for Finance, Yale University, United States.

Li, M., Zheng, H., Chong, T. T. L., & Zhang, Y. (2016). The stock–bond comovements and cross-market trading. Journal of Economic Dynamics and Control, 73(C), 417-438.Available at: https://doi.org/10.1016/j.jedc.2016.10.007.

Li, X.-M., & Zou, L.-P. (2008). How do policy and information shocks impact co-movements of China’s T-bond and stock markets? Journal of Banking & Finance, 32(3), 347-359.Available at: https://doi.org/10.1016/j.jbankfin.2007.04.029.

Park, K., Zhongzhen, F., & Young, H. H. (2019). Stock and bond returns correlation in Korea: Local versus global risk during crisis periods. Journal of Asian Economics, 65(C), 1-18.Available at: https://doi.org/10.1016/j.asieco.2019.101136.

Shahzad, S. J. H., Raza, N., Shahbaz, M., & Ali, A. (2017). Dependence of stock markets with gold and bonds under bullish and bearish market states. Resource Policy, 52(C), 308-319.Available at: https://doi.org/10.1016/j.resourpol.2017.04.006.

Skintzi, V. D. (2019). Determinants of stock-bond market comovement in the Eurozone under model uncertainty. International Review of Financial Analysis, 61(C), 20-28.Available at: https://doi.org/10.1016/j.irfa.2018.12.005.

Yang, J., Zhou, Y., & Wang, Z. (2009). The stock–bond correlation and macroeconomic conditions: One and a half centuries of evidence. Journal of Banking & Finance, 33(4), 670-680.Available at: https://doi.org/10.1016/j.jbankfin.2008.11.010.