Assessing Bitcoin and Gold as Safe Havens Amid Global Uncertainties: A Rolling Window DCC-GARCH Analysis

Abstract

We examine the roles of Gold and Bitcoin as a hedge, a safe haven, and a diversifier against the coronavirus disease 2019 (COVID-19) pandemic and the Ukraine War. Using a rolling window estimation of the dynamic conditional correlation (DCC)-based regression, we present a novel approach to examine the time-varying safe haven, hedge, and diversifier properties of Gold and Bitcoin for equities portfolios. This article uses daily returns of Gold, Bitcoin, S&P500, CAC 40, and NSE 50 from January 3, 2018, to October 15, 2022. Our results show that Gold is a better safe haven than the two, while Bitcoin exhibits weak properties as safe haven. Bitcoin can, however, be used as a diversifier and hedge. This study offers policy suggestions to investors to diversify their holdings during uncertain times.

JEL Codes: G1, G11, G12

Keywords

Gold Bitcoin safe haven COVID-19 Ukraine War

Introduction

Financial markets and the diversity of financial products have risen in both volume and value, creating financial risk and establishing the demand for a safe haven for investors. The global financial markets have faced several blows in recent years. From the Global Financial Crisis (GFC) to the outbreak of the pandemic and uncertainty regarding economic policy measures of governments and central banks, the financial markets including equity markets around the world were faced with severe meltdowns. This similar behavior was observed in other markets including equity and commodity markets, resulting in overall uncertainty. In this scenario, the investors normally flock toward the safe-haven assets to protect their investment. In normal situations, investors seek to diversify or hedge their assets to protect their portfolios. However, the financial markets are negatively impacted when there are global uncertainties. Diversification and hedging methods fail to safeguard investors’ portfolios during instability because almost all sectors and assets are negatively affected (Hasan et al., 2021). As a result, investors typically look for safe-haven investments to safeguard their portfolios under extreme conditions (Ceylan, 2022).

Baur and Lucey (2010) provide the following definitions of hedge, diversifier, and safe haven:

Hedge: An asset that, on average, has no correlation or a negative correlation with another asset or portfolio. On average, a strict hedge has a (strictly) negative correlation with another asset or portfolio.

Diversifier: An asset that, on average, has a positive correlation (but not perfect correlation) with another asset or portfolio.

Safe haven: This is the asset that in times of market stress or volatility becomes uncorrelated or negatively associated with other assets or a portfolio.

As was previously indicated, the significant market turbulence caused by a sharp decline in consumer spending, coupled with insufficient hedging opportunities, was a common feature of all markets during these times (Yousaf et al., 2022). Nakamoto (2008) suggested a remedy by introducing Bitcoin, a “digital currency,” as an alternative to traditional fiduciary currencies (Paule-Vianez et al., 2020). Bitcoin often described as “Digital Gold” has shown greater resilience during periods of crises and has highlighted the potential safe haven and hedging property against uncertainties (Mokni, 2021). According to Dyhrberg (2016), the GFC has eased the emergence of Bitcoin thereby strengthening its popularity. Bouri et al. (2017) in their study indicate that Bitcoin has been viewed as a shelter from global uncertainties caused by conventional banking and economic systems. Recent research has found that Bitcoin is a weak safe haven, particularly in periods of market uncertainty like the coronavirus disease 2019 (COVID-19) crisis (Conlon & McGee, 2020; Nagy & Benedek, 2021; Shahzad et al., 2019; Syuhada et al., 2022). In contrast to these findings, a study by Yan et al. (2022) indicates that it can function as a strong safe haven in favorable economic times and with low-risk aversion. Ustaoglu (2022) also supports the strong safe-haven characteristic of Bitcoin against most emerging stock market indices during the COVID-19 period. Umar et al. (2023) assert that Bitcoin and Gold are not reliable safe-havens. Singh et al. (2024) in their study reveal that Bitcoin is an effective hedge for investments in Nifty-50, Sensex, GBP–INR, and JPY–INR, at the same time a good diversifier for Gold. The study suggests that investors can incorporate Bitcoin in their portfolios as a good hedge against market volatility in equities and commodities markets. During the COVID-19 epidemic, Barbu et al. (2022) investigated if Ethereum and Bitcoin could serve as a short-term safe haven or diversifier against stock indices and bonds. The outcomes are consistent with the research conducted by Snene Manzli et al. (2024). Both act as hybrid roles for stock market returns, diversifiers for sustainable stock market indices, and safe havens for bond markets. Notably, Bhuiyan et al. (2023) found that Bitcoin provides relatively better diversification opportunities than Gold during times of crisis. To reduce risks, Bitcoin has demonstrated a strong potential to operate as a buffer against global uncertainty and may be a useful hedging tool in addition to Gold and similar assets (Baur & Lucey, 2010; Bouri et al., 2017; Capie et al., 2005; Dyhrberg, 2015). According to Huang et al. (2021), its independence from monetary policies and minimal association with conventional financial assets allow it to have a safe-haven quality.

Bitcoins have a substantial speed advantage over other assets since they are traded at high and constant frequencies with no days when trading is closed (Selmi et al., 2018). Additionally, it has been demonstrated that the average monthly volatility of Bitcoin is higher than that of Gold or a group of international currencies expressed in US dollars; nevertheless, the lowest monthly volatility of Bitcoin is lower than the maximum monthly volatility of Gold and other foreign currencies (Dwyer, 2015). Leverage effects are also evident in Bitcoin returns, which show lower volatilities in high return periods and higher volatilities in low return times (Bouri et al., 2017; Liu et al., 2017). According to recent research, Bitcoins can be used to hedge S&P 500 stocks, which increases the likelihood that institutional and retail investors will build secure portfolios (Okorie, 2020). Bitcoin demonstrates strong hedging capabilities and can complement Gold in minimizing specific market risks (Baur & Lucey, 2010). Its high-frequency and continuous trading further enrich the range of available hedging tools (Dyhrberg, 2016). Moreover, Bitcoin spot and futures markets exhibit similarities to traditional financial markets. In the post-COVID-19 period, Zhang et al. (2021) found that Bitcoin futures outperform Gold futures.

Gold, silver, palladium, and platinum were among the most common precious metals utilized as safe-haven investments. Gold is one such asset that is used extensively (Salisu et al., 2021). Their study tested the safe-haven property of Gold against the downside risk of portfolios during the pandemic. Empirical results have also shown that Gold functions as a safe haven for only 15 trading days, meaning that holding Gold for longer than this period would result in losses to investors. This explains why investors buy Gold on days of negative returns and sell it when market prospects turn positive and volatility decreases (Baur & Lucey, 2010). In their study, Kumar et al. (2023) tried to analyse the trends in volume throughout futures contracts and investigate the connection between open interest, volume, and price for bullion and base metal futures in India. Liu et al. (2016) in their study found that there is no negative association between Gold and the US stock market during times of extremely low or high volatility. Because of this, it is not a strong safe haven for the US stock market (Hood & Malik, 2013). Post-COVID-19, studies have provided mixed evidence on the safe-haven properties of Gold (Bouri et al., 2020; Cheema et al., 2022; Ji et al., 2020). According to Kumar and Padakandla (2022), Gold continuously demonstrates safe-haven qualities for all markets, except the NSE, both in the short and long term. During the COVID-19 episode, Gold’s effectiveness as a hedge and safe-haven instrument has been impacted (Akhtaruzzaman et al., 2021). Al-Nassar (2024) conducted a study on the hedge effectiveness of Gold and found that it is a strong hedge in the long run. Bhattacharjee et al. (2023) in their paper examined the symmetrical and asymmetrical linkage between Gold price levels and the Indian stock market returns by employing linear autoregressive distributed lag and nonlinear autoregressive distributed lag models. The results exhibit that the Indian stock market returns and Gold prices are cointegrated. According to the most recent study by Kaczmarek et al. (2022), Gold has no potential as a safe haven, despite some studies on the COVID-19 pandemic showing contradictory results. The co-movements of Bitcoin and the Chinese stock market have also normalized as a result of this epidemic (Belhassine & Karamti, 2021). Widjaja and Havidz (2023) verified that Gold was a safe haven asset during the COVID-19 pandemic, confirming the Gold’s safe-haven characteristic.

As previously pointed out, investors value safe-haven investments in times of risk. Investors panic at these times when asset prices fall and move from less liquid (risky) securities to more liquid (safe) ones, such as cash, Gold, and government bonds. An asset must be bought and sold rapidly, at a known price, and for a reasonably modest cost to be considered truly safe (Smales, 2019).

Therefore, we need to properly re-examine the safe-haven qualities of Gold and Bitcoin due to the mixed evidences regarding their safe-haven qualities and the impact of COVID-19 and the war in Ukraine on financial markets.

This work contributes to and deviates from the body of existing literature in the following ways. We propose a novel approach in this work to evaluate an asset’s time-varying safe haven, hedge, and diversifier characteristics. This research examines the safe haven, hedging, and diversifying qualities of Gold and Bitcoin against the equity indices; S&P 500, CAC 40, and NSE 50. Through the use of rolling window estimation, we extend the methodology of Ratner and Chiu (2013) by estimating the aforementioned properties of the assets. Comparing rolling window estimation to other conventional techniques, the former will provide a more accurate representation of an asset’s time-varying feature. This study explores the conventional asset Gold’s time-varying safe haven, hedging, and diversifying qualities during crises like the COVID-19 pandemic and the conflict in Ukraine. We use Bitcoin, an unconventional safe-haven asset, for comparison.

Data and Methodology

We use the daily returns of three major equity indices; S&P500, CAC 40, and NSE 50 from January 3, 2018, to October 15, 2022. The equity indices were selected to represent three large and diverse markets namely the United States, France, and India in terms of geography and economic development. We assess safe-haven assets using the daily returns of Gold and Bitcoin over the same time. Equity data was collected from Yahoo Finance, Bitcoin data from coinmarketcap.com, and Gold data from the World Gold Council website.

Engle (2002) developed the DCC (Dynamic Conditional Correlation)-GARCH model, which is frequently used to assess contagion amid pandemic uncertainty or crises. Time-varying variations in the conditional correlation of asset pairings can be captured using the DCC-GARCH model. Through employing this model, we can analyse the dynamic behavior of volatility spillovers. Engle’s (2002) DCC-GARCH model contains two phases;

Univariate GARCH model estimation

Estimation of time-varying conditional correlation. Let X(t) be the multivariate return series of interest:

X_{t} = μ_{t} + H_{t}^{\frac{1}{2}} ε_{t}

(1)

H_{t} = D_{t} R_{t} D_{t}

(2)

In Equation (1), X_t = (X_1t, …, X_nt) refers to the vector of past observations. The multivariate conditional variance is represented by H_t, while conditional returns are presented by μ_t, and ε_t is the standardized residuals. R_t is N × N linear dynamic correlations matrix and D_t refers to the conditional standard deviation of the return series based on a univariate GARCH (1,1) model with $\sqrt{h_{i i, t}}$ on the i^th diagonal, where i = 1, 2, …. N.

The DCC specification is given in the Equation (3).

Q_{t} = (1 - ψ - ζ) \bar{Q} + ζ Q_{t - 1} + ψ δ_{i, t - 1} δ_{j, t - 1}

(3)

Where $Q_{t} = [q_{i j, t}]$ is a (N × N) time-varying covariance matrix of standardized residuals $(δ_{i t} = \frac{ε_{i t}}{\sqrt{h_{i t}}})$ ; δ_i,t, δ_j,t are unconditional correlations, and ψ and ζ are non-negative scalars such that $ψ + ζ < 1.$

Here, we employ the methodology developed by Ratner and Chiu (2013) and Baur et al. (2017a) to analyse Golds’s and Bitcoin’s capabilities as a safe haven, diversifier, or hedge against fluctuations in equity markets. For this, we first obtain the pairwise dynamic conditional correlation between equity returns against Bitcoin and Gold. Next, the DCC-GARCH estimates (DCC_t) are regressed on dummy variables (D).

\begin{array}{l} D C C_{t} = β_{0} + β_{1} D (r_{a s s e t} q_{10}) + β_{2} D (r_{a s s e t} q_{5}) + \\ β_{3} D (r_{a s s e t} q_{1}) \end{array}

(4)

D shows the extreme movements (fluctuations) and it equals one if the returns on assets (Bitcoin or Gold) hit a certain threshold. This is provided by the first, fifth, and tenth percentiles of the return distribution. β₀ is the intercept term in the equation. If β₀ is positive and significant, it indicates that Bitcoin and Gold act as a diversifier against other assets. If β₀ is zero, Bitcoin/Gold will be a weak hedge against other assets. If β₀ is negative and significant, Bitcoin will be a powerful hedge against other assets. β₁, β₂, and β₃ are the partial correlation coefficients of the first, fifth, and tenth percentiles of the return distribution, respectively. Bitcoin/Gold will perform as a weak safe haven against other assets if β₁, β₂, or β₃ are insignificantly different from 0. However, if β₁, β₂, or β₃ are negative and significant, Bitcoin/Gold will perform as a strong safe haven.

We extend the method of Ratner and Chiu (2013) by estimating the aforementioned regression in a rolling window so that we can see whether Gold and Bitcoin could act as safe havens/diversifiers/hedges during periods of crisis. We estimate the rolling window regression over a window of length 60. The rationale for selecting a 60-day length as a rolling window period is that it will show the median-term holding of safe-haven assets (two trading months) and is also widely used in the literature (see Demir et al., 2018).

Results and Discussion

The outcomes of the parameters under the DCC-GARCH model for each of the asset pairs selected for the investigation are shown in Table 1.

Table 1.

DCC-GARCH Parameters for Asset Pairs.

DCC Parameters	NSE 50-Gold	NSE 50-BTC^a	S&P500-Gold	S&P500-BTC^a	CAC 40-Gold	CAC 40-BTC^a
ρ	–0.065 (0.224)	–0.008 (0.821)	–0.095 (0.053)	–0.006 (0.986)	–0.227 (0.000)	0.029 (0.520)
α_DCC	0.020 (0.122)	0.055 (0.102)	0.027 (0.199)	0.039 (0.121)	0.049 (0.057)	0.038 (0.091)
β_DCC	0.971 (0.000)	0.595 (0.000)	0.831 (0.000)	0.831 (0.000)	0.823 (0.000)	0.803 (0.000)

Note: ^aBTC denotes Bitcoin.

First, we look at the dynamical conditional correlation coefficient, ρ.The rho value is negative and insignificant for NSE 50/Gold, NSE 50 /BTC, S&P500/Gold, and S&P500/BTC indicating a negative and insignificant correlation between these asset pairs, showing Gold and Bitcoin as potential hedges and safe havens. The fact that ρ is negative and significant for CAC 40/Gold suggests that Gold can be a safe haven against CAC 40 swings. The asset pair CAC/BTC, on the other hand, has possible diversifier behavior with ρ being positive but statistically insignificant.

Next, we examine the behavior of the DCC-GARCH parameters; α and β. We find that α_DCC is statistically insignificant for all the asset pairs, while β_DCC is statistically significant for all asset pairs. β_DCC quantifies the persistence feature of the correlation and the extent of the impact of volatility spillover in a particular market’s volatility dynamics. A higher β_DCC value implies that a major part of the volatility dynamics can be explained by the respective market’s own past volatility. For instance, the NSE 50/Gold’s β_DCC value of 0.971 shows that there is a high degree of volatility spillover between these two assets, with about 97% of market volatility being explained by the assets’ own historical values and the remainder coming from spillover.

Thus, we see that the volatility spillover is highly persistent (~0.8) for all the asset pairs except NSE 50/BTC.

The results above show that the nature of the dynamic correlation between the stock markets, Bitcoin and Gold is largely negative, pointing toward the possibility of Gold and Bitcoin being hedge/safe haven. However, a detailed analysis is needed to confirm the same by employing rolling window analysis, and we present the results in the forthcoming section.

We present the rolling window results for S&P500 first. We present the regression results for Gold in Figure 1 and Bitcoin in Figure 2

Figure 1.

Note: Areas shaded under factor 1 represent significant regression coefficients.

In Figure 1, we examine the behavior of β₀ (intercept term), β₁, β₂, and β₃ (partial correlation coefficients). The intercept term β₀ will give an idea about whether the asset is behaving as a diversifier or hedge. Here, the intercept term shows significance most of the time. However, during 2018, the intercept was negative and significant, showing that it could serve as a hedge during geopolitical tensions and volatilities in the global stock market. However, during the early stages of COVID-19, we show that the intercept is negative and showing statistical significance, suggesting that Gold could serve as a hedge during the initial shocks of the pandemic. These findings are contrary to the results in the study by Tarchella et al. (2024) where they found hold as a good diversifier. Later, we find the intercept to be positive and significant, indicating that Gold could act as a potential diversifier. But during the Russia-Ukraine War, Gold exhibited hedge ability again. Looking into the behavior of β₁, which is the partial correlation coefficient for the tenth percentile of return distribution shows negative and insignificant during 2018. Later, it was again negative and significant during the initial phases of COVID-19, and then negative in the aftermath, indicating that Gold could act as a weak safe haven during the COVID-19 pandemic. Gold could serve as a strong safe haven for the SP500 against volatility in the markets brought on by the war in Ukraine, as we see the coefficient to be negative and large during this time. From β₂ and β₃, the partial correlation coefficients of the fifth and first percentile, respectively, show that Gold possesses weak safe haven properties during COVID-19 and strong safe haven behavior during the Ukraine crisis.

Next, we examine the characteristics of Bitcoin as a hedge/diversifier/safe haven against the S&P500 returns. We present the results in Figure 2.

Figure 2.

Note: Areas shaded under factor 1 represent significant regression coefficients.

Like in the previous case, we begin by analysing the behavior of the intercept coefficient, which is β₀. As mentioned earlier the intercept term will give a clear picture of the asset’s hedging and diversifier property. In the period 2018–2019, the intercept term is positive but insignificant. This could be due to the large volatility in Bitcoin price movements during the period. It continues to be minimal (but positive) and insignificant during 2019–2020, indicating toward weak diversification possibility. Post-COVID-19 period, the coefficient shows the significance and positive value, displaying the diversification potential. We see that the coefficient remains positive throughout the analysis, confirming Bitcoin’s potential as a diversifier. Looking into the behavior of β₁ (the partial correlation coefficient at tenth percentile), it is positive but insignificant during 2018. The coefficient is having negative sign and showing statistical significance in 2019, suggesting that Bitcoin could be a good safe haven in that year. This year was characterized by a long list of corporate scandals, uncertainties around Brexit, and tensions in global trade. We can observe that throughout the COVID-19 period, the coefficient is showing negative sign and negligible during the March 2020 market meltdown, suggesting inadequate safe-haven qualities. However, Bitcoin will regain its safe-haven property in the coming periods, as the coefficient is negative and significant in the coming months. The coefficient is negative and shows statistical significance during the Ukrainian crisis, suggesting strong safe-haven property.

Only during the Ukrainian crisis could Bitcoin serve as a safe haven, according to the behavior of β₂, which displays the partial correlation coefficient at the fifth percentile. Bitcoin was a weak safe haven during COVID-19 and the Ukrainian crisis, according to β₃, the partial correlation coefficient for the first percentile (coefficient negative and insignificant). According to the overall findings, Gold is a stronger safe haven against the S&P 500’s swings.

This result is consistent with the previous studies of Triki and Maatoug (2021), Shakil et al. (2018), Będowska-Sójka and Kliber (2021), Drake (2022), and Ghazali et al. (2020), etc.

Next, we analyse the results for CAC 40. We present the results in Figures 3 and 4.

Figure 3.

Note: Areas shaded under factor 1 represent significant regression coefficients.

Figure 4.

Note: Areas shaded under factor 1 represent significant regression coefficients.

Gold is a good hedge because the intercept term is β₀ negative and important during the COVID-19 and Ukraine War. Looking at the graph for β₀, we see that it is significant most of the time, except in the period between 2020 and 2021 and during the start of the Ukraine crisis. β₁ is negative during the initial stage COVID-19 period, which could indicate weak safe-haven behavior. It later regained its position to become a strong safe haven during the Ukraine War, when it turned negative and significant, but it further moved on to become insignificant during the later periods. Looking at β₂, we see that it is negative and insignificant during the COVID-19 crisis; therefore, it is a weak safe haven. It becomes negative and significant during the Ukraine crisis, therefore showing strong safe-haven properties. This is similar to what was observed in the case of β₀.

From the movement of β₃, we see that it is showing a positive sign but is statistically insignificant in the year 2018. However, it was negative and insignificant during COVID-19, therefore acting as a weak safe haven, while it became negative and insignificant during the Ukraine crisis, showing weak safe-haven attributes. In general, we may say that Gold has weak safe-haven properties considering CAC40.

Here, the intercept term β₀ is negative and insignificant during the 2018–2019 period. This indicated that Bitcoin performed as a weak safe haven throughout this period. Later we observe that during the initial phase of COVID-19, the β₀ term is positive and significant and then it shows a negative sign. This may indicate that Bitcoin is not a very good hedge. β₁ was negative and shows statistical significance during the initial period of 2018, meaning it served as a strong hedge during the stock market dive. This period was marked by other events such as geopolitical uncertainty such as the US trade war with China, the slowdown in global economic growth, and the concern that the Federal Reserve was raising interest rates too quickly. It was positive during the COVID-19 and the Ukraine crisis, but insignificant, showing weak safe-haven properties. β₂ and β₃ were negative and significant in 2018, emphasizing Bitcoin’s strong safe-haven properties during the period and it became negative and insignificant during the COVID and Ukraine crisis. Therefore, Bitcoin possesses weak safe-haven property. Based on the results above, we can conclude that Bitcoin’s safe-haven qualities for CAC40 are weak.

Next, we analyse the results for NSE50. We present the results in Figures 5 and 6.

Figure 5.

Note: Areas shaded under factor 1 represent significant regression coefficients.

Figure 6.

Note: Areas shaded under factor 1 represent significant regression coefficients.

Figure 5 shows that the intercept term β₀ is negative and insignificant during the 2018 period, that is, it was a weak safe haven against other assets. However, during the early stages of COVID-19, it turned positive and significant, indicating that Gold served as a diversifier. Further, the intercept value was negative and significant during the Ukraine crisis; therefore, Gold could perform as a strong hedge in that period. The β₁ coefficient is positive but insignificant in 2018, meaning that it is a weak safe haven in that year. Further movements show that it became negative and significant during COVID-19, but negative and insignificant during the Ukraine crisis. Therefore, Figure 5 indicates that Gold is a strong safe haven during the COVID crisis and a weak safe haven during the Ukraine crisis. Looking at coefficients, β₂ and β₃, they are positive and insignificant in 2018, and negative and insignificant during COVID and the Ukraine crisis, therefore possessing weak safe-haven property during all three periods considered. As a result, we can say that Gold showed weak safe-haven characteristics during the Ukraine crisis and good safe-haven characteristics for the NSE50 during COVID-19.

Like in the previous analysis above, we look at β₀, which shows a negative sign and is statistically insignificant in 2018, meaning that Bitcoin is a weak safe haven in that period. Later, it showed positive sign and statistical significance during the initial phase of the pandemic, showing diversifier properties. During the Ukraine crisis, we see that the intercept became negative and significant, therefore becoming a strong hedge. β₁ is positive and significant in 2018. This means that it acted as a strong diversifier as compared to other assets. It further turned to become negative and insignificant during COVID-19, therefore acting as a weak safe haven. The β₁ value turned negative and significant during the Ukraine crisis, therefore indicating its strong safe-haven properties. The β₂ graph shows that it has been negative and insignificant during the year 2018 and the Ukraine crisis, meaning it possesses weak safe-haven properties. However, it was negative and significant during COVID-19, meaning it was a strong hedge during that period of market turbulence. We see that β₃ is negative and insignificant during COVID-19 and the Ukraine crisis; therefore, we can affirm that it is a weak safe haven. Hence, we may say that Bitcoin exhibits weak safe haven, but strong hedging abilities to NSE50.

Concluding Remarks

In this study, we suggested a new method to evaluate an asset’s time-varying hedge, diversifier, and safe-haven characteristics. We propose a rolling window estimation of the DCC-based regression of Ratner and Chiu (2013). Based on this, we estimate the conventional asset’s time-varying safe haven, hedging, and diversifying properties during crises like the COVID-19 pandemic and the conflict in Ukraine. For comparison purposes, we include Bitcoin, a nonconventional safe-haven asset. We evaluate Gold and Bitcoin’s safe haven, hedging, and diversifier properties to the S&P 500, CAC 40, and NSE 50 variations. We use a rolling window of length 60 to estimate the regression. From the results, we find that Gold can be considered as a better safe haven against the fluctuations of the S&P 500. In the case of CAC 40, Gold and Bitcoin have weak safe-haven properties. While Bitcoin demonstrated strong safe-haven characteristics during the Ukraine crisis, Gold exhibited strong safe-haven characteristics during COVID-19 for the NSE 50. Overall, the findings indicate that Gold is the better safe haven. This outcome is consistent with earlier research (Będowska-Sójka & Kliber, 2021; Drake, 2022; Ghazali et al., 2020; Shakil et al., 2018; Triki & Maatoug, 2021). When it comes to Bitcoin, its safe-haven feature is weak. Bitcoin, however, works well as a diversifier and hedge. Therefore, from a policy perspective, investing in safe-haven instruments is crucial to lower the risks associated with asset ownership.

Policymakers aiming to enhance the stability of financial portfolios might encourage institutional investors and other market players to incorporate Gold into their asset allocations. Gold’s strong safe-haven qualities, proven across various market conditions, make it a reliable choice. Gold’s performance during crises like COVID-19 highlights its potential to mitigate systemic risks effectively.

Further, Bitcoin could also play a complementary role as a hedge and diversifier, especially during periods of significant volatility such as the Ukraine crisis. While Bitcoin’s safe-haven characteristics are relatively weaker, its inclusion in a diversified portfolio offers notable value and hence it should not be overlooked.

Further, policymakers may consider how crucial it is to monitor dynamic correlations and periodically rebalance portfolios to account for shifts in the safe haven and hedging characteristics of certain assets. Such measures could help reduce the risks of over-reliance on a single asset type and create more resilient portfolios that can better withstand global economic shocks.

For future research, studies can be conducted on the estimation of the rolling window with different widths. This is important to understand how the safe-haven property changes across different holding periods. Further, more equity markets would be included to account for the differences in market capitalization and index constituents. This study can be extended by testing these properties for multi-asset portfolios as well. We intend to take up this study in these directions in the future.

Footnotes

Data Availability Statement

Not applicable.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Akhtaruzzaman

, Boubaker

, Lucey

B. M.

, & Sensoy

(2021). Is gold a hedge or a safe-haven asset in the COVID-19 crisis? Economic Modelling, 102, 105588.

Al-Nassar

N. S.

(2024). Can gold hedge against inflation in the UAE? A nonlinear ARDL analysis in the presence of structural breaks. PSU Research Review, 8(1), 151–166.

Barbu

T. C.

, Boitan

I. A.

, & Cepoi

C. O.

(2022). Are cryptocurrencies safe havens during the COVID-19 pandemic? A threshold regression perspective with pandemic-related benchmarks. Economics and Business Review, 8(2), 29–49.

Baur

D. G.

, & Lucey

B. M.

(2010). Is gold a hedge or a safe haven? An analysis of stocks, bonds and gold. Financial Review, 45(2), 217–229.

Będowska-Sójka

, & Kliber

(2021). Is there one safe-haven for various turbulences? The evidence from gold, Bitcoin and Ether. The North American Journal of Economics and Finance, Elsevier, 56, 101390.

Belhassine

, & Karamti

(2021). Contagion and portfolio management in times of COVID-19. Economic Analysis and Policy, 72, 73–86.

Bhattacharjee

, Das

, & Kumar

(2023). Evaluating the symmetrical and asymmetrical linkage between gold price and Indian stock market in the presence of structural change. NMIMS Management Review, 31(4), 288–297.

Bhuiyan

R. A.

, Husain

, & Zhang

(2023). Diversification evidence of Bitcoin and Gold from wavelet analysis. Financial Innovation, 9(1), 100.

Bouri

, Azzi

, & Dyhrberg

A. H.

(2017). On the return-volatility relationship in the Bitcoin market around the price crash of 2013. Economics, 11(1), 2.

10.

Bouri

, Gupta

, Tiwari

A. K.

, & Roubaud

(2017). Does Bitcoin hedge global uncertainty? Evidence from wavelet-based quantile-in-quantile regressions. Finance Research Letters, 23, 87–95.

11.

Bouri

, Shahzad

S. J. H.

, Roubaud

, Kristoufek

, & Lucey

(2020). Bitcoin, gold, and commodities as safe havens for stocks: New insight through wavelet analysis. The Quarterly Review of Economics and Finance, 77, 156–164.

12.

Brenner

, & Galai

(1989). New financial instruments for hedge changes in volatility. Financial Analysts Journal, 45(4), 61–65.

13.

Capie

, Mills

T. C.

, & Wood

(2005). Gold as a hedge against the dollar. Journal of International Financial Markets, Institutions and Money, 15(4), 343–352.

14.

Ceylan

Ö.

(2022). Hedging Effectiveness of the VIX ETPs: An analysis of the time-varying performance of the VXX. In Handbook of research on new challenges and global outlooks in financial risk management (pp. 384–401). IGI Global.

15.

Cheema

M. A.

, Faff

, & Szulczyk

K. R.

(2022). The 2008 global financial crisis and COVID-19 pandemic: How safe are the safe haven assets? International Review of Financial Analysis, 83, 102316.

16.

Conlon

, & McGee

(2020). Safe haven or risky hazard? Bitcoin during the COVID-19 bear market. Finance Research Letters, 35, 101607.

17.

Demir

, Gozgor

, Lau

C. K. M.

, & Vigne

S. A.

(2018). Does economic policy uncertainty predict the Bitcoin returns? An empirical investigation. Finance Research Letters, 26, 145–149.

18.

Drake

P. P.

(2022). The gold-stock market relationship during COVID-19. Finance Research Letters, 44, 102111.

19.

Dwyer

G. P.

(2015). The economics of Bitcoin and similar private digital currencies. Journal of Financial Stability, 17, 81–91.

20.

Dyhrberg

A. H.

(2015). Hedging capabilities of bitcoin. Is it the virtual gold? Finance Research Letters, 1–6. https://doi.org/10.1016/j.frl.2015.10.025

21.

Dyhrberg

A. H.

(2016). Hedging capabilities of bitcoin. Is it the virtual gold? Finance Research Letters, 16, 139–144. https://doi.org/10.1016/j.frl.2015.10.025

22.

Engle

(2002). Dynamic conditional correlation: A simple class of multivariate generalized autoregressive conditional heteroskedasticity models. Journal of Business & Economic Statistics, 20(3), 339–350.

23.

Ghazali

M. F.

, Lean

H. H.

, & Bahari

(2020). Does gold investment offer protection against stock market losses? Evidence from five countries. The Singapore Economic Review, 65(02), 275–301.

24.

Hasan

M. B.

, Hassan

M. K.

, Rashid

M. M.

, & Alhenawi

(2021). Are safe haven assets really safe during the 2008 global financial crisis and COVID-19 pandemic? Global Finance Journal, 50, 100668.

25.

Hood

, & Malik

(2013). Is gold the best hedge and a safe haven under changing stock market volatility? Review of Financial Economics, 22(2), 47–52.

26.

Huang

, Duan

, & Mishra

(2021). Is Bitcoin really more than a diversifier? A pre-and post-COVID-19 analysis. Finance Research Letters, 43, 102016.

27.

, Zhang

, & Zhao

(2020). Searching for safe-haven assets during the COVID-19 pandemic. International Review of Financial Analysis, 71, 101526.

28.

Kaczmarek

, Będowska-Sójka

, Grobelny

, & Perez

(2022). False safe haven assets: Evidence from the target volatility strategy based on recurrent neural network. Research in International Business and Finance, 60, 101610.

29.

Kumar

A. S.

, & Padakandla

S. R.

(2022). Testing the safe-haven properties of gold and bitcoin in the backdrop of COVID-19: A wavelet quantile correlation approach. Finance Research Letters, 47, 102707.

30.

Kumar

M. A.

, Swathi

, Pallavi

T. A.

, & Bavana

(2023). Volume progression and price–volume relationship of commodity futures: Case of bullion and base metals. NMIMS Management Review, 31(4), 265–274. https://doi.org/10.1177/09711023241230463

31.

Liu

C. S.

, Chang

M. S.

, Wu

, & Chui

C. M.

(2016). Hedges or safe havens—Revisit the role of gold and USD against stock: A multivariate extended skew-t copula approach. Quantitative Finance, 16(11), 1763–1789.

32.

Liu

, Zhichao

, Wei

, & Wang

(2017). GARCH model with fat-tailed distributions and Bitcoin exchange rate returns. Journal of Accounting, Business and Finance Research, 1(1), 71–75. https://doi.org/10.2139/ssrn.3666106

33.

Mokni

(2021). When, where, and how economic policy uncertainty predicts Bitcoin returns and volatility? A quantiles-based analysis. The Quarterly Review of Economics and Finance, 80, 65–73.

34.

Nagy

B. Z.

, & Benedek

(2021). Higher co-moments and adjusted Sharpe ratios for cryptocurrencies. Finance Research Letters, 39, 101543.

35.

Nakamoto

(2008). Bitcoin: A peer-to-peer electronic cash system. Bitcoin. https://bitcoin.org/bitcoin.pdf

36.

Okorie

D. I.

(2020). Could stock hedge Bitcoin risk(s) and vice versa? Digital Finance, 2(1), 117–136.

37.

Paule-Vianez

, Prado-Román

, & Gómez-Martínez

(2020). Economic policy uncertainty and Bitcoin. Is Bitcoin a safe-haven asset? European Journal of Management and Business Economics, 29(3), 347–363.

38.

Ratner

, & Chiu

C. C. J.

(2013). Hedging stock sector risk with credit default swaps. International Review of Financial Analysis, 30, 18–25.

39.

Salisu

A. A.

, Raheem

I. D.

, & Vo

X. V.

(2021). Assessing the safe haven property of the gold market during COVID-19 pandemic. International Review of Financial Analysis, 74, 101666.

40.

Saxena

, & Villar

(2008). Hedging instruments in emerging market economies. Financial globalisation and emerging market capital flows. BIS Papers, 44, 71–87.

41.

Selmi

, Mensi

, Hammoudeh

, & Bouoiyour

(2018). Is Bitcoin a hedge, a safe haven or a diversifier for oil price movements? A comparison with gold. Energy Economics, 74, 787–801.

42.

Shahzad

S. J. H.

, Bouri

, Roubaud

, Kristoufek

, & Lucey

(2019). Is Bitcoin a better safe-haven investment than gold and commodities? International Review of Financial Analysis, 63, 322–330.

43.

Shakil

M. H.

, Mustapha

I. H. M.

, Tasnia

, & Saiti

(2018). Is gold a hedge or a safe haven? An application of ARDL approach. Journal of Economics, Finance and Administrative Science, 23(44), 60–76.

44.

Singh

V. V.

, Singh

, & Ansari

(2024). Bitcoin as a distinct asset class for hedging and portfolio diversification: A DCC-GARCH model analysis. NMIMS Management Review, 32(1), 7–13.

45.

Smales

L. A.

(2019). Bitcoin as a safe haven: Is it even worth considering? Finance Research Letters, 30, 385–393.

46.

Snene Manzli

, Alnafisah

, & Jeribi

(2024). Safe haven ability of energy and agricultural commodities against G7 stock markets and banking indices during COVID-19, Russia–Ukraine War, and SVB collapse: Evidence from the wavelet coherence approach. Discrete Dynamics in Nature and Society, 2024(1), 2587000.

47.

Syuhada

, Suprijanto

, & Hakim

(2022). Comparing gold’s and Bitcoin’s safe-haven roles against energy commodities during the COVID-19 outbreak: A vine copula approach. Finance Research Letters, 46, 102471.

48.

Tarchella

, Khalfaoui

, & Hammoudeh

(2024). The safe haven, hedging, and diversification properties of oil, gold, and cryptocurrency for the G7 equity markets: Evidence from the pre-and post-COVID-19 periods. Research in International Business and Finance, 67, 102125.

49.

Triki

M. B.

, & Maatoug

A. B.

(2021). The GOLD market as a safe haven against the stock market uncertainty: Evidence from geopolitical risk. Resources Policy, 70, 101872.

50.

Umar

, Bossman

, Choi

S. Y.

, & Teplova

(2023). The relationship between global risk aversion and returns from safe-haven assets. Finance Research Letters, 51, 103444.

51.

Ustaoglu

(2022). Safe-haven properties and portfolio applications of cryptocurrencies: Evidence from the emerging markets. Finance Research Letters, 47, 102716.

52.

Widjaja

, & Havidz

S. A. H.

(2023). Are gold and cryptocurrency a safe haven for stocks and bonds? Conventional vs Islamic markets during the COVID-19 pandemic. European Journal of Management and Business Economics (ahead-of-print).

53.

Yan

, Lei

, & Wang

(2022). Bitcoin is a safe-haven asset and a medium of exchange. Axioms, 11(8), 415.

54.

Yousaf

, Plakandaras

, Bouri

, & Gupta

(2022). Hedge and safe haven properties of gold, US Treasury, Bitcoin, and Dollar/CHF against the FAANA companies and S&P 500 (Department of Economics, Working Paper Series No. 2022–27). University of Pretoria.

55.

Zhang

, Zhu

, & Xu

(2021). Has COVID-19 changed the hedge effectiveness of bitcoin? Frontiers in Public Health, 9. https://doi.org/10.3389/fpubh.2021.704900